<%BANNER%>

Acoustic measurement techniques for suspended sediments and bedforms

University of Florida Institutional Repository
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20110115_AAAADH INGEST_TIME 2011-01-15T19:38:19Z PACKAGE UFE0008389_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 1706 DFID F20110115_AACGKC ORIGIN DEPOSITOR PATH mouraenko_o_Page_051.txt GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
c2425ef84b6a01bbeedaa04def8344ec
SHA-1
4e1d078ccbf18595d992b0b5a9278e4df22bb7b2
1967 F20110115_AACGJO mouraenko_o_Page_030.txt
be81699ab9847af5d2ec3e53f49d8b89
f247dd2fa63d41edcad61ef19476c19bd695ca59
1474 F20110115_AACGIZ mouraenko_o_Page_003.txt
14b7fa2754493561ea0477dd28b533c6
107d83932fcea9da4864c622a0d8443512a8427a
354372 F20110115_AACFHA mouraenko_o_Page_013.jpg
125f07518f65cc47d63879c0422b1774
08fbacb5bbdad0de8491538ddc012ac4b455c37e
198940 F20110115_AACFGM mouraenko_o_Page_026.jpg
2ec02dc83a48e3e295858922782e3ae8
c39667be0f687a53c9cb8852d902dff616dab026
1901 F20110115_AACFFX mouraenko_o_Page_095.txt
da14da60af32e5a3cbb69b09d6452a1b
cd174f233de7d6514376434486c8431bf941fa09
1949 F20110115_AACGKD mouraenko_o_Page_053.txt
3a9e038b7e0ec70cde5c20eb3bef6993
3c4d00b5c8a10c99d5860c49eac69972f9024ae3
47145 F20110115_AACFHB mouraenko_o_Page_152thm.jpg
671f8fdaaef3a291d1e7c1d48d788cfa
9cf052adc5ae5da74b187a2f55233c990711defc
231368 F20110115_AACFFY mouraenko_o_Page_059.jpg
1ebeb4b41f92ab3701395e6fe06dc5df
71aba511fc4c29635a9adac3b30774de273f53df
1917 F20110115_AACGKE mouraenko_o_Page_056.txt
04eea3a972b3b19e1514bb568187f264
505bd033db5253a6030e6e2bad4fc7fdcd4414b3
1272 F20110115_AACGJP mouraenko_o_Page_031.txt
b17627fb87295428fec3a7a6d7aa8c3d
feaa4eca84ba9ed2c4145960fdb30e4f6790f20e
1488 F20110115_AACFHC mouraenko_o_Page_049.txt
dce110ab8c9381866342224146363d7c
909ab391cfe3705d23e100aa98cf0bab12c20f45
52060 F20110115_AACFGN mouraenko_o_Page_093.pro
ad4833df6f605e6b65c9a9b7a200c506
18cc79303ba82f352ab5e67429459dd065062b61
862266 F20110115_AACFFZ mouraenko_o_Page_088.jp2
cc6b613cc52c313237955be23d7fad75
bdf59fb6b0f1a5b249785ac2d0758ff37e014624
1938 F20110115_AACGKF mouraenko_o_Page_057.txt
4a83764080c0042d3f0b2353d0f06595
7d5e0eea7c81ebff27fd06c9f64b1a55f1d835c9
1549 F20110115_AACGJQ mouraenko_o_Page_032.txt
1ad18fad3e24d7e20f3e64b2a0a2ad37
48288e250802bcdda52396d123a8688d31a320c1
86569 F20110115_AACFHD mouraenko_o_Page_119.jp2
0c4f6a32ea66a6c6d5c5c834b00306ad
4924bddd0bf001a1a18db71d203b976ac713b86f
56956 F20110115_AACFGO mouraenko_o_Page_166.pro
079f96719c8cb56876fa2136fa645843
d46aa56f1d674278f07bea48090eded5a61f2bd9
2002 F20110115_AACGKG mouraenko_o_Page_059.txt
9358b19b6cb257b332f7342a85a56472
06afd50cca29871bd0f35ff9ad1dd8d1aca73d55
1744 F20110115_AACGJR mouraenko_o_Page_033.txt
967440495b097338e192372916d3db50
e8a68a7e4a3dd5bef96eb5d7398c6f52aafc7a5e
25271604 F20110115_AACFHE mouraenko_o_Page_098.tif
6e1342a842c8f7b2a294151985dc52f2
b11ad4dc93659bd39c89cbe4a9d9cc10732a8346
181110 F20110115_AACFGP mouraenko_o_Page_160.jpg
7510c3ebe88f31a60f1fd84576c696d7
2552cee2112f446a96c6baabd5287c3dacafe5d4
2287 F20110115_AACGKH mouraenko_o_Page_060.txt
9c113b68764ed6a3f1ed40b32fde89d5
7d3a5533eaff558f5ee76d1acc9e621afe67328d
1546 F20110115_AACGJS mouraenko_o_Page_035.txt
25a61688ed7f70e95c9868ae2c521b64
61dea8ba7723892cfe5ac1ae5e578e0dc2120d5b
21722 F20110115_AACFHF mouraenko_o_Page_160.pro
1b4560d85527b4e033d0d267406a2aee
b8b789f6edd10044d853fa7e1711f48de3f73c51
F20110115_AACFGQ mouraenko_o_Page_061.tif
c28a0c7ee26bf8971654dd10ad9ee059
e7c261f6469829ac8c2040901f877553c1aa98e9
1669 F20110115_AACGKI mouraenko_o_Page_061.txt
0d4fab5fb8a44ac0144099dc3dcad6b7
c746b8b6270955b2e7f6f5f118a21f380f131225
1711 F20110115_AACGJT mouraenko_o_Page_036.txt
bd2a0c4a0e2a4b0f5802765b4733016d
5516b057605c337ab8c8fba6aa5d800d4ed19d2e
F20110115_AACFHG mouraenko_o_Page_044.tif
42a2d5ae6b448fbb09df6cfd724615f2
4f331aa1362225fa940e8b020131ad7ecd099bbe
161744 F20110115_AACFGR mouraenko_o_Page_054.jpg
3b71407df9d6d9dc1d775be21c080988
fe3b182316dc330720c3e2746dfbb79759ea4e12
1580 F20110115_AACGKJ mouraenko_o_Page_062.txt
211be662b9d616e442ddda628da483b4
bef6812dd5ec2753decb7d9b251d9f6d6c334ba5
1489 F20110115_AACGJU mouraenko_o_Page_037.txt
d16d717035678a35ed65a6081cb6b706
898840750543ffb1fe227af86bf22fcf37c8ee22
1555 F20110115_AACFHH mouraenko_o_Page_052.txt
0d7e76909db92064d758bcff971292d1
9dc7404dea922ff4b7810439b7a9ff263d2c6b73
1959 F20110115_AACFGS mouraenko_o_Page_120.txt
ad5c8f68278fdba895600b100844c90d
6b1e412a424fae3fd74a09b70481e0b3877a6891
2839 F20110115_AACGKK mouraenko_o_Page_063.txt
cc4af8165712df08a4772e6ed8d250cc
f6767a92facbb1fed2a9b6c462905a5c6c56956d
F20110115_AACGJV mouraenko_o_Page_038.txt
9ef65ed55208adcbbdc8a1b07113d348
183fe5c050e2fb04123d2f85533e925d878577e1
46699 F20110115_AACFHI mouraenko_o_Page_061thm.jpg
0df936acd046f57893edd24af23c9646
7d3c429e56e26bcc8ab9f4a9cb56f529c382fd65
49416 F20110115_AACFGT mouraenko_o_Page_166thm.jpg
2760f5d494f070185a7d42d6ab694fbb
7aece1e54a4040cfdddc92e4ee9365bed02bc3cc
2403 F20110115_AACGKL mouraenko_o_Page_064.txt
5a270021ad5c00fd8412688416557b55
d834a41fdc581dc46f1386abb04d1b80d914baf6
493 F20110115_AACGJW mouraenko_o_Page_040.txt
8338b39bdad4b1644fa68f7f0a2f3d2b
98a1b6eb827ffd657b3365bb03bc7f45b08ba60b
91215 F20110115_AACFHJ mouraenko_o_Page_005.pro
9fbbb03c3df22637e0b6d72907b7f307
24f18248b534e17c4ff43d50924f95c7c7884cae
F20110115_AACFGU mouraenko_o_Page_024.tif
21cb695939211a80a8c0fb5a3fe8713c
790151b8066b57ca9f87f59adaf90e0d0e3c7d92
1332 F20110115_AACGLA mouraenko_o_Page_084.txt
f52514af64116e156b7ef208827e08b8
c0326535389a9821d8f28f4429a004958745e499
2278 F20110115_AACGKM mouraenko_o_Page_065.txt
5465e45bed4c2bd5e149621429e69436
ac5220c663a8da5961a0c6a02dfbca1349de7e2e
814 F20110115_AACGJX mouraenko_o_Page_041.txt
bebe8729a04d7e6e79df2e77f81e016f
4fed2ed8cb6ebb180e65ef90e8e140e6047ba09e
F20110115_AACFHK mouraenko_o_Page_093.tif
afe710773c1e31ac4100264f579655a9
c84cc293b9ae23ec91913816f3e35555df6d93e7
1051950 F20110115_AACFGV mouraenko_o_Page_077.jp2
e410bfd7cff14514f050b35bccfcb6d2
90e8290c3beb336e5787d52127b982eb50dca56b
1327 F20110115_AACGLB mouraenko_o_Page_087.txt
80973067f279e771024ecf63a99ba54d
d6e5ddba4f5554fa3eed234992bfc3438ec1e032
1822 F20110115_AACGKN mouraenko_o_Page_066.txt
b53d155d3a8e42bffaf653626c076b53
7a7145d9a67e78dba3999d5d3663be42de6b8f47
1461 F20110115_AACGJY mouraenko_o_Page_042.txt
a26abe4a53716fc7f7eab23ee9399bb6
4551714a5b1b26af67ec5973245c2a3a5ff25c1d
F20110115_AACFHL mouraenko_o_Page_130.tif
1444eba932eb21b842b703a58ba228c0
353ff59b73d327dde4c16d94b24124d819fdc89c
69444 F20110115_AACFGW mouraenko_o_Page_048.QC.jpg
41c131fecef998ef1ab6e2313af03418
bad1d482c99549f096ac35ae8ddf419c469bdfa5
1437 F20110115_AACGLC mouraenko_o_Page_088.txt
f9dc43e48a55050f0dff44412ad5b51a
9fc49f540051bd6017915caa6efb13d53c0aa098
1808 F20110115_AACGKO mouraenko_o_Page_067.txt
049b8f65e28af3cf513781d5e22588c4
a5b07c58ce581ad603b7f4f00d4f6659d8376fb2
1732 F20110115_AACGJZ mouraenko_o_Page_043.txt
6a69e08c6acf153534d55e19a53161c7
ac23ddb568e213d46fc43f19a4247aa55a7b1b43
28252 F20110115_AACFHM mouraenko_o_Page_029.pro
a1462a4b8d1e1022c8b0da686d1d92db
9fea4f279daa0e8054cc53c027db5689919e80b2
51696 F20110115_AACFGX mouraenko_o_Page_063thm.jpg
fc866e88603a70eabb04f0875617771b
a6f5de04b8f37676557dda07e17aeda6798150a0
F20110115_AACFIA mouraenko_o_Page_096.tif
c52ce8a485eced787025f682069c0058
388b87877a56422599da845531cead963f150a71
1621 F20110115_AACGLD mouraenko_o_Page_090.txt
a4d8d82346d9dfe53093480a25fa3e21
46d6d3467c03a96be8307b882f9a96b060a96e69
1873 F20110115_AACGKP mouraenko_o_Page_068.txt
cd5406ce503ae161e1341ea2289acdf4
d943de78158c76a518ef27d9c05349d24faaef3f
101582 F20110115_AACFHN mouraenko_o_Page_055.QC.jpg
77d68870f675419eafcb28f830cfcf25
81f3442e7da46e50d8530804de1dac70b25a07a5
90278 F20110115_AACFGY mouraenko_o_Page_091.QC.jpg
e6fd9b1e3c1592e975ce5b9a7320859e
ce35737120f9011a00749e3d9e7fc1453c8f12b4
1964 F20110115_AACFIB mouraenko_o_Page_121.txt
5f7060e02e4b6804dc604abc30a8d175
aeff3aebcaaaf674fff524b440690886f66e517d
1973 F20110115_AACGLE mouraenko_o_Page_091.txt
ecfe437ff7f4b2515c84ad07ef68d35e
8716520087c9113d261a172e346bd023fadf025f
87635 F20110115_AACFGZ mouraenko_o_Page_123.QC.jpg
ca5047403c511992bba0e5e25aad2db7
41deeb4a2298bab2730ff8752530c091507fe88f
1884 F20110115_AACFIC mouraenko_o_Page_069.txt
f2f2a5a49c024914aed637507a0c07c9
e5042c3b086b32fac9595a9c890ae8b4c70c2d79
2595 F20110115_AACGLF mouraenko_o_Page_092.txt
452aaaf0442001ded050898828867661
fc40f26687d90290102ec0c89b1e587739edf9b7
1418 F20110115_AACGKQ mouraenko_o_Page_071.txt
c5d19c859ae2785cec7e0e3b11790d0f
ea45d7267179d59f583bbce2ad39b6a26acc7376
1051915 F20110115_AACFHO mouraenko_o_Page_140.jp2
097326139a801a02fba7f320b6b61fcc
06f2c0caaa222095cb15a9ed4a37ef8a4536a7a2
1051972 F20110115_AACFID mouraenko_o_Page_118.jp2
00a0c29a046557125006f6d08643e86b
db37741b1ca5ca2ef82aa0b34594b91c2ec5af74
2052 F20110115_AACGLG mouraenko_o_Page_093.txt
f9dc99d38d02266cf2d405656e99804d
f7ad2cd64e7709095f6bf92e3b7bf7eb534b79f7
2015 F20110115_AACGKR mouraenko_o_Page_072.txt
425d212c887e5fbf927826a437b2d145
d105e3c5d4abbefb1aaf2190730604567d091bd4
241988 F20110115_AACFHP mouraenko_o_Page_126.jpg
234555ddafe55b7d26bc2765ff97e962
33f64ff1ad12ee4a68fcf8f4503e1f976f0efc59
49213 F20110115_AACFIE mouraenko_o_Page_111.pro
35e3cfc3474b193f406a75fee6426a66
8ac264f2665dd85d841b6e39a8549c698ba41aae
2184 F20110115_AACGLH mouraenko_o_Page_096.txt
3812c09633f0aa5e198c312fe0fd2803
4da4f453789e3f2d9446ce89e05f138c309b1cfe
F20110115_AACGKS mouraenko_o_Page_075.txt
addbc5c043c0b0cc4f6d1e1c2a514673
883cb833e9dc108e2ce3b78d6cc9aa5ac0c4c2b3
204597 F20110115_AACFHQ mouraenko_o_Page_104.jpg
9001c2178e37ab82e8c0111a052ea423
2d0b3687fcfa98750923334dd2759a8cca6c5c73
78629 F20110115_AACFIF mouraenko_o_Page_158.QC.jpg
4039b55172653d6ea26ef6161f60a2e9
c949b7a1e848d270ce3496ab11092158dd1e5a98
2016 F20110115_AACGLI mouraenko_o_Page_098.txt
8a0677373177e2af4ec1b32f70da54e7
b75476885010b7b850de5b2b42ac68a2f59dd243
3101 F20110115_AACGKT mouraenko_o_Page_076.txt
294c2f0e100489e0bc7345c7753ee94e
93ebfc0eeb47c843d45b993cac373642f04319f8
1053954 F20110115_AACFHR mouraenko_o_Page_051.tif
642ff1580b839670db9eddc721b893a2
db82addcc21f698798c9d6f02f6abaafffdff93f
476 F20110115_AACFIG mouraenko_o_Page_058.txt
ee5e7c854ea2a527d2ff89c84990b74f
4bfee0badaebdc3df356f69c6a52af49c7504ba1
2412 F20110115_AACGLJ mouraenko_o_Page_101.txt
b528606aaca57d7ceaa5950426933c86
c3ba780d5bb12d99070b66a5385f98659a030abe
2419 F20110115_AACGKU mouraenko_o_Page_077.txt
074c64d9dff7f30641cbf1a44aec732e
bafaf28d8ed45523873e44cd189ddef886ef2231
2175 F20110115_AACFHS mouraenko_o_Page_097.txt
01b037f3c503eea93092ff5c7d34feb6
52b6bbe6bf45da17f07553f1de4f01468d85645a
1051980 F20110115_AACFIH mouraenko_o_Page_055.jp2
61b59830cdb8b94aaa6ab70b6428b420
40f9c5988505d5e726d2ae2b191d02be0d6a159a
1781 F20110115_AACGLK mouraenko_o_Page_102.txt
83a03d9609c6ccb609f1368d27d3b9ab
0f72ba629ed5a41a11287da02b41b27ce7233999
1752 F20110115_AACGKV mouraenko_o_Page_079.txt
5b16d93a4f5537436f33e3df56e09644
3986d44e0a9bd71c75c409c3d794824aab94cbcd
200570 F20110115_AACFHT mouraenko_o_Page_083.jpg
6af207c63caa0d2a400b99f123a7433c
2c23bb96ccb9d2c756e8fc4e0e183b223d78ddb0
47702 F20110115_AACFII mouraenko_o_Page_096thm.jpg
ab56e4dd38b0a5e7f8f2a3310eb76215
594ae6433e08344457557c03ebac01e4102e5545
1693 F20110115_AACGLL mouraenko_o_Page_105.txt
13703d24de54e5fceb1aebf6417cb97a
236f58684bfe4f1c75f4fdd36196af0d9c383bc7
1641 F20110115_AACGKW mouraenko_o_Page_080.txt
155592510e31c3efd0e749706488d247
e5f26485ce9a8de4745a7d1654f208ccf2d35c4b
44037 F20110115_AACFHU mouraenko_o_Page_053.pro
29bd534717beb8cd81a54cd82f3cc53a
663b9560cd29fbcf9ade8cedd23762df67e46655
236100 F20110115_AACFIJ mouraenko_o_Page_128.jpg
f5ceab926ff0686ea242e18c9ac2f6cf
fdc8ec9ca17be9389596c4105ea6f3968e0c0ddf
1793 F20110115_AACGMA mouraenko_o_Page_133.txt
1cedda1abb20508df05a38eb78a2e944
7462931b4e526cf115b2986066a76d23dce22dfe
1878 F20110115_AACGLM mouraenko_o_Page_106.txt
5185ee7373cc2faf229c87d8bda69fec
ac5aa25143e06ffe9744a65a8584e8e3e0ad6b0f
F20110115_AACGKX mouraenko_o_Page_081.txt
e1f5099ddd3b9fafbd9b722a7f3a9e51
e329116ef56fb3094dfaab4e808120e787288561
F20110115_AACFHV mouraenko_o_Page_099.tif
fc033a3963a23d07c42869176758b4d5
ae75245005d3ae32c477ccff4299b76b241a419f
2124 F20110115_AACFIK mouraenko_o_Page_021.txt
5bc1650c721e1c5577f3926317b8e8fe
4117bc989044929f7ba267fda3fca902f5bdafdb
1617 F20110115_AACGMB mouraenko_o_Page_135.txt
c239e5b3f7cc7512a58a8bd600c9580e
10818377278f6fa43b0a4cbf57caf1ea97425b83
2001 F20110115_AACGLN mouraenko_o_Page_107.txt
be9fe82fca62f1dc995f1410553cafec
234eedbf1c154ec67ff12af10013f301b8d32a2a
2120 F20110115_AACGKY mouraenko_o_Page_082.txt
ba9d29d31800687a2cf6d39dc8b86cd9
0574dfe47d16a94d0f87b75c61935392e2d86e26
2473 F20110115_AACFHW mouraenko_o_Page_078.txt
91f639dd48fe84cfc1b9d7d331704abe
01ff9550c752afbc61f56f5880467add5dce875b
767360 F20110115_AACFIL mouraenko_o_Page_105.jp2
9adad8cc22a9c0e5b2f477d3e25c62f7
a9e9e05762b552f590239dceaddc665a9ed78923
1823 F20110115_AACGMC mouraenko_o_Page_137.txt
817cd0fcbcd92dc2c2e39a317305b301
087ff153d3ab36d7a1817a2b7896a8898c762674
1692 F20110115_AACGLO mouraenko_o_Page_108.txt
c7b4bdcd26b13dd012667fbf698e390a
c91714e7a8ac4dabc1292ba9bfc3b63b3d181adf
1886 F20110115_AACGKZ mouraenko_o_Page_083.txt
a423f82f33502d861a7f9f1f88d33110
098ea1526845ae6e510b6beb45507d6f77b4e60e
1995 F20110115_AACFHX mouraenko_o_Page_022.txt
87743d3950120187a9575d2d24f499d3
eee1f2c3c76245db8f1b145975ef97c88f58c0a2
46468 F20110115_AACFJA mouraenko_o_Page_147thm.jpg
c44be6cfcf6baa61ba2860c6c57bb402
1705e598484793d699ac007eb9e8124abd16cfab
174331 F20110115_AACFIM mouraenko_o_Page_028.jpg
82efe40bb47235535d0f5be406c029c0
7dc43170f3ecd4ac664b42e3d8b0e844a3a2fc43
1647 F20110115_AACGMD mouraenko_o_Page_140.txt
c10469ff8fe2bc2c3a0652ff40a1ed8b
0e6e60c2e1c1714c0e9d4fd7c408255e5ca53a27
2165 F20110115_AACGLP mouraenko_o_Page_109.txt
0e480356328ff2a3387ad20b846aef6d
f6a508f3e7412facdbb3a9f1df15b77c2cbb115f
71203 F20110115_AACFHY mouraenko_o_Page_047.QC.jpg
88f46fa5c93e878056829eb588c3b0e6
752aafcf7f89db352885b920bc189e9c32de2946
31584 F20110115_AACFJB mouraenko_o_Page_135.pro
d14c3fd13202340619293f8d733a3583
562e725f7e8cfe501585223af3032f824d1878c9
81572 F20110115_AACFIN mouraenko_o_Page_090.QC.jpg
2d572acc00b1185252db5fc81f1da649
8e97e9d6c9025d2beccbaf6aec1ebc787c61fcf7
933 F20110115_AACGME mouraenko_o_Page_141.txt
ccfbba9fdfac9d4a96f7b80c96894209
dd07371493dbca8ff7f487de96c8095987d41f6d
2360 F20110115_AACGLQ mouraenko_o_Page_113.txt
590eb5d908909fdd50ab9c938e829124
d600f701b051bda6cdc0ed214e37de8f2d0f0a4f
64116 F20110115_AACFHZ mouraenko_o_Page_007.pro
7fcc7194a8e05aae37c354859d689700
ba417d2d7e1594fe2bc1db76d73983b77cf08fd0
49046 F20110115_AACFJC mouraenko_o_Page_107thm.jpg
4184e47681c795c8d36f8c6a04048b0c
34a6aef05ae83cac076565d2d88cb131e9342a24
256110 F20110115_AACFIO mouraenko_o_Page_071.jpg
4e4d32a04697fd64fa70c39c8ba1105f
a7294c107cb5de6fa7bda5a46a27d8c7a2a072e6
1860 F20110115_AACGMF mouraenko_o_Page_142.txt
59a4efe617d85cd76b15fd55d8946864
d8d8d44e0bb3fada9c453c0b8fef8c528907ffae
18865 F20110115_AACFJD mouraenko_o_Page_019thm.jpg
1431bb4577727c31c6232a63db57c4e5
f301a2b841a363321d1cba8ed6015f4e2af90def
1862 F20110115_AACGMG mouraenko_o_Page_144.txt
01b8ff170a2ca6222c5bde2475c5cea0
fb4771a98d0af9caf8dea337c8047a0f9024969a
2599 F20110115_AACGLR mouraenko_o_Page_114.txt
60ece4db4f799b08ee91935e240db17d
4678a5d2c89a520c1b66c4a2cee7219d9879d9f1
49705 F20110115_AACFJE mouraenko_o_Page_077thm.jpg
c29b31a5e04b96b058e06aab619443f1
9b6e7c10ca59527b16fdd16f5db5cafde635e302
34907 F20110115_AACFIP mouraenko_o_Page_138.pro
6184256b80af38106c23f8f81aa6ae10
b729014e049f3f1f03ed9217c8ecced8d96c97ad
1854 F20110115_AACGMH mouraenko_o_Page_145.txt
8778be7913479ccfdfe461f2cd29b5c1
1b1091a6132813ac7d88c247cc23098a356a185b
2113 F20110115_AACGLS mouraenko_o_Page_115.txt
9cf54f49eb92201ee34afae4461155d8
76fd1bf0409dac45448e61e1f436f0851d4710eb
78277 F20110115_AACFJF mouraenko_o_Page_056.QC.jpg
0f4e8409085de2a75d618d7564de564e
9be8c383aca9067a690b3c142b2a3d4894648e5b
1108 F20110115_AACFIQ mouraenko_o_Page_017.txt
8c80f6671cef13baf42aaf93bb52017d
f0da4aa3efcea4c0b5f56cda790dec44dce60650
2009 F20110115_AACGMI mouraenko_o_Page_146.txt
35cfad0b2151189b8f338bab927b5d99
1f3d8b70e97758be781aa90b4e0cd9940afca4d1
2118 F20110115_AACGLT mouraenko_o_Page_117.txt
cc78c4771d66edf4d8931e5a30fe45df
401766bd33701095ce9e4d897121334d06ab8c91
909994 F20110115_AACFJG mouraenko_o_Page_153.jp2
acb8d3a2e1300bd40cf5c51dc9a9e9f0
e9c96e08c64f7880f9edfe77acf58ae63d167cf4
34138 F20110115_AACFIR mouraenko_o_Page_090.pro
98a92263256ecc4391e9cd7c733a32c6
4418d161deda5613edf6590fe449c25be157647c
2065 F20110115_AACGMJ mouraenko_o_Page_147.txt
d0a0b3b29ee3bca0271371628d2c6b21
837f0ca06f839aceb460d3b3b3190d67bbe15377
1402 F20110115_AACGLU mouraenko_o_Page_122.txt
b90a181fa56e7ecb68eae6b7c161b98d
1066cbfd56a4d66d8c4d088f91c1ef2ff40fbd4c
88899 F20110115_AACFJH mouraenko_o_Page_028.jp2
f713e2bf2fedfab94b1f75fbccbb141c
218bb3ac6ff757540bd82dba8cbf6b9d7ae6e752
1051973 F20110115_AACFIS mouraenko_o_Page_060.jp2
761256822d759fe61f156462e6ed8258
5e66294e9938ed40fdf4010d3421e27de14f81db
1442 F20110115_AACGMK mouraenko_o_Page_149.txt
c843fc3da37f51c35c31de527f0afdd4
07c22193b2c510c4b54e1f7c6c12a44b4178b779
1827 F20110115_AACGLV mouraenko_o_Page_125.txt
32e20e3ed20d0f76a7afa6c1989f45cd
d3e387824a766c6eaa4292a5e142d8f45ea9db94
234773 F20110115_AACFJI mouraenko_o_Page_089.jpg
640964974101f7cd6f6b770dacf056d4
05edc0413b4a966fdf33405cd051b2edb8fedff4
947455 F20110115_AACFIT mouraenko_o_Page_053.jp2
51f8541f8a13069a813e70be6bbf6235
7b1a88fd78bd365008e03da6668d53223ffef6f4
1500 F20110115_AACGML mouraenko_o_Page_150.txt
afb91d3e18bdbb17d7b592af674f30ae
7e624a1ca7af5647b177e3e6c91b7ffb0e1ebc1a
1791 F20110115_AACGLW mouraenko_o_Page_127.txt
21d7f04c59308409d08bf2d04cdf7b4a
e0df198a4964b5df24d08107f2725666c4031ae1
F20110115_AACFJJ mouraenko_o_Page_169.tif
746c3aff76cacd08f668fa3860bc0dfb
340cfe994bec62ef6727e1e77c2cc4a2a70a8175
76229 F20110115_AACFIU mouraenko_o_Page_116.QC.jpg
5c0f0d366fe970556fdfd1ca3db08d13
532dcf55de478f8a4441600671317661a90dc7ce
18960 F20110115_AACGNA mouraenko_o_Page_003thm.jpg
314e13aa29ee50cbfd38bb01cac1129a
dad33095be04d3cc8cc3eb1016d0b591457ac304
1929 F20110115_AACGMM mouraenko_o_Page_152.txt
15a65d61869a8db366842e121f5728ec
06878782df29b4b39b6d014b0e140473fb4793d6
2099 F20110115_AACGLX mouraenko_o_Page_129.txt
2580d34b7058fc7caffd608408cb2ad1
19def01dd0eab931ee8bf2476e4f62b9cab5658a
1759 F20110115_AACFJK mouraenko_o_Page_094.txt
5db13e07eeb0e3470e6986a1a8766051
578a5d57263943ed7dc5aea0e80bcb41a39589f3
F20110115_AACFIV mouraenko_o_Page_166.jp2
fa3f882e878d0efad0be2127ee9dae09
64690296d547b9392fcb8422ebcfebcd58ad2796
59415 F20110115_AACGNB mouraenko_o_Page_003.QC.jpg
7e7d201adaff5704321071d7c6af98f2
3130f9ada012e0f36d64151d929d405d0e782532
1947 F20110115_AACGMN mouraenko_o_Page_154.txt
3843129da764fe0b4f4ac597f69296ef
9aa87f16e435809bbba8bca7f2d733210efd9db1
1982 F20110115_AACGLY mouraenko_o_Page_131.txt
e1e9bb057a27940874a685c8ff1bc53b
f665009e34f0258ef123c4acc9621bb540ff57da
850719 F20110115_AACFJL mouraenko_o_Page_108.jp2
1bb8b2a7dcb80038f909a51f3ced096f
3691df022d0c19c232e5800188f5a023b7a4ffb1
26200 F20110115_AACFIW mouraenko_o_Page_165thm.jpg
5f980d63c73143bd7e59adb2a1db3d1b
af4537527c54c8efe4518539147d1b77fcd169fa
47064 F20110115_AACGNC mouraenko_o_Page_004thm.jpg
61e22e7f0f04983e2daa1ad7b2e52aa1
5b7ab913b4ead4fee32776254e50324442d3bac9
1905 F20110115_AACGMO mouraenko_o_Page_155.txt
8ba8c7043d22ab1885a1a19c3ee8995c
21ef5cb10ffe17f4e00fd8e0ca74ea306af3a1da
1948 F20110115_AACGLZ mouraenko_o_Page_132.txt
34056461e6be076e1733aaf57e8f4c18
d1996d12af60a8e6e7a082d9a64cd58cb463ed75
1051975 F20110115_AACFKA mouraenko_o_Page_167.jp2
0efd00fb86b13dbb87891affaeaa6394
a038d8f096df32c42e346e0fe964ea7e9d2dd863
1051966 F20110115_AACFJM mouraenko_o_Page_004.jp2
c6d7912d4ce7e49e71a68b9b06ec8120
4060c16ee9842340cfd1b67b8967d4f2eb24cd22
79822 F20110115_AACFIX mouraenko_o_Page_121.QC.jpg
20cfd67113a11542b53c3edf5419148c
10bcc569a1db4a1ffe6a6336dc12d04cdb191214
44157 F20110115_AACGND mouraenko_o_Page_006thm.jpg
bf113d23f03debcf0b6a183711ee798d
42ebc2410a05e7b5909845e4f27294388477706b
1103 F20110115_AACGMP mouraenko_o_Page_158.txt
67768f5fd5e7b254214db824429fc61e
458e5037f03aa209a92e7ff021bea1204e180478
153378 F20110115_AACFKB mouraenko_o_Page_018.jpg
5843015b2454d71abc23661d268ad8d3
be8f064a74450cc6fd56238f0c98cee876d4eac8
45833 F20110115_AACFJN mouraenko_o_Page_082thm.jpg
05715237f727d9405b08d210e48694a8
278abece91e8107e432a5b04f4a484cf786881ff
3991 F20110115_AACFIY mouraenko_o_Page_128.txt
1baa272db4b5a9e58526b7e0450d6c4e
8f4a501c0f52a8fa44dd113c4668c3c0fd9e478c
107418 F20110115_AACGNE mouraenko_o_Page_007.QC.jpg
ea46d9789432238d6bad9ed61d159a2d
a5747c1fbebe18bb5df1cdd109bffd57d587dbe1
1222 F20110115_AACGMQ mouraenko_o_Page_159.txt
0c5d99070711adb78d5d42ef19026627
b5a6df2036a5eb70a7ef073faab5e2cbaf477215
58642 F20110115_AACFKC mouraenko_o_Page_169.jp2
ab4a2a12158920a97a1fa2c2fbe85fa2
c00886a34b5bca9201fd88e0b4a43f7f0cbae51d
19832 F20110115_AACFJO mouraenko_o_Page_050.pro
51cf94a492ca0f3fb4f6b90c46ee9bd1
c8674d54abec98bf1032029d131c139087e9c185
185240 F20110115_AACFIZ mouraenko_o_Page_020.jpg
c7fa565b2077f2015e36538905e9514f
96de6177e492326b57bbc3c0ac0b3cddbc07ae03
132956 F20110115_AACGNF mouraenko_o_Page_008.QC.jpg
c483b7cf39d045345dbbeba30e6f8312
183da694b86f1c9abb5af3b8357718d909af7248
1109 F20110115_AACGMR mouraenko_o_Page_161.txt
bf837f0918971b316970b4b7ff07ae3d
01e7b12884eb36b1ee6c645d476cbf6fc6121781
86695 F20110115_AACFKD mouraenko_o_Page_165.QC.jpg
fc7c7543edd4c87a0c79fdea4d81cf3c
d7ac5ecadcc51ec617e8ce12cf8b3326c6dd25b2
23975 F20110115_AACFJP mouraenko_o_Page_168.jp2
7a901da00f3e61897b333881f706f890
e276e34a875d15b1d068d7138702361ff5c5dc92
57518 F20110115_AACGNG mouraenko_o_Page_009thm.jpg
a692603dfac023569d4382387c3728c2
1d7e0594c05973e9af2503aacc771fd2afea45ed
F20110115_AACFKE mouraenko_o_Page_004.tif
33f709bf9562d98073dbd902635b5bb4
df6de31a153e9a336cd07e66d08678b1134b80ee
132444 F20110115_AACGNH mouraenko_o_Page_009.QC.jpg
cbe413168b74349fc979bb26e3902a29
cf2c8fefa483296cc81d0b2f200605cac16e50ed
1150 F20110115_AACGMS mouraenko_o_Page_162.txt
0348995761dd72019dc369af2866ccd7
0c6c1b38071c2dccac9e217ccbb3c23921dc35a7
60827 F20110115_AACFKF mouraenko_o_Page_025.QC.jpg
b10a33f47cdef447887ff5cd620166cc
489e86ef94bc2e989a479bf46741ef4b1dc97e50
231627 F20110115_AACFJQ mouraenko_o_Page_097.jpg
3e28f27733a2c396253674bedb919614
0e165cad024dd81851a90c041e585978c50eaa53
56494 F20110115_AACGNI mouraenko_o_Page_010thm.jpg
93fa3c44cd327b410d9f3aa51e2d54a5
2d88770462cc47aaf92037503ddac027efbe141c
2319 F20110115_AACGMT mouraenko_o_Page_166.txt
03b6e12d00d071f5adf4e5a6e9635648
c41dc96cda543b14f4c8eeaa136308ecaf96c238
1090 F20110115_AACFKG mouraenko_o_Page_160.txt
b57cbad328fd6991501a04318764e4df
b6772c86b3d611cde50410954beb261b7d9b1136
51310 F20110115_AACFJR mouraenko_o_Page_077.pro
5d89ce3f13954b9fa8bad0348b32f080
18f91d1f9471ed1bf145356d8d2acf7874ea8e63
131095 F20110115_AACGNJ mouraenko_o_Page_010.QC.jpg
9157c83dd0e742828bf67b4fd6ba53da
eaa8e7de260e7c820e2b7fe2257a21e671f562a4
445 F20110115_AACGMU mouraenko_o_Page_168.txt
4e94489abc959025e72a1f64608c1e0c
89f8d50b10b2ac3c679c84fe91e7d9d24e337097
1041276 F20110115_AACFKH mouraenko_o_Page_075.jp2
0c16d1336f47502b314983994b2844d5
7618d7bcb694113008a1f61ab575f087e5338af8
1922 F20110115_AACFJS mouraenko_o_Page_074.txt
1f6a2bbfc044afb6ab8a4afbb897a4de
d72dbcab51dde37caa9f33fdbcc8630d50db4832
53732 F20110115_AACGNK mouraenko_o_Page_011thm.jpg
7991157ef3193e2c129258ae92e7374d
011897656a77be27c425568d391ffbf183b05b38
1047 F20110115_AACGMV mouraenko_o_Page_169.txt
42b3dc4f05054f2cfd1947c5ec276c07
9b53b6d419b98f9f39cac6853a0a70e535336663
745038 F20110115_AACFKI mouraenko_o_Page_080.jp2
e3f0a330f0f26fdd8a80b8f2abf2918b
fb41fea46581f1d24ed0eda37ce81a18176aac23
F20110115_AACFJT mouraenko_o_Page_043.tif
ac214198e19e2d8cb4b5f6fe13543b0f
912e92d5c6dcde1977a95c11de4a1374898434e6
122264 F20110115_AACGNL mouraenko_o_Page_011.QC.jpg
814b38910cb362951d5502c7f29b3c78
1a91b750a1670fbfd7d4a4f349334592462725d9
7197 F20110115_AACGMW mouraenko_o_Page_001thm.jpg
19ddd1b718af2a670c190a4748181cc4
2dd9550e4af00f6612f63f227381711798a43a30
F20110115_AACFKJ mouraenko_o_Page_144.tif
38b2f1f03c50cbb25598aa5dae992700
1f50351833ccdb0454c661ec02db20971b9e6f19
92026 F20110115_AACFJU mouraenko_o_Page_064.QC.jpg
e77c1694174029130d183b0c12737edd
fbb6437c5b9eef04c23b11a557a7b1a539b114ce
45377 F20110115_AACGOA mouraenko_o_Page_026thm.jpg
293787cbc20f28609d2deebc770e4481
fb02538c89d83f8cb677ad9061fe5a3411f2ff6a
55842 F20110115_AACGNM mouraenko_o_Page_012thm.jpg
9066e1128f1f444178f454481adc8143
31423c96ca79128fc13e851c10103c7fb3909312
2514051 F20110115_AACGMX mouraenko_o.pdf
6fd09ec48a86948bfbb7ff35d9b3162f
f9c81c374463ca7a832d1bc27eed68a2b886b168
1051920 F20110115_AACFKK mouraenko_o_Page_148.jp2
0f3760573ded2ad9f3153da584a0f2c6
444ef9005ca98ff1af2f69bfdbe6aeebe3d2dd64
86087 F20110115_AACFJV mouraenko_o_Page_021.QC.jpg
a2da6c54a34a657b45a438c39a7d10bd
3eccdd75bbb86de40c5e2f78645b918b08b9702f
89053 F20110115_AACGOB mouraenko_o_Page_026.QC.jpg
edec97186bc9d0df18e23650f0ed3989
b7342845a99ace42dd1b92466f86bfb9759dc594
135143 F20110115_AACGNN mouraenko_o_Page_013.QC.jpg
5cb890a0be074e2deca781ed312b0d2a
8cb28c70f837f60e14e36b01321bbd4a4a5ac0d0
2689 F20110115_AACGMY mouraenko_o_Page_002thm.jpg
ff91eacf3312d0d049473b20af399ba1
e6c9ea011d9df10538c754d2d37d96a36d1ac7f4
43813 F20110115_AACFKL mouraenko_o_Page_020.pro
81c699052c60643d133ef2c5912aed56
376d1bce43b56e84f085410b295680970e517aeb
F20110115_AACFJW mouraenko_o_Page_148.tif
6c9adf5dc727144d5f7652304f288f7e
d4c9c7601c54578abcb555b03a7629079225d725
21260 F20110115_AACGOC mouraenko_o_Page_028thm.jpg
6364e7252ac580ca3866c2210b0b4954
a4a26efdcf738fd99ba10ea56a3ad488fa1bb10c
36722 F20110115_AACGNO mouraenko_o_Page_014thm.jpg
d2fc9f3e88e3b666786d5116db1f1c60
e5c0070d22c717db4ff0e9e9cd83422dae4fd824
4751 F20110115_AACGMZ mouraenko_o_Page_002.QC.jpg
8a963064a46e2d1e30e09ce15a16ca89
257f0d0d9daf6ba3b7f0a163bafc61e19153ef35
F20110115_AACFKM mouraenko_o_Page_166.tif
5d818e6462616cd152018e5ccd9a4adb
cc56dc1a58f84fc5b1f87b235d085ba920392497
F20110115_AACFJX mouraenko_o_Page_058.tif
571d5474bf156ba312b5fd4b67a335b7
2c9c6744aeb72a44ce239514653144cb81ea0253
148385 F20110115_AACFLA mouraenko_o_Page_019.jpg
20f9a505a8a3afe93ded6c7fc30761ee
fb2b934f906ce0f081b621cf91ea814087967397
42007 F20110115_AACGOD mouraenko_o_Page_029thm.jpg
923bd71ef9aa69670d80cbe1200f49f0
efac966758074d7c4d60d4ec2a600909aa261959
61067 F20110115_AACGNP mouraenko_o_Page_014.QC.jpg
1953c6e5370a6b53fb8310584a2fb8ba
2f8d46b204c8632853886bfc4e7c39acd0e8be22
1051962 F20110115_AACFKN mouraenko_o_Page_146.jp2
351fa4afa3582713b34e55cf6ba41595
d7329d1d1402a479cf9c2cfe3cf603836cd27665
91302 F20110115_AACFJY mouraenko_o_Page_053.QC.jpg
413388543a411b03443dd151894615d7
9c532c1b856778991efbaf6c0940a156383449fb
43551 F20110115_AACFLB mouraenko_o_Page_069thm.jpg
567c2173b1f3112dbc4f5f2f18f7d62e
52e9b68a5261adfdfffd29324f73db400e609b46
49238 F20110115_AACGOE mouraenko_o_Page_030thm.jpg
e5f6fb823e54774bd62b5ec7ade0db88
396dad0d9649906c0306f0cdbb8da2c98e153dec
17550 F20110115_AACGNQ mouraenko_o_Page_018thm.jpg
0f20716b1fb09b997c10c7e2bee825ac
1122a56ad7b7ae20e562be14f25923ce1a67ae79
50484 F20110115_AACFKO mouraenko_o_Page_100.pro
8855bd9150d7bdab6d132aa26f0f8a96
5e1374c28ccdac77ebde16378510cf4d003c80dc
48903 F20110115_AACFJZ mouraenko_o_Page_085thm.jpg
cbe9c18e73cb85421c6fe6d085a28369
437bb41526d9da3fba4aaed5030466df0db62e24
1051983 F20110115_AACFLC mouraenko_o_Page_071.jp2
adf2b9b6ff77130b609d6e08f35962d1
8265385f0aac73d659038b420b7fc089f082849d
82904 F20110115_AACGOF mouraenko_o_Page_031.QC.jpg
671d5c61974f8bf55795ee624c90a391
9446a286978c6234b121aa3659a475d9cffcfc67
56003 F20110115_AACGNR mouraenko_o_Page_018.QC.jpg
29fb35ece019cbf2ecad9e7782d49f08
77595a461090d60134cfca6401274c5d032e4307
50467 F20110115_AACFKP mouraenko_o_Page_140thm.jpg
f1333cacb2262cba3a174acddf8ae285
218b6dec3fe895e0dd1f8866bda91efadf468be0
70981 F20110115_AACFLD mouraenko_o_Page_167.pro
e8e7e2f4aed43a5fcc122ba3404fc657
67bfa9c59c082c7abf92c17aec0e8311d94451aa
79805 F20110115_AACGOG mouraenko_o_Page_032.QC.jpg
94614db3c2f5ad2fedb52b2c112bff51
5b24961b9f352a2e4dabfc504caa0077a5f4c16b
57743 F20110115_AACGNS mouraenko_o_Page_019.QC.jpg
5341a40629d0ea6474281d5cdd3351d2
e40caef6bf4d89acd6492ab8c407b36f6ec11c54
755750 F20110115_AACFKQ mouraenko_o_Page_040.jp2
5ab1acaf23ced937ea376c1c4af1d9eb
dd22c01467cae3bb09548f99e9317528615010fe
160023 F20110115_AACFLE mouraenko_o_Page_003.jpg
2809d89b231604d933f3a792ec18c76b
149fccc3719d6d2b1e9e3d5dd657b70476eb38c0
45214 F20110115_AACGOH mouraenko_o_Page_033thm.jpg
16be9487d466b2783adb7d8390a794d3
14adc91038aa99773cdf8d854407f5726ee611fd
31374 F20110115_AACFLF mouraenko_o_Page_116.pro
e9beba87e7fa27b1df0b3697b993f6eb
09b3df1625c397f2e603142a570beca1dc3e527b
76670 F20110115_AACGOI mouraenko_o_Page_033.QC.jpg
67b0c82e02d78cd8296ec98cbf46b188
fd2bb085e31493f6c8b32b4e85d565228b6899a0
21444 F20110115_AACGNT mouraenko_o_Page_020thm.jpg
82540f5d94b1f336f807a5104c0805f8
9d707aa633bedfc3108b9ebf98692775a45a522c
48350 F20110115_AACFKR mouraenko_o_Page_133thm.jpg
95e0fd14b9c038fb25270c35ce55cc52
91b3511e71c0b2d7adad515da7618b4433b89f7d
45062 F20110115_AACFLG mouraenko_o_Page_038.pro
4eb60324856dfdd2ea176ffb93b1ff2b
f590ac24d91b453d21c9d028179058a51c3b33ef
102174 F20110115_AACGOJ mouraenko_o_Page_034.QC.jpg
3c4ad8d51c3adc3e3523ee2786c466d8
e270e866d7e9a37f15a69cbe9b6057d29e5477ce
70615 F20110115_AACGNU mouraenko_o_Page_020.QC.jpg
e0e402abcbf0f8fe29636157aad32cff
3760424ec1ea9bfda8fed15d329650ce2a2bf853
244634 F20110115_AACFKS mouraenko_o_Page_132.jpg
25c5824f6048d9949327f56b9bb13a66
e725e131f6b51c189118b91d9fefd116f8a68bde
1051978 F20110115_AACFLH mouraenko_o_Page_066.jp2
5efdea34959c67c3db9ad58496250c93
f8363e67b95017bb64a996c2d437b30a898a9fda
76685 F20110115_AACGOK mouraenko_o_Page_035.QC.jpg
a30983cfe602f4e9dd200119e40543ab
7f59a4130dbfd4d39012d9a57cdd6ab7324dfa50
25089 F20110115_AACGNV mouraenko_o_Page_022thm.jpg
ff2f610672d6d0a0a0546969d31fb791
a17eb88e0dd4b82605b7082940133491653d1db2
22103 F20110115_AACFKT mouraenko_o_Page_144thm.jpg
1993be039aebb2b44b59f7c54ea7c4a2
50301cd57d85debaba0944da8a3667cad046bec8
47725 F20110115_AACFLI mouraenko_o_Page_149.QC.jpg
51622766f677e406bbf229330ff3cb00
d2c8db8c38ff405241609b07a4de5f29e43c5586
46176 F20110115_AACGOL mouraenko_o_Page_036thm.jpg
2d4e8f0afea852a69449d17f376cb913
c52f9b7d1ebcf930ad49aff495a949b61c9f2242
81000 F20110115_AACGNW mouraenko_o_Page_022.QC.jpg
43148c9901876a39fc94f9eb54f0a8e5
9abed5f95e8348bb443f1821436d2667e91d3848
93358 F20110115_AACFLJ mouraenko_o_Page_125.QC.jpg
34550fceb9f2c797c9804aa906c73803
ba431d23572883591097d592979bc3e170705bc7
45571 F20110115_AACFKU mouraenko_o_Page_108thm.jpg
77215805fbf2d6e1ea9dc424aa36943f
fce30abdcf91bf014d2f13ed9314bfe4cf830247
45530 F20110115_AACGPA mouraenko_o_Page_051.QC.jpg
69d0d9b71bfa7764ec98bfd215ebfa4f
990540c1663559e86f2ded3c56c1b47dfd960622
88004 F20110115_AACGOM mouraenko_o_Page_036.QC.jpg
e994a2a80201880936a651c142c94169
877ace3eba6c50ce02234b0a73b04202a355f90b
50921 F20110115_AACGNX mouraenko_o_Page_023thm.jpg
eb7baa2e398c795e8d98e2c87f538ef3
7089bacddf70faa505b22899e993a6df99d787a3
19767 F20110115_AACFLK mouraenko_o_Page_103.pro
ec0edbc80a6d23f024a92909e32435d1
c17f97ee136b928512d686d5200a1203c9094354
1594 F20110115_AACFKV mouraenko_o_Page_029.txt
6692582fcd6ed3fe37f76c49ea187a4c
8cab2a16245e1bde4601f77437129030d84a7049
42384 F20110115_AACGPB mouraenko_o_Page_052thm.jpg
7e437b815d977bf2b0fcf283d78b9c40
15986ec74b0ef4cc889c780338a3c6d08a2c923c
78973 F20110115_AACGON mouraenko_o_Page_037.QC.jpg
d7315369911fbf2a67d13c3d95f617e3
544bba706c809ce5ea619aa7a3832031848d85f0
51042 F20110115_AACGNY mouraenko_o_Page_024thm.jpg
d5058dafebd91ccabb0980881561450d
360874184954da519de00fabc3b88395d09c1d1f
39988 F20110115_AACFLL mouraenko_o_Page_072.pro
e148ad6900caf21c7108ec16e72f0d5d
bef4ff69a7a67a203ef6f27554f00f81bc08afe5
107281 F20110115_AACFKW mouraenko_o_Page_005.QC.jpg
700bc58f420a5cdcf0826ca26030440e
384c1ecfbf23bdd64d4e38336f77dad3440dedbb
46995 F20110115_AACGPC mouraenko_o_Page_053thm.jpg
42d854758a13e46596159550db72c3c1
49a7e692eb7a545c2757425a9a120e01fbffe081
47119 F20110115_AACGOO mouraenko_o_Page_038thm.jpg
cf8f018cc9048666d13609f0c5906458
87f125c5d45278bb9653b8b3a0dee4378ef0d574
112230 F20110115_AACGNZ mouraenko_o_Page_024.QC.jpg
b9e3e84853c945eb231e9bbc8c3ef486
e6df1d14cd36833ff918662dfc993fe5937ba985
103161 F20110115_AACFMA mouraenko_o_Page_059.QC.jpg
fd0f727410c5b5be8bfb4a217158655f
4081fc93bc41f7804472dc18bdef71390318bcb1
F20110115_AACFLM mouraenko_o_Page_140.tif
63bcd4370a0f75422f62d372855ced01
f9ea1ee9fed14c24e05d65fe2b320dfbd9ae64e1
F20110115_AACFKX mouraenko_o_Page_028.tif
9a459e7a0786d5bf4a302347402e96c5
df794f4f9f6e390f0ef861727aee3b04d510dd50
42561 F20110115_AACGPD mouraenko_o_Page_054thm.jpg
842ce594b6ef8b5399d640a3f3a15284
42ded4671f4e13a7ce75ae609bdf9bffb4486268
93373 F20110115_AACGOP mouraenko_o_Page_038.QC.jpg
b96eb655b2723fd94c4ed8d38383393b
f14105e5b5e73c280ef5209154a17efb60dcb983
90734 F20110115_AACFMB mouraenko_o_Page_078.QC.jpg
160d67073dbbff4cabe3d5ac24ac3dcf
6eaf14f70536362688c2770ce2cf52d43f68eca4
968 F20110115_AACFLN mouraenko_o_Page_014.txt
9dd4227dd784abfb1ed7434e6284df9f
9e2747c879d7f6d059045a2516f800ea4301b4ce
F20110115_AACFKY mouraenko_o_Page_052.tif
c998ee2896629534c19f3c6532e660c0
bb36f77a50d8f5ed3eaa2522cb12af766835c865
76287 F20110115_AACGPE mouraenko_o_Page_054.QC.jpg
7d6ee161785f8897a7afc085e9de111a
62f602cf1537f4fa9da372299e764a8cd2eb471b
51052 F20110115_AACGOQ mouraenko_o_Page_039thm.jpg
c0bb818066f422490b624e52282e1b2a
c342b454c430534fd2c0a4f710b32fdb8ec0dc34
F20110115_AACFMC mouraenko_o_Page_147.tif
6cd44aa0520c3a3391d7fcc4f16bedf7
41da67ef50068db6cb200bc6c9a83bf428a21fe1
77766 F20110115_AACFLO mouraenko_o_Page_120.QC.jpg
a49a95c3ad76ccd245ea45a2bf098f13
e9865e672fe412a59675c83df1692784a7d7a43c
191316 F20110115_AACFKZ mouraenko_o_Page_082.jpg
f734d12a8a8f837f130327f451b6f7ae
ce6e5f7fa9601a04dd69d741990883ac20e92f36
24973 F20110115_AACGPF mouraenko_o_Page_056thm.jpg
77fdfc20056074c80cbf489d4c8e522c
e3fa47c14e597531abbce89b376147d7af7bed6c
106996 F20110115_AACGOR mouraenko_o_Page_039.QC.jpg
2dbafb413c85be13ee4264fdd7638273
ee6fe8abec762c48d7b773a456e8c6876fd6c3ca
1670 F20110115_AACFMD mouraenko_o_Page_047.txt
5a00970a6ed46db88f8270a28a453653
316ee91bc76c845612a1a605bc73cc3232b4ddeb
1051986 F20110115_AACFLP mouraenko_o_Page_161.jp2
4279d9ad34dbd748c9c69c81167bcb23
77285bc8c2cc563e96dba45923c589f335426ed0
48504 F20110115_AACGPG mouraenko_o_Page_057thm.jpg
1c04ae4acd770d6cfd524579b0aef84b
a7fba349ba2b56bcb30b40746c36c9e5065619cb
72261 F20110115_AACGOS mouraenko_o_Page_040.QC.jpg
83a4105dee9b9c737ef71dceb9e228a1
37c9d360a2021f4046151707d88e6fea63756305
211425 F20110115_AACFME mouraenko_o_Page_094.jpg
2486aa6b3132c1944d0ce677ab6bc598
5d6cd255c671b4757883cb002b3591a7917c0c60
F20110115_AACFLQ mouraenko_o_Page_041.tif
d63c43f4fecd7f0a70d08700d1683140
e41a893421bc130762504ffc73b151bf992e236d
100863 F20110115_AACGPH mouraenko_o_Page_057.QC.jpg
6e0c3f10688bf9a817471553cc31ed9a
4a11a67a9f13b06bbe0e5efb046001928d75d33f
79074 F20110115_AACGOT mouraenko_o_Page_041.QC.jpg
bc0679234e3af6db8b402880047960c6
545b4d8890307c6b8fdb5fd8b11d4241abb0487c
43644 F20110115_AACFMF mouraenko_o_Page_035thm.jpg
8d84b446295841473462dc31c3c952f4
c71112ef0c652cb3f7ec0c121a57a4a8e2bedb17
37463 F20110115_AACFLR mouraenko_o_Page_127.pro
165023d3a3ab5857d66f6ef4a92b6295
185edc185759c419d6e586b5b2571e66af1fa014
9542 F20110115_AACGPI mouraenko_o_Page_058thm.jpg
9a73e221f5a7d64a12d6b995c4517d59
ee6bf173067bfaaef739d51857531841687acfd3
100638 F20110115_AACFMG mouraenko_o_Page_132.QC.jpg
62c94181dae914a45615dfc3df176979
5ec0b1bd5322a06a93fa46636603d5c296fb0df9
21780 F20110115_AACGPJ mouraenko_o_Page_058.QC.jpg
54b84e6124182737474ce01184f7b70c
6115ca72805fb814a0ba03488c8ad41e3352b8f4
75127 F20110115_AACGOU mouraenko_o_Page_043.QC.jpg
dec0c1e20b46bb48becbb86b817490a9
81d3de1f15d5417966c44756261dd405d9b2ebd7
F20110115_AACFMH mouraenko_o_Page_022.tif
d86b5e5ca75a0231723b79378cb1e2e7
6b58c5fb5ad064ed800987384b4dbbbb1bc26d19
33724 F20110115_AACFLS mouraenko_o_Page_136.pro
0b183837427688c1aac875e993ca7b94
dfe33efb579fceadf0d043396921736a4ca39045
49768 F20110115_AACGPK mouraenko_o_Page_059thm.jpg
764b473c1c1bf67454e084695b8b89f7
e41a496d0e40fb0d2974c0498aa9dd8a9707a892
43159 F20110115_AACGOV mouraenko_o_Page_044thm.jpg
ae3ce777faf06bea627b9ea5b8dd5b65
c78a40617710f67149b1cbd79964d5f657e73234
1051963 F20110115_AACFMI mouraenko_o_Page_011.jp2
76e0f45df918fce336f19e9266f2a056
4d0ce90a734dddc7a0417607b41278b2c746e983
42460 F20110115_AACFLT mouraenko_o_Page_097.pro
c453c2b21b727d2c973bf5274e44af6f
3e353377077dddd308132f7df173b0f60d6e789c
48502 F20110115_AACGPL mouraenko_o_Page_060thm.jpg
43be046d94ef7bf91769a295ee7214fc
b1504b74d7e8c06ff9c033c0043a34a0e7e4de6d
74498 F20110115_AACGOW mouraenko_o_Page_045.QC.jpg
67ab7c681dbef6564bd9c8f3bb7c6163
f57bc67068dfa240e7d16def12bc6e53ec02176e
68331 F20110115_AACFMJ mouraenko_o_Page_029.QC.jpg
8f23ffb9a268ff2227a40128c4f50185
fbae71706ee3654cf969b36285b1e719da95057e
F20110115_AACFLU mouraenko_o_Page_031.tif
3147bc33295cb580d6d8e2693fb1adbb
0800d9b4d26bcf0463767b36e161bac4544b89f1
96297 F20110115_AACGQA mouraenko_o_Page_074.QC.jpg
e9b5a3c7de14b28b7fc19f4886564b23
771abbd09bda6e1e5d9508e2539fab12cebdb51a
94103 F20110115_AACGPM mouraenko_o_Page_060.QC.jpg
7e83858300b5e851f7bdb99d1a413a30
75c6031889024847f97e1aa4af903fd1e490cb06
44797 F20110115_AACGOX mouraenko_o_Page_046thm.jpg
09dc487e61cc7d689ad0fef48d3338d5
5ebcb6fade8b000267a76174014ab3401c95a09b
1983 F20110115_AACFMK mouraenko_o_Page_085.txt
8edcb5ee34275c77d1812d4f7a0dec52
43da1eaa363c3551357950083023a0d09330321b
179844 F20110115_AACFLV mouraenko_o_Page_031.jpg
cf6f1f02a3eeb7cea03e4d83e02b8012
3e4ff42aa1ed9447ee1cc309e7261f3dbe8cbbba
50034 F20110115_AACGQB mouraenko_o_Page_076thm.jpg
524426664f145e5cb90ec8264da11a61
f2059e57aa979f2b3353eb98d74f4a3108614716
49870 F20110115_AACGPN mouraenko_o_Page_062thm.jpg
335090fadc64aeb19f2fef1a20c253d3
d537ba34b24c60f0b5c3775eafb0472b03ed1f7e
42795 F20110115_AACGOY mouraenko_o_Page_047thm.jpg
57dbe3c2cadc03069d045e0aa0916887
ceef485ec40e66caf56265532bac42f2575ab57c
22893 F20110115_AACFML mouraenko_o_Page_017.pro
34ca513abf7905834af6072eac367091
d0b3f63177c4809a5ff8a7cfc435d2d11b15b4e6
41646 F20110115_AACFLW mouraenko_o_Page_161thm.jpg
b148c08745c5687e803bf8785b05e9ac
1251e115945d127350f00adfaaf8fb2186605f61
103630 F20110115_AACGQC mouraenko_o_Page_077.QC.jpg
711939e8a7b66e79912f7d6316fe748c
d8521cf72bbf30a1d5a8f49687c18812b77cdbaa
110398 F20110115_AACGPO mouraenko_o_Page_063.QC.jpg
aacc8eb0ac28325f4feb88275ec6645c
b1c53a399a669ab0c2f25e348f8f9722cabe2ecf
15777 F20110115_AACGOZ mouraenko_o_Page_051thm.jpg
a9829b587307105125a68c204a0ebbe4
6052c5b721f09937d01e491feee833fbe4cd6617
45921 F20110115_AACFMM mouraenko_o_Page_143thm.jpg
68fd0ef522a9b4f8574852c812492d95
f1289ee44deb922dc32fd6d3277e70d17e29d7aa
212502 F20110115_AACFLX mouraenko_o_Page_101.jpg
24a8124470508d8f40eac83f1cf07608
0038759c833974d2e27ee558723e3434361d326d
F20110115_AACFNA mouraenko_o_Page_084.tif
0e8ac5331406eded5e74fd428dd85c53
292b3fee0a068f7aec569630798e3abce8b94269
47723 F20110115_AACGQD mouraenko_o_Page_079thm.jpg
1c1b41b6c1f1f215e9c9b0b3781fdf51
ac1b300e68292efadb8ab54b6deaae13df666fbe
49148 F20110115_AACGPP mouraenko_o_Page_066thm.jpg
3b4975f85f0a15fe2aa9683fbb19e439
488ebeb23eb245650145c16451b0748c9ea35aa2
886128 F20110115_AACFMN mouraenko_o_Page_073.jp2
d6503a7899f1fde6ccb0d73900b3b761
ee5ce8dfc11093cf65b4fd6c7147fcef4143b291
130330 F20110115_AACFLY mouraenko_o_Page_012.QC.jpg
074ed6e3583bb43d645a147984d2c862
2b055ec2ed86efdf553a9fa6c90ea8e8a710d489
42137 F20110115_AACFNB mouraenko_o_Page_144.pro
2f31bf7c8a54580f4e06c05c2d4631db
a0fd7b1b630b9ea80b074fd261cc79a4231c1668
86390 F20110115_AACGQE mouraenko_o_Page_079.QC.jpg
3d8b0d767e8877f0f0c931eb810e4884
b38e5ff82551f16c4fea5b9cb85ec06c913ab21f
95763 F20110115_AACGPQ mouraenko_o_Page_066.QC.jpg
1191f8735f8787085a42470aaa256503
aa59d6a440f36b20e0f04bddfbbb38d960c4ea6e
51762 F20110115_AACFMO mouraenko_o_Page_005thm.jpg
8c292b23e723e0b803254922839e5234
23c331101a2ceb412035eec43b408d3cbcd3b95d
2166 F20110115_AACFLZ mouraenko_o_Page_116.txt
7af933ca1424fa9b96a4750e6e1d328d
a1e345fb540b719d3de50062158f83f00b82348b
24637 F20110115_AACFNC mouraenko_o_Page_120thm.jpg
8b4e57fbba53ff040782b23fce38eeed
c8e12cbda3ff404fd56d97c8f4bb72c0d8bf2614
83065 F20110115_AACGQF mouraenko_o_Page_080.QC.jpg
ef42403751c5fcc53db33704741c334d
a6b25b666a235949f297b9dce7e156e332972796
47736 F20110115_AACGPR mouraenko_o_Page_067thm.jpg
0c1c475f657e443b9e13577bec880f58
a77a4c109be2db684b90ee93fada5b1e05912216
1965 F20110115_AACFMP mouraenko_o_Page_089.txt
20f603c9306de294c3396e29c6a22ddc
405dda5ddd9b616d6207189189fa1be2eb91fd43
1423 F20110115_AACFND mouraenko_o_Page_045.txt
27aa67aa496212362fdd90b362b7dca5
499c78f3494b24faa5c7cd0da35daf4b3f32da03
45264 F20110115_AACGQG mouraenko_o_Page_081thm.jpg
6e3ecff5cb88a6c80a28fcb1bb3bb778
5946139ca6caa3cf9437d5f153e93038b87a06d7
48580 F20110115_AACGPS mouraenko_o_Page_068thm.jpg
b5cd09250f621a5faa419b3cc988d8bb
1a7f6d99ff72cfbac989eee89732ae3ca9089fb7
1051965 F20110115_AACFMQ mouraenko_o_Page_093.jp2
f09b7826c478a46af80a7d995cabef2d
e8939a48864d4173be664c88fd28592e8ca9c605
89037 F20110115_AACFNE mouraenko_o_Page_103.jpg
11f466f9ca865d2fa243fa8c365beb16
0db658f8b1f40449af1c4c69879734a4633adda9
80971 F20110115_AACGQH mouraenko_o_Page_082.QC.jpg
087728afce7b3beed3ed79f3e230a310
5969f4fae9591efaccc34bf3524c9031e2b90d1b
95048 F20110115_AACGPT mouraenko_o_Page_068.QC.jpg
a305ffb1ab6a3bec66f26e7e9289688a
47f88da07349d2994f6292e6cfb4defbdff85950
186501 F20110115_AACFMR mouraenko_o_Page_157.jpg
51018eedc7585e6e3d03d9d298788376
66b9890251e3aadb4955bee0a8bb1761a43de183
F20110115_AACFNF mouraenko_o_Page_139.tif
cdf518235fa98543dea8cb4898437579
2b82202fa4e6298b12358cef834566adcd1c5aa0
46936 F20110115_AACGQI mouraenko_o_Page_083thm.jpg
fd7f0122bf3593f62b72cb82f2d67e9b
4d5909ae9cab8c83f0eb67b4d860e626eb58f2df
83804 F20110115_AACGPU mouraenko_o_Page_069.QC.jpg
0ddf82cccf385f96974882d3b84c15b6
b2af7ae71f28c80c790fc519705bc46c7562aa2b
F20110115_AACFMS mouraenko_o_Page_119.tif
428925b381f70b0268110cb7bf5968eb
7f6283061f89da63dcb8e10e25b49041c4c34d0a
F20110115_AACFNG mouraenko_o_Page_132.tif
6c9e888e0a7b735a84c12bbbc7071461
07513678c3268003caa88a84dd50fbc55747c3d2
95180 F20110115_AACGQJ mouraenko_o_Page_083.QC.jpg
e8599af9c7480070d274aeb69e04fb4d
08cdcec5ab72e953449cd686296d3a86511c34f9
49973 F20110115_AACFNH mouraenko_o_Page_089thm.jpg
b842100b130097e14ea6a3c118f81659
fc28a2233e942c114c050c49c5b9a2f8c9442e27
45316 F20110115_AACGQK mouraenko_o_Page_084thm.jpg
65e289910389a20786efe3e69420f944
96e9a427a3acad1671124098fd5c97b00c8b2a49
45703 F20110115_AACGPV mouraenko_o_Page_070thm.jpg
a994fdc6397b38ab1689e2fa7a42b8d0
73f3c9a62d2abaa898abfcdb37cf907554e61219
846489 F20110115_AACFMT mouraenko_o_Page_083.jp2
440d35a8d61bd00c9c04096f93e608e3
e51e88f06ed648fa2d2d9b971bf5668cdc8b9e19
204042 F20110115_AACFNI mouraenko_o_Page_120.jpg
d1fdebc8554a0efb23b15a740801a73a
e5b2e2a7bee937cf6f56305cc9fa6b5c23386c6a
83769 F20110115_AACGQL mouraenko_o_Page_084.QC.jpg
6b8b458310a0468f7a3f5c9904706321
8a4898585b59ab9fafbeb77bea71512afe529936
85508 F20110115_AACGPW mouraenko_o_Page_070.QC.jpg
4f7ed0ebaba1c471f005f0a9bf25eedc
6fb89be479a3e26d7e87d46d26c9ed077377312a
F20110115_AACFMU mouraenko_o_Page_033.tif
5de3e4633a4f7043d830c28a2dfd6fac
66df41a2225b6e281805570e976d3e16515c9f68
107591 F20110115_AACFNJ mouraenko_o_Page_062.QC.jpg
2e4aeccda47284fbeb1e1beea9b4f5cc
1bb98d7a9045348b19a2ef4b98f90e5e1ef77ab9
48403 F20110115_AACGRA mouraenko_o_Page_099thm.jpg
1f65c8f11828503f2f21fee12069765f
8db5eee211e3a3ba9068c106227ac233cff68475
101099 F20110115_AACGQM mouraenko_o_Page_086.QC.jpg
4e8f33f25b293d72f339422737703aa7
8bd36e4c2cc4aa19a99eb014d781fd395e8fcc9c
44105 F20110115_AACGPX mouraenko_o_Page_073thm.jpg
9455ddbb0f717b3e17c5d2f0c5ed7f35
a5cf4e94e8883a44ed729f359b0b460778219c3a
354350 F20110115_AACFMV mouraenko_o_Page_008.jpg
00cc9a7681b6a4e4576f28bf72fa3b55
6cb6c369ef724bade0cec4f2221441729656b546
1051946 F20110115_AACFNK mouraenko_o_Page_134.jp2
4bfec98e82c81bc862e7faa4a8aa1c2b
bef657f3d25023697847395d0c0a799ce3616540
101013 F20110115_AACGRB mouraenko_o_Page_099.QC.jpg
d34c9d7d98645a2170cfa7c2ec18f1a0
f2e31dcd6ad50f8867ed39ac6a839a0b2069fddc
84712 F20110115_AACGQN mouraenko_o_Page_088.QC.jpg
8348e7325c4a2dd1b4ef600f967fba5c
0f6dbb946956f7558dba2fc84d29d8c4230b7bc1
86101 F20110115_AACGPY mouraenko_o_Page_073.QC.jpg
c34b2f942325b287c75347bdef0942d9
58cc13b838de69340bbfdfcc270d9812f629b91a
F20110115_AACFMW mouraenko_o_Page_110.tif
dc142b94ac901d3907f0c9671115740b
c286b3da3887227a9958d04244f31d2dc56bfddb
48076 F20110115_AACFNL mouraenko_o_Page_055.pro
0ae6c64dc2ddade1e0fcc51b73a64c49
026ae6607ae9d1fca82c4ec55f58a5c111de0d5e
105789 F20110115_AACGRC mouraenko_o_Page_100.QC.jpg
4dd0c9ca69d032632458880da0e67bec
0bcac26292f6d142e3bfc5d44f712f6aa5d1b339
46502 F20110115_AACGQO mouraenko_o_Page_091thm.jpg
a00d4e2e6aa4847628d910d7a6bfd93a
cb248814c40a91d5f7352b3bdc3d13b12cad2b94
50183 F20110115_AACGPZ mouraenko_o_Page_074thm.jpg
f38bc7f2661c776021f5d92a9064d17c
203621dd75ca0539e67dcd22ccdb3724aca66b1c
203320 F20110115_AACFMX mouraenko_o_Page_098.jpg
d28c253f8e5f2bc3f83e1509b1288d1b
79d90ec3d8829e2e47e71d710598fd4dd3126d65
41603 F20110115_AACFOA mouraenko_o_Page_085.pro
64ddfef0543369af0de12b8544fc95b3
d0031e917d45bafd0851302d8c0cdd9d884d1325
2004 F20110115_AACFNM mouraenko_o_Page_039.txt
5671781685571d0500ac517c9a37cb99
b8906596a57964dc95a44d294c8ac7fbb1ce9ef5
89778 F20110115_AACGRD mouraenko_o_Page_101.QC.jpg
38d3a91a5926c1c855f7e6b2a4076c98
f5542d420fc13f41fb4e5ac6d152ebc5085eb10c
110909 F20110115_AACGQP mouraenko_o_Page_092.QC.jpg
a9496858a49cddd2fcec1c2322167f27
89299a2a94e2699ae4fde4b6dbf92c5f6bf0ec03
76612 F20110115_AACFMY mouraenko_o_Page_157.QC.jpg
e2c6068f9f1379ff9add22707c0e8315
eb363d0dae61b2318da8f30745227e76f43786bc
55259 F20110115_AACFOB mouraenko_o_Page_109.pro
c87a1a65b1b8c5620ab665197cddb4f0
a164aa261c20f61feaa69af1cbecec3089df7b16
53699 F20110115_AACFNN mouraenko_o_Page_023.pro
d91a104f8abca56b17f09e1e339b44e1
0bdf33a7183a3776e3df21680210514a614989f7
96968 F20110115_AACGRE mouraenko_o_Page_102.QC.jpg
d647adbff7defd077c0fb26fb7cdf4d3
6771e9f69c84e6547a953ce22f7173596426671b
51691 F20110115_AACGQQ mouraenko_o_Page_093thm.jpg
8e581b4f5de99eac4b1808c4c1971a8f
71102ebe7e7a7dcc9dfd5996c8c92390e06765bf
1051984 F20110115_AACFMZ mouraenko_o_Page_109.jp2
c66d1abba25e75234a23ea50cf7fcd8b
c2a10b30c052f6747fc48d1a2c669f971f0f4bc9
34216 F20110115_AACFOC mouraenko_o_Page_137.pro
39d374f8e2e1d6709d2daf0a3be41d71
17fe2d3c168dfdd6173f378ea18fb3f784879611
112538 F20110115_AACFNO mouraenko_o_Page_022.jp2
ad966732b9dc0f46954b5ba81b8b8bb9
0d8a5e272292bf0a192841522936698ccbeb568b
31888 F20110115_AACGRF mouraenko_o_Page_103.QC.jpg
853a5fd299c15f9220bd8d43f879287c
b07d91e776659cc9c6c18ace33ba263536efdb33
105632 F20110115_AACGQR mouraenko_o_Page_093.QC.jpg
950d0d3f6aee918b8ad17a19c17b16df
37513d373693150cb8d00165667fe11831b60f2d
F20110115_AACFOD mouraenko_o_Page_107.tif
4bd5bf38dc020ce655f70319a0dd1d0a
bd5b73cff4013e7585159b3f936ff793606c77de
42305 F20110115_AACFNP mouraenko_o_Page_028.pro
33166fe2f9d2d49edb0386ea7d22b3d1
db403578ae63a5ff5d5f5f160052cf0d131cd97c
46693 F20110115_AACGRG mouraenko_o_Page_104thm.jpg
650ad97a8640f2acaa8e25ef8a45f49d
ea25b4de5db7dc35a34845b61baf0956c51702ea
48539 F20110115_AACGQS mouraenko_o_Page_094thm.jpg
9e40361575a7a1d631d20f513ac21d8a
fc579a5627257136b664f0e6432b0d8f1fcf2cc6
103679 F20110115_AACFOE mouraenko_o_Page_085.QC.jpg
1c4b18226cb610db2dff230e4cc026f2
6e179a04ab5ad5c96fabe78ac775aa5a371e99b0
19680 F20110115_AACFNQ mouraenko_o_Page_159.pro
a2b94234c3137b2ff93481d85b58b4be
f0c410f9a7cd72df19205e42d016502eab335f1f
45343 F20110115_AACGRH mouraenko_o_Page_105thm.jpg
728b326b025c283cc72bcde7162ea67a
0ba0878ae89a2cf98b64e931d08b2bdd70370467
97797 F20110115_AACGQT mouraenko_o_Page_094.QC.jpg
1fadf458a3ed0556b07b9adc1a84eab3
beaaaac6f827fc8be2058862f5871cb43d0c5c7c
1038322 F20110115_AACFOF mouraenko_o_Page_091.jp2
12664971de0203fdb5cb5ae5382a0f5a
97161237ae92f9aabd15ad73032642f1bd87cdb1
F20110115_AACFNR mouraenko_o_Page_090.tif
ea00b6277419de572ccdcd8a0f6ae164
b2c650ad4c3b92d501556ee5572996d22d37f9de
46095 F20110115_AACGRI mouraenko_o_Page_106thm.jpg
db04e11c59e49bbd7ef31de856d16d1d
c796f471c390cf3482e9536c7b4f433e97d159d0
48055 F20110115_AACGQU mouraenko_o_Page_095thm.jpg
8cb2325f0f5250f9fb317be0386b9ae3
35baca0f9d9b65e2eee62a33e6570d4dde1037be
252283 F20110115_AACFOG mouraenko_o_Page_114.jpg
e8ebbd7f91080f13411f71c823271018
61a2bffb1d4ec6bb63782dc8a053006e6230e922
19314 F20110115_AACFNS mouraenko_o_Page_163.pro
2520b473a835cc4eeff0a3f031c27a92
2419db3d90bac6dd14c985d85bce3d3eb0ec7678
84839 F20110115_AACGRJ mouraenko_o_Page_108.QC.jpg
537744c69e956f25a9330d372bcc4bc3
3a313b7068e54e26ed04aee71db9648ed4814b57
95991 F20110115_AACGQV mouraenko_o_Page_096.QC.jpg
ae93a2e7ed3b6e2cbf537186c054801c
d05a615215e7dd44d7b53ce9bc1febb6480e6cd0
35805 F20110115_AACFOH mouraenko_o_Page_018.pro
5eb143c76e1bbd363b81bbcdf00ecee5
f99dea5ef9a1f3fa47a1857f993269c304cedc29
628 F20110115_AACFNT mouraenko_o_Page_002.pro
dc27a72f57d8e2f72e354c4955ba7e4c
49cd96ec2c0d522880362a0bf6345b6f68b3de6b
51584 F20110115_AACGRK mouraenko_o_Page_109thm.jpg
dd2b4da053efad8a243e6c135546e018
22c5347d3ae9cd748b954c2951bb3e94c8d576ff
F20110115_AACFOI mouraenko_o_Page_138.txt
75780e03995211f7d5068481f2a79f6d
c1645200e262d557fcf2123abcad58f0617968d0
110038 F20110115_AACGRL mouraenko_o_Page_109.QC.jpg
bb3956c4e12ba50b55b0a4c57d44b7ce
9cc060a6ba28646b18a7a4aad14d6613a498a379
48842 F20110115_AACGQW mouraenko_o_Page_097thm.jpg
3073a5a2e174b381a263dbaf1763d1e7
0eb6e41994ea7057f15c953c95e5c7e01914cff1
122133 F20110115_AACFOJ mouraenko_o_Page_015.jpg
f51ad480c8670af19789d0ef1e8bfc14
7e471ffaa53a92e77dff809b449f1cf08568fd01
91481 F20110115_AACFNU mouraenko_o_Page_134.QC.jpg
93b62860ea95de5b16171bc073fb443c
d06bd41ed5ba5ee2b7b1f6360a7b78877c1228bd
48895 F20110115_AACGSA mouraenko_o_Page_124thm.jpg
9a64f9ef0e08d2ea4da8a8d4f03f4cb7
2f2af27e99114bbd5bbcbe21eae53f66956560c6
22455 F20110115_AACGRM mouraenko_o_Page_111thm.jpg
74bc64aafc9d406b4454348af5be6de9
6f9f69bac8586b2452e4da62c6e432833936c40a
98115 F20110115_AACGQX mouraenko_o_Page_097.QC.jpg
66ffed97d0de08c4b201ea5ca2b03838
0efccba9d4ca7f025362d5c9305eaa8bdc16087b
203411 F20110115_AACFOK mouraenko_o_Page_061.jpg
dc9799f59c3a12d48fb0b4224c33b1a9
78ddede88799ed02c19b8e4dbf2858630e3e805c
48899 F20110115_AACFNV mouraenko_o_Page_102thm.jpg
58af1c1e4b6e355ff85dbd8222491f7c
a50a38d3594403c759ae59b1e3e2361b5079b67a
100769 F20110115_AACGSB mouraenko_o_Page_124.QC.jpg
b813b12fb150538feccf07943b33f36f
7b32d1ae79262c436ba0f086bff2975630abe44d
78656 F20110115_AACGRN mouraenko_o_Page_111.QC.jpg
6b3a648815fb7ecb643ed124410cec22
9d8b19e16a9f069b33b30dea780f6c034342e029
46565 F20110115_AACGQY mouraenko_o_Page_098thm.jpg
c75e55ddc631971b9b698d0113daecc4
47866fd130b31ee4474e07e22ab81c2e8ec2ba52
63550 F20110115_AACFOL mouraenko_o_Page_015.jp2
b0096b5896a37b07df7b4a38067854a3
2e49137a0153480c68a724a30b329716aa6263cb
1201 F20110115_AACFNW mouraenko_o_Page_163.txt
778465b103f33036523530763d75e962
f64a720ef99d9dd45e93a6f71a0f65f7ff217d19
48509 F20110115_AACGSC mouraenko_o_Page_126thm.jpg
21cad8fe4a44f8aa413cb1d616a677c2
daeaa9e9610978774fdc87e91b6f6bb9957101ee
47437 F20110115_AACGRO mouraenko_o_Page_112thm.jpg
4c703c6dd659a340469b9d08d5101c6b
b0759889cbd3a3723c68ae555f4aff7ece6e49bd
89191 F20110115_AACGQZ mouraenko_o_Page_098.QC.jpg
d919061591eafa9838e81e3718094d3f
36674dcdca44611f35e3f11db6f2f3be52da3641
272633 F20110115_AACFPA mouraenko_o_Page_166.jpg
2c0c575ab384cb427cfc3e36c6cfabf0
1122590b9801a714ee24b88ef641fb7aef809b38
F20110115_AACFOM mouraenko_o_Page_050.tif
21ae99efc083f358247103471d13c7d4
fc43b37e5386478acf65890e88ee9df08df683b6
42262 F20110115_AACFNX mouraenko_o_Page_078.pro
97048e11f2d0c23c4701b4b38eea5f18
7f76a5808e9367279e16d0a87da459508092c8e9
48767 F20110115_AACGSD mouraenko_o_Page_127thm.jpg
47385c4d7b43581cb3edd2df1dcb19a2
1e07766ff9d7ac0e586a72d53523c6f0585fa3aa
77135 F20110115_AACGRP mouraenko_o_Page_113.QC.jpg
7749512a0bc27ce07d6f5aee4460878d
c72e269d1861d04d4d664644092849ae062bb515
1051940 F20110115_AACFPB mouraenko_o_Page_039.jp2
2970d8d4941c1bc7342bbfec9d34cc05
e770b8faab4806a4bda262bffc4e5e61a3d32747
37037 F20110115_AACFON mouraenko_o_Page_069.pro
d0551a300be876989b9a3702e53ad362
80e9ae6b4f6f504a6af59da6c204dde5ee9396dd
815 F20110115_AACFNY mouraenko_o_Page_103.txt
6461342409642c4e63ac56b7edc48b17
21d78337c028880c9b6e8971ebf83ad448b6bb85
99409 F20110115_AACGSE mouraenko_o_Page_127.QC.jpg
7f1cde50bddfa18df6227db8cde0d740
4100402e49ff8b8c127e275cceab5204edc92c4c
47127 F20110115_AACGRQ mouraenko_o_Page_114thm.jpg
fa1c39ab6beec77873be1d785a03071d
a0a30b3564a3f626ba2d1ddef7beb2f73c4695ad
47996 F20110115_AACFPC mouraenko_o_Page_125thm.jpg
665a62cdf4d65191fbae175fc3c4bf6a
216556b99b20612619014ac9aa10844a3729d522
50292 F20110115_AACFOO mouraenko_o_Page_100thm.jpg
948f9404d8d94515c200f4b6e2de7b94
0dda9a58fbb80e5bd4502156191514c8a217771e
250644 F20110115_AACFNZ mouraenko_o_Page_024.jpg
a4d581c8ee31493708d7d7362fb746e3
2c4793a5f6245d70ca544c98177ab9e89ed12e17
48047 F20110115_AACGSF mouraenko_o_Page_128thm.jpg
4ddbcd4408f499d46ee0df78a9c9e973
b26c84aebf31651b219f376ccdca27e725364a79
98699 F20110115_AACGRR mouraenko_o_Page_114.QC.jpg
35c1ce37632eba2646ecb0d39a275491
813c61495c2284cc7f89d77715ecc5aea94ed2dc
50700 F20110115_AACFPD mouraenko_o_Page_086thm.jpg
3d1cc5dae9a13be022b42a5a067394d7
c510f9d614662f83ebae805552aac2364a9ec233
38175 F20110115_AACFOP mouraenko_o_Page_110thm.jpg
44d8497d9112b6ec60e0ac5dd160e743
3fa5f4c8666e9693a5550b6077c03ddd9b340ca0
95683 F20110115_AACGSG mouraenko_o_Page_128.QC.jpg
1c7c2620e0ae6b86a460e1157c73bfc5
04ff31c39a29961cbbbc6671783a2db141bb4dd5
91619 F20110115_AACGRS mouraenko_o_Page_115.QC.jpg
8396f81db6497233a630cfeca11b070a
f008663954874216934fbda098377ee7d738079c
1051985 F20110115_AACFPE mouraenko_o_Page_128.jp2
9088abc14223a11b98ff13814f33bba2
7fde431bcab7c048d2381baaca77db33f5d807e4
1712 F20110115_AACFOQ mouraenko_o_Page_015.txt
1b974ec8be96b8461172a3add4ff9552
d16349875006bbfea2d46437158d991b1748ec32
97787 F20110115_AACGSH mouraenko_o_Page_129.QC.jpg
f035c81183efd3c8367d3f5421c96d2a
0b3ae7ac813d30d1d5d320bffa5093279489196a
41817 F20110115_AACGRT mouraenko_o_Page_116thm.jpg
d06d37b8be984cbc576fe6ab267b5a48
1abc96b696c5d87efac4dc221c342b7cd2ae2f58
2028 F20110115_AACFPF mouraenko_o_Page_139.txt
4a00c578cffdc792f3d201de167f8b66
3e618c12917734741b216ae57853f20470b432f6
90825 F20110115_AACFOR mouraenko_o_Page_104.QC.jpg
d0727040f768c68a93f47b4812c1d2c7
fff60988050225216900a9090d8638509919103b
103926 F20110115_AACGSI mouraenko_o_Page_130.QC.jpg
8548a8dcf8f3be672d80ae8d1779072d
a16483005622ceddf6cc17b074ee187fd7080d00
51549 F20110115_AACGRU mouraenko_o_Page_117thm.jpg
fc289c169719124b80527b22aaf9a8d1
fcc960174bd28f7a74ccb6a5dbce390828ea209d
F20110115_AACFPG mouraenko_o_Page_137.tif
dd25e93d0dd55aa0c11d3e853dbd232b
a5e159078a7b730eede0a7d1ef29355fc7b877b2
99574 F20110115_AACFOS mouraenko_o_Page_135.QC.jpg
a20feb22af0c839f102dae9df4046336
3b3b123cc2fd15335b7a01f09f04e9edbf83671f
51145 F20110115_AACGSJ mouraenko_o_Page_131thm.jpg
77b02e5e345593f290f153247c672324
9b7f2de024e0e76ac65a20b9a8ef246c893ce22a
46461 F20110115_AACGRV mouraenko_o_Page_118thm.jpg
1a8f992f51e9bc97af9b0bbf79cd5662
6aad15fc99523de8bf1933c2bd367e35f92d73a9
178909 F20110115_AACFPH mouraenko_o_Page_032.jpg
702020c006053de95ac2a20a7eaa730b
a948d22b87314ca9adc168d6b3f4ab5f1cb293d7
113103 F20110115_AACFOT mouraenko_o_Page_155.QC.jpg
825555ba190a0d078477199040407510
f5b50c75d95a6d54b5ee48e0b72db787afef428a
104488 F20110115_AACGSK mouraenko_o_Page_131.QC.jpg
a85c7c8bc7c92119a8ebb072935f81f9
9330b20ff147d1571770fd3745e341f09a6c36e0
58841 F20110115_AACGRW mouraenko_o_Page_119.QC.jpg
3664d4f6afadbfbee4cb27c4ab533bc1
428003527161174f3888bc3d4ae5c669d679ef3b
246508 F20110115_AACFPI mouraenko_o_Page_111.jpg
bd1296d7f4ecda04bada469a12b6b1e1
2a99e234ad59076998b41f5a756c6e2cc4e93b1b
49713 F20110115_AACFOU mouraenko_o_Page_034.pro
76f43a3c593d36279977dd26f28558da
ec83d65818b1c7adda2266bf410097a7e1b8b0bc
48882 F20110115_AACGSL mouraenko_o_Page_132thm.jpg
f0bb3eed291017deb1056cd9f0133af9
fcceb5e4803cdb7a9dbf98a041c65ca499603fb9
89705 F20110115_AACFPJ mouraenko_o_Page_061.QC.jpg
7e9d5e88078612acfb6f0c5a2f08767b
49e17487922a574fbdb595199535ee49088c81b9
49922 F20110115_AACGTA mouraenko_o_Page_145thm.jpg
e460d18245ba82722cfbe575dc44eab0
47020576a6119b00f0c99a1b10affe3a70513237
F20110115_AACGSM mouraenko_o_Page_133.QC.jpg
9b5020a09d97173b422a00587fdd45cc
fb340b6fd9cad1c76b34496010304756ffc71fc4
24636 F20110115_AACGRX mouraenko_o_Page_121thm.jpg
714148ff08d04213b82cc8306b26ca98
3f742af43f40b535dc27c0b9f0b6b1d87cacb3ac
F20110115_AACFPK mouraenko_o_Page_063.jp2
9e6b174887598650f6cdf9f549f986d3
5614f5e81deb67c6db748905ea3deb33ff66447d
176409 F20110115_AACFOV mouraenko_o_Page_144.jpg
751becf23c4fe8de5b2a8fc8c330ae8c
e37fa6ee53a202d60d8353d8db925b0429729b16
105593 F20110115_AACGTB mouraenko_o_Page_145.QC.jpg
c2cd68e74a6a635a546957c9a46f0df3
8e75f85281c90cbfd3a6cde0d2d1c45e6bed7b27
48926 F20110115_AACGSN mouraenko_o_Page_134thm.jpg
28f3676ad12ebbf7bfc19a9b35dd74da
fef0e0ae4a273cf2a5526576ca315565425bd4c4
18302 F20110115_AACGRY mouraenko_o_Page_122thm.jpg
81543ecb9d6777780e496da11a54f87c
35ae324576faa58fe1fb8c9065512afdb2cec3f0
105046 F20110115_AACFPL mouraenko_o_Page_071.QC.jpg
d7ca5398eb3574b8752601bcb07fa478
dcafb7e7fe73d175d6289e05aacea9f0e79b9b96
F20110115_AACFOW mouraenko_o_Page_063.tif
63149ff289552de036ca4167c18ef290
dc22aabaf1a5213df01b0b8b517e189d4f7531d5
48221 F20110115_AACGTC mouraenko_o_Page_146thm.jpg
ebe56f5672260af7faafbe2bdd007822
d3ef690567f406f3ffb7bc49b382ffc143046607
50405 F20110115_AACGSO mouraenko_o_Page_136thm.jpg
b7e4ac9405444edbb7df86d16d5637b7
d79e54817d5821d84e934f40911e8e7e840e75e4
45317 F20110115_AACGRZ mouraenko_o_Page_123thm.jpg
5f10e062b8e5a7efafa904422c34fc71
364591e99c58e578af5541f1763b8b98ad9911c0
47861 F20110115_AACFPM mouraenko_o_Page_064thm.jpg
1dbcaf8192f84797b4c83e8eecf0c406
2551999eb248397c5365da1f7af00286d3c6a214
F20110115_AACFOX mouraenko_o_Page_009.tif
152fc969a7082c3bc7a4eda872a6c8f8
be4c0d8bf49433c03b5b72a26424eb09385f0162
9835 F20110115_AACFQA mouraenko_o_Page_168.pro
073a9f1d50464537fb0e4d619108d92b
64eef186162a60048e10abde4b24c7d78acb6996
91801 F20110115_AACGTD mouraenko_o_Page_147.QC.jpg
a5b56a79a7cd96a94bbec2cd01a4ca3e
18516aec69db7cd4fe5898ed15a605f514811aaf
102900 F20110115_AACGSP mouraenko_o_Page_136.QC.jpg
9a480e3d6c3180f4e10476030071c6ce
283b539a5493af6d1f72e47d2a6ed9b73d02dd26
42816 F20110115_AACFPN mouraenko_o_Page_129.pro
0b0c1c1f9ae30627f894d4c0b7cf1c2a
f1dc6c6c92c967e312ce72573311c9497e47cd35
2433 F20110115_AACFOY mouraenko_o_Page_111.txt
49c1595103e78a8a44537b9a63e95031
1261fb3816bbe757b448f4f0e20b3059e8ebf72e
63429 F20110115_AACFQB mouraenko_o_Page_087.QC.jpg
83b36880d3afbf564f06c8f01b77b67d
b7a17c7f1ca29de2695660d3eec4e3d2e65fe6ad
44334 F20110115_AACGTE mouraenko_o_Page_148thm.jpg
0696b249979ba469ff449e6d6eaa834a
89ad2ea3ac8449ce0859d072f57ce7805873ab99
50243 F20110115_AACGSQ mouraenko_o_Page_137thm.jpg
c28c2053f39da8af3eb62a49f6b45f7b
527d115a73641cc02b11f77a0a1a1a1e62753121
1643 F20110115_AACFPO mouraenko_o_Page_006.txt
18db7e6ce7bf35004153dc965212ea3b
bb91567ae87d866d064d74389de6c7d4a99b63c4
1879 F20110115_AACFOZ mouraenko_o_Page_134.txt
c298ab233f24b7640514ac6de9def26d
992f88e343f1b20ad6209dfda708a89a81f77297
1051938 F20110115_AACFQC mouraenko_o_Page_155.jp2
f4b24648bbef887868131a615dedbd7f
213c736d80bcd0cc6859c63d25b81142c1cce2df
16599 F20110115_AACGTF mouraenko_o_Page_149thm.jpg
4af66b71a7d881761018f33163f91dfc
3a61816558984c12b98da3295321c4e58bb136aa
99795 F20110115_AACGSR mouraenko_o_Page_137.QC.jpg
9a71e017445640d5754b4ef5e5b6b133
d9daa968dbf850ffb261338c84ee7cabc4335587
46195 F20110115_AACFPP mouraenko_o_Page_030.pro
f6ac8f2010cfe2c7aefa4581df3eafa2
97f411c711a8281a4b5fb97a1e78b970a1e5e894
220366 F20110115_AACFQD mouraenko_o_Page_030.jpg
2511ef4339cef2aedc646ecde08483ff
c8c236bcf658ab322231263bec5d7759ab1d68d9
42756 F20110115_AACGTG mouraenko_o_Page_150thm.jpg
170abf887fb7102406d60bddf79ce886
39406d592c758ae2361c2fef87fd1c31f63dcd55
50859 F20110115_AACGSS mouraenko_o_Page_139thm.jpg
09033aed0bb0bf068965694dc2b5082d
23fbd3776edf785b4ef23611656fd23f93261f42
1636 F20110115_AACFPQ mouraenko_o_Page_153.txt
54e76d6e169a2eda607256e7bd83ef33
dcb23e228c14cddaa4e6f3f781ddd8843362a8b5
42193 F20110115_AACFQE mouraenko_o_Page_049thm.jpg
207dad0f545c30845301c66997b2eb37
e5a6baf966cdace2f779acf39179287854ccacb0
73465 F20110115_AACGTH mouraenko_o_Page_150.QC.jpg
70c983f141f745261563b09066d8484f
2c9b45f75775514fb41073f4f7257f3d4c682de7
105393 F20110115_AACGST mouraenko_o_Page_140.QC.jpg
f22374765436672bb53f7d44fdcbae54
5fa99cee8619e01f0b4b401160e7239d88d2cda2
102711 F20110115_AACFPR mouraenko_o_Page_113.jp2
508699db544d3a21725a2f707a42fc3f
1321d11c0afd73c8dadc6476f429f7e77407628b
F20110115_AACFQF mouraenko_o_Page_155.tif
98324862e11ccd506a2abbfd2665ee68
ef7bf8c85e8ef7f079a9988b1329d259f8dcbde1
42065 F20110115_AACGTI mouraenko_o_Page_151thm.jpg
dba694acf93babb755ca976d1317ebf9
343a5ceccd60a5e7621f50bd55c769ce879ff5b9
38473 F20110115_AACGSU mouraenko_o_Page_141thm.jpg
c43d0bffc4200cee1d8337920b17a581
87e37ac4efa15a7da73175b4c661675e14426908
994962 F20110115_AACFQG mouraenko_o_Page_068.jp2
74ba87a82ba09766bd13688e97b74530
96f465ea99cd09759ca4f655b224bb51676e77b3
173206 F20110115_AACFPS mouraenko_o_Page_105.jpg
52a43b531e3e6f2674701fa559ab0292
0bd31ec276f6ea311a43d1d16ffd5c575aaeb142
78166 F20110115_AACGTJ mouraenko_o_Page_151.QC.jpg
090897e91fe153edd33d1bc0f5e5d753
6e0ad748dbf5e3e092d538700a58966aef370e6f
63435 F20110115_AACGSV mouraenko_o_Page_141.QC.jpg
f73c173c0c39654f5a2d86e10a0a40f5
6b8980f623435ecbaa9726f3973408afab85b6f7
80866 F20110115_AACFQH mouraenko_o_Page_016.jp2
b8496083df9341a26de148ebcd9e75c4
ca2f4e992dd96d238f70eefa9fdec6ab5db41cb3
639316 F20110115_AACFPT mouraenko_o_Page_043.jp2
c390e7aa03dcc21f094a8c48d00de003
325afa69057f1e637c6f0ffb7928a2d726401d89
90844 F20110115_AACGTK mouraenko_o_Page_152.QC.jpg
a2185be12727b7fb4c764ebf3b08c2fc
bf05970f7b317ef230c19f7b25dd3e9eddbb311c
48032 F20110115_AACGSW mouraenko_o_Page_142thm.jpg
50a9d4bed138dbd8c7dbc28f8503eabb
3964639c27ef613515979f348ae446670eb43e70
F20110115_AACFQI mouraenko_o_Page_123.tif
f64fef044f17734aeafd6b872f2bdf1e
e8eb76064606c7e52a6cc624439f52ef1eb47ebf
340169 F20110115_AACFPU mouraenko_o_Page_012.jpg
b86681280aef61b14a49cab835111a24
1aa3b32d4ae668eb76a2763fbe2fc675a9ce150b
89536 F20110115_AACGTL mouraenko_o_Page_153.QC.jpg
39eb0becc342e6c22f4566022c2aead2
67531cefd7c9e72a87cd2afed6c97e3d4397956c
98948 F20110115_AACGSX mouraenko_o_Page_142.QC.jpg
d9471666c00d897b215cb1b6395b08b3
bebb5044f747f18b659372794c40d7671794164c
1051976 F20110115_AACFQJ mouraenko_o_Page_124.jp2
2df07bc1c58c54d750168f5c2bf7387c
ae89a406ed6d8ac1c6721429f25144a84b708a43
35304 F20110115_AACFPV mouraenko_o_Page_070.pro
bbacae12da63f3e1d8d92e64c5d5cd05
b9a7e5cce66f2169e0b71e57cf0460ed096724df
16425 F20110115_AACGUA mouraenko_o_Page_168.QC.jpg
3ecc6caff0eda3d016e5dbcb0353977f
1a390d672794abb20a2404b88dab333ae4af88c7
111745 F20110115_AACGTM mouraenko_o_Page_154.QC.jpg
f0ec3b62b57afae899877827f27a20eb
d72db7beb523b8622845fd0edafa064dc8b362bd
882108 F20110115_AACFQK mouraenko_o_Page_079.jp2
90c94a152146308d0b1fc18cb32c8ae5
28b9ff32ef5c59b85bd8e308568b0f77cc8a4a08
44984 F20110115_AACGUB mouraenko_o_Page_169.QC.jpg
c342a90b9e2f1fb61ee1b64881e2b59f
7eeb70c8ffa1ed4fb883271fcc0e55b55d11234b
45290 F20110115_AACGTN mouraenko_o_Page_156thm.jpg
624584fe439a1b503a9c80270a4bcdc0
e278249dfdae0dc806d65d3431e3f58f049c4fef
79936 F20110115_AACGSY mouraenko_o_Page_143.QC.jpg
c66920b94ba5e88834463d274b3046c7
ecf12729316f471ef23e14b74a8e391f88bacc3c
5563 F20110115_AACFQL mouraenko_o_Page_168thm.jpg
9ae076eb1afbaddd9d414a65fe2e86a3
0b7f4d97aa56f038519940e89ebb56ccab633f55
F20110115_AACFPW mouraenko_o_Page_162.tif
d745c220c93165d2080cf6ddc64f963e
5201646fbbb557a40fd84a3731c4aee98fb67d78
197502 F20110115_AACGUC UFE0008389_00001.mets FULL
c360865c2331286e1536996faa46c5bc
912a4d6a9665b1c7702468cdbde4ac8b6c99ce2d
91897 F20110115_AACGTO mouraenko_o_Page_156.QC.jpg
1fb92efdf8a601422241a61324dc7f8c
bf1341612529547ce412619931716c3d735108ad
65192 F20110115_AACGSZ mouraenko_o_Page_144.QC.jpg
edfeb7b4615564175b55f8e5c5ac817a
f74387fa0285846fe24cee39e21edc1fcf1c3921
F20110115_AACFRA mouraenko_o_Page_003.tif
e97c2de93a440b0bcb23ffb9100a0303
d650659920644510349d54ca3c5a082a51108eb3
200030 F20110115_AACFQM mouraenko_o_Page_106.jpg
f952a7bb25b9dd96bdfc144ccbfd2f72
d4b9af4790a57dd3c636f47a5f887f7c1f691e7a
64797 F20110115_AACFPX mouraenko_o_Page_110.QC.jpg
69acd1707980b1515b844792e04bfc75
53911f9f07b72a92671d6e218dd45aef732bdde5
41408 F20110115_AACGTP mouraenko_o_Page_157thm.jpg
1bc3195ae309eadc8defd005431b5016
a74eec5b1235a938f145f7f52a22d1fd17e68ff5
175534 F20110115_AACFRB mouraenko_o_Page_046.jpg
50e60034d3b198b63b3c117e5d4d3993
233d6766a94a86b8d7a5f24be7efec86e1742029
98941 F20110115_AACFQN mouraenko_o_Page_146.QC.jpg
dbd25dd755e2e1f0a9cb1d0898c1f844
c51a9e42e790da7ee724996c0e9b978fc2b4f4e3
51490 F20110115_AACFPY mouraenko_o_Page_092thm.jpg
5a1465d6e848554438dd9c21f0f879a1
94a2f72d82c0db04b35d139d2bacf57a11951a1f
41665 F20110115_AACGTQ mouraenko_o_Page_159thm.jpg
92d376a8d5599b684a2d9cc0efee786d
563bb6e0b05dbb9b07fa081b515d0ddac8a99d0f
129007 F20110115_AACFRC mouraenko_o_Page_014.jpg
345c640493b1b3f84f748dafb598e050
33df5566a3a232c61aecca44b67583fbd94df2ed
40638 F20110115_AACFQO mouraenko_o_Page_125.pro
66cb9f701ee905c115ec8a10318b9285
80d630d90c9da4850b2c40df962423b17d5d8ab1
1051977 F20110115_AACFPZ mouraenko_o_Page_076.jp2
e34ef0d5ab11de08da6600695e05698f
b5f8d02bce446dbf8571c61a8329de88a4eb5a42
F20110115_AACGTR mouraenko_o_Page_160thm.jpg
5648ed54bb1d825f5bffa3bacec17f27
70a3b91a6147fe96f006fce721690c4f15ef8b17
41910 F20110115_AACFRD mouraenko_o_Page_043thm.jpg
35167569ead2fc75611fe865f74537da
a5d1df59a4d70413bb2189e612cc6a3f32ebac98
35278 F20110115_AACFQP mouraenko_o_Page_019.pro
d09b39dc4576a1f51e68b5af4f64464c
6bdcdfb715354468b0fff2b184d51b89605f6a06
76419 F20110115_AACGTS mouraenko_o_Page_161.QC.jpg
f3ad4d68b5871d2b3500464d9ca76de6
66a49a95b7a5b22573c354c28c495d2cf27ada0a
219689 F20110115_AACFRE mouraenko_o_Page_066.jpg
5b36d51a977defd3fd389846bdf4be56
464b7816cf833c28f970dd542ddc1b9361d92d8c
F20110115_AACFQQ mouraenko_o_Page_060.tif
97b1ad193ad95ec5fd2b69f392e92077
3b2f2d88eb65ac56d948adaea77d9bba4250fbc6
41532 F20110115_AACGTT mouraenko_o_Page_162thm.jpg
d569239ecd89d0f6b1e10f184047558a
adcf9218fba90aa1dba42f134b1ab1aac2f672cf
50991 F20110115_AACFRF mouraenko_o_Page_130thm.jpg
f207da26c4661fafd723d450688117e7
871bf683e0fbaaa105934622009d02af9febf8e7
38452 F20110115_AACFQR mouraenko_o_Page_106.pro
c3420351215bd228b9f74c489265474c
eadd88f8630e203b30d1a2a037513014bbd1d0e5
78251 F20110115_AACGTU mouraenko_o_Page_162.QC.jpg
7f9594a02acc205f5bb4cc049dc663c0
85f6c5d8a7f094fee8a841d1790f341c06736a45
41358 F20110115_AACFRG mouraenko_o_Page_158thm.jpg
180bb65c142be0326d5f07adbf732ba2
db161f92c02e39bc1805a692428bc6787333b833
F20110115_AACFQS mouraenko_o_Page_102.tif
5372d3c34c629290b9d3fd74703c90f1
363680ae6f6fecca5677228243f2b333bd4dc77f
76183 F20110115_AACGTV mouraenko_o_Page_163.QC.jpg
b170dae20cd7b7e91dc337d7c79a6e14
0038cfcc3c119026236da1f88c9901e2eeb95dd9
254934 F20110115_AACFRH mouraenko_o_Page_063.jpg
bf73c0827cecc90c2a4067c060996e14
7cfbf7af456cd06cd21295879c3a17e28329c7d4
196553 F20110115_AACFQT mouraenko_o_Page_152.jpg
719430121fd70701f8a24beea4f3e2e8
e1eaa4b5fcb93246ec6a6f39516cc198ba5b4b32
69910 F20110115_AACGTW mouraenko_o_Page_164.QC.jpg
4176433af5439815c6c9cdba281ee27b
cc87e1b26a9e9acb5b41c2f499ee3bcec5a87e26
2364 F20110115_AACFRI mouraenko_o_Page_118.txt
e26d061cb7dde453b8f6960a673f95c0
ed7a51b0c76a433dc8d175cc63ac5fa62aeda4bf
F20110115_AACFQU mouraenko_o_Page_027.tif
483e6aa896b39508eea1241f6468c6ec
1c81b408855aeaff8f8cee59103fc90218135b71
104100 F20110115_AACGTX mouraenko_o_Page_166.QC.jpg
3f9a25fca0231bb2e294dfd3bad9cc55
e6d1cfb97713b49fee0170ce0a8e304a25df8ad6
29234 F20110115_AACFRJ mouraenko_o_Page_052.pro
bbd939dd2dbc864b978450264144198d
7d4460c2a2db68e81a96cf76b1f0ef961573b40d
90696 F20110115_AACFQV mouraenko_o_Page_118.QC.jpg
f9cadf3f8aa3ce4394a0d9b30a95b862
9c1883592e97d6eab0c1f292f49e63af1071557b
53844 F20110115_AACGTY mouraenko_o_Page_167thm.jpg
f47e2bab2493bf8469c664bf900b92df
9f931ec260662c6fed87912a7a5dcf1b166ea985
38112 F20110115_AACFRK mouraenko_o_Page_131.pro
40c09b60c31b8eec5db8586a2fd1db35
971e6fd13bb6576cc428d275be6404bfb34935d9
F20110115_AACFQW mouraenko_o_Page_092.tif
393263d3eb95b8b16f07055cfbe495de
3025edd9ffd31f733609250b52ad1d24b0fc92b7
2548 F20110115_AACFRL mouraenko_o_Page_165.txt
bed9e18dcd073e519e1b394d4102cba8
2471621a27291d0b9277d322cf2d58a291b145e3
121249 F20110115_AACGTZ mouraenko_o_Page_167.QC.jpg
b778c9e04306ee1e5574e9705991e6ea
44478171b440a98aeae14b14d7279ca29f748340
97024 F20110115_AACFRM mouraenko_o_Page_030.QC.jpg
52f3e260cd9df044e59979f7bdcfe439
ffa235da074ed0779faf263dcbaa75412c6d8067
F20110115_AACFQX mouraenko_o_Page_167.tif
18867b56eec9e18daed133710c64a317
f6e329252a9bf8e1f65742bd76da894d9edce044
116791 F20110115_AACFSA mouraenko_o_Page_021.jp2
7d1d01982a74028091841ce87ae39241
d09cb145e88f0f4191915f09cb1a393270f2fde6
F20110115_AACFRN mouraenko_o_Page_008.jp2
21952e32261c6f2ba9bf5ea39677cbef
8bc212e68fdc54c7cc1f082386b34e0116f8efe4
48430 F20110115_AACFQY mouraenko_o_Page_114.pro
384f3183001176b1f09d39e488156d2f
b71ebf1d413144d37d056bedce8c46fbabcf8404
1051959 F20110115_AACFSB mouraenko_o_Page_024.jp2
e23d7812da590db3d86de6b3b26449c2
96dfb442b03ff1c9e7722dd86515c2e01c3761dd
F20110115_AACFRO mouraenko_o_Page_103.tif
c4178959a25c02dec95d53eee492c3be
c7f35074a5b398cd44a4e9130f04154772ead57d
38150 F20110115_AACFQZ mouraenko_o_Page_132.pro
8a92eaabc55f6b2929372bc75830d2f9
70492b9f2e06525fd1519dba083d03fd65a782c2
F20110115_AACFSC mouraenko_o_Page_151.tif
279052d9756892f6d5e120ace21a6fe5
08f8d28986b26593425e52d9a4ac8fd32bb67859
51433 F20110115_AACFRP mouraenko_o_Page_155thm.jpg
a9b9cdaea33cee97e56fde20c439f710
bb1267407f6818181e67e1a27137a63d8b50d797
4048 F20110115_AACFSD mouraenko_o_Page_002.jp2
be99f7da2d97b52521431b629dcb0e0f
34c847b92362db8618495b4fd51fbbee4d3e68b4
106168 F20110115_AACFRQ mouraenko_o_Page_139.QC.jpg
ae51c1c321ec1fff8a0553ede9530eeb
fc0614c4e526a8275f4a8f4fa6813114f6462b32
56892 F20110115_AACFSE mouraenko_o_Page_027.QC.jpg
572bbdc0189c9fb75fabd49b8d731860
2c2c1ed8d637cce9f3fecd0382da4c6f629f77fa
99705 F20110115_AACFRR mouraenko_o_Page_072.QC.jpg
e7de498243dad7883a85b068111a27b0
38fe17d356301ab66ef4574e1483028355ffba2b
87750 F20110115_AACFSF mouraenko_o_Page_148.QC.jpg
d42ab0f3e9c3289e7775c2c1938c302f
baaf83d70b3f455f2aed0a160bb6ba026bf7cb29
1051964 F20110115_AACFRS mouraenko_o_Page_163.jp2
a98466da3eaff4a3eededa0a9d86239e
8fa27afdca0674d9b6c079b2986c0549f622adff
185386 F20110115_AACFSG mouraenko_o_Page_108.jpg
0db979625422b63b98c8c4c4f53bc919
ad241a941ccae3417508367006acf2dbf858eba3
F20110115_AACFRT mouraenko_o_Page_165.tif
985d0fdcc5a4ef4d4a5c55dbb54d7f3d
46d9353dcfcd433c32024bcae47d3dcfc76b8dd3
1051969 F20110115_AACFSH mouraenko_o_Page_133.jp2
91bc73c292bf21a529af0e846943b01d
7b188c0b00e092d11555ea3513ea25829dbc1f9b
F20110115_AACFRU mouraenko_o_Page_021.tif
83f814c36ac2fbdf3da766880a2f5f65
d351906650615c3d4da3f901fb5e0023ef394230
21988 F20110115_AACFSI mouraenko_o_Page_158.pro
bee71d16b16e5f3c6acdf2cbb008ce96
a081eca08b76a0cad35b530719cfeb671814fa36
89474 F20110115_AACFRV mouraenko_o_Page_067.QC.jpg
effaf15e36e08026c4b129d8d2df1994
6a9e1d4f9b570446c8c215ac0cea14c7d5aeaac2
74 F20110115_AACFSJ mouraenko_o_Page_002.txt
fe0c182300a10cadd4e2a5a69d5c1170
cbcc37118e1cacd7046f0d4a57da4bf17c3614ec
28827 F20110115_AACFRW mouraenko_o_Page_031.pro
8d2e4a25be484d7247680e13033a6e44
b98f1bc0b694a859fa9a67884a3c9bba87ebea9d
F20110115_AACFSK mouraenko_o_Page_111.tif
71875024f0c0df33f6c308b622b3880c
dfc21c728e343f64b671c2ffc1e1bf7563426270
232147 F20110115_AACFRX mouraenko_o_Page_055.jpg
6f30807b17c27bfebfbbc1678756850b
38e83bfece101f77076554b518fc3d74982402ff
43620 F20110115_AACFSL mouraenko_o_Page_015.QC.jpg
398f1e7a32821ed79c44fe2f2e914559
9158558e46c25319954f383337c423bad5b8452c
17817 F20110115_AACFTA mouraenko_o_Page_016thm.jpg
9997f696baa1812c8ec46af141436df8
b50c5302b90147aa827f81e3498228ae0d516835
F20110115_AACFSM mouraenko_o_Page_073.txt
a1e5146749c35e1fe4765fd1106f9e61
5e3d23d1b2d2567c920422c4efc4b1a4200bb321
1581 F20110115_AACFRY mouraenko_o_Page_070.txt
53abbbe3bb25b7765d9cbf7cef726317
216d4822d3faf13eeee4a790d5cbfc423f01bf7b
56643 F20110115_AACFTB mouraenko_o_Page_124.pro
cf86cfc92800a4693a291cbc9526ec1a
33e47cfa07ccd6644491286ef44d50777dd130d3
F20110115_AACFSN mouraenko_o_Page_034.jp2
1afece8a1a4919925cbf3607271d4625
aa8aa95b5a901b7178c846a162353187fa5e5756
43732 F20110115_AACFRZ mouraenko_o_Page_130.pro
ef306fcb6a17e3614c8cb13fb77000ef
bf17c62baa85333cb267decc3b645cb9e71839e7
25594 F20110115_AACFTC mouraenko_o_Page_021thm.jpg
f88e3f8f56bd147310c042e5520af58d
f5fb6add6c13883a47d81bd6a16acac0ed29573d
919706 F20110115_AACFSO mouraenko_o_Page_026.jp2
2c1d6f2debe1b72a3938219543e3d9e5
5a5953db7a921e1b7f8a48d15bb160a808c43f01
42923 F20110115_AACFTD mouraenko_o_Page_032thm.jpg
30aff7c2ff1d8e997bfed3a22ce1b3b4
0c24b123da2157c90d5e18eaf5bbac7e42b2585f
819069 F20110115_AACFSP mouraenko_o_Page_143.jp2
2b8b5287f1a1c21916da23e93096a9cb
7d448814789b70863d9c3fee7b798677a8b2b680
274726 F20110115_AACFTE mouraenko_o_Page_139.jpg
9016b6d028cabc21d43dd0f9803f4b48
dfa584291dc1d4be6462217a1f395af1f386dfd1
161981 F20110115_AACFSQ mouraenko_o_Page_116.jpg
55590081702ab2a32f56d4c1f5e934ef
264eea09a06ee8674a5b9ad79be0edf152840222
78185 F20110115_AACFTF mouraenko_o_Page_044.QC.jpg
178b389ee714d3161c69bf0b574c64bf
849913e3f4414042d33510a971c9995d09e1683c
49436 F20110115_AACFSR mouraenko_o_Page_034thm.jpg
afab7bf1b0c9a71d2bf61b2ac4820001
50a30b7f11edccbc3958bdc4a917bd48b0c9202d
20904 F20110115_AACFTG mouraenko_o_Page_113thm.jpg
8b40328f2a0ed4e6a063944085cd7cb1
a9cb96776d48940aa5499a2c0a26661f6ab444a4
F20110115_AACFSS mouraenko_o_Page_032.tif
953be3b7dd7f7be1a3225974420d6c9b
9602f082dc00806a516308e8435f89fa147b936f
182618 F20110115_AACFTH mouraenko_o_Page_158.jpg
2081bc81d2fe2260de3afeebf385c462
30e03b42aadae9b9bd9034a1b9016faebedb68e3
F20110115_AACFST mouraenko_o_Page_153.tif
f42682be441c703c92b1f3ec8649e816
15eea32ec681e0598ba7a5fb3884be660919d78d
1355 F20110115_AACFTI mouraenko_o_Page_050.txt
05320076f1f44d0bb89a2627cdf7e443
76cde05cf54536cc101749c9c16ae3a6e3e0a6c8
1610 F20110115_AACFSU mouraenko_o_Page_018.txt
a2f7c48643a604f8dcdea289f8036301
3d1c92c37c1f7f355fbc558c22921339df4539a9
158268 F20110115_AACFTJ mouraenko_o_Page_033.jpg
2b9cfc735c90343fd3ebf494b824e15f
7cc98daec0ab0b3c490c2f3b84ca5d96a9ea3869
F20110115_AACFSV mouraenko_o_Page_076.tif
15ac396e316f9de518dbc5923caf6b80
50a81681d605a6246f74299e767a0c74928d2da4
47254 F20110115_AACFTK mouraenko_o_Page_075.pro
2b53fdae4a06fddc36d83e093a8016b1
f6460e3f829e521ce2affa83507ddcea872c56b8
43560 F20110115_AACFSW mouraenko_o_Page_037thm.jpg
627ce9851175498a8fe6072010a0c0a5
a47ee01270b50242b75f272811bb434fe5202a23
46118 F20110115_AACFTL mouraenko_o_Page_115thm.jpg
778aa84e4edd4ecbe23a05aaf913a751
555b937c94425f225baad2df7ca6888e64571ab1
148084 F20110115_AACFSX mouraenko_o_Page_122.jpg
35f849d880a64fb47ff13ec7189c435b
e604881f018b431a831633c39b7b5df2a66e35f0
F20110115_AACFUA mouraenko_o_Page_034.txt
fd03deba89290952a07d7219e2c8a545
393cf7f85423c9e2ae103c30981b3ab887c0a650
27829 F20110115_AACFTM mouraenko_o_Page_051.pro
4835451c6e20e8ea62dea3e4f59207ef
570d1196b095b843ffdc8ca1bfebcc05d05bad1f
43104 F20110115_AACFSY mouraenko_o_Page_042.QC.jpg
19f1302cdd7d6e38243f01f43bcd28fe
3a9bec71f337c1b861d73596df0250edeaab8a05
1190 F20110115_AACFUB mouraenko_o_Page_110.txt
c1b20791c28c0c32eb88cc4ef5f25b96
4a2a192f62bb52cf0a9259316ee15590df14d32d
50032 F20110115_AACFTN mouraenko_o_Page_059.pro
02848c34855bf23d33178e4d046c781e
d0267161e1416ea7a48f89b2cb20c7294204dfb8
1051971 F20110115_AACFUC mouraenko_o_Page_117.jp2
9d8dcb38a2922ef1ba45297a19b2da7e
2303c78bc56eaf4b5fb2fe254db6774609d538ee
99978 F20110115_AACFTO mouraenko_o_Page_089.QC.jpg
2e4ab4c60ad952ec0f4810b10532e762
94a9af457fe29d1907666d85c36eb2be9ea1550f
F20110115_AACFSZ mouraenko_o_Page_040.tif
aa45c1a8488263451032dc1cd433ce2f
1614c0926fa9811de8a60d7291b4c7678275fad9
1737 F20110115_AACFUD mouraenko_o_Page_143.txt
906a81f7c6fc4adaceb0ce873434bd41
e06dfcc81e9278ea387d5838004d8a51a591df38
56149 F20110115_AACFTP mouraenko_o_Page_016.QC.jpg
509623a7ef9b2b0fe77e923ebf3ef65a
7fc6489eca2a990e67c772cad1b4ff9e989f4083
692952 F20110115_AACFUE mouraenko_o_Page_150.jp2
3c8dfb39c680e91e964425f8d0f61a8e
d109e47811b59ada542f7a535067cca2bc884f6e
88388 F20110115_AACFTQ mouraenko_o_Page_106.QC.jpg
25146d39d345a70441189187372b221d
25b6b701f4cec37b9c028ed2261db175ccc680f0
F20110115_AACFUF mouraenko_o_Page_100.tif
032d98fc82b1cb5dc099ede26acae26a
f9e585f724b4d6cf71ff3b0ebd47e9143995369f
44616 F20110115_AACFTR mouraenko_o_Page_090thm.jpg
a8765d84931660b819b585ae5b65c18f
005fa0b420824c2dd3b3098f83b3bfd67b432104
1051970 F20110115_AACGAA mouraenko_o_Page_085.jp2
181b2b2aa8f78be18acbd8e51023ba92
4f95e292553405b07250791f5d0771631800587e
66002 F20110115_AACFUG mouraenko_o_Page_011.pro
f443324e99b885e5e699ad143c54301f
bb29547200593c3db5f8d9c7602cd5882578441b
F20110115_AACFTS mouraenko_o_Page_145.tif
ec06a422bd55b89a3b6c417868c9c789
b4a514d4fc4258e8c5adb8ac4e6d976a3d039c16
1034196 F20110115_AACGAB mouraenko_o_Page_086.jp2
57052d5edddeae5d40d53a5c5096b350
f2410f39947358bed546d7abde0c79eaff66b06c
25984 F20110115_AACFUH mouraenko_o_Page_035.pro
f04ae7b0ffdef5d28ccf89e1f01439d9
de74f46e93088c54b4e916876496b76f9fc3bc24
85510 F20110115_AACFTT mouraenko_o_Page_081.QC.jpg
4b2f672d92d24c121ff88e78a3a48a24
842c5f52b109a7d4423ad290e1dc4ef70268ca96
546895 F20110115_AACGAC mouraenko_o_Page_087.jp2
433645c1e65ad0698bcc1bbde9e0348a
ef54b90781284d02fa2fe09105659cc10131b70f
1939 F20110115_AACFUI mouraenko_o_Page_055.txt
b92c361496b4f5b7b5e77062d7b81a0c
561115ba96e6c8a9673605fc8de232607f6db1c9
38712 F20110115_AACFTU mouraenko_o_Page_079.pro
f69ce438481d5488bec5335e415ff0a0
00fed7995d05a4435549381ad36cb0852914f744
F20110115_AACGAD mouraenko_o_Page_089.jp2
2433010199902148688a8673032eba12
b414e67832486842a21ad10a9d49cc4d8500ca4c
F20110115_AACFUJ mouraenko_o_Page_059.tif
bdfb5e5b1fc41338399e97a45daa0cd3
a26a5c3bd26e17e1621d9cbc3fd2848d2d45a70c
104390 F20110115_AACFTV mouraenko_o_Page_065.QC.jpg
ccaacae556c679e1ed659dceb81617d9
0767541d09f3432cbd339675a4b0cb02e477ef1c
756969 F20110115_AACGAE mouraenko_o_Page_090.jp2
000568289324b6687ceaca7c9834e0bf
ab8c743b1d47a12a2d4dc9524a4dc7fc87ad8e65
1051948 F20110115_AACFUK mouraenko_o_Page_072.jp2
02cbfb7341ab85a75e90816047a9c721
43b1291782a210701153b5195e7023b8e250970e
1672 F20110115_AACFTW mouraenko_o_Page_046.txt
828db4295b0f5e1b3f29c946e4d6c7da
f40210dd410664d0984b324ba49017f0e3e0ee65
F20110115_AACGAF mouraenko_o_Page_092.jp2
0af45c80169fba1ee662d019e8527c44
1462ff4c23a7499c8de895fd737a6a9d49b8285d
46418 F20110115_AACFUL mouraenko_o_Page_078thm.jpg
409874bff30605ab9be0115741b411c2
97970a3c3f994e47c93f21249b0e80ea1ddb6cb9
1904 F20110115_AACFTX mouraenko_o_Page_136.txt
5c9cb740f3fc270825b6c08cd3ec8e9d
4eb7c0d59c21520ee64c9f1257987bfec97417e4
94559 F20110115_AACFUM mouraenko_o_Page_095.QC.jpg
21dcabbff301c4b44df04056fa5dc39c
de41e0e2e6b1f5bbf53f5126fee4c4743a8d11f9
27110 F20110115_AACFTY mouraenko_o_Page_049.pro
c49918651686aeb35de1e090da4f3933
8be40243bd0a5165544ff09a5718da7e494280ee
F20110115_AACGAG mouraenko_o_Page_095.jp2
7c86c679a0ad40f8ba260567d6aa6988
6400563f8167e2659c997302c912e526e0a4f847
293779 F20110115_AACFVA mouraenko_o_Page_007.jpg
2b34bd67ad8b4b4a7bf48029e77fb967
f027694dca4a770d84e67d34d923ce1dfd368b7f
F20110115_AACFUN mouraenko_o_Page_056.tif
6f534a16906fab16c3c3cab75827d1b9
72ef40c3890a82212fc38cb51f3b0aeb1d1f79b9
F20110115_AACFTZ mouraenko_o_Page_163.tif
0de2f7f1ff213e35c41c92a9a2bca295
0256f031f6a0188ca31c03b563b191a866416389
F20110115_AACGAH mouraenko_o_Page_096.jp2
5b85695e8098ee608869005234000b88
685958c95a437a3e64eb4b19e9d35d4f97284c7f
378582 F20110115_AACFVB mouraenko_o_Page_009.jpg
619a2439c2d634e8d91282f9b55ef683
9f5438d547bf9f5f265901ba30b7195bd262a09d
16970 F20110115_AACFUO mouraenko_o_Page_001.QC.jpg
398957f3622c9245e1f472ecbc2b1f93
1d3771cf020186db62c7d5e9aee95613fdaf34c5
1051979 F20110115_AACGAI mouraenko_o_Page_098.jp2
dd9a30658f0af5001de0a8eb2d6b6be7
782666bbfdcb5940d251af990d1c5007705b1291
352535 F20110115_AACFVC mouraenko_o_Page_010.jpg
a43a552d7099a8f855f224c7080491ac
2417891d200ec56a78771a99827de2ee2546cedd
F20110115_AACGAJ mouraenko_o_Page_099.jp2
f21d3dbdf5882bd0b83ed87016bc2d6e
68dbd8ca9eef0d9f2151d6b24343fcc70036ba66
308327 F20110115_AACFVD mouraenko_o_Page_011.jpg
bb239df8259958777d80622a66171a76
dc75e42c66ad0b108849a68c80d37fee14f382de
14044 F20110115_AACFUP mouraenko_o_Page_015thm.jpg
52254a1f4ca51775c5990146731c3513
12a8c041c4d3a51d8574b9b8128e0b7398b2da1f
F20110115_AACGAK mouraenko_o_Page_100.jp2
d149ae372d65c053b544b706c6d037ff
f6e7fbb852777ab11d6955440f4e948c2bf76cc0
156616 F20110115_AACFVE mouraenko_o_Page_016.jpg
7b9afa968c39c063782530e72499fd65
e116d7056902e8e820a56dacba2cbdef808e0056
246240 F20110115_AACFUQ mouraenko_o_Page_093.jpg
bae2fbbd33ef5a8de2c2ec0908d3fef6
ec2fc32d6a8f8e5e32ec917f8cd39b31a5ea8a60
1051982 F20110115_AACGBA mouraenko_o_Page_125.jp2
067b8920623a2c06555eaf4055fa59a2
39cbe28de7cccac48ea63f626d002810b297434a
F20110115_AACGAL mouraenko_o_Page_101.jp2
d796489f056406e902474bd6755fe202
639f0853f165078eaa25ca104a00a57ef36e9037
82242 F20110115_AACFVF mouraenko_o_Page_017.jpg
0de67aedc8b6ad89f759c3887a4ac0d5
0be3244606fc355210c1ed4609cc63edf101a8ad
F20110115_AACFUR mouraenko_o_Page_131.jp2
c0f8b969dd173f4776f8bd724ec8ec4a
30d3caf2eac9bf2e0a8128a029bd63e4ca8d2aa0
F20110115_AACGBB mouraenko_o_Page_126.jp2
3d61ce5b6210944310d90e2e563d9484
fe01683cdb30e33fd7e9d658a737a89b7b11dd48
F20110115_AACGAM mouraenko_o_Page_102.jp2
e3b7ab902b9a350290f557721ec96fda
0785c2b64a9f1176960c853575c2294a79da511a
223497 F20110115_AACFVG mouraenko_o_Page_021.jpg
c53f12747ff44aaaf144df4c14453cf6
d4418c217d9e91248ec7d3b72d593c0dd24b37c4
256208 F20110115_AACFUS UFE0008389_00001.xml
503ef2514802e7ee88a5672277317b3e
8afb4829ab2a979fd299e8c081586e81b4aa2ea7
1051981 F20110115_AACGBC mouraenko_o_Page_127.jp2
a46570f4e25a0720b70dde5817158cc5
6fa4dbf6b107ecfb5ae34da76b0faa210c6fa8ba
45237 F20110115_AACGAN mouraenko_o_Page_103.jp2
6d426fc7037b305387a7c00882420fc5
38ccad51329bed88557c0146c7ecc41daef355d8
213116 F20110115_AACFVH mouraenko_o_Page_022.jpg
e6984645f9c4f0b50e84c06b4c14405d
0056278cf9d44ec61d9a3be5e669f8048fa312e4
F20110115_AACGBD mouraenko_o_Page_129.jp2
4ed6f83d24fd851fa1e5c5429e496399
f81744b00742aea034a71aa3fe86b39f9f463fed
948429 F20110115_AACGAO mouraenko_o_Page_104.jp2
23cbe1da503dc039c7411e69faf4e13a
401a9274d39841c429affd5f9a2d50100e9b0bae
246986 F20110115_AACFVI mouraenko_o_Page_023.jpg
3192dfad65c576bd47870baf63658694
5be0a129f221f58831f9a6dde6a42ce20e48e6ed
1051947 F20110115_AACGBE mouraenko_o_Page_130.jp2
e48cadd31dd239e42d7d6bdba49804a7
45af10ea02598cfdd2789e17584561045a5bafd9
897280 F20110115_AACGAP mouraenko_o_Page_106.jp2
d6abf152ad1b266e4d5afb48e1ec32b7
e4fc9bfa5cfd24ca8bb50e3dd373aee25f2c43cc
147500 F20110115_AACFVJ mouraenko_o_Page_027.jpg
246a84db54a191f18113da9fee7ab16f
e052528f94d49570da07fea091356641cd0017e1
52320 F20110115_AACFUV mouraenko_o_Page_001.jpg
1fa562d61c23e51a295b622cd8c26e2d
361fbcfae56550983be3fc9997f0324c982a0d7f
F20110115_AACGBF mouraenko_o_Page_132.jp2
9d46fc14ade16ea6ae6efe5ac3e8020a
4d5f089a05d7cd95e5c39becdc5e73edfc5b92d9
F20110115_AACGAQ mouraenko_o_Page_107.jp2
f79c03c44d2cecd1cbb58c8bb2950c89
0388989671fea453468ece5dbc5a455498ec7e97
142909 F20110115_AACFVK mouraenko_o_Page_029.jpg
e55dec7f0f5cd9cc62e695e524860113
0c5e94999117b052e80c675fad5677fe4e845498
11671 F20110115_AACFUW mouraenko_o_Page_002.jpg
88684a97bb1000fa76de38c6a996e630
4bebed1c93047e2c7deafd8a1889eac92daf0081
F20110115_AACGBG mouraenko_o_Page_135.jp2
3870d0a012e8cf04d8136491fda69820
ea94391562fcea8276cd6a6ec064ea9c102dca4a
563263 F20110115_AACGAR mouraenko_o_Page_110.jp2
7bdfb130566e489d220728fa6096ea4a
6aaad2354643c310845f4d06d631539823c6e28d
232470 F20110115_AACFVL mouraenko_o_Page_034.jpg
b0e069aa5e1a48c6aa52b7af792921cd
72d5dcf43a8b30927b37f37328ab6c86f0c40146
282100 F20110115_AACFUX mouraenko_o_Page_004.jpg
30fd151e345e5860639ebb830357daa7
127c7acb851230e85e62d7fde12c2cbcc2ea73aa
207989 F20110115_AACFWA mouraenko_o_Page_053.jpg
d699e92dae028ed4ea7e0f061365efef
a4e1e372a4d22647b73c287ea37172d0d7ab73d6
110644 F20110115_AACGAS mouraenko_o_Page_111.jp2
81d55d4da4f8ce0c2ca10ae40fbec92e
29751a9b383bea4f2b1a9345f76803f6b21a4d89
179821 F20110115_AACFVM mouraenko_o_Page_036.jpg
0f1db51dab4bde9433e0b0a4d3ba90ff
deeec925c029e7c1569db28993e3b9e0f0507b77
319115 F20110115_AACFUY mouraenko_o_Page_005.jpg
4972c65edade444ffb6740588cf8168b
1f7be3a15012222ed4ffc3763daf60491ca8ed24
F20110115_AACGBH mouraenko_o_Page_136.jp2
489ecf89bfb73eca4513718008d0baa4
6f34618d486b95ce7d2b6869968707294e98a120
204403 F20110115_AACFWB mouraenko_o_Page_056.jpg
3f47578c180a8a3f801157d8d6e03d4f
f10f6eac5df8dea2f65469d7d735f58b7e1fac64
F20110115_AACGAT mouraenko_o_Page_112.jp2
948412bf5eedd38cf53a88680bc8a508
fa601d59fb8e9d83fe6af5ad5fa259b398558165
165484 F20110115_AACFVN mouraenko_o_Page_037.jpg
ff66c695809e838df3613296504c92fb
a0a1e90d588d45c815b6e8705706b6b4cf7a0e1a
185576 F20110115_AACFUZ mouraenko_o_Page_006.jpg
af4161e54d375473a3d2f02c2e399e7d
e8e8a34a17eb385f852c79f30095b10899fbe420
F20110115_AACGBI mouraenko_o_Page_137.jp2
2f984c90e3c4e038c8640e52d2466a22
5747a807f6f16f301310ca11edb655ce687e1946
212034 F20110115_AACFWC mouraenko_o_Page_060.jpg
82984507226ad6f59fc15fba1cf8c111
af468f116b8ecce2d1d5db1a7fb299b71e626409
1051924 F20110115_AACGAU mouraenko_o_Page_114.jp2
d4d6d363dede0f3fa04a81803ab92a2d
5e3a4d41838538a798006918dedbeb6dd72ac059
236621 F20110115_AACFVO mouraenko_o_Page_039.jpg
a03b7d9f9d333bd113110fde82d4e4bd
eb33c7681e1892aa774a0d729872171209961047
F20110115_AACGBJ mouraenko_o_Page_138.jp2
31c9a7488fc942418a86d729d8ba73bd
242dbbafd72fdf0afd4950a6e6f4171de091f242
254557 F20110115_AACFWD mouraenko_o_Page_062.jpg
aea39cc665c84604aa3f34283c961d5b
9abaabe36cacc0d602d2ca7eb68401794d5fc51a
F20110115_AACGAV mouraenko_o_Page_115.jp2
3fa9a7ec67cd15e97ea0803f40715150
d138eb3e778332aed554952407553f29ec36f6e9
160950 F20110115_AACFVP mouraenko_o_Page_040.jpg
e1193fbd20130626ce25de5c8c8e8128
5187db3ad62f6cd1141352eaff336c9d51b5faec
F20110115_AACGBK mouraenko_o_Page_139.jp2
4aa74b99196ade24f48114c91f66ddae
994c31f3d406e0877a5988d4dde78e424aaed129
210500 F20110115_AACFWE mouraenko_o_Page_064.jpg
29ed02dc6adb8f1642ec9af72eaf83fa
ce1d29066db5734bc240e0a1a8bf532a36b0fea6
791738 F20110115_AACGAW mouraenko_o_Page_116.jp2
9d8c0802c6c45198eacdaaef20bc2d57
b88a7454b428c52cc0b06add095b2dab82fe0587
184792 F20110115_AACFVQ mouraenko_o_Page_041.jpg
2313b35a717f1d2cc93e8e479c0363d9
cb9ae62741639f2b5d19ed9c876b28b68a323b2c
F20110115_AACGBL mouraenko_o_Page_142.jp2
55853a5b4dc855176bd951d0d1c68ff8
ee2d498609a98158ea3cebb9f4560cb639be551e
233288 F20110115_AACFWF mouraenko_o_Page_065.jpg
1c33e0efce3328493044061c796dc3a7
92161b7d485f45ca66d5ff05870292bf5815ad1d
106316 F20110115_AACGAX mouraenko_o_Page_121.jp2
1306f091a3d27f9d1bc51abbae6ba2e9
ae5e3d4201226757b7351f686b68a0c74f910b92
126424 F20110115_AACFVR mouraenko_o_Page_042.jpg
43ee38d9275a1f823a5da24672fd472c
022ad64b591186fbdae694dcf7152920b89e91e0
F20110115_AACGCA mouraenko_o_Page_005.tif
b564629251b5987b4413da9d1f03fd67
da6204353a72e5433ea35e6818d99c45c2ef61f9
200249 F20110115_AACFWG mouraenko_o_Page_067.jpg
6ad2dfa84a5e2b3540599b521faf7306
e0b6dab5f86bc4b15feaacc4ca9af074bb6fcf56
78698 F20110115_AACGAY mouraenko_o_Page_122.jp2
3e24f8c51936c2e4e81308420ef3474b
4b898a4705e611eddc9c8441af8a338892e76981
148563 F20110115_AACFVS mouraenko_o_Page_043.jpg
aa2eb48f4bd6d0a4e03fa27b1940f947
4c64afd965d1db9bf5ac1b2d77b985acb7b273d9
F20110115_AACGCB mouraenko_o_Page_006.tif
4d4ef67aaac707775f54f0d35ce57f4e
9d246b517775d96e701bce2a77646a88a007e192
90470 F20110115_AACGBM mouraenko_o_Page_144.jp2
bca431e7cac0a533fe83c7107c8ac571
b4e246736cb1f12fb3db0a12e124d06a32a00343
208653 F20110115_AACFWH mouraenko_o_Page_068.jpg
b9e69ee0c2cc756e7e9913112beb53af
55679164416a326aa1badacfa77629118ca17544
F20110115_AACGAZ mouraenko_o_Page_123.jp2
d775565f249f9c63273f181b4180c1e9
3394ca308fe096d197e6cf04d627258655f36e15
163762 F20110115_AACFVT mouraenko_o_Page_044.jpg
31d031ecde97965cb573c96d668ef340
ceb31050c10956751b90c8f6e579f36ff820a78b
F20110115_AACGCC mouraenko_o_Page_007.tif
490087d50f7ee04591461470d01732fe
70fa34d5d715ac409cebcc10eb5e1286425e2511
1051956 F20110115_AACGBN mouraenko_o_Page_145.jp2
f13721447966615e92727ad8075c0461
e86f2d6152ad900d376a39e7cc9b73559889a0dd
175802 F20110115_AACFWI mouraenko_o_Page_069.jpg
4b5d534bbfed2bca7a02ef24869995f1
144717d0a21fb7f4ec5fdfc1d262378aa92a78c8
149793 F20110115_AACFVU mouraenko_o_Page_047.jpg
767064445ee716f39a7ef9f60044419a
fde4cdc10a2d31165b185fdf45822380b0e9add6
F20110115_AACGCD mouraenko_o_Page_008.tif
51a822421809c18d7d0eb64f412fdcf1
e8901f4c8fefb862bc83f657568a0f0c10807224
F20110115_AACGBO mouraenko_o_Page_147.jp2
76a1485a1488e2830127ed59599d9bbe
8ec83ef714548666c52bd54297a81f8d5542b657
176894 F20110115_AACFWJ mouraenko_o_Page_070.jpg
50a6b4d3bc9d270244009716a0e15041
04e5b57e94b52588827d7d29006ea423a336e84b
151705 F20110115_AACFVV mouraenko_o_Page_048.jpg
cd98abe46a1c78002eace33cb90b36e1
bf8579fb41ba9d1897f289c5245030fd0c2bdab3
F20110115_AACGCE mouraenko_o_Page_011.tif
268b89cb8ebb2540d23d6d92e4527977
5bce66c86d44591f6850022ae47302097c0884e8
65198 F20110115_AACGBP mouraenko_o_Page_149.jp2
976efcc15cd2513cbf9cfaf077827c52
4b4b0a37440cd0890cdd245df8ef3198ac107a9c
229726 F20110115_AACFWK mouraenko_o_Page_072.jpg
88e4c9647fe299a4d0ecc3075361aa92
448e92405379de07e29d81358840f73078b7cd61
151314 F20110115_AACFVW mouraenko_o_Page_049.jpg
115e5ca4f9f6e22d921a08b923805a39
44008dde7976d8d3f302ac5dd2f9515334a00de4
F20110115_AACGCF mouraenko_o_Page_012.tif
204c406a56d456b036ff08f22902c935
7036d12fab6e546051035a4750d8b8c48f6002c0
663001 F20110115_AACGBQ mouraenko_o_Page_151.jp2
9175d050c8666796447a44395d6bb393
5447c8c2ac98a07eba04d14850ebf85313bcab9b
178287 F20110115_AACFWL mouraenko_o_Page_073.jpg
e9c17ec72be19014f6796018caeb51dc
07883d54d471a513af2764435b006a2e5225910c
93654 F20110115_AACFVX mouraenko_o_Page_050.jpg
9fc1cececc63ffd9f94b0d0d2b59a346
f540cb924a0c62922f58225f02b868001de6e6b3
F20110115_AACGCG mouraenko_o_Page_013.tif
c0c0acdb43721b8a30c0e71ca8f2f810
3e6e0b3d090c299499910edb06cb5d9a2cfea3d5
1051949 F20110115_AACGBR mouraenko_o_Page_154.jp2
6ce40fe12f8c7d00c32677e3deb8182c
09f9318fa463fe1fa0712c393e4bdf7e2be7f8b6
220788 F20110115_AACFWM mouraenko_o_Page_074.jpg
cae844ea0653363d75f8a567802817cc
cc011ee1e1277531113daa32e418c89729004106
117136 F20110115_AACFVY mouraenko_o_Page_051.jpg
771c27cda25e3816a66c173d507a0ca2
b9702390e74ca84e8d47e8867aa9e39839f335cd
F20110115_AACGCH mouraenko_o_Page_014.tif
68aecd230bb7feb31b1f57cd6c1250c6
df68ba388871168ab92dd990d2c1f6cfe40efe98
232495 F20110115_AACFXA mouraenko_o_Page_100.jpg
96f5bcfb806e3953570baaa0ade0ca16
98bca9b81baa0e9243ba79297c89834c9d2d75fb
F20110115_AACGBS mouraenko_o_Page_156.jp2
2c57dd98c631a860b60ae9ec98d237f8
2315b01fd78c095e9057301e625f6c5fad41a5b1
237401 F20110115_AACFWN mouraenko_o_Page_077.jpg
03ad36c2d5240a633bba07f864d3bc1b
7b802d873181cd7d33d8ad46c3720854d28b7b7c
160979 F20110115_AACFVZ mouraenko_o_Page_052.jpg
501765a24b0a5d1f111b58a6b7217ee7
182ed0de5a0c52ff56bbee84e0365257052bc2ca
240551 F20110115_AACFXB mouraenko_o_Page_102.jpg
c6881ecc46233811024d09a48ebe8618
82f28766661b2f9adc7950453c0d040f18328c58
F20110115_AACGBT mouraenko_o_Page_157.jp2
42f5d2ab42b104f8d52dd71e20049759
0ce8555b78cfad2069d406170366807208019b22
204832 F20110115_AACFWO mouraenko_o_Page_078.jpg
e5ac6b59724be50740d89f9828b51162
3948f5f693f886ee0e20f316ca43e72178158c0b
F20110115_AACGCI mouraenko_o_Page_016.tif
ad0e0ab1928667a97d31b5cc851c18fa
be5cb56e52992fe08ec5602df572adbc5a432f3a
226833 F20110115_AACFXC mouraenko_o_Page_107.jpg
ac6b6a73f147edcf54456ce6f0e8a1c2
a4f64b630d0fb49974672562a5c76cdfab3a89a5
960305 F20110115_AACGBU mouraenko_o_Page_158.jp2
ffde34fd552dd03378378c557219009f
1b1e211b86e042e5c734288ea43fbfc6274a867e
195718 F20110115_AACFWP mouraenko_o_Page_079.jpg
33a87ea3ef3d0c02e922a7ba629ffea1
5aba72907b4b6133b6c7a2b8df86a8aea24134ec
F20110115_AACGCJ mouraenko_o_Page_017.tif
6000ed270f8f0a49e0af698aaf79347f
5770334dee2d846bb2740758cf545f2a4f16a79f
249619 F20110115_AACFXD mouraenko_o_Page_109.jpg
918114bc142ca0ea3251950d35ad27d4
32657c75d80833c289376203e4c958c076cc190b
952259 F20110115_AACGBV mouraenko_o_Page_160.jp2
fb57f9575874ceb4ba09893778bbfef8
db92740b7c605b08ce603fa9cb7c03f6d5c68574
164090 F20110115_AACFWQ mouraenko_o_Page_080.jpg
66d55d9af2599b53be12b50fbe7d9961
d4482d89b393d9a65484e47159514319eb1ac95b
F20110115_AACGCK mouraenko_o_Page_018.tif
6be0b8c6e24790e367de22c6970d05fb
f006e4f3ebab3d15f5e6467ab50d76f1c329b02e
144251 F20110115_AACFXE mouraenko_o_Page_110.jpg
61a01dadeafef328125be58aa8b8b775
1a1c23496a21d8292dbe0398641aabe519008b11
941921 F20110115_AACGBW mouraenko_o_Page_162.jp2
6173dec3fa80b3d833a0ab9b2c6a8c8b
e1029a429e117cdbcf35646da2eaf1974e6024af
192065 F20110115_AACFWR mouraenko_o_Page_084.jpg
32d7441f79f4f37f9888a256896141b7
14299634048c70a3e1f5e47a87ab7896448a41be
F20110115_AACGDA mouraenko_o_Page_047.tif
a1a5ba55055fa96f4bb43bf973c690b9
93936f30e3e8e80d55ac0ac5888fb7647781e7f1
F20110115_AACGCL mouraenko_o_Page_019.tif
79b56a0e01b9a3ee0026db438f1b9ee5
aff22d3b4fbdd79034a5fd2d6d02e17de8f52507
256662 F20110115_AACFXF mouraenko_o_Page_112.jpg
2e9188211c12d0f1546a45ee307fc77f
dd41470470e20a51c6cdac1913fa577066a800ac
114742 F20110115_AACGBX mouraenko_o_Page_164.jp2
275a250bd8fb479efd00276163f3af47
ed8f79418e4a6242442293fa302dda0edb2e4043
228958 F20110115_AACFWS mouraenko_o_Page_085.jpg
2bf67b08b6544a0dc7ec9b011a84811c
a3a4b49d59469d067217a68d1298e7af9820aaa7
F20110115_AACGDB mouraenko_o_Page_048.tif
ca9718fbdd1825b260cde2e9e1f7f4e3
301843330a2992f29f2f9bb19da2d78409dc14bf
F20110115_AACGCM mouraenko_o_Page_020.tif
07bf76d5de43bdaf97084ea1c648359d
8a0e7f95434c4c63f1f39826493cee39991bb0de
209520 F20110115_AACFXG mouraenko_o_Page_115.jpg
767b2d06a18819025cd2a1494872a110
69fec5807e583359640256b1b3efcef3ffb5669d
F20110115_AACGBY mouraenko_o_Page_001.tif
9a15975d2303cfce8ec618631f696624
5e8619d7e5d2219f1f0a4a4be844ef913753d5f1
216017 F20110115_AACFWT mouraenko_o_Page_086.jpg
3ee3dc3402e103bc6724a8e30a57aadf
cd8e52609c0c233d73d6295ccb3204c6d448cae7
F20110115_AACGDC mouraenko_o_Page_053.tif
fc2fec1f29f4442f8d7bf1a7296733b3
cabf3b03b0d418c7843be09c7cd88394ac22478c
F20110115_AACGCN mouraenko_o_Page_023.tif
e85b0d474e99d09948952d03cca5f464
c4fe37809499e0fbcacb7e018bccb044deda2465
245721 F20110115_AACFXH mouraenko_o_Page_117.jpg
55b9ee13a30b38bb8aa453797fb17350
53383e1c19ff16f11462afe32fd79988bf1f7ba9
F20110115_AACGBZ mouraenko_o_Page_002.tif
a99754e41929650744b706aa636a1895
614bda65a550c9707cae91b183a131ba1410fc5f
126200 F20110115_AACFWU mouraenko_o_Page_087.jpg
89598bf6fdaf85afb85cfeb8a7963ed7
64c872e9a93cbb2f233d01e47cfb2e9197eda56d
F20110115_AACGDD mouraenko_o_Page_055.tif
52ba4196b47829689094315470ce879f
f7e691d29d239646bdfc5cc9b3e621894d0a0737
F20110115_AACGCO mouraenko_o_Page_025.tif
dec8f8cd11861df19a53b6e002f44604
fc35407b195b3151e051f21d1666f897e2b34870
214598 F20110115_AACFXI mouraenko_o_Page_118.jpg
fd1f35e271dfc5690df8a40514065280
3680b619fec4bc7babaceee60fe9bd10c5d0ff10
185409 F20110115_AACFWV mouraenko_o_Page_088.jpg
a63e724744d63396d124ba3e696d7b77
af7fd2269ad899874b0b040133fdd725dfab64e8
F20110115_AACGDE mouraenko_o_Page_062.tif
c8d32967435f7c4388b26ee984d7f6d1
b282a492030f3762b6ba8b1d49f0456502a8a829
F20110115_AACGCP mouraenko_o_Page_026.tif
fdc69cc8c50560a9c062473e9a1dfa88
61eee92e1ef8b2117279c3ff6c51ee33f2a45d7d
168717 F20110115_AACFXJ mouraenko_o_Page_119.jpg
8a883e0ff4313cae65a11972fe2c26f0
dd52cf167387a3a884db56856a1622aea28840ac
172831 F20110115_AACFWW mouraenko_o_Page_090.jpg
9bbe3d777f74c7289ffcfe2310f1ed5a
d81a41fe70d0bec070eeab5a0b70763e90a64d54
F20110115_AACGDF mouraenko_o_Page_064.tif
2c805e4dbfd6314edaaca9e114cc13d0
2c589748e0596e5b2cbaf250f4093199c1bf7949
F20110115_AACGCQ mouraenko_o_Page_029.tif
53c883972ec27bf46dca3ec9c2d0af31
d8db5062500be4bedfcd354761e10d76d0913730
206372 F20110115_AACFXK mouraenko_o_Page_121.jpg
f4b80ae5500e45f17d3bd321374665d7
73baf42917988a9352426b8083250e67e2e92446
204179 F20110115_AACFWX mouraenko_o_Page_091.jpg
f0f4265138e96bec0d5538305d1c67c6
b90fa4f9d91d8e7c93244899ae133863761caaa9
F20110115_AACGDG mouraenko_o_Page_069.tif
59beffb122b225e8393839850bda3846
db313a32286ed82174ecf25983e55fb5c020b305
F20110115_AACGCR mouraenko_o_Page_030.tif
7f0c53990109fa011129166e45ba36ca
2cea970e93f431ff7399a90f0dfcf14d74372ec8
205293 F20110115_AACFXL mouraenko_o_Page_123.jpg
305f9ad62ae28cb077b060aa54aade52
78ffa8fcf45512ef93ac06cf949c6099e1aca2fd
272482 F20110115_AACFWY mouraenko_o_Page_092.jpg
32ad6c11812301a19a3772cd722a72d7
adf70e2049caa1bdfed524355307f9b65a80f377
F20110115_AACGDH mouraenko_o_Page_070.tif
abc0d4b5619e9b5f020ba10e25b1e501
c82610061c61caf84f9169b67083923c7bb379ab
183111 F20110115_AACFYA mouraenko_o_Page_143.jpg
fd595b6e9e201b09af49d4ab87417c39
acdd539797de6469d87f6f96637414af2b11c899
F20110115_AACGCS mouraenko_o_Page_034.tif
db1644e7ae0c6e28cdabd2ac14a46816
14181e12aeb3098c16d8eaa6d078c81a5777ebce
244416 F20110115_AACFXM mouraenko_o_Page_124.jpg
e202e593874e5f3857c4056650512091
512e9490f8a22c4f47c6bfd94a27018d713fea5c
248446 F20110115_AACFWZ mouraenko_o_Page_099.jpg
25706302bcb9b6ffa2a1cc985674d512
abffab787e7c7eb6705fdfc304cf767f45d0eed9
F20110115_AACGDI mouraenko_o_Page_071.tif
5fd0a5dd13c845eb05853135f88927cc
56394272e953620c1aa82abbd239dde7c6622ee8
231653 F20110115_AACFYB mouraenko_o_Page_145.jpg
fde7b19d2bdb14daf48a4042e58771dc
51499a3cb47991d0c4e13800ee8c393110d6e26e
F20110115_AACGCT mouraenko_o_Page_035.tif
727d7bd8875f12445da55fc50be3fa81
5106ab61a9852be4ed5b67550b515ec3c0e42894
220206 F20110115_AACFXN mouraenko_o_Page_125.jpg
002843cc12867fe9aba554685581c2b6
0dde83385d8ed33fa9dfcb5e45c68b4c2254b9ee
205106 F20110115_AACFYC mouraenko_o_Page_147.jpg
6194f2f1a5791001f4ac5eaae1312f55
437caf8be6bcb1ce07fa49f1577b688d3079ebe2
F20110115_AACGCU mouraenko_o_Page_036.tif
897c0e049be5430bfaa627ffd88ff817
15ce97818119200bb504f9ae470803a58841d521
237900 F20110115_AACFXO mouraenko_o_Page_127.jpg
0e0e1c4455bc99640a8750a502dcd3f7
9adda8365431cbcc8eeb53ea5ea590780914c3cb
F20110115_AACGDJ mouraenko_o_Page_073.tif
18604f2867e689a04a1e384c0d5b1a0d
c0f6f3bee507a05588a56a3a098d76b8703e2c81
160588 F20110115_AACFYD mouraenko_o_Page_150.jpg
300068dc4d071370f854a745a35880f4
7f26980d0797e30290883c2f250304b8a3a5c5a7
F20110115_AACGCV mouraenko_o_Page_038.tif
dea1f09dcd2a4c8d87eceac4722bf63c
e6be4728a1a3da5200e2bc452d33af1ff21c807f
240565 F20110115_AACFXP mouraenko_o_Page_129.jpg
910528c229f8e82863fbe9cc8709a5fd
12918eb4fb0683a940221e5c3cb20ee53d12caf6
F20110115_AACGDK mouraenko_o_Page_074.tif
11a081254ffdf99c8bfd548bb44f4070
3ff73a8ab4fcfe5c763650d306aa79f48b8f4062
159194 F20110115_AACFYE mouraenko_o_Page_151.jpg
09e8c063ad59438be7a24079359fb97f
74eef4bcfb2a5003a958547d780c3d29e6b18d87
F20110115_AACGCW mouraenko_o_Page_039.tif
917e95ae895620c65db4f4ea98824226
37b0193178a4a234c3de2e189beb11d4a2c5111e
265327 F20110115_AACFXQ mouraenko_o_Page_130.jpg
5cac4f2fc32512d8c921e59e9c761258
1f0e68a66b1dafe88e20b6fceb22b7a2f16f0669
F20110115_AACGDL mouraenko_o_Page_075.tif
1239dda68adb92c5ad6bbb009e23f6e5
e58aeffa089228735399811bff8c144a9cc678b6
195946 F20110115_AACFYF mouraenko_o_Page_153.jpg
a05c8903a6a25c451dc5a843b989c25a
0f2bcc38578bbc33d60b3aea2312d77055ee873f
F20110115_AACGCX mouraenko_o_Page_042.tif
b170071c0ea722d899cba3325e42c6e2
969acb2a54d36e715fb42869b87f0d5aa7bb27b7
262708 F20110115_AACFXR mouraenko_o_Page_131.jpg
aca794b4e9005d26e0a0d73622683370
2acf2b13961c404ceb490838e2f825604d8307b8
F20110115_AACGEA mouraenko_o_Page_104.tif
a0057b71eedc16bdd1af4a590ca8a530
78c1999a2e3a342825e9bfeea52443cc1b844175
F20110115_AACGDM mouraenko_o_Page_077.tif
0353cb89fd77758ffa24750f5e9a4d41
cf0a4a4353f885bf4e73742d96079c94b2303835
272793 F20110115_AACFYG mouraenko_o_Page_154.jpg
5869f0ec17c3bd7dea294a4aab14758b
40cd49222fd86fcb041b38e3d3b62a406d0a71ef
F20110115_AACGCY mouraenko_o_Page_045.tif
b364676c1a15f342f346913a8704c3fd
afeafb89303592b3b301eb8585d4f31bb9dfd9c3
223514 F20110115_AACFXS mouraenko_o_Page_133.jpg
5b141e75832a22bb4a41089759cc34d5
83ef6fc8c7ba476ce20431a972f0fdef839506c7
F20110115_AACGEB mouraenko_o_Page_105.tif
91e5c291f9c4e2cae5c11d6e544f057b
4bcfaeee5d4d913814f537f08d91a291aa0daf5e
242222 F20110115_AACFAL mouraenko_o_Page_146.jpg
58d173d2cc804ddf0971646ca71ab239
38446c22d0037774ea0e254e08c6eb5cc1a8738a
F20110115_AACGDN mouraenko_o_Page_078.tif
d48b45f8ceba18ba368ac1917584c9aa
61b5c36ca5fe4d4be5ca12e17b3ff4e518b355d9
265531 F20110115_AACFYH mouraenko_o_Page_155.jpg
97d6243505bbd3f11c85036e84d75c0a
55e791e369175a7db340221362daf5fd69d4b3ab
F20110115_AACGCZ mouraenko_o_Page_046.tif
b5e1a75cd3d851cc46746f89b3975f8b
f7605a64de6b409d87b2b35406f474b52b4ac1b6
218972 F20110115_AACFXT mouraenko_o_Page_134.jpg
e48de434f9a7ba7afb7360b36420ff2d
2b09deb04c4411323da2aa0419dd70b4795f9292
50617 F20110115_AACFBA mouraenko_o_Page_022.pro
70a97ae5f015f78923025dea9af989de
b22ae8819d782e58502215f75118d5020c64723b
F20110115_AACGEC mouraenko_o_Page_106.tif
40b8c2204f3883ec36809624fca16808
188f125eb787cb0ab18d97a21ef5572e4e4a02c2
72808 F20110115_AACFAM mouraenko_o_Page_049.QC.jpg
c1ac3a9cbff735381d8ed2b1822bc589
3be8bf986698b3608db90921ec5ca27d62036666
F20110115_AACGDO mouraenko_o_Page_079.tif
53f1a433d04f1541e3ee66a6d150cad1
0349f1f0afde424188e4988ae7bfe838d7a94ce7
206737 F20110115_AACFYI mouraenko_o_Page_156.jpg
9d931b55cb0e784fda6daea54a739f44
b4f1182f57440f93d79d0d5e4f102e5a17ca983d
257467 F20110115_AACFXU mouraenko_o_Page_136.jpg
f9500cec782e1480c593b34b8955efc3
d9efc0a658c8d81848735237a15aa778bdc63b26
210012 F20110115_AACFBB mouraenko_o_Page_075.jpg
9f870400002ab0009b6ee4833d4cee20
52a8888a465840b6c5c79b2b3a84ef58b29aaa94
F20110115_AACGED mouraenko_o_Page_108.tif
dd771664ea13ebf5ad3af0ae5ddfba5d
3c5fd01c8cbb171c9a987e358d4194d8b14adf48
F20110115_AACFAN mouraenko_o_Page_067.tif
5a56dd0462538e889e9d163c97259979
a9a8935f42631bd703e18db85009868236709f0c
F20110115_AACGDP mouraenko_o_Page_080.tif
f38fcb5229c418c57bf0a327270eb4a8
a5417f290cbf57e27fe0c61a1a677e987eb77be4
186054 F20110115_AACFYJ mouraenko_o_Page_159.jpg
158ab69f49520b03498eab04e83e1a89
08a7e75d10ea1ac39037a4d125a79eb21fe0d469
245573 F20110115_AACFXV mouraenko_o_Page_137.jpg
d21258d8a6117c4233e7ce6d525c4388
874484b1ca048ea55b1d8cc5a9107f01453f1aa2
F20110115_AACFBC mouraenko_o_Page_066.tif
555c7c8b3957125b8937141511dc35d2
54b90d3d6e8ce6dc7cccf7fb09278a7bc7e66fc7
F20110115_AACGEE mouraenko_o_Page_109.tif
784bcf590314056d0417053d36bd0bd4
d9c6025337ff1145adbfee092b4a8a5fe78ec200
30383 F20110115_AACFAO mouraenko_o_Page_048.pro
dd1d6888af31bc4298d9b78eeb4d24cc
18662abf169ec3fad856f24b9a70138af11e3e88
F20110115_AACGDQ mouraenko_o_Page_081.tif
afccba5e2c07994d08ab5e7ba5390751
305b51d6e8b0af2ecdb7d950c82425a5073ad7cd
186131 F20110115_AACFYK mouraenko_o_Page_161.jpg
f9930f2533784357b51406283120ef7f
38352c547850b857322ea6d085e5824f85894ced
266378 F20110115_AACFXW mouraenko_o_Page_138.jpg
86201a0d05391732c493fb3ee82c7892
2181d63d46786b97910a27fe1dcc158a39a44e4a
41837 F20110115_AACFBD mouraenko_o_Page_040thm.jpg
5c21e53e68beaef7bf33d6a5d0825e27
1c1e04ad304b66e862de87d6a1abf25b527c1b83
F20110115_AACGEF mouraenko_o_Page_112.tif
c7708f4ac2d2a391b4010d6665233cee
0728c0e27b77c035a3c8d34ea0e60db8cf618ccb
135274 F20110115_AACFAP mouraenko_o_Page_165.jp2
a1daad8288ed6bb70d6f76fcb8656fb0
4712cf210ad89b02d1fc887bc98d69e5b9c7f80c
F20110115_AACGDR mouraenko_o_Page_083.tif
5178d3a93ae117db8245ddd98e6b2c32
b2ae40050eb16a7cfd2d365ff20d95984122471a
186202 F20110115_AACFYL mouraenko_o_Page_163.jpg
9d3b409f93eb71071dd08301351e52ad
b6404656fd152adf218d78b0c81ab33dabaa95aa
262734 F20110115_AACFXX mouraenko_o_Page_140.jpg
6f141687ef15be927a04b4d05673a714
1fd0a999722b65223986658eaf6d77113ca13dc8
246502 F20110115_AACFBE mouraenko_o_Page_135.jpg
e398a33d2ffd26b28d611030bde56b13
31191271849e6e8e98f15a94dd3ae293b43dc7b7
F20110115_AACGEG mouraenko_o_Page_113.tif
3f4201d1755dc97abcf8890fdb067d9c
a478b6fede3db302d460f7e78d4d6fa54a5f31d1
228068 F20110115_AACFAQ mouraenko_o_Page_057.jpg
3b3c07df66f7b51e424f4db7d25abe7b
9a7b3c105f0749988a541625a8f716dc2d258561
486951 F20110115_AACFZA mouraenko_o_Page_025.jp2
46995cca7cfb1bc75121114371ba7dd9
e678319aafad8911e60ab0ea2f56b3e99629aabc
F20110115_AACGDS mouraenko_o_Page_085.tif
559bde07fd49a4f0d575b28a27fb731e
9a825a08bd55bcfc8c6d009cae405c0e9bd68880
332972 F20110115_AACFYM mouraenko_o_Page_167.jpg
582d106463a1c15c2b7ab06f7a705d57
8e0110811f8f25a9c604c38cba08b0d12b6960ba
132066 F20110115_AACFXY mouraenko_o_Page_141.jpg
646ef23985b86029808467d13b8b23e1
3566ef764bd5fb5d6eb15549ca6fceb4f0e2a946
3205 F20110115_AACFBF mouraenko_o_Page_124.txt
960d271a2c9cc913b4ebfde8fe1d51de
fe9c6b1517f9831b523f30e63b3c855713e95c7a
F20110115_AACGEH mouraenko_o_Page_114.tif
e310acadd0fdbbf159210dcca7152b72
4d55c388a0e54df085214435faf75424620b2eb0
76426 F20110115_AACFAR mouraenko_o_Page_008.pro
2c3e34007ed47124bf0698724dba0e12
2467a592530b9c5e836d0214e279edcc36ba4a4b
78658 F20110115_AACFZB mouraenko_o_Page_027.jp2
2d3b6040b4132987fe53654d5d256214
1dae63c126443e02ca93a5f043df6bd0565ba68c
F20110115_AACGDT mouraenko_o_Page_086.tif
6dc0a189d881009a00e3e16306487972
3541dba32274f401b68a93d8673678cdd549ca91
52727 F20110115_AACFYN mouraenko_o_Page_168.jpg
e1f7f3d628690ec7369408ec472a9ac0
9ad179cd13d3fc81334e1832dbc92d646bbffc73
222130 F20110115_AACFXZ mouraenko_o_Page_142.jpg
e78b4001b125336c089a7ea281781941
202a033e4862868d43bbb8c404ba7a41ca022f4c
60639 F20110115_AACFBG mouraenko_o_Page_058.jpg
5ef3ad14f2516523993368d0165e00c8
5e33e74a26c829865513960a1f1ec9c2a54a8347
F20110115_AACGEI mouraenko_o_Page_115.tif
b1e187729bb57084c97ada2f650a9434
ed7d44f31a41307695a540ec82ea3ed897034d4e
49441 F20110115_AACFAS mouraenko_o_Page_075thm.jpg
20f2974308bd814ebe4cff8cf3d792c7
c8c5592dba0f6f2c1319a18d2d46b9bdbd7e05c9
740699 F20110115_AACFZC mouraenko_o_Page_032.jp2
bbadb3ea629717da80a87eebf2dfd71a
9de8027faab9788d742c22fdba8b6d637ce00a7b
F20110115_AACGDU mouraenko_o_Page_087.tif
a33732e318bb59345cdc43d1469c3250
5bc5115d8d9ada432ab258c45ca4ad5a9f4551cb
23789 F20110115_AACFYO mouraenko_o_Page_001.jp2
65b52da59fcc9e1beb71c5888625a532
d9a86adc7d6ee10429cf8cb39add3f9d9ba0cc67
56675 F20110115_AACFBH mouraenko_o_Page_008thm.jpg
f153762023354df9cda77c11f9bc643a
16b4f0bb0a14abfcdbb56d3c6f948ad11c489b8b
F20110115_AACGEJ mouraenko_o_Page_117.tif
dfb9e12f386419cae299ae0a77d5cf0a
31267c5c75d2ac213f1dbe217a17e2a3f5b3a15b
10438 F20110115_AACFAT mouraenko_o_Page_017thm.jpg
f910bb3d406af66b2385525c56cc1c83
2b03bac0e0323d496f96235b23784e52a40ea88f
815425 F20110115_AACFZD mouraenko_o_Page_036.jp2
4095e98473f36b25ac7545f746be8765
b2894a19e1eb0a6b0ce657fe5372630629377a92
F20110115_AACGDV mouraenko_o_Page_088.tif
c0f7dd1feee6040ade8f9742d0624ab1
d6bd4725b7e3e51c8109f7bea2f44d29254f6305
81295 F20110115_AACFYP mouraenko_o_Page_003.jp2
77b3e4daf0f9d252b9c2b6d4107badf1
2919f2df11c9754db3bec44b706e513d3ea36151
F20110115_AACFAU mouraenko_o_Page_168.tif
3e34eb43466c4c1bca3b842159658294
2bbd9ded89e004c6ba980be9fa220a52759e43ce
948005 F20110115_AACFZE mouraenko_o_Page_038.jp2
bb545ddeb41f60cac173eec4159c904b
90d71e8b8d21ff96afecb47188cd1ac1e285699a
F20110115_AACGDW mouraenko_o_Page_091.tif
78280fa5ae6cdeddedb173460ab8f6ff
a7b263ed54c6fa73feca4170d7c2e136ada3d1a0
F20110115_AACFYQ mouraenko_o_Page_005.jp2
9bb20e3b45b171fbf546ed950104e1e0
dc631122bc344d688b6b9859f8b83c79815577cc
F20110115_AACFBI mouraenko_o_Page_015.tif
fd80a5ba1288ba41d76d7f27bf26937a
877cf7eeec6aff834cf82cec80005b255d722300
F20110115_AACGEK mouraenko_o_Page_118.tif
95ee7874b49383f6e7625a96b6b55384
8d67c554bd41143c648217ebb722e5fa383dcff8
F20110115_AACGFA mouraenko_o_Page_143.tif
bb8170dae607f6f99900aa86123da33d
fb162ba850ecffcc3f982a3a9793571ef1317ca5
204999 F20110115_AACFAV mouraenko_o_Page_148.jpg
9373730d71228d64b7d5acf3760fe199
9969e0a82864287dda7e88d442839ab64f54ca1d
974532 F20110115_AACFZF mouraenko_o_Page_041.jp2
3a81b73765f34dddd2b09b826df4e239
27016091ed871231958e35f1c3b405cff114ce2d
F20110115_AACGDX mouraenko_o_Page_095.tif
543f1644c78f5bee0bc3d35d3e3f6124
c5fdbe5804d35198412961048d1dcfabc2f5f78f
1051967 F20110115_AACFYR mouraenko_o_Page_007.jp2
653159e5a7eb00e012a338e6a7ebca61
b654d06a52e980a53271f19c9155ed06a64d950b
49028 F20110115_AACFBJ mouraenko_o_Page_089.pro
d7ff1b58dc93afaa8766caf12805d0e9
4b169ba4a82be379f694beff5b9d94dd071265c0
F20110115_AACGEL mouraenko_o_Page_120.tif
2efef8ab1b16fed25d8165fdf8650c34
493f471516b757fbf098343ca36b2c904a4e55c1
F20110115_AACGFB mouraenko_o_Page_146.tif
004433d520cdb254a3e0daff3c2af5c9
9d81f505bc09a4dcddd8dd878950b07baebe3728
28375 F20110115_AACFAW mouraenko_o_Page_058.jp2
9790e73d696426812c3aea479f232056
fefb56f77815e27beae28b59ad85d0f697f1546c
64089 F20110115_AACFZG mouraenko_o_Page_042.jp2
b8e16f9a72a58391e5c31c4d786add2d
e1b81966e7fe7dcc114f286b6d873e8d9feec189
F20110115_AACGDY mouraenko_o_Page_097.tif
d8a4e5903df3718ec2601db4a96e743b
c413fe946eceb71036b82e282eac98cecb9f2c65
F20110115_AACFYS mouraenko_o_Page_009.jp2
73cabbe251e8a894688f1e98753af259
d08997eb600fcecf20e291f0f14aeda203d24afd
707859 F20110115_AACFBK mouraenko_o_Page_033.jp2
68ba77a85686091b250ad3376b6e8464
03f18de4e3bae03c0736eb08f2346fffd57e1a4f
F20110115_AACGEM mouraenko_o_Page_121.tif
e905290d2a08b341bf66ae5253ef7197
afd708e955ae565f173e8ad95855cfa3d9eb0a28
F20110115_AACGFC mouraenko_o_Page_149.tif
f0e31d24387f31b4e49d9cbef50cdfe2
26b8e72ac8f9c5245af05bda904fe9b918899a82
F20110115_AACFAX mouraenko_o_Page_065.tif
b2133189ae5f6ec92dee7607ce146946
9ccbb5c7afa6b25dfa327966f5e403668f769b4f
811490 F20110115_AACFZH mouraenko_o_Page_045.jp2
c0234c3ba212464b7f6796637f59d060
3364176f68cf4589deae4ad1bbaa110142532159
F20110115_AACGDZ mouraenko_o_Page_101.tif
4a3f307c2c14dc5b7695c65d37163719
b1101825a9d524ec0a865b4d2456ba539e561cac
F20110115_AACFYT mouraenko_o_Page_010.jp2
0e6ab14c75bd901104ddb271df439da3
6e3c938f62dde71ec1e475b1c0a763ff987ea6e2
1411 F20110115_AACFCA mouraenko_o_Page_019.txt
0a762b8b3830755b718210fbf7a126d3
8bb7758e83aca43620553dad5d05235a457a68d3
F20110115_AACFBL mouraenko_o_Page_054.tif
11fc6b42c06b4a946302e75f20ffa3bb
8dca91de953f6e0ddc7ed390e215363ae2a1eec8
F20110115_AACGEN mouraenko_o_Page_124.tif
2451d2f5c03406f02b7aba59ddd15d63
337588215edcfdab6d10c78301092b80d3087d68
F20110115_AACGFD mouraenko_o_Page_150.tif
2af642012350fcd2bf9d821b31f3d37f
fc31e2d7212064e0bbabd05ce837eedef604ce8a
49951 F20110115_AACFAY mouraenko_o_Page_135thm.jpg
7c6da63cd147dd522f47698b84061ce1
e6eb4c04a7e20bbd84d9f1b4fd483294197bbc04
641687 F20110115_AACFZI mouraenko_o_Page_047.jp2
22918792b2d38eb17b5ea7c1adbb183b
56388e75abd8440c018a660c1c80e6f102f7a69b
F20110115_AACFYU mouraenko_o_Page_012.jp2
df5a2b0dc0a6501b8148938e9e118f7e
2f618ad0d8f152eaa619d8046a343a8237baff5b
F20110115_AACFCB mouraenko_o_Page_157.tif
9f31de1facde5e8fcb38db58e7b8d530
7ee3087807cdaef941a6b92196b0f0bba6e6ae7d
98494 F20110115_AACFBM mouraenko_o_Page_107.QC.jpg
998cbd46eaed55210a90a673f68b96af
23f5cba30da76eee8079b523b4105eab0c2d2c03
F20110115_AACGEO mouraenko_o_Page_126.tif
6c55e28cd76d9e2abe56b6c23f9cf7f7
ea5bb20dd08dcca027d190ae7380480785398b94
F20110115_AACGFE mouraenko_o_Page_154.tif
918bc6b251dc66b2e6b43d10a1e5bcf5
20a15d9f60964b65de9a96b8fc974d3768f597dd
107746 F20110115_AACFAZ mouraenko_o_Page_117.QC.jpg
e48a19f4936370e96734dd31cf996021
4a9ef7e229abad8b0dfa23f205f6b3c3664e2a3c
665238 F20110115_AACFZJ mouraenko_o_Page_048.jp2
76ea9f7ddf63ef0fe56827bbe17a61af
f09c1d37001dcd578881df2d6e35e8e6720a8505
915083 F20110115_AACFYV mouraenko_o_Page_014.jp2
d8459fc996803bb4f8b18390f8230fad
47fc7931c3b505ea9b8596faaafe5bd699e5224e
105539 F20110115_AACFCC mouraenko_o_Page_076.QC.jpg
f926db7bbff9ab76e5be686ee91ce126
87992bbc5d3e246966ff81edd985b983f4e1762c
172771 F20110115_AACFBN mouraenko_o_Page_035.jpg
ae81d966eb642fe50ad752c96f7c4c86
d22a238338a239510e217fbbd9d2cd8a593aab4a
F20110115_AACGEP mouraenko_o_Page_127.tif
4e23087f744db7ac1c6a46210f71349e
8eecca0df841e0f94ea96c8057951a46d083c9bd
F20110115_AACGFF mouraenko_o_Page_158.tif
5633ccf181ca08d3275b7e6652244ae7
1875406a380a871cf1d3e3097823f633ad57ec6d
631969 F20110115_AACFZK mouraenko_o_Page_049.jp2
6f688b4b8671ade4b908c095113ceecd
7c807354a7346401eab9a80df38957a2761bd7c4
42801 F20110115_AACFYW mouraenko_o_Page_017.jp2
a60098e9ddbabb8d22f4015a26c5f1f9
882a7d13f4232965a11cb6ef2e9bf35287df8693
29409 F20110115_AACFCD mouraenko_o_Page_017.QC.jpg
6f67e54d9c15dbc118b80ece56e24213
d5c8782a36ba1c7c0ef156c72e18b699b04a3826
58064 F20110115_AACFBO mouraenko_o_Page_122.QC.jpg
3325eed6a02f70b53717cb741abff9d3
e4a504c70106762c744e1cd2b9488d050751bbd3
F20110115_AACGEQ mouraenko_o_Page_128.tif
8411f945f2b88ba5a334149ee23fff2c
96a794436554e4773e9c4a5dee5762cd2f8c54d2
F20110115_AACGFG mouraenko_o_Page_159.tif
83ab76d00de7a8a72a3be768751496e7
2b80dade4b51b7cf5f0e300b44bfbcf61e3e03e1
F20110115_AACGER mouraenko_o_Page_129.tif
3140619ad7f0f82ad9ab5ad50b5a47ab
99c32a4759c145aa05e1ed6a33ec630afb59f7d7
48145 F20110115_AACFZL mouraenko_o_Page_050.jp2
6bf4e2c327973dba2609cc4a97939616
deb772b456010a42e36836eb2f8f1e2d1c88c225
80209 F20110115_AACFYX mouraenko_o_Page_018.jp2
ab6268aaa7afa001dc04483877d0e7a1
a06d58d09a891a71fdaee893ae17589b0b6c7b0c
1006046 F20110115_AACFCE mouraenko_o_Page_094.jp2
e05c4d5a26c32a3f7fdcb8ae258fa061
33584ee6f598fbb9e1439462004ce4c63cb900ce
51309 F20110115_AACFBP mouraenko_o_Page_138thm.jpg
e579e803e672efa7fad9a665a4c9ca73
eb4d506faa7768000b3064949ae40deddab24460
F20110115_AACGFH mouraenko_o_Page_160.tif
3d28ff9e8ed6060b5297167e088a6600
a3de32456e28b26f848c3c563cd93104d24989ed
F20110115_AACGES mouraenko_o_Page_131.tif
355021346867daebbaa6c5e428ebedda
a03c9ade3a3596a7dbc90cd8106d538dd59a4ce9
78771 F20110115_AACFYY mouraenko_o_Page_019.jp2
2dec8e9b57b88e5a21a5b2e5a9823fb6
eaad8a3e65b5a84fa6f1dd095f8f9d4ba2043468
102909 F20110115_AACFCF mouraenko_o_Page_004.QC.jpg
1343c0ea97117deb3c538c8f0eb6bebb
4953231b0f4346bb8177fd0731e6627731dbbcf4
45622 F20110115_AACFBQ mouraenko_o_Page_153thm.jpg
d8718deac29d30d58a56c6e996566218
064b3205f5999711c0adafd4dd8e8493b24323d2
60992 F20110115_AACFZM mouraenko_o_Page_051.jp2
86be6579365e62d944b07158d2be37a7
688af83d70b32af76d7352e8dd785f0c1d08ed91
F20110115_AACGFI mouraenko_o_Page_161.tif
246e0ec6770d372d36d78a92724f1875
742bb2c2ec2a5858754a5467444d336f29a16b03
F20110115_AACGET mouraenko_o_Page_133.tif
fa7b4ad75f0bd131db73c96ea2b993ec
c8cbc3921c370f7ea70ac036acc8bfbafe040f49
97009 F20110115_AACFYZ mouraenko_o_Page_020.jp2
4bead9c6b91b8857ad7f0f12f188644f
b1d250cc1da8306a8500cdb67edc3662a394a950
35137 F20110115_AACFCG mouraenko_o_Page_122.pro
91f5aa4f422f04dadb28c121c4c4666b
56fe01f5ca55a23de47ea4440d0db53007ce4525
93519 F20110115_AACFBR mouraenko_o_Page_075.QC.jpg
b8b0f3b1041fd70fc161bbe6ab851c8b
57595322ce910092c19c5096b36a86cfe1229da0
762235 F20110115_AACFZN mouraenko_o_Page_054.jp2
9a9fa59a9aedfe79bc7a3bffa725eb65
fc0269187c1cad0805835521023417b2bcac79d0
F20110115_AACGFJ mouraenko_o_Page_164.tif
fa88780a3f25fe3705a79b6562236630
5dab326722a0d3365d62cf62d278285726685d83
F20110115_AACGEU mouraenko_o_Page_134.tif
8502e4c56e5bd01d9e51e7b80f7afac0
e6799331be9b2b3f4f25621e6e97dbdc075809be
115788 F20110115_AACFCH mouraenko_o_Page_169.jpg
30a4e527bf77f2487072025583a7d315
a6c9374999ac65318599ae41ccb1ce88cc0e212b
729612 F20110115_AACFBS mouraenko_o_Page_035.jp2
55d3cab1f1f9f461c4fdabc280114639
efd39e14bf0072aa6c18ce8eb8a264b45d732aa1
105531 F20110115_AACFZO mouraenko_o_Page_056.jp2
dc4f3d917ce02b1fb7a9521c3f495368
64a5e36f6e8e5cd8fe5cc2aeff296966a9f7a6e4
8085 F20110115_AACGFK mouraenko_o_Page_001.pro
36d83077ebd515cc872ad56b334a11d8
f1c0350a8b9ebb5ff3085fc26e6e65be63f69372
F20110115_AACGEV mouraenko_o_Page_135.tif
92575273855aa6b84f79fefa81d36a34
fd2dd3da493a4c2a9f7e7d7bfaca05210656afbb
100201 F20110115_AACFCI mouraenko_o_Page_126.QC.jpg
7c7300c604e8f7ccd2f8e8ee2ac88b13
bf93de7b00ea7f34edc6d10f5221157ab447e9fe
44560 F20110115_AACFBT mouraenko_o_Page_080thm.jpg
2c81103d6da839433a02f716d378ae89
bc01f7c91faeacef8dd9d44ede7bab02f724b047
F20110115_AACFZP mouraenko_o_Page_059.jp2
932bbc0322c15d29eecc9528f2a1918d
84e3614f2a8c955e06729afbb8e11c40fd99b266
F20110115_AACGEW mouraenko_o_Page_136.tif
0bf69f78426e6cb02965a02ba6ea46e2
f79ed530d85a83a34f00789c286b79a17f798881
43140 F20110115_AACFBU mouraenko_o_Page_048thm.jpg
a0fc5fad5b942110b2c5e8b61ffd8b60
82d5dcf6f3f3e228ef4f07fd136129eb01d8ef63
F20110115_AACFZQ mouraenko_o_Page_062.jp2
537f6cd0d5431edc0dfa6a5987c2639e
5f1dc80dbe695acd9166dcaa513334ff2dc72576
32683 F20110115_AACGGA mouraenko_o_Page_037.pro
fffee00de57f9dd4d36ed6779c8911f1
c903b2c5fb21185edd9bfa3745890200a62022b7
36441 F20110115_AACGFL mouraenko_o_Page_003.pro
762bb3ad4dc68482cdc99ee7dfee1a72
6cd3105efb9adb1a80bc868a81daf4ed116de369
F20110115_AACGEX mouraenko_o_Page_138.tif
f77c543aa4d09b4217f7a41a2398dde5
ea25ca3b4913170d791b62f5bdfd8a021d2ee4ee
45077 F20110115_AACFCJ mouraenko_o_Page_031thm.jpg
e819cc26b81a1429f7f0ede26daf1a1a
313cbb4f941d3dcd365f19db20afad80f235d55c
2621 F20110115_AACFBV mouraenko_o_Page_099.txt
500af18914e9da8be8ceeaa7877d2db7
d7b63018e9176fd403a654c32116039bd9e3705f
F20110115_AACFZR mouraenko_o_Page_064.jp2
280142c64c02176eac458d2be43a107a
276b18fa8ec8d34112f431d585d94e2a8a77d7e1
50904 F20110115_AACGGB mouraenko_o_Page_039.pro
28d97ade11bae615cefcd44da1080b2f
ae991c14aff00f7d0834877a42f7356bf3673be5
39374 F20110115_AACGFM mouraenko_o_Page_006.pro
0677ba7c4b824dfac17c9d631f529e5b
b76ef52a5516cc2307bec8b661347226895f3c19
F20110115_AACGEY mouraenko_o_Page_141.tif
b7d6610b7b181fe2dfd32bba6e498d25
e70e78138bfd12cdffd3ce85b7fbfddd82e540d1
606951 F20110115_AACFCK mouraenko_o_Page_029.jp2
ff1c38722821433f9e9acaca5c205eb0
36677080ccf28516beeb2307885bfa18143f7057
48861 F20110115_AACFBW mouraenko_o_Page_129thm.jpg
0d0894d4005ffca435de56784ac0fd12
9262993be3e190e9b670e3ea7aebea7cbbb471cf
F20110115_AACFZS mouraenko_o_Page_065.jp2
4ed631e1a17689e6010cf5c09e8aedd3
1ba3e98ccdde017ce63a4948d5aac5ca4691b3cc
10751 F20110115_AACGGC mouraenko_o_Page_040.pro
100923f42e852be81e581ea5378bca23
a2c9a09aaaa064f806f0d7f646e69bd7840d41d6
79313 F20110115_AACGFN mouraenko_o_Page_009.pro
63cb5590956b4be58ed8e2600826d762
884ba67d108638ce670b7c48afccf6d4ebb201e9
F20110115_AACGEZ mouraenko_o_Page_142.tif
e682f6b2c621cbde8876e2329c7b614e
03fa56619f4303acc4b33523b690b29984761a6a
34688 F20110115_AACFCL mouraenko_o_Page_105.pro
bf032fda7d64ef66439f0bae1a3b89e2
deeebd66e4dd44ef1a9bea64605754226e62dd00
1708 F20110115_AACFBX mouraenko_o_Page_151.txt
1d6ccc4f61d45365f06da42b2324acac
767aa6b1fec20925b4ca5a3db829a81a76d633f1
960638 F20110115_AACFZT mouraenko_o_Page_067.jp2
7624b9e8361ccc8531e2358b504d58fb
5ebe3613c638957655a3773df7052d0b0bf3776e
205945 F20110115_AACFDA mouraenko_o_Page_164.jpg
ced0c55d38847f1ab886736f25549c11
cf36a9148fa5653ad7618285a1d6188af320497a
14974 F20110115_AACGGD mouraenko_o_Page_041.pro
d57d95e50bc822d32b85ff7a5472666f
300bdb92a99ab9a74d25595c14188034a5cb79a2
73949 F20110115_AACGFO mouraenko_o_Page_010.pro
0295138696bc061c70c28e5ab95403d1
0ce15dd7d9ba2563409bcf7be25c81fe90dae9e6
F20110115_AACFCM mouraenko_o_Page_037.tif
6a57684fa368dedf2272a8f51cf72b58
a88e93ff7f9b53b00eae063a8524fa75b5d93205
31853 F20110115_AACFBY mouraenko_o_Page_140.pro
ec176fc335ef26b9ca63a58f1e2c6472
8cffc7df7c74061d0a95a3e2917a7b4a46234824
759375 F20110115_AACFZU mouraenko_o_Page_069.jp2
896166a23ad266d004cfe40d10b22e68
4598706d34cc701dd1ee3e4fd48a444207956611
49671 F20110115_AACFDB mouraenko_o_Page_065thm.jpg
fce1888b07dc1be7a82aca1f190bc02f
3daaf593c1df6875cadf1e5354c607ef5b90050d
27553 F20110115_AACGGE mouraenko_o_Page_042.pro
0329e97df81c2a2ec81e135da382cdef
637c35b3c3cc493f87f64a986f9d068eeae211de
65595 F20110115_AACGFP mouraenko_o_Page_012.pro
31feab6c93a5463bdd6c74cf8e9511ae
9f466306f77917745741d3f9ee303a624ca17785
74883 F20110115_AACFCN mouraenko_o_Page_105.QC.jpg
3fa1d945a187596ebde5f2e0c036dd05
f9cba5109af08e71287a61b3f4fc26a7a679e2c3
1206 F20110115_AACFBZ mouraenko_o_Page_157.txt
1ed49e6fee6424151fe1522500ab5d29
2a12659c6d0ebcbceebd9da3e02adcc3e957fee7
831932 F20110115_AACFZV mouraenko_o_Page_070.jp2
f89da3058afd64faa34face43f5f91e1
6e41b474ac6f98f3ab2d77d2b8a637fd286517e2
47190 F20110115_AACFDC mouraenko_o_Page_101thm.jpg
06e390bf61fa3ee06c3ace491a0400c6
35c0dc9b51dc3f7c312ae31c2af1037ba8321a85
28910 F20110115_AACGGF mouraenko_o_Page_043.pro
96f02a855bbeeb15157e5180301397b4
984327f53667499243bc60de8f3d3ffe15b8c084
70076 F20110115_AACGFQ mouraenko_o_Page_013.pro
5d9eb155feb7e467362c2236c69958e2
c446d9f47521a0060a1b7e60657dda9ee916abc8
1742 F20110115_AACFCO mouraenko_o_Page_104.txt
12c332b0dba521a0ca27d51d83f2f3da
0193234e49afad7443e9e2c7cd302d88fc21252b
F20110115_AACFZW mouraenko_o_Page_074.jp2
c1b06432727101f1d6fe14cee9855e7a
3dbc60768c55dab6674ce31a9d17da825a38f27d
264348 F20110115_AACFDD mouraenko_o_Page_165.jpg
d7fb6de1f3b8f545547cb3121c1b6b21
9edda2249a6d17bbceab455e5563edf096a9b74f
29528 F20110115_AACGGG mouraenko_o_Page_044.pro
ecf4f6f4b9205100461886ff179be6f3
ef7f4c8c2d4917f828890921e3808543698394dc
23906 F20110115_AACGFR mouraenko_o_Page_014.pro
4d36ce2a358c2ee6b5b232e6fe388b68
0d9e99c91a30e40271a07a51ba157dc2c7ecc128
44314 F20110115_AACFCP mouraenko_o_Page_045thm.jpg
463b028689a20cbffe873f6f1af3a004
12489796b34b57720fd8f9d45f45a9565d7f3d48
F20110115_AACFZX mouraenko_o_Page_078.jp2
7c0d999c3b50e99ffc521bbd9e40e672
2a7db46933ef7811f5eeb589593130366f98b651
F20110115_AACFDE mouraenko_o_Page_057.tif
08128ca72bc432bd1e447db1271146a5
afa381b81a00bf0c764f7b0c6fd8e81bf5892951
28891 F20110115_AACGGH mouraenko_o_Page_045.pro
51d932c31346658e4133fc2ae163609c
223110547343fc3b7c32b8388b041429e581d0d5
46276 F20110115_AACGFS mouraenko_o_Page_016.pro
8c2715e3bd877a77a7d087a229a07a11
00b08a814956f453a76a115581094dc318034f09
F20110115_AACFCQ mouraenko_o_Page_068.tif
8a5160e7921b292099b74f6e1a793a4f
fa17a6caeacf8c223bf370fdabb2a0e35f77740b
843631 F20110115_AACFZY mouraenko_o_Page_081.jp2
70ccc147cd2fa84103eec28a60f07dc4
8105308b443cbc53bbc42b42ea7d49df035d911f
11181 F20110115_AACFDF mouraenko_o_Page_103thm.jpg
e00dc4fdbe24e4b60c0227aa922236c9
27402707252f7eb1689f531c30891e92fb6e0fd3
35158 F20110115_AACGGI mouraenko_o_Page_046.pro
462fd18ca905b78f7229f7e7ce3bc687
6f6b9c2a603249f4ffee133b804639928d0d0ab6
54021 F20110115_AACGFT mouraenko_o_Page_021.pro
32dc857e86d34ade3b77284b0c15dc72
e73a4364e470aefe43c81f0f2230531cbf15df76
F20110115_AACFCR mouraenko_o_Page_097.jp2
ec551c14486577b250ee5783d9874a2e
fb9b8f84312f85ad662fbc5a2a11684600992cad
1011803 F20110115_AACFZZ mouraenko_o_Page_084.jp2
1895e8b6310f593385d62a5449689c02
3b5e8e763ae1dd9fde2e53e3cf62c9e706812473
854185 F20110115_AACFDG mouraenko_o_Page_082.jp2
e26d20698533c1ab0954c472c349b8b3
c86fbcfacc46512c421121259587d735ad96a42b
29745 F20110115_AACGGJ mouraenko_o_Page_047.pro
3adc7c8264230c46395267dadb44ea75
aae7c0d97f6fee9a134414bb0e6f26ade9b85d1f
53975 F20110115_AACGFU mouraenko_o_Page_024.pro
fccbe948f7113aaa2a1806918b572baa
7afdc69c5bb47dd4acb6b8a3108c704ddc5c8a5d
14491 F20110115_AACFCS mouraenko_o_Page_169thm.jpg
24747b5602a32c320464c21cb229bbdd
48183e510da0875a2532f92d61ec63df7eeb2b72
F20110115_AACFDH mouraenko_o_Page_086.txt
07879314887670cf6cd6fe1e393859aa
75f5a8d5ae4ed25b25dea29f4541cf723dc145ae
27555 F20110115_AACGGK mouraenko_o_Page_054.pro
32704481da43ad074a5b4b581ea46c5d
16b9a81c1a977d4e72049262545cbb2d4a8b4ce7
21174 F20110115_AACGFV mouraenko_o_Page_025.pro
2d72682cf962f9c81d453c56918827be
db1bcf585f5f4667e61fe2f34c6fe373e454d6cc
48668 F20110115_AACFCT mouraenko_o_Page_055thm.jpg
8b37ccdd51e8f012e84eaffe611bdb55
36dd19900608b31537fc684b5e065c7bbeee21c7
43706 F20110115_AACFDI mouraenko_o_Page_088thm.jpg
860226198bf0a965b6e98820f50508a1
419c89d7aac3d6eddeb4c728dedbe2e57cade42e
48172 F20110115_AACGGL mouraenko_o_Page_057.pro
15c8c1611c93e5598c2209f9c22a4e57
32b4dc21090615a4fd3113c0e2e597ab58695a4a
41743 F20110115_AACGFW mouraenko_o_Page_026.pro
5eaab49744a5cdb900af48d7f361af1f
74dc47a8f4730c292cdd72782f87a6def39790d7
32911 F20110115_AACFCU mouraenko_o_Page_134.pro
19a0cebfa873cd431e96750b282d41ca
6e9e31bf450be2896fb2e6abe530959197b00d12
76064 F20110115_AACFDJ mouraenko_o_Page_159.QC.jpg
db812c0c3cccb2203944638665c300e8
2002b4ffe6d4e367660a943dcdcc6068a48bec62
36814 F20110115_AACGHA mouraenko_o_Page_081.pro
cebf213389efdc77e1abde52cf052e10
09652ac4757ad72ab7109fd970feadcbbf68fbf6
35719 F20110115_AACGFX mouraenko_o_Page_027.pro
632f7b4e7648dec001cfd218405773ee
cea69be2886bde82f9c6f7638f6d06bebcba86c0
77672 F20110115_AACFCV mouraenko_o_Page_004.pro
718ab1fe4a67bbe6595dd6f2be70dc71
5a72fc689b5891fa8ba6ba9d2730cd95a4120cb6
41717 F20110115_AACGHB mouraenko_o_Page_082.pro
24f83d52c3d4b7dce537eca953a4637e
9b5b35947c2e0075876ba5ea668e2dffe74f426b
4303 F20110115_AACGGM mouraenko_o_Page_058.pro
7645b3562ce831ba5e90cfeb4add5d6b
eee304aecc3ca6a219359386592eaa8ca17f11aa
32958 F20110115_AACGFY mouraenko_o_Page_033.pro
4697a4270af84bb36acae80d28edfa0e
1608fe2398b5c1f41c9ab4a2041eba9491b1faba
743997 F20110115_AACFCW mouraenko_o_Page_037.jp2
4b8bd2b834b1845b0f35d8365b2ae80e
5b70af07b447e9ea9a52e73d40dbe0fe24c3b73b
917326 F20110115_AACFDK mouraenko_o_Page_152.jp2
19febfab58a77e95cb017c4e9f95aaf3
561293e1e3f17f04c23114e4f9185141d35488e6
42093 F20110115_AACGHC mouraenko_o_Page_083.pro
cdfa5f546184488949a7d0d41e45ff85
de85c49e2d5e75f7c29a2ea8d773435ce197e9f9
46511 F20110115_AACGGN mouraenko_o_Page_060.pro
5311f678a3980c519ab1aaca399e348c
0adb62a640ef556f6f053f2839b2818a84506471
37327 F20110115_AACGFZ mouraenko_o_Page_036.pro
76d318f27094e89b6cc598fb8cf6b10e
918d8b5bb5540608741858fc85558b398d123f50
698102 F20110115_AACFCX mouraenko_o_Page_052.jp2
e6363adc7dab118c357257ef05d7ec10
26f19f83308f92ea9102b96545447f3fda69d350
1991 F20110115_AACFEA mouraenko_o_Page_100.txt
08d14f30fc6ea6c4ab79cd9401fd8738
a6dedf82f0a6bd233f3bb3db79b29d0018d4d8d4
39657 F20110115_AACFDL mouraenko_o_Page_087thm.jpg
93090635a2256f176ffbe56394dd790b
a7f8be59331b6cb9a524ca33db8d8cc68b0e8b81
45726 F20110115_AACGHD mouraenko_o_Page_086.pro
fbc991005fb26a6a299d8b9f56cc9721
3a05f056c424f360e4f14d58742639f4f9dc5566
37217 F20110115_AACGGO mouraenko_o_Page_061.pro
9c651b9fa572977c5a280c170c823a1e
3e46631a062b60b1dcb7d20e26188b0260d4b374
3247 F20110115_AACFCY mouraenko_o_Page_126.txt
a73655417fae6659082e598181ac0a2d
aed57ceabd96c5312b8a4ab669d8f349f085869c
100129 F20110115_AACFEB mouraenko_o_Page_112.QC.jpg
c9afbc9fd1f69b0b0a213dc537eb4c9f
90160db82369a340eab8b2c63d0741a8ae9432f3
18701 F20110115_AACFDM mouraenko_o_Page_027thm.jpg
c2ad2c350a84d2a7b7b75b39054d8195
1f20c471f65b31ca65ff0979f0d7a4ede39a4e73
21530 F20110115_AACGHE mouraenko_o_Page_087.pro
6ade7c307f6338255fc6a3313f83f83b
5e64f8f311f52881f62d2be119fc6435ccfee129
33788 F20110115_AACGGP mouraenko_o_Page_062.pro
4780e9943b3a8e059438b3b0268a953a
8c9e3b83265c9b7b261e087f5b24ba63efce2455
2850 F20110115_AACFCZ mouraenko_o_Page_167.txt
bef8cc17e097d4ca69e47c19a425dcde
f31501ea227674eebbae1e1874b15bf0a3c1bfd1
30480 F20110115_AACFEC mouraenko_o_Page_084.pro
f7e8ab1bd94e5962c6b53ac4e3cd34b5
141c41936a6722221b9ef469a1fb4432f98de29c
1631 F20110115_AACFDN mouraenko_o_Page_119.txt
ae14fbd64ba223bdf2167f73eaa1c7a2
bc1480a488ebd7a691e9abac8b1e8bfcba79d13e
28184 F20110115_AACGHF mouraenko_o_Page_088.pro
417421c0f9aa030aff8ab13b137d9eff
e57f6367092d7ccade848d04608dd7c1388f08ce
57408 F20110115_AACGGQ mouraenko_o_Page_063.pro
669f87218f122d82976377627796d80f
df86918e301de9aa214105540c721fc8361af020
3667 F20110115_AACFED mouraenko_o_Page_005.txt
48523966ecab993c7685f26ebebd94bf
cb450f50dbaf5037d05f939f3a18825f96dc320e
187841 F20110115_AACFDO mouraenko_o_Page_045.jpg
5995717bca42089b61b84a9ffd5b8b42
08cfc897bdb50dc42b12903abaa51b830fe6f18a
39346 F20110115_AACGHG mouraenko_o_Page_091.pro
1f2d76284db9e384525364fa852704d6
ad31637b48ef5a18c7aeed4e1c083c61ba6d9d15
43757 F20110115_AACGGR mouraenko_o_Page_064.pro
c6e6a8c7751eede59e6effcdab6a2c57
87fc1dfa6152581cf7916dc0afe2722f0c525acd
12792 F20110115_AACFEE mouraenko_o_Page_050thm.jpg
8df2350bb9add4511fbc8830116ceae7
189ad0997e282f8efadb6944446c03eb76a40980
62649 F20110115_AACFDP mouraenko_o_Page_028.QC.jpg
9bee5f3659383391009aeb1ac59d8b88
c5b679b2816bc3332c3a9ca7a498a65d8748bfcc
55432 F20110115_AACGHH mouraenko_o_Page_092.pro
98150cbb0826e08f96dec79f40958788
9761c37930cdf9f7edb9894662a1f7646f0f39df
50931 F20110115_AACGGS mouraenko_o_Page_065.pro
acc87b878e9a64c51b21a83a63173d18
54982014400742b38eb9f9fae22dda0d7074f25a
78997 F20110115_AACFEF mouraenko_o_Page_006.QC.jpg
5ffdb3c866af8baa3ad4fbc5391724fd
ab422c9e26b33935f00e9c8f2db30624704267fd
1024897 F20110115_AACFDQ mouraenko_o_Page_030.jp2
f5debb1d56798971e15e9998f1605022
9889ab6168ff0c1b52eaec66a1fb528a28c926b1
43703 F20110115_AACGHI mouraenko_o_Page_094.pro
556213fd70ee81eff32419645152e963
48fcda899447a51fbdf009af5265da9790a531c7
33846 F20110115_AACGGT mouraenko_o_Page_066.pro
6f7a16d975b3ece787e1359929e64c51
822566b0abecb1c0d7f5e3dbc66d653a5ccd0ad9
51241 F20110115_AACFEG mouraenko_o_Page_154thm.jpg
07b97abea388fa82e029e608faedd915
80faeefec387001b76aef05e3a30bd346a9f929c
20546 F20110115_AACFDR mouraenko_o_Page_119thm.jpg
4bdb6e6a6a61d19de950c5c9afcf648d
55c510e0c49a19e8066e953cadc882cf95922615
38070 F20110115_AACGHJ mouraenko_o_Page_095.pro
92a12d8495082b19c086d83b81ae5d38
82259cee92574fc860e63ea12914d89e365e9f7a
42776 F20110115_AACGGU mouraenko_o_Page_067.pro
654328064c6f71bdec791ac58518a6a0
22f741e23f6456a0e25ca9a8a9fa2b93b3252bad
3048 F20110115_AACFEH mouraenko_o_Page_008.txt
ec35db5c067085b383dce5c328182a72
071f6b7400bcdc66c67b768070f1c530a6c398b5
226732 F20110115_AACFDS mouraenko_o_Page_096.jpg
2da9eec1b8036c9782f260858f1c3656
e98763b115133088aa969583ef7ca077efc12249
42824 F20110115_AACGHK mouraenko_o_Page_096.pro
a1cece7eeb63014cf321ad6a0ffc7fa7
8fb55a7f1a019e8fb3cd02e34a030050a19bd6c8
44711 F20110115_AACGGV mouraenko_o_Page_068.pro
bed2d6f10fb7d1a590c98e0af3e8379b
782ae41fba4665a629a83718f1a8395fbb3bf415
2189 F20110115_AACFEI mouraenko_o_Page_164.txt
155619554c57dcd83f5d83b6e808215f
3bbbd560f5dd2dc1a57be5c663736b130759c7dd
F20110115_AACFDT mouraenko_o_Page_049.tif
c7a1d4c193b317384225302fdf0805f7
6428db8e6e8c115e58325428146e2cb28100a01c
36707 F20110115_AACGHL mouraenko_o_Page_098.pro
08891a02a46e19f3781661b20c1b31c5
d273cedcfa2285712b996c24ddf602bc4ebfa3d3
26955 F20110115_AACGGW mouraenko_o_Page_071.pro
d47db6a197834d94e6c88a2a65b839f9
edefe5c6acdbae57a41422d354b62adb76b831c4
F20110115_AACFEJ mouraenko_o_Page_116.tif
0850934b0d5ffc3ed641a2c7d18fb874
97bd6aaf874fcc702dcd58445b413f7e9463ee53
39043 F20110115_AACFDU mouraenko_o_Page_119.pro
1b05146d42de308b9ec4d1ae502ddcb8
ee1c979882faeee052475550f343d7b4c8aa27b2
43367 F20110115_AACGIA mouraenko_o_Page_123.pro
ee8ef3d18a6e982a8d23fd3e95cab483
4593714ef3ab00a04607852b5522e3105850ee5e
51138 F20110115_AACGHM mouraenko_o_Page_099.pro
1071b7960eae09db76bad0736cd48886
ac270f197ff9ffc0e3c09bd94d5daaa2df94df36
46964 F20110115_AACGGX mouraenko_o_Page_074.pro
a32a1566c77eb7295941db366d5b8abc
37aa021029acf06c2f0ed5e694b616f1de26392e
F20110115_AACFEK mouraenko_o_Page_072.tif
d4b6b5602a09a83e23e809e96bfa37d1
21a00c0bff92c103a8d092b2ba3122cc52faaa27
F20110115_AACFDV mouraenko_o_Page_010.tif
a78e69649be77326cc491b341751f8e5
2081d74045ff1d33e90dab6e433a9dde559d0fa8
51327 F20110115_AACGIB mouraenko_o_Page_126.pro
26dd2a3611c6593a9dd27484df0ca630
dcae0e034ec1773bddaf1c818bbaf56672836eb5
54699 F20110115_AACGGY mouraenko_o_Page_076.pro
82805abb2799c4f3b79ed416794413cc
05af8088bf58bad07a18518a177c98ef90f8295b
F20110115_AACFDW mouraenko_o_Page_057.jp2
c428c12ef9804c6beecbd79c784f64f1
e00954e3366453630632c9eb41c7837876180ff8
59667 F20110115_AACGIC mouraenko_o_Page_128.pro
a555628643005b498fa3faba4555e3e7
03e25e602fd152c67a3dcb15a684a46aca9147d2
35964 F20110115_AACGHN mouraenko_o_Page_101.pro
3987f5c1298870e751beeed547f94f46
8806abe35992f215e8fe7d7d03df5c5fb31a8c47
33044 F20110115_AACGGZ mouraenko_o_Page_080.pro
4da5383e3f1daa924276eb32153f2bea
ce7ce825c3532847d546e033c28af9de3e10c2e8
50957 F20110115_AACFFA mouraenko_o_Page_007thm.jpg
42beae741bf62b664d327e5e4dd6a581
297aadfddff5bce16f6332fab4e57d13239460a6
F20110115_AACFEL mouraenko_o_Page_094.tif
2e0a117f7aa9e47623ce8fb49643185a
257fcf5f76ecfdcf71809db56894c645661c0a37
104118 F20110115_AACFDX mouraenko_o_Page_138.QC.jpg
aa1c3aeb3b15ff954af209bf284d7e64
39a7273513de75ac011052390819331678a9eff4
33687 F20110115_AACGID mouraenko_o_Page_133.pro
e221747d2006a9803262a56b54e4811a
8176b44ac7c3bda8ae7d71b76e07623e95101d13
34701 F20110115_AACGHO mouraenko_o_Page_102.pro
c436453b04903a428a1c258c6383f682
dcc0d3991f64246e42c03cd6dd222feaa483c596
F20110115_AACFFB mouraenko_o_Page_089.tif
9da6876a6b5f78d1ef437ae45e4373da
18051d51bffedc3a6b1d38a9ba5bdd7db82b1d4c
F20110115_AACFEM mouraenko_o_Page_152.tif
9158dfb3d66e7c8590c6754201783a3e
08cca906bff0d61a7e56021dce3c696c90519b70
779748 F20110115_AACFDY mouraenko_o_Page_044.jp2
a4fac09aaf4e12e610200a135ba46b55
04b8714ca9b24efeb5b41434c7af4e84998e97aa
38666 F20110115_AACGIE mouraenko_o_Page_139.pro
ed02d8158121d11ccf0c23b7ec450aa4
d173788d357091013ba03a3e8e9f0d4ecb67498a
41177 F20110115_AACGHP mouraenko_o_Page_104.pro
acb39b850cee226da70a3f08d1054482
cd9bbef66a2da498051ec0554233ae71639074d9
118281 F20110115_AACFFC mouraenko_o_Page_025.jpg
ea70ea80b30ee7527a18f4af1e762839
2f814e6fbd3064c611e35d4adef2213e99c10cd8
33954 F20110115_AACFEN mouraenko_o_Page_050.QC.jpg
dae5fa76f861ded12dd381b089cbf458
0b41665325dbdd4482bf6303dd89d27268d93936
F20110115_AACFDZ mouraenko_o_Page_023.jp2
33c58889f2044695c47c9df32caec489
71e5bd3a1c8c079e22cf81a954396ed183e40d08
15378 F20110115_AACGIF mouraenko_o_Page_141.pro
562256b29dbb1757bf770a0841287336
9cb75da2a8b34a8dc4ce169fafdbe26d0f7ddd33
48906 F20110115_AACGHQ mouraenko_o_Page_107.pro
de167b78280fa818ad87425dd88d17c4
7c2a0ab057519096d48f48fbf2dc92f4212c8250
770587 F20110115_AACFFD mouraenko_o_Page_046.jp2
5cc4f799a57dbae1d2d5ff6947fe5fa7
0b05afd5dbafe30fc231702c4c78ece812277691
74650 F20110115_AACFEO mouraenko_o_Page_052.QC.jpg
ec1a7799611493c2408464d2a1b18f13
2b0b9f75d9df58afbf808fb65d88df37dbe85bca
45669 F20110115_AACGIG mouraenko_o_Page_142.pro
4d9b96a412d201dd31fd3be09916721f
ad232dc8415aac0acfc464071522412b7fae9443
35091 F20110115_AACGHR mouraenko_o_Page_108.pro
6baea668e4ca61c3ced1610b4c9358d8
23ceaee92dba50384f29de43a975c0af0e6c0716
136036 F20110115_AACFFE mouraenko_o_Page_149.jpg
21be66618de6ed65fe952246a931ad66
b601ae8a4117d407d8e7d749b515814ff7d4b448
F20110115_AACFEP mouraenko_o_Page_156.tif
f48c116ff463e12f67ec15ab77044976
ccee5a3031607cb756260ba9feb683dcefbd7e6d
37341 F20110115_AACGIH mouraenko_o_Page_143.pro
c069557d33dc2e6ca1a678331988768f
10eb240bf08f6c11f36025fd56b82edab14aad27
28755 F20110115_AACGHS mouraenko_o_Page_110.pro
8877fa80b61c699ed3c8970e98e380d8
1b76a3b5e057dae3baa5c2f96545abea1cb48483
27023 F20110115_AACFFF mouraenko_o_Page_032.pro
c419c6a4481673560e9eccac48c6856d
0c5b3aabd5d62f454a37791c03e3da2ad5f09dff
F20110115_AACFEQ mouraenko_o_Page_122.tif
ce2863e8a56a9625fd01fb641260cd19
953b4d3b3ddb81f9d7d8edad0a9c6d10df18d127
47113 F20110115_AACGII mouraenko_o_Page_145.pro
8bac5514f01fcdd127cfe68bb0e209fa
82ed98172d4571bd131437ae38e5b4d54b529cca
53405 F20110115_AACGHT mouraenko_o_Page_112.pro
7b832c008de7a884a35988590d14292a
e03894dea9a4fd288121b1e5aff09b7cd9758e1d
50151 F20110115_AACFFG mouraenko_o_Page_071thm.jpg
11cf197aa3c0bdad6ff6ceaaeff204a0
fca95210e4033af85a6379ffc84ff4937c0d4cde
56126 F20110115_AACFER mouraenko_o_Page_013thm.jpg
dfa48c3123fc4a3655c9280bf2965e3c
bf4cbed8f3addba4cb596871db688ad5f80ac34b
41875 F20110115_AACGIJ mouraenko_o_Page_146.pro
c221efd7878ebfcb69e0fcac7bd95536
0ddf21d4f88c9fc0343ef721f7e854ff096f9835
47266 F20110115_AACGHU mouraenko_o_Page_113.pro
f112b4ba964121a6b9b5a4e657f9c605
c3a617985a4dec9911ce7aaefa25810ef9691f09
30628 F20110115_AACFFH mouraenko_o_Page_073.pro
7647f88b89054c74d3cffcc0641ee12c
43dfd8ea90e59702bd2fcd09e5ecfa9558473963
2948 F20110115_AACFES mouraenko_o_Page_148.txt
2292d3f47a2b7719b01d6205d5b49f70
08035c93409fe5589277477613eff92b91fea658
40443 F20110115_AACGIK mouraenko_o_Page_147.pro
4d7ad9bfb78269165b98c7917f471377
61bd1fb56a8804c3c5851c11139fce602cba446a
38389 F20110115_AACGHV mouraenko_o_Page_115.pro
7c924d26936a838a11119f152d579fc5
be77e0a287ae66fd5b469ce5acdf56b685cb250f
F20110115_AACFFI mouraenko_o_Page_159.jp2
3ccca4a608961f80879a252cc4befbc4
aa765df3d8039f07254a0214521c41205ecd911a
33896 F20110115_AACFET mouraenko_o_Page_155.pro
bb82e041adbf8e3e10351d16576afa7b
6b589c233e188b080bfd41f47f2b6dfd1ddda7f0
43198 F20110115_AACGIL mouraenko_o_Page_148.pro
1c06010437beb58fea38640ee52a97ce
fa72f8f1e1650fe2cfe33705e394583de6a7e6c3
53100 F20110115_AACGHW mouraenko_o_Page_117.pro
4be337b0462841548307d7c296d93dbf
f9c1924a7f233112198ef177d0aaa9577a5689b1
F20110115_AACFFJ mouraenko_o_Page_006.jp2
0297a38080121bef03c908df7dc8a228
5f9f516c714b97086fed37b4568ab48407a5cc5a
181882 F20110115_AACFEU mouraenko_o_Page_162.jpg
6ed1dd540934089bbac9f088f112e334
ea037aa4ac3a32a937cf93ea4e84ecf32cfccd87
3240 F20110115_AACGJA mouraenko_o_Page_004.txt
69c776948155676e046bb89550a40ed1
8a3d7cc31f9ada80067691699b71e9884e2fff7f
32373 F20110115_AACGIM mouraenko_o_Page_150.pro
126473326dc0dfda82a40d7e8d799870
f4df357274388b5c294a5091083980e963e49899
38197 F20110115_AACGHX mouraenko_o_Page_118.pro
bf0d9d86dc2064c49172e0a1de84430d
c0fb8b9f047a624efd8e979f805a6e74d6bf178a
36668 F20110115_AACFFK mouraenko_o_Page_015.pro
177c719deb95ecf92c058a8b9de5dc7c
cd1042a1d180d7c9477123e5c1f209719943a14f
2923 F20110115_AACFEV mouraenko_o_Page_112.txt
8faf0f4db88cb8a30faf3dde8c8b4c9a
d247ce3bca2cfc214ea6292b7f8ef2fae052647f
2578 F20110115_AACGJB mouraenko_o_Page_007.txt
f4785ee6cad15ca080a2491119f5ee5a
e1be5318d43fd0345f58a6a6d4eab9955b64c99b
30601 F20110115_AACGIN mouraenko_o_Page_151.pro
a7a994db64281e81919b93f854cc0c34
a4d39976989f4d1e8793aab4deb0c6e0c91eb54a
49495 F20110115_AACGHY mouraenko_o_Page_120.pro
531a993e42307aa43d915b798e3d4d4b
c488ae79b709ce9f4b5130359144f9c3c095232f
77198 F20110115_AACFFL mouraenko_o_Page_046.QC.jpg
9693742ecc9e1a606726e8657057ad82
cbeadf5f6cefed1b593e986487a36219397911c6
78117 F20110115_AACFEW mouraenko_o_Page_160.QC.jpg
a8a9c0bcd9dba40d9cdd0fdb7233fd4d
cb454742128f74af4f277f6a1f35b2de7eb2567a
3155 F20110115_AACGJC mouraenko_o_Page_009.txt
ebb99aec18cb4a226f19ae7b18ed005d
b49c407537379f725c1ce91e42266f52e7b42212
49776 F20110115_AACGHZ mouraenko_o_Page_121.pro
d67bd2f60082887a11ab092b8688be32
c64d355522962b691fe62525b7fecebce564193c
215186 F20110115_AACFEX mouraenko_o_Page_038.jpg
0843f8be499101a63393c0060cf6a99f
7fd359e5b7e23b158d939d6aef4e4741f131c576
106752 F20110115_AACFGA mouraenko_o_Page_120.jp2
4a71c6f0fcace1d43980f75d38c64dbe
7eddded7f6e2f47237f58542952b00f64ab34419
2627 F20110115_AACGJD mouraenko_o_Page_011.txt
c539dcf735ee1f5402e4780a2694a5ce
4d3968dc56d623f9ac2b9c7bb72da55d090ab74e
42036 F20110115_AACGIO mouraenko_o_Page_152.pro
c9c686940aecb2fd22e3ff79c8dd3e19
500cae2787ac0d985678ac2269c1c4381503f455
258939 F20110115_AACFFM mouraenko_o_Page_076.jpg
a28983b312b694faccf5b675c95b1a8b
985f6bd88eb025b40b250e50e17c09a72f8830fc
29054 F20110115_AACFEY mouraenko_o_Page_149.pro
d075cba5bc5b0508227d4a8ac40d8376
77ec8308a5f844bf1cbe759a1b9b8bdbb11230fa
43220 F20110115_AACFGB mouraenko_o_Page_041thm.jpg
02e52373208290e60b29ae0b56cb36d1
84d27929923bf2b762a0ac97c82a8cc4309c38f2
2614 F20110115_AACGJE mouraenko_o_Page_012.txt
051bbe31f7e14f99305fd0bd3be4ab67
f6cdbb1c68dd71bb3033e858dd4f620a24c506cf
40787 F20110115_AACGIP mouraenko_o_Page_153.pro
fa44900a38d84296bd0741b425f72a3a
688d58df97ad4affd69ab8d707aec0354df43dad
1535 F20110115_AACFFN mouraenko_o_Page_054.txt
be376b8f7dd0472a80676bdf46767955
98563611eb9565f562bc9cd5af067030c93ba7b7
F20110115_AACFEZ mouraenko_o_Page_125.tif
82082e1bbb5dbff2e27b1bf74663968a
c2217dbef881c4f674a15653ac7385be6c58a28b
F20110115_AACFGC mouraenko_o_Page_123.txt
db0983dc7f4e399a84aeaf9215077260
30e8b3a5568cea134be859c40f949da4072be5cd
2780 F20110115_AACGJF mouraenko_o_Page_013.txt
e552d55ec69b096aa4d363e76b86521b
beb368ee2b023715aa4fc80869eab2ca51103fe1
37915 F20110115_AACGIQ mouraenko_o_Page_154.pro
f56f70746088e2a3ff634d51488ae908
3c8cce5fef295b9fc37e39ad0c8f53a68f7bae37
928489 F20110115_AACFFO mouraenko_o_Page_061.jp2
47a3e92655f648ad5e39bc1ba8783beb
c0c3f41c5640549a5566188bee75809b7435ba25
37862 F20110115_AACFGD mouraenko_o_Page_025thm.jpg
67497cd4a5bde6ed8bc8878a0dfbf4ad
f4088501ae01729b3e5286adf1010b7e5673c801
2103 F20110115_AACGJG mouraenko_o_Page_016.txt
a93a1492ab5617a864d3fff9807a17b3
0b27ea3e38dbbec3ae8dc4c3a339ac57684f51c0
32157 F20110115_AACGIR mouraenko_o_Page_156.pro
f301e09afeb6ab890fef270b1ec0ebe6
7e3a0036adc38ea98cbe2e07bbf82a94875f0a71
49054 F20110115_AACFFP mouraenko_o_Page_072thm.jpg
4827df0e1bd006ccab33fd7d3cfe42fc
16d586b7182d14f35be955d7021d4b4450328230
2946 F20110115_AACFGE mouraenko_o_Page_010.txt
84c0c06fdf46ea7f97a925bf95dac09e
87c4b8993fa68c7843e8bd32b4db6861840a8a17
1805 F20110115_AACGJH mouraenko_o_Page_020.txt
0d607dd97006a8613360da446ea6cc01
62788ca9106800579ee25deb800530154f79ab88
19484 F20110115_AACGIS mouraenko_o_Page_157.pro
8ea643fe0fcf51c7bfef0625c5442f56
fd4c9f94d830abd601dfd54ba65a712b0e3c290f
111112 F20110115_AACFFQ mouraenko_o_Page_023.QC.jpg
0e96746d441492f82595b9c0b6f0f8d9
d8132981368452b2ab236d7f55689f35f879753a
240710 F20110115_AACFGF mouraenko_o_Page_113.jpg
7c3d00e6b5d96e1cbc409b2da1d850e1
7f43aa23c66831b764db7c21f530ae5c529200c5
F20110115_AACGJI mouraenko_o_Page_023.txt
0b223eb3fc1d85339a3010cbe6dd3de7
820c6273dae40729dc0d757eaa6b97276c4c339a
18483 F20110115_AACGIT mouraenko_o_Page_161.pro
19bd75b4fcb773ed59d1764e80f1c1ce
9469bbab6151faa662d4c4631acaa47a0edbd37e
183020 F20110115_AACFFR mouraenko_o_Page_081.jpg
98f43176dee169b0a15a2f90516ae7da
9f0b36f004e30fccc170b5a95714041e8ed6cf77
1659 F20110115_AACFGG mouraenko_o_Page_156.txt
07c0ced1d3e7f311c6ac4538cc2ab271
9513bba865583eddcd7d8959ee23e216761d245b
F20110115_AACGJJ mouraenko_o_Page_024.txt
852ff55f6891cff3945b0b350938a9b6
a71387c045054faeef1410cbe0201703f0641272
22694 F20110115_AACGIU mouraenko_o_Page_162.pro
f0c1e2d93de42294a679ad92bc36037f
ba5190c77a755a745cdafe41f61c489d070fc875
16368 F20110115_AACFFS mouraenko_o_Page_042thm.jpg
38554628d840017ac9bff57230fedd6b
2618e0616f9e0f45968c1a8d5df8ce1bdaacbdb8
665778 F20110115_AACFGH mouraenko_o_Page_141.jp2
315cfe9b340e35d0e30606bc8decd881
ce6db89da6d3fbdf4d0ff7e5bfdaa23d16087e49
848 F20110115_AACGJK mouraenko_o_Page_025.txt
037a8288937278d5a65d656c14ad160e
893680ac4cd4853c833e682d356ef0ccb36fcfca
52222 F20110115_AACGIV mouraenko_o_Page_164.pro
a35f1d7f12539cee245b83e23c4b1d17
4e167f3a552632e149ea4457e71dfa5e5cf86b1b
F20110115_AACFFT mouraenko_o_Page_013.jp2
24227ac022df81197e6aa33cfde3725b
4c8cfdc7297969133455c3417e24b0ad8b506954
22510 F20110115_AACFGI mouraenko_o_Page_164thm.jpg
6909ee05e33af502ecd7369d468f0c73
c726fbb02c58765ddd5faf8ef4d52f09448253f4
F20110115_AACGJL mouraenko_o_Page_026.txt
43c291a65d51e314e68a84fbc32d5033
80332a1b1ccc03f7c12902be0cfa346d648464b7
62945 F20110115_AACGIW mouraenko_o_Page_165.pro
60284829aef481172efb00508981095d
6d927d3d223524e4fbc1ccab1f84bcdcc1ede82d
216348 F20110115_AACFFU mouraenko_o_Page_095.jpg
e413f78f7273f67f7041280a21060298
6f8c238f7a331c4b64670521eec42838fad28bd4
41507 F20110115_AACFGJ mouraenko_o_Page_163thm.jpg
4f139d18985f0c154aae339ec806222a
aeb55b8cb939786e0b532aa889d6e32b61b9ff7a
1589 F20110115_AACGKA mouraenko_o_Page_044.txt
48e10656b552b7a01ae2fb943765580a
2a499a319a58c8671df85dde6868906fa96c21e6
F20110115_AACGJM mouraenko_o_Page_027.txt
ccc1fbb98e8c8e21927ff0e26ba747e8
404c034fa0c70cfcd70cfa8cbb4291fa4082961c
25884 F20110115_AACGIX mouraenko_o_Page_169.pro
6eb4aad0d7fa97d3e9741276158765be
7913bf4e948df90f2d285c2efb38df5b1a6f1000
48584 F20110115_AACFFV mouraenko_o_Page_056.pro
139c33ec2a4d85f68494789dfea314b6
30ce5c6c522eac2b140495de7d7c44a400019ad3
2309 F20110115_AACFGK mouraenko_o_Page_130.txt
aa505dbf4a46331e47d69b4c48b7595a
a79b6db4fb87c5a7e73bb9965fc5694bcd4c13e4
1660 F20110115_AACGKB mouraenko_o_Page_048.txt
79fe413ac8021dd8746026152c622431
4baf59eb66b7f4681dae5f9d9b37dbbc59311ced
2039 F20110115_AACGJN mouraenko_o_Page_028.txt
9c155f9f111682fbe2939a1da5528112
0e1ffaa2e886dc99b88313ae528f66ce8ab2ad79
414 F20110115_AACGIY mouraenko_o_Page_001.txt
b54e4e6c5c2250d7ee00633a16819fda
bbbd5e103ba356572437a7e27e1f142a81a8bd2d
F20110115_AACFFW mouraenko_o_Page_082.tif
bbae291bc061689e6c310ee0b403d37a
f9e18a76d4257a2a7384576313b7b8eeb01bd0a3
819768 F20110115_AACFGL mouraenko_o_Page_031.jp2
9a58f113b574e2a0040066e7c349ea42
93762c892f6c50b6aa2d5734edbf584f7d593d1c



PAGE 1

ACOUSTIC MEASUREMENT TECHNIQUES FOR SUSPENDED SEDIMENTS AND BEDFORMS By OLEG A. MOURAENKO A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2004

PAGE 2

To my family.

PAGE 3

iii ACKNOWLEDGMENTS I wish to express my appreciation to my chairman, Dr. Andrew B. Kennedy, for the freedom granted to me while pursuing my research interests, and for the guidance and support. I also wish to thank the members of my graduate committee, Dr. Robert G. Dean, Dr. Robert J. Thieke, Dr. Donald N. Slinn, for their intelligence and comprehensive supervision. My special thanks go to Dr. Renwei Mei, the external member of the committee, for his interest and involvement. I thank Dr. Peter D. Thorne for the guidance and great discussions on various aspects of acoustic measurements. I also thank Dr. Christopher R. Sherwood for providing the acoustic backscatter system and technical help. I thank my father, Alexander Mouraenko, for his patience and love. My special thanks go to the staff at the Coastal Engineering Lab, especially to Sidney Schofield, Jim Joiner, Victor Adams, and Chuck Broward, for their help, support, and technical guidance, without which this research would not have been possible. I thank Jamie MacMahan and Justin Davis for their help and endless support, with special thanks to Brian Barr for his exceptional knowledge, great vision of life and numerous discussions. This work was supported by the ONR Coastal Geosciences program and the National Science Foundation through the National Ocean Partnership Program (NOPP).

PAGE 4

iv TABLE OF CONTENTS Page ACKNOWLEDGMENTS.................................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii KEY TO SYMBOLS........................................................................................................xv ABSTRACT...................................................................................................................xvii i CHAPTER 1 INTRODUCTION........................................................................................................1 2 MULTIPLE REFLECTION SOLUTION FOR ACOUSTIC BACKSCATTER........7 Acoustic Backscatter Equation.....................................................................................7 Inverse Problem for Sediment Concentration.............................................................11 Multiple Layer Approximation...................................................................................11 Numerical Realization of Multiple Reflections...................................................15 Determination of Sediment Size and Concentration...........................................17 Number of Paths..................................................................................................18 Effect of Multiple Reflections.............................................................................19 Special Parameters for Acoustic Backscatter Equation..............................................23 Form Function and Normalized Cross-Section for Elastic Spheres....................23 High-Pass Model for Quartz Sand.......................................................................24 Modified Model for Elastic Spheres....................................................................25 Modified High-Pass Model for Natural Sediments.............................................26 Sound Attenuation Coefficient............................................................................27 Lognormal and Normal Distributions of Sediment Sizes....................................28 Averaged Form Factor and Normalized Cross-Section Terms...........................30 Circular Transducer Directivity and Half Intensity Beam Width.......................32 Near-Field Correction Factor..............................................................................34 3 CALIBRATION OF THE ACOUSTIC BACKSCATTER SYSTEM.......................36 Calibration Chamber...................................................................................................38 ABS Tests in the Chamber.........................................................................................40 Tests in Air and Still Water.................................................................................40

PAGE 5

v Homogeneity of Sediment Suspension and Effect of Chamber Walls................41 Results.................................................................................................................48 Estimation of System Function...................................................................................48 Approximation by 2nd Order Polynomial...........................................................49 Calibration of the ABS1......................................................................................50 Calibration of the ABS2......................................................................................53 Sensitivity Analysis For Estimation of System Function...........................................55 Dependence on Sediment Sorting.......................................................................59 Dependence on Transducer Radius.....................................................................60 D e t e r min a tion of P a r a met e r f .................................................................................60 Estimation of S y stem F u nction for A B S 1 ...........................................................62 Ana l y s is of Re l a tionship between and f ...........................................................67 Fun-Pump Tests......................................................................................................70 Sample Sediment Si z e D i stributio n .....................................................................71 Prediction of Concentration and Sediment Size Distribution..............................73 Parameters of Inversion.......................................................................................74 Results of Inversion for Concentration................................................................76 Results of Inversion for Sediment Size...............................................................81 4 MEASUREMENTS OF TARGET POSITION AND BOTTOM ECHO REMOVAL ALHORITHM............................................................................................................85 A Simplified Bottom Echo Model..............................................................................85 Tests with a Steel Wire...............................................................................................88 Tests with a Water-Sediment Jet................................................................................96 Result of Bottom Echo Removal................................................................................98 5 CONCLUSIONS......................................................................................................100 APPENDIX A CALIBRATION TEST RESULTS..........................................................................104 B EMPIRICAL MODEL FOR ACOUSTIC BACKSCATTER..................................123 C 1D ADVECTION-DIFFUSION MODEL................................................................130 Numerical Solution...................................................................................................132 Vertical Sediment Size Distribution.........................................................................133 D RESULTS FOR FUN-PUMP TESTS...................................................................137 LIST OF REFERENCES.................................................................................................145 BIOGRAPHICAL SKETCH...........................................................................................150

PAGE 6

vi LIST OF TABLES Table Page 2-1: First 24 Catalan numbers..........................................................................................19 2-2: Beam width for a circular piston transducer (ta = 5 mm and c= 1500 m/s)...........34 2-3: Near-field range nr for a circular piston source (ta=5mm and c=1500m/s)..........35 3-1: Locations and concentration levels in tests for 1.08MHz transducer.......................42 3-2: Locations and concentration levels in tests for 2.07MHz transducer.......................43 3-3: Locations and concentration levels in tests for 4.70MHz transducer.......................43 3-4: Sediment sizes..........................................................................................................63 3-5: Estimation of parameters and f for different sediment sizes..........................64 3-6: Values of x ka for test cases................................................................................64 3-7: Concentration levels, mean and standard deviation of the mixture for “fun-pump” tests.......................................................................................................................... .72 3-8: Coefficients of 2nd order polynomial for system functions.....................................75 4-1: Tests with 0.33mm steel wire...................................................................................91 C-1: Advection-diffusion model test parameters...........................................................133

PAGE 7

vii LIST OF FIGURES Figure Page 2-1: Example of the scattered sound within the narrow beam.........................................12 2-2: The multilayer approximation for a possible sound pulse traveling path. The time interval corresponds to 2 c....................................................................................13 2-3: Generation of a binary tree for numerical realization of multiple reflections..........16 2-4: Difference between intensities modeled with one and three reflections for selected profiles of concentrations: a) constant profiles, e) linear profiles, b–d) and f–h) percent difference.............................................................................................21 2-5: Difference between intensities modeled with one and three reflections for selected profiles of concentrations: a) profiles of constant concentration with “bump”, e) profiles of constant concentrations with “spike”, b–d) and f–h) same as in Figure 2-4.................................................................................................................22 2-6: Form function and normalized cross section for quartz sand (blue) and spheres (red): a) form function, b) normalized cross-section................................................25 2-7: Modified function for selected values of parameter .......................................26 2-8: Circular piston directivity function, 2D............................................................33 2-9: Near-field correction factor, 1 r, for ta =5mm and c =1500m/s.......................35 3-1: Calibration chamber—side view..............................................................................39 3-2: RMS voltage profiles (solid lines) with scatter intervals (dashed line) from the measurements in air and still clear water: a) 1.08MHz, b) 2.07MHz, and c) 4.70MHz...............................................................................................................41 3-3: Calibration chamber—top view. The jet system is shown with transducer positions for the sidewall proximity tests................................................................................42 3-4: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for 1.08MHz transducer for sidewall proximity tests....................................................43

PAGE 8

viii 3-5: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for 2.07MHz transducer for sidewall proximity tests....................................................44 3-6: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured along centerline with 1.08MHz transducer..............................................45 3-7: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured along centerline with 2.07MHz transducer..............................................45 3-8: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured along center line with 4.70MHz transducer.............................................46 3-9: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured near the chamber wall with 1.08MHz transducer....................................47 3-10: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured near the chamber wall with 2.07MHz transducer....................................47 3-11: Estimation of sediment size distribution used for the calibration of ABS1 by: a) sieve analysis, b) fall velocity..............................................................................50 3-12: Estimation of the system function for 1.08MHz transducer: a) ratio of measured and calculated voltage profiles, b) estimated system functions, and c) inverted sediment concentration.............................................................................................52 3-13: Estimation of the system function for 2.07MHz transducer: a–c) same as in Figure 3-12...............................................................................................................52 3-14: Estimation of the system function for 4.70MHz transducer: a–c) same as in Figure 3-12...............................................................................................................53 3-15: Estimation of sediment size distribution for sand used in calibration of ABS2 by: a) sieve analysis, b) fall velocity..............................................................................54 3-16: Estimation of the system function for 1MHz of ABS2: a–c) same as in Figure 3-12...............................................................................................................54 3-17: Values of fitted parameters ta , and with corresponding errors in inverted concentrations for the 1.08MHz transducer of the ABS1..........................57 3-18: Same as on Figure 3-17 for the 2.07MHz transducer of the ABS1.........................57 3-19: Same as on Figure 3-17 for the 4.70MHz transducer of the ABS1.........................58 3-20: Estimation of system function v K for 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12.............................................................59

PAGE 9

ix 3-21: Estimation of sediment size distribution for Case 4 used for calibration of ABS1 by: a) sieve analysis, b) fall velocity........................................................................63 3-22: Estimated system functions and their means for Cases 1–4 with =0.5 (dashed) and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz transducers of the ABS1........65 3-23: Ratios between individual system functions for Case 1–4 and mean system function with =0.5 (dashed) and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz transducers of the ABS1...........................................................................................66 3-24: Fitted parameters and corresponding parameters f for Cases 1–4.................68 3-25: Values of parameters f and for Cases 1–4 for different frequencies...............68 3-26: Values of and ka as functions of ka with approximated functions ka for Cases 1–4 (solid lines)........................................................................69 3-27: Sediment size distribution for Sand 1, Sand 2, and 1:1 mixture..............................72 3-28: Predicted with 1D advection-diffusion model profiles of concentration and sediment size distributions for maximum concentration of 0.8g/l at times 0–70s: size distributions at a) 30, b) 50, and c) 75cm below the free surface, d) sediment concentration profiles, e) profiles of mean sediment size in phi units, and f) profiles of standard deviations in phi units............................................................................73 3-29: High-pass model for backscattered intensity and its modified version used for “funpump” test: a) form functions, b) normalized cross-sections...................................76 3-30: Inversion by unmodified high-pass model concentration profiles for “fun-pump” test with maximum concentration of 0.4g/l: mean profiles (thick lines) and 1.08MHz profiles (thin lines)...................................................................................77 3-31: Inversion by unmodified high-pass model concentration profiles for “fun-pump” test with maximum concentration of 0.8g/l: mean profiles (thick lines) and 2.07MHz profiles (thin lines)...................................................................................78 3-32: Inversion by modified model concentration profiles for “fun-pump” test with maximum concentration of 0.8g/l: mean profiles (thick lines) and 2.07MHz profiles (thin lines)................................................................................................................79 3-33: Comparison of concentration profiles inverted by modified high-pass model with profiles calculated by the 1D advection-diffusion and empirical models (labeled as “ABn”) for the “fun-pump” test with maximum concentration of 0.8g/l at 10 second intervals........................................................................................................80

PAGE 10

x 3-34: Inverted profiles of mean sediment sizes for “fun-pump” test with maximum concentration of 0.8g/l by unmodified high-pass model..........................................82 3-35: Inverted by modified high-pass model profiles of mean sediment sizes for “funpump” test with maximum concentration of 0.8g/l..................................................82 3-36: Comparison of inverted profiles of mean sediment sizes by modified high-pass model with profiles calculated by the 1D advection-diffusion model for the “funpump” test with maximum concentration of 0.8g/l at 10-second intervals..............83 4-1: A sample of voltage profile with definition of “A”-point........................................87 4-2: Setup for the tests with steel wire.............................................................................89 4-3: Root-mean-square voltage profiles for the tests with 0.33mm steel wire. Tests 1–16 with the wire location (dashed line) and the bottom location (solid line)................92 4-4: Same as in Figure 4-3 for tests 17–31......................................................................93 4-5: Root-mean-square voltage profiles with subtracted bottom echo profile for the tests with 0.33mm steel wire. Tests 1–16 with the wire location (dashed line) and the bottom location (solid line)......................................................................................94 4-6: Same as in Figure 4-5 for tests 17–31......................................................................95 4-7: Root-mean-square voltage profiles (solid) for tests with 2mm steel wire with the wire locations (dashed) and “ghost wire” reflection locations (dash-dotted) from a) 1.08, b) 2.07, and c) 4.70MHz transducers..............................................................96 4-8: Setup for the tests with the sediment jet...................................................................97 4-9: Root-mean-square voltage profiles (solid) for tests for different water-sediment jet locations (dashed) from a) 1.08, b) 2.07, and c) 4.70MHz transducers...................97 4-10: Time series of suspended sediment concentrations estimated with a) bottom echo and b) bottom echo removed; c) free surface elevation...........................................99 A-1: Case 1: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer............................................................................104 A-2: Case 1: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer............................................................................105 A-3: Case 1: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer............................................................................105

PAGE 11

xi A-4: Case 2: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer............................................................................106 A-5: Case 2: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer............................................................................106 A-6: Case 2: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer............................................................................107 A-7: Case 3: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer............................................................................107 A-8: Case 3: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer............................................................................108 A-9: Case 3: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer............................................................................108 A-10: Case 4: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer............................................................................109 A-11: Case 4: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer............................................................................109 A-12: Case 4: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer............................................................................110 A-13: Case 1: Estimation of the system function for the 1.08MHz transducer initial value of parameter : a–c) same as in Figure 3-12.......................................................110 A-14: Case 1: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................111 A-15: Case 1: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................111 A-16: Case 1: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................112

PAGE 12

xii A-17: Case 1: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................112 A-18: Case 1: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................113 A-19: Case 2: Estimation of the system function for the 1.08MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................113 A-20: Case 2: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................114 A-21: Case 2: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................114 A-22: Case 2: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................115 A-23: Case 2: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................115 A-24: Case 2: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................116 A-25: Case 3: Estimation of the system function for the 1.08MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................116 A-26: Case 3: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................117 A-27: Case 3: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................117 A-28: Case 3: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................118 A-29: Case 3: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................118 A-30: Case 3: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................119

PAGE 13

xiii A-31: Case 4: Estimation of the system function for the 1.08MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................119 A-32: Case 4: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................120 A-33: Case 4: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................120 A-34: Case 4: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................121 A-35: Case 4: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12.............................................121 A-36: Case 4: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12...........................................122 B-1: Calibration coefficients for 1.08 MHz transducer: a) measured voltage (dots) and fitted by curves B-3 and calculated coefficients b) A, c) B and d) n .................127 B-2: Calibration coefficients for 2.07MHz transducer: a–d) same as in Figure B-1.....127 B-3: Calibration coefficients for 4.70MHz transducer: a–d) same as in Figure B-1.....128 B-4: Concentration profiles converted with the empirical model from the measurements with a) 1.08, b) 2.07, and c) 4.70MHz transducer..................................................129 C-1: Case 1—zero diffusivity: sediment size distribution at distance a) 0.25, b) 0.5, and c) 1m; d) mass concentration; e) approximate mean, and f) standard deviation....135 C-2: Case 2—constant diffusivity: a–f) same as in Figure C-1.....................................135 C-3: Case 3—linear diffusivity: a–f) same as in Figure C-1..........................................136 C-4: Case 4—parabolic diffusivity: a–f) same as in Figure C-1....................................136 D-1: “Fun-pump” test results for concentration profiles with maximum concentration of 1.2g/l: a–h) same as for 3-33..................................................................................137 D-2: “Fun-pump” test results for mean sediment size profiles with maximum concentration of 1.2g/l: a–h) same as for 3-36.......................................................138 D-3: “Fun-pump” test results for concentration profiles with maximum concentration of 1.6g/l: a–h) same as for 3-33..................................................................................139

PAGE 14

xiv D-4: “Fun-pump” test results for mean sediment size profiles with maximum concentration of 1.6g/l: a–h) same as for 3-36.......................................................140 D-5: “Fun-pump” test results for concentration profiles with maximum concentration of 2.0g/l: a–h) same as for 3-33..................................................................................141 D-6: “Fun-pump” test results for mean sediment size profiles with maximum concentration of 2.0g/l: a–h) same as for 3-36.......................................................142 D-7: “Fun-pump” test results for concentration profiles with maximum concentration of 2.4g/l: a–h) same as for 3-33..................................................................................143 D-8: “Fun-pump” test results for mean sediment size profiles with maximum concentration of 2.4g/l: a–h) same as for 3-36.......................................................144

PAGE 15

xv KEY TO SYMBOLS a [m] sediment radius ab [phi] sediment size range ,, abc coefficients of power-law curve ,, ABn parameters of empirical model ta [m] transducer radius b [m] root-mean-square width of bottom echo return B c, pressure correction factor c [m/s] speed of sound in water flC Courant number lc [m/s] speed of compressional wave in spherical particle nC n th Catalan number tc [m/s] speed of shear wave in spherical particle d [mm] sediment diameter t D transmitter directivity function E modulus of elasticity f [MHz] sound frequency f nondimensional form function f x nondimensional form function h filter impulse response function (1)nh spherical Hankel function of first kind I sound intensity 1 J first order Bessel function of first kind n j spherical Bessel function of first kind k [rad/m] sound wave number s K scattering parameter t K system parameter v K system function s L complex acoustic scattering length M [kg/m3] concentration of suspended sediments a M [m] mean sediment size N number of scatters bn [1/m3] density of scatters

PAGE 16

xvi 0P [Pa] reference sound pressure at 0rr i p coefficients of best-fit polynomial for v K L p lognormal probability density function N p normal probability density function rms P [Pa] root-mean-squared backscattered pressure t P [Pa] far-field sound pressure P [Pa] total pressure an receiver r [m] range from transducer transducer receive sensitivity 0r [m] reference range br [m] range to the bottom location b R bottom reflection coefficient i R reflection coefficient nfr [m] near-field range aS [m] sediment sorting, standard deviation of sediment sizes t [s] time T [C] water temperature ijT transmission coefficient vT voltage transfer function V [m3] gated volume 0V [volt] signal from mirror-reflected surface bv [volt] magnitude of bottom echo signal bV [volt] recorded bottom echo return envV [volt] amplitude envelope of recorded voltage rmsV [volt] recorded root-mean-squared voltage s V [volt] recorded signal from suspended scatterers W [m] reference beam diameter at r=1m mainW [m] reference main lobe diameter at r=1m s w [m/s] sediment settling velocity x ka z normalized distance [Np/m] total sound attenuation coefficient s [Np/m] attenuation coefficient due to suspended sediments w [Np/m] water attenuation coefficient f form function correction parameter normalized cross-section correction parameter modified scattering correction function 0 scattering correction function

PAGE 17

xvii r [m] thickness of insonified layer relative error 2 M variance of estimated concentrations s [m2/s] sediment diffusivity sediment attenuation constant [rad] latitude angle from central axis of transducer [rad] half intensity beam width main [rad] width of main lobe [1/m] sound wave length [phi] mean sediment size correction factor, calculated from data [kg/m3] water density A acoustic density s [kg/m3] sediment density t [m2] total scattering cross-section [phi] sediment sorting, standard deviation of sediment sizes [s] sound pulse duration phi scale normalized scattering cross-section near-field correction function D integrated beam pattern

PAGE 18

xviii Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy ACOUSTIC MEASUREMENT TECHNIQUES FOR SUSPENDED SEDIMENTS AND BEDFORMS By Oleg A. Mouraenko December, 2004 Chair: Andrew B. Kennedy Major Department: Civil and Coastal Engineering The acoustic backscatter system provides a nonintrusive method for measuring of profiles of suspended sediment size and concentrations. Therefore, an inversion model is required to convert the recorded series of intensities into parameters of sediment suspension. The measurement precision highly depends on the choice of the inversion model. In this work the high-pass backscattering model is modified to allow accurate inversion for both sediment sizes and concentrations. The modified correction function with a single parameter is introduced for the form function and normalized cross-section. The parameters can be found from the calibration of the system in the chamber with generated homogeneous suspension. To account for the difference in shape and mineralogy, a set of two parameters can be determined for a particular sediment. Their use greatly improves the accuracy of mean sediment size measurements obtained with the three-frequency acoustic backscatter system.

PAGE 19

xix The inversion model was applied to data obtained in the chamber with settling mixture of two sediment samples. The results were compared to the predictions by the 1D advection-diffusion model. Good agreement was shown for concentrations and sediment size measurements, with the error for sediment size approximation lying within the standard deviation of size distribution. The accuracy of the concentration measurements was different among the transducer frequencies, and was found to depend on the sensitivity of the transducer to the particular sediment size. The sediment sizes were estimated by minimizing the variance between concentrations, determined from different frequencies, and found to depend on the accuracy of estimated concentrations. For correction of recorded voltage data with applied time-varying gain, the system functions were found to be better approximated by second order polynomials. A multiple reflection model was developed based on the multiple layer approximation. The numerical simulation with the initial uniform concentration profile of 3g/l shows that the effect in intensities from the multiple reflections can be up to 14%. Also, a bottom echo removal algorithm was developed in order to eliminate the effect of high intensity scattering on the concentration measurements close to the bottom.

PAGE 20

1 CHAPTER 1 INTRODUCTION The application of acoustics for the study of sediment transport and development of prediction models for morphological changes of coastlines and bathymetry has been widely recognized and developed during the last few decades [ Thorne and Hanes 2002]. By providing remote, no intrusive methods of measurements of flow velocities, bottom topography, and profiles of suspended sediment concentration, acoustic instruments have proven a reliable way of obtaining high resolution data during field and laboratory experiments [ Hanes 1991; Hanes et al. 2001; Hanes and Vincent 1987; Hanes et al. 1988; Jansen 1979; Libicki et al. 1989; Traykovski et al. 1999; Vincent and Green 1990; Young et al. 1982]. The ultrasound at megahertz frequency range is used for the measurements of sediment suspension parameters within a meter range above the bottom [e.g., Smerdon et al. 1998]. The system consists of a sound transmitter and receiver. The properties of watersediment mixture along an insonified path can then be estimated from the changes in sound attenuation and from the intensity of scattered sound [ Morse and Ingard 1968]. Two configurations of such systems are usually applied: bistatic, when the transmitter and receiver are spatially separated, and monostatic, when the same transducer serves as a transmitter and a receiver. The later configuration measures the backscattered intensity from the scatters in suspension sound. The acoustic backscatter system allows measurement of both suspended sediment concentration profiles and distribution of sediment sizes [e.g., Crawford and Hay 1993;

PAGE 21

2 Hay and Sheng 1992; Schat 1997; Thorne and Hardcastle 1997; Thosteson and Hanes 1998]. Compared to other techniques, acoustics are a nonintrusive way of measuring suspended sediments with a high temporal (in the order of a second) and spatial (in order of a centimeter) resolutions [ Wren et al. 2000], with comparable accuracy [ Hanes et al. 1988; Thorne and Hardcastle 1997; Thorne et al. 1995]. Since the acoustic backscatter system is an indirect method of measurement, an inversion algorithm is required for converting measured backscattered sound intensities into parameters of suspension. The acoustic backscatter equation [e.g., Medwin and Clay 1997, p.353] provides the basis for the development of such an algorithm. The difficulties arise from the dependence of a backscattered signal from the properties of scatterers, their concentration, directivity of the sound beam, and also particular electronic implementation. Semi-empirical approximations of these parameters [e.g., Tamura and Hanes 1986] can be used to convert the data into concentration profiles. Although a better description of the parameters can be found from theoretical approximation of sound scattering properties of natural sediments. The acoustic properties of scatterers depend on many parameters, such as size, shape, mineralogy, distribution of those parameters in a sample, speed and direction of motion, their concentration, relative position to the sound transmitter and receiver, and others. Only a limited set of properties can be considered for the inversion. The simplified model for acoustic backscattering allows a user to obtain the most important properties for the study of sediment transport, while others can be approximated by general expressions. Only suspended sediment concentration, mean sediment size and sorting are considered to provide the necessary description of the water-sediment suspension, while the lognormal distribution of sediment sizes in a sample is assumed. The scattering pa-

PAGE 22

3 rameters then are approximated by two functions—the form function and normalized cross-section. The choice of these two functions determines the accuracy and reliability of inversion model. The following works define the necessary framework for development of the inversion model: [ Downing et al. 1995; Faran 1951; Hay 1991; Hay and Mercer 1985; Hay and Schaafsma 1989; Neubauer et al. 1974; Sheng and Hay 1988; Thorne and Campbell 1992; Thorne et al. 1993a]. A multiple frequency setup needs to be used in order to estimate sediment sizes. Schaafsma [1989] used a bistatic configuration of variable frequency transducers, which covered the frequency range between 1–100MHz, to measure concentration and sediment sizes of a mixture of two sediment samples. He showed that concentrations could be obtained with 20% error, while the sediment sizes can be resolved within a factor of 2. For field measurements monostatic configurations are usually used. A system consisting of a set of three or more transducers with frequencies 1–5MHz is able to estimate the mean sediment sizes along the insonified column [ Crawford and Hay 1993; Hay and Sheng 1992; Schat 1997; Thosteson and Hanes 1998]. The accuracy of the estimation highly depends on correct approximation of scattering properties of the sediments, as well as on the accuracy of estimations of suspended sediment concentrations from each of the frequencies. It has been shown that the accuracy and stability of estimations of sediment sizes can be achieved by averaging among a series of consecutive profiles, which leads to significant loss in temporal resolution [e.g., Thorne and Hardcastle 1997]. Several problems regarding measurement accuracy were addressed by the current dissertation. Their resolution improves both the accuracy and the temporal resolution of the measurements, and provides a basis for future improvements.

PAGE 23

4 Usually, only measurements at low suspended sediment concentrations (<10g/l) are performed. Therefore, the multiple scattering of the sound waves can be ignored [e.g., Sheng and Hay 1988]. Although it is true for natural sand, in the presence of stronger scatterers, like air bubbles, or at higher concentrations, as in the region near the sand bottom, multiple scattering can become important. Here a multiple reflection model is developed and the effects of multiple scattering on resulted errors are shown. The acoustic properties of different natural sediments vary from sample to sample. By using the approximated parameters, such as form function and cross-section, obtained for a particular type of scatters like glass spheres or quartz sand, the inversion algorithm can lead to erroneous results if the actual sediment differs from the approximation in mineralogy and shape [ Thorne and Buckingham 2004]. Therefore, acoustical properties of the actual sediments need to be well defined before the inversion process can be done. The detailed study of each sediment sample, obtained from the experimental site, for mineralogical content, distribution of shapes, and overall acoustical performance is an even more difficult problem, which even in a case of positive solution cannot guarantee correct results, since the sediment content can change during the period of measurements. The current work describes a new scattering model, which modifies a correction function originally presented by Thorne and Buckingham [2004], with a technique to determine necessary parameters. The description of the acoustic backscatter equation, an implementation of the multiple layer solution, and the newly developed model are discussed in Chapter 2 The calibration of an acoustical backscatter system is performed in order to obtain system transfer function, which describes the dependence of sound intensity and recorded voltage levels. Different methods of calibration can be applied, such as by using thin-wire

PAGE 24

5 targets [ Sheng and Hay 1993], calibration with a particle-laden turbulent jet [ Hay 1991], or in a calibration chamber with a homogeneous suspension [ Tamura and Hanes 1986; Thorne et al. 1993a]. The time varying gain (TVG) is usually applied for the amplification of low level sound signals, arriving from the farther distances from the transducer. With the proper correction for the TVG, the system function should be uniform with the range and therefore, can be determined from the calibration. In fact the system functions found from the calibrations are not always uniform with the range and may depend of particular sediment sample used for the calibration [ Thorne and Buckingham 2004]. The variations in system constant can reach 25%, which result in a corresponding error in estimation of suspended sediment concentrations. By providing proper corrections to the system functions developed here, these errors can be lowered. The solution of the acoustic backscatter equation for suspended sediment concentration can be obtained explicitly with some limitations [ Lee and Hanes 1995]. Further development of the method [ Thosteson and Hanes 1998] allowed us to estimate the profiles of sediment sizes. The implicit method [ Thorne et al. 1993a] provides a more flexible way of solution, although it can become unstable; when the concentrations become large and the sediment attenuation becomes significant, even small 5–10% errors can result in unbounded profiles of concentrations from the iterative solution [ Thorne et al. 1995]. If the concentration profile is approximately known, for example, from the pump sampler measurements [e.g., Thorne et al. 1993a], the correction for sediment attenuation can be performed more accurately. In Chapter 3 calibration of the acoustic backscatter system together with a technique of evaluation of the system parameters and the parameters for the correction functions is described. Also, the results of a series of laboratory tests are shown and compared

PAGE 25

6 to the 1D advection-diffusion model predictions for sediment size and concentration profiles. The bed echo contains important information, which can be used to improve accuracy of the acoustic backscatter system [ Thorne et al. 1995]. Also, the intensity of the bed echo at low concentrations is much greater than the intensity of the sound backscattered from suspended particles. Because the acoustic backscatter system requires averaging and filtering of the input signal, signals from sand bottom interchange with the signals from suspended sediments. The algorithm for separation of these signals can provide one with the information on bed location and allow to determine concentrations of suspended sediments in the vicinity of the bottom. The algorithm is discussed in Chapter 4

PAGE 26

7 CHAPTER 2 MULTIPLE REFLECTION SOLUTION FOR ACOUSTIC BACKSCATTER Acoustic Backscatter Equation The acoustic backscatter equation is derived from the time integral of squared pressure backscattered from the objects in an insonified volume with following assumptions [ Medwin and Clay 1997]: 1. The objects in the insonified volume are alike, which implies that they are of a similar shape, size and mineral structure, and can be substituted with a population of objects with average properties. 2. The distribution of the objects within a gated volume is random and uniform. The gated volume is a volume of fluid, insonified within the short time interval—time gate. 3. The time gate is assumed to be large compared to the sound pulse duration. In practice, however, the time gate is often chosen to be equal to half of the pulse duration [e.g., Smerdon et al. 1998]. 4. The attenuation of sound within the gated volume is ignored. The far-field sound pressure from a circular piston source in iso-speed medium is then equal to [ Medwin and Clay 1997] 0 0()r ttr r P tDPte cr (2-1) where tD is the transmitter directivity function given by 2-65 0 P t is the sound pressure at the range 0r along the central axis, is the total sound attenuation coefficient given by 2-42 c is the speed of sound in water, and r is the range from the transducer. After the transducer emits a sound pulse, it travels through the volume of water with suspended sediments. A portion of the sound is scattered by any particle in the direction of

PAGE 27

8 the receiver. The pressure at the receiver is equal to ()r s rrtL P tDPte r (2-2) where rD is the directivity of the receiver, s L is the complex acoustic scattering length, and the factor 1 r accounts for the spherical spreading of the sound wave. The complex acoustic scattering length describes the scattering properties of the particle. It is a function of angles between incident and reflected waves relative to the particle, and the sound wave number k. For spheres it can be rewritten as [e.g., Sheng and Hay 1988] 2safx L (2-3) where a is the particle radius, and f x is the nondimensional form function with x ka In general, the form function is a complex valued function. For scatterers of simple shapes, like cylinders and spheres, it can be found analytically [ Morse and Ingard 1968]. But for particles of irregular shapes, such as sand grains, approximations are used to describe the amplitude of backscattered form function [ Crawford and Hay 1993; Neubauer et al. 1974; Sheng and Hay 1988; Thorne and Buckingham 2004; Thorne et al. 1993b]. It relates to the form function as 0, f fx (2-4) The total pressure at the receiver, P is a sum of all backscattered pressures from individual particles. The squared pressure amplitude can be calculated as

PAGE 28

9 2 ** ,,,, 111NNN ririrjri iij ji P tPtPtPtPt (2-5) where N is the total number of particles, and asterisk denotes complex conjugate value. The second term is the sum of the cross-products of the pressures. In the case of a homogeneous suspension it can be shown that this term approaches zero either with increasing the number of particles or by ensemble averaging of the total backscattered pressure realizations [ Libicki et al. 1989; Smerdon et al. 1998; Thorne et al. 1993a]. Previously mentioned assumptions 1 and 2 define the homogeneity of the water-sediment suspension. In practice, both methods of averaging are used to obtain the root-mean-squared (RMS) pressure profiles [e.g., Smerdon et al. 1998]. The RMS backscattered pressure for a given range r is calculated as is the integral over the gated volume of the total backscattered pressure squared, which for the gated time equal to sound pulse duration, can be written as [ Hay 1991] 22 22224 00 41sinh 4r rmsbDaf B P rnrre rB (2-6) where 4 42 2 00r c r cr c P tdt is the reference pressure, bnNV is the density of the particles in gated volume, 2 rc is the extent of the sample volume in r, and implies expected value over the sediment size distribution. The term sinh B B with B c is the correction factor if the attenuation over the gated volume is large [ Hay 1991; Thosteson and Hanes 1998]. The integrated beam pattern is calculated as 2 2 22 00,,sinDtr D Ddd ! (2-7)

PAGE 29

10 which for the monostatic system, when transmitter and receiver are located in the same point and their directivities are the same, can be approximated as [ Thorne and Hardcastle 1997] 20.96 2D tka 10tka # (2-8) where ta is the transducer radius. The density of the particles can be calculated by approximation of the particle shape by spheres with radius a Therefore, it can be written as a function of particle mass concentration, M as 33 4b s M n a (2-9) where s is the density of sediments. The root-mean-square voltage, recorded by the acoustic backscatter system can be written in the form [ Thorne and Hanes 2002; Thorne and Hardcastle 1997] 12 21r rmsvrmsstM VTPKKe r (2-10) 12 22 31s saf K a (2-11a) 1230.96 16tv tc KK ka (2-11b) 00 vv K Tr (2-11c) where is the transducer receive sensitivity, vT is the voltage transfer function of the system, is the near-field correction factor given by equation 2-67 t K is the system

PAGE 30

11 parameter, which is a constant if all other parameters of the system are fixed, or a function, if the time varying gain (TVG) is applied. It can be determined by the calibration with known concentrations and sediment properties [ Hay 1991; Thorne and Hanes 2002; Thorne et al. 1993a]. Inverse Problem for Sediment Concentration The parameters of the suspension such as sediment concentration and sediment size distribution parameters can be evaluated by solving the inverse to equation 2-10 problem [ Thorne and Hanes 2002]. For the concentration, the solution can be obtained from the equation 2 4()r rms stV M rre KK (2-12) It cannot be solved explicitly, since the reflection parameter s K and total attenuation coefficient are functions of sediment size distribution and concentration. The solution by iterations is usually applied. As can be seen from the equation, the solution for the concentration can be unstable, since it has positive exponential dependence on attenuation. Multiple Layer Approximation In the derivation of the equation 2-6 for RMS pressure, the multiple scattering between the particles was ignored. Multiple scattering for the suspension of sand can be neglected for suspended sediment concentrations of less than 10 g/l [ Sheng and Hay 1988]. At higher concentrations, which can be found at the ranges close to the bottom [e.g., Dohmen-Janssen and Hanes 2002], or in the presence of stronger scatterers in the fluid, such as a wire or bubbles, multiple reflections can affect the measurements. After the transducer emits a sound pulse, it travels within a narrow directed beam. Part of the sound wave is scattered by the suspended particles, and it may be rescattered

PAGE 31

12 by other particles or reflected from the transducer surface multiple times before the backscattered sound is recorded by the receiver. The receiver detects only the sound arriving from the direction within the narrow beam, which is defined by the directivity of the transducer ( 2-65 ). Therefore, the sound scattered backwards in the direction of the incident sound will most likely remain within the beam and be detected by the receiver (see Figure 2-1 ). Table 2-2 shows the approximate beam widths for transducers of different frequencies. r r D ( ) 3dB Figure 2-1. Example of the scattered sound within the narrow beam. The RMS pressure ( 2-6 ) is calculated as the integral of the backscattered pressure over the time gate interval, which is again assumed to be equal to half of the pulse duration. During each time interval, the sound wave travels the distance 2 c. Therefore, the whole range from the transducer surface can be represented as a set of layers with thicknesses 2 c (see Figure 2-2 ). The center of the n th layer is located at distance

PAGE 32

13 2nc rn (2-13) where n =0 corresponds to the transducer surface. 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 0 1 2 3 4 TimeLayers T R A N S D U C E R leading edge trailing edge Figure 2-2. The multilayer approximation for a possible sound pulse traveling path. The time interval corresponds to 2 c. The sound waves can travel along many possible paths within the beam and scatter from the layers of suspended particles. In the multiple layer approximation all the paths begin at the layer 0 and pass between the layers with possible changes in direction. The mean squared pressure at the transducer at time nntrc is equal to the sum of mean squared pressures of the sound followed all possible paths 122 1nC rmsrmsi i P P (2-14) where the nC is the total number of the paths given by equation 2-26 Obviously, the length of every path in the summation should be equal to nr.

PAGE 33

14 Let the sound pressure corresponding to an individual path at level i be i P With spherical spreading and attenuation of the sound, the backscattered pressure from level j becomes j jiij ijR P PT r (2-15) where ijr is the distance between layers, ijT is the transmission coefficient, and j R is the backscattering coefficient. The backscattering coefficient defines a part of sound which is scattered from the layer of suspended particles in the direction of the incident sound wave. It is proportional to the number of scatterers within the layer and their scattering length ( 2-3 ). Using the same assumptions as for the acoustic backscatter equation, the coefficient can be written as 22 22 33 42 4jj jDj s jaf M c R r a (2-16) where the term 224jaf approximates the scattering length of the scatterers in the layer j and the term 22Djrc defines the insonified volume at layer j The transmitting coefficientijT is equal to ijijr ijTe (2-17) where the attenuation coefficient is calculated using the trapezoidal rule as 1122j i ijij (2-18) Therefore, the RMS pressure corresponding to a path of length r is equal to 2 2 22 ,1 2ijij ijr j rmsti r ijR PPe r$ (2-19)

PAGE 34

15 where ,1 ti P is initially transmitted pressure ( 2-1 ) with the near-field correction factor applied ( 2-67 ). The multiplication is made over all sub paths, whose total length is ij ijrr Numerical Realization of Multiple Reflections The main difficulty in numerical realization is to account for all possible paths that sound waves can travel along. In the newly developed algorithm the binary tree is used to solve this problem. The generation of the binary tree is possible, because from any layer a sound wave is assumed to be traveling only along the directional beam in two directions—up or down. For the given number of layers, the binary tree will look like it is shown in Figure 2-3 The nodes of the tree correspond to the centers of the layers, and their numeration was done in a way that simplifies the generation of the binary tree. The root of the tree is located at the layer corresponding to the transducer surface. There is a set of parameters associated with every node. They include the indices of the nodes for possible up and down directions, as well as calculated reflection ( 2-16 ) and attenuation ( 2-42 ) coefficients. The same reflection and attenuation coefficients are applied to all nodes located at the same layer. After the binary tree is generated, a fast algorithm of tree-walk is used [e.g., Knuth, 1997]. The walk over the binary tree starts with the node at the transducer surface, where the reflection coefficient is assumed to be 1 and the attenuation coefficient is equal the value for plain water. This implies complete reflection of the sound wave from the transducer surface and zero suspended sediment concentration. After the first step “down” is made (from the node “1” to “2” in Figure 2-3 ), neither reflection nor attenuation coefficients are known at layer 1 (node “2”); therefore they are initially assigned with the val-

PAGE 35

16 ues 1 and 0 correspondingly. The intensity of the sound at layer 1 becomes 02 21 11rIIe r % (2-20) where 1 I is the initial intensity of sound emitted by the transducer with the index corresponding to the node number, and 1 is the near-field correction factor calculated at rr 0 1 2 3 4 5 6 0 1 2 3 4 5 6 Bin numbersLayers 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 T R A N S D U C E R possible direction possible path Figure 2-3. Generation of a binary tree for numerical realization of multiple reflections. A step “up” is made next (from the node “2” to “8” in Figure 2-3 ), and the calculated intensity of sound arriving back to layer 0 becomes 082 21rIIe r%% (2-21) where “primes” are used for the intensities calculated with initial values 11R% and 10% The calculated intensity 8 I should be equal to the measured intensity correspond-

PAGE 36

17 ing to bin one and recorded at time ,1m I This provides us with an expression from which the concentration at leyer 1 can be found, and hence, corresponding reflection and attenuation coefficients 1,1 2 2 1 8m rI Re I % (2-22) After the reflection and attenuation coefficients for layer 1 are determined, the intensities along all the paths within the layers 0 and 1 can now be calculated and accumulated according to the equation 2-14 The next step “down” is made (from the node “2” to “3”) and the procedure is repeated until the last layer is reached. The expression to find the concentration in general form becomes 2 ,, 2i i ir mii nII Re I % (2-23) where ,i I is the sum of intensities calculated for the paths which arrive at the transducer surface at the time corresponding to bin i. The node number is found as 1 1 2ii nni (2-24) with n number of layers. Determination of Sediment Size and Concentration Three parameters need to be known at every layer to calculate reflection and attenuation coefficients: suspended sediment concentration M and parameters of sediment size distribution, a M and aS. If suspended sediment concentration is the only unknown parameter and mean sediment size, a M and standard deviation, aS, are known somehow (for example, from additional measurements or assumed values are used), then equation 2-23 can be solved

PAGE 37

18 via iterations [Thorne et al., 1993a]. To determine the sediment size parameters and the suspended sediment concentration more than one measurement with different sound frequencies are needed. The techniques were developed to obtain concentration, M and mean sediment size, a M by using 3 transducers, while standard deviation, aS, is assumed to be known [Hay and Sheng, 1992], [Crawford and Hay, 1993], [Thosteson and Hanes, 1998]. To obtain all the parameters— M a M and aS—a system with six transducers can be used [Schat, 1997]. More often the 3-transducer setup is used. It usually requires sorting, aS, of the sediments to be known. The mean sediment size can be estimated from the minimization of the variance of sediment concentration approximated by each of the transducers 3 2 2 1aMak k MMMM (2-25) where 3 13k kMM. Number of Paths The number of possible paths grows very quickly with the number of layers n For a specified n the total number of paths is equal to (1)nth Catalan number [ Weisstein ]. The n th Catalan number is equal to 2! 1!!nn C nn (2-26) For illustration the first 24 numbers are given in Table 2-1 The asymptotic form of n th Catalan number is

PAGE 38

19 324n nC n (2-27) which for the number of layers n = 120 gives 68 1207.5810C& Table 2-1. First 24 Catalan numbers n nC n nC n nC 1 1 9 1,430 17 35,357,670 2 1 10 4,862 18 129,644,790 3 2 11 16,796 19 477,638,700 4 5 12 58,786 20 1,767,263,190 5 14 13 208,012 21 6,564,120,420 6 42 14 742,900 22 24,466,267,020 7 132 15 2,674,44023 91,482,563,640 8 429 16 9,694,84524 343,059,613,650 Because of the large number of the paths, the execution time becomes very long even for a small number of layers. Therefore, not all the paths can be included in calculation. The number of reflections which a sound wave is exposed can be chosen as a criteria for the path to be included in the calculation. The number of reflections for each path varies between 1 and 12 nn where n is the number of layers. The reflection coefficient for the layer of scatterers is determined by equation 2-16 To approximate the order of magnitude of the coefficient, the following values can be used: r=1m, =10s, c =1500m/s, D =0.015, s =2650kg/m3, M =3g/l, a =0.2mm, for which R "0.01. Two reflections result in attenuation of incident pressure on the order of 10–4, three—10–6. Although, the number of the paths can be very large, the sound waves of small amplitude will be absorbed by the suspension. Therefore, the number of reflections can be limited to 3 if the number of layers is large ( n 15). Effect of Multiple Reflections Four tests were performed to analyze the relative effect of multiple reflections on measurement of sediment concentration. The ratios of intensities calculated with one and

PAGE 39

20 three reflections using the model described above are plotted in Figures 2-4 and 2-5 For uniform concentration profiles (see Figure 2-4 a–d), the effect of multiple reflections is almost uniform with distance. The difference reaches its maximum of 12–14% at 3g/l, but at 1g/l it is only around 5%. For the concentration profiles, which linearly change from 0g/l to maximum values of 0.1–3g/l with distance (see Figure 2-4 e–h), the difference is even smaller—on the order of 1–3%. This is probably the most common situation for field measurements, when the concentration near the transducer is much smaller than it is near the bottom. However, it is possible for a cloud of sediments to be injected by a turbulent eddy into the water column with the concentration at higher elevations to become greater than the concentration near the bottom. Figure 2-5 a–d shows the results when uniform concentration profiles were modified by a jet-type profile with maximum concentration of 2g/l, located at around 30cm distance from the transducer. The difference becomes more visible, especially at the ranges just after the jet location. Compared to uniform concentration profiles, the intensity increases around 1.2 times. In the next test, a wire-type response was modeled with results shown in Figure 2-5 e–h. For the profiles at 0.1 and 0.5g/l there are “ghost” peaks present, which correspond to the second reflection of sound between the transducer surface and “wire”. This situation can be modeled in the lab with a strong reflector, like steel wire (see Chapter 4 ). Although the second reflection is present in Figure 2-5 b–d, there are no such “ghost” peaks in the tests with the sediment mixture jet (see Chapter 4 ). In the field, air bubbles can be present in the water column, and since the bubbles are strong reflectors the second reflection from them can affect the measurements at longer ranges.

PAGE 40

21 0 1 2 3 a).M, g/lConcentration profiles 0 5 10 15 b). 1.08MHz% 0 5 10 15 c). 2.25MHz% 0 0.2 0.4 0.6 0.8 0 5 10 15 d). 4.75MHz%Distance 0 1 2 3 e).M, g/lConcentration profiles 0 0.5 1 f). 1.08MHz% 0 1 2 g). 2.25MHz% 0 0.2 0.4 0.6 0.8 0 2 4 h). 4.75MHz%Distance Figure 2-4. Difference between intensities modeled with one and three reflections for selected profiles of concentrations: a) constant profiles, e) linear profiles, b–d) and f–h) percent difference.

PAGE 41

22 0 2 4 6 a).M, g/lConcentration profiles 0 5 10 15 b). 1.08MHz% 0 10 20 c). 2.25MHz% 0 0.2 0.4 0.6 0.8 0 10 20 d). 4.75MHz%Distance 0 5 10 e).M, g/lConcentration profiles 0 10 20 f). 1.08MHz% 0 10 20 30 g). 2.25MHz% 0 0.2 0.4 0.6 0.8 0 10 20 h). 4.75MHz%Distance Figure 2-5. Difference between intensities modeled with one and three reflections for selected profiles of concentrations: a) profiles of constant concentration with “bump”, e) profiles of constant concentrations with “spike”, b–d) and f–h) same as in Figure 2-4

PAGE 42

23 Special Parameters for Acoustic Backscatter Equation Form Function and Normalized Cross-Section for Elastic Spheres The theoretical far-field form function for elastic spheres can be expressed as [ Thorne and Campbell 1992] 02 121n s n n f xnb ix (2-28) where 11213111213 22223212223 3233323300nb (2-29a) 2 1 tsn x jx 2 n x jx% 2(1) 11 tsn x hx (1) 21 n x hx% 2 12214tnllnlnnxjxxjx%, 22 lnl x jx% 322nllnljxxjx% (2-29b) 1321tntntnnxjxjx% 231ntnnjx 2 332212tnttnt x jxxnnjx%, where tt x xcc ll x xcc c is the sound speed in the water, tc and lc are the speeds of shear and compressional waves in the spherical particle, and s are the fluid and particle densities, n j is the spherical Bessel function, and (1) nh is the spherical Hankel function of first kind. The derivatives of the functions can be found by the recurrence relation [ Abramowitz and Stegun 1965]

PAGE 43

24 1 nnnn g gxgx x % (2-30) where g x is either a spherical Bessel or Hankel function. The normalized scattering cross section is defined as 22ta (2-31) where 2 ts L is the total scattering cross section of a single particle of radius a For the spherical particles it can be found to be 2 2 02 21sn n x nb x (2-32) Equations 2-28 and 2-32 involve infinite series, but only a finite number of terms can be calculated. The following approximate expression can be used to calculate the number of terms sufficient for the resulting accuracy of 610O for 50x( 6101.22466.4232Nx (2-33) High-Pass Model for Quartz Sand The modified high-pass model for backscattered intensity [ Crawford and Hay 1993; Johnson 1977; Sheng and Hay 1988; Thorne et al. 1993b] provides simpler expressions for the form function and normalized scattering cross section for quartz sand 4 ,0 21f h f K x fxC K x (2-34a) ) ) 12 1222 0121exp1expxxxxC++,,& (2-34b) 4 4 3 24 4 31hKx x x Kx (2-35)

PAGE 44

25 where 10.25, 11.4 x 10.5+ and 20.37, 22.8 x 22.2+ f K =1.1, 0.18 K and x ka. The functions given by equations 2-34 and 2-35 are shown in Figure 2-6 together with corresponding curves for quartz spheres ( 2-28 and 2-32 ). The values of the velocities of shear and compressional waves in quartz spheres were taken to be tc =3760m/s and lc =5980m/s. 100 102 0 0.5 1 1.5 2 a). k & af 100 102 0 0.2 0.4 0.6 0.8 1 1.2 b). k & a Figure 2-6. Form function and normalized cross section for quartz sand (blue) and spheres (red): a) form function, b) normalized cross-section. Modified Model for Elastic Spheres Based on the spherical model, modified expressions were developed by Thorne and Buckingham [2004] to better describe the scattering properties of natural sediments. The following empirical function was introduced 3 0 30.53.5 3.5 xx x x (2-36) and applied to the low-pass filtered spherical form function and normalized cross section in form

PAGE 45

26 ,0,, s fs f xxf (2-37) 0, s sxx (2-38) The parameters describe the deviation of the modified model from the spherical model due to irregularities in shapes and sizes, which occur in natural sediments. While the parameters f and are generally independent, their values are close and usually change between 1.3 and 2.2. The value of =1.9 can be taken for both of the parameters to provide a reasonable fit with the data [ Thorne and Buckingham 2004]. It can be seen, that the function 0, x is greater than 1 for all values of x It implies that the values of modified form function and normalized cross section are greater than the original ones calculated by spherical model. 10 1 100 101 102 0.5 1 1.5 2 = 1.0 = 0.5 = 1.0 = 1.9 k & a Figure 2-7. Modified function for selected values of parameter Modified High-Pass Model for Natural Sediments Similarly, the high-pass model can be modified to better model the properties of particular sediments. Therefore, the original function 0, x ( 2-36 ) was currently

PAGE 46

27 modified as 0 3 0 3,,1 0.53.5 ,1 3.5 x x xx x ((2-39) The function x is equal to 1 for =0.5. The lower limit 0 implies x 0 and is equal to 30.55.251.753.85". The graphs of the function x for selected parameters equal to –1, 0.5, 1, and 1.9 are shown in Figure 2-7 The modified high-pass model becomes ,,,hfh f xxf (2-40) ,hhxx (2-41) Sound Attenuation Coefficient The sound attenuation coefficient defines the exponential rate of decay of sound intensity with distance. It depends on temperature, salinity, pressure, presence of suspended particles and microbubbles, and varies with sound frequency [ Richards 1998]. The total attenuation coefficient can be written as a sum of the attenuation coefficient in clear water, w, and the attenuation due to scattering by suspended sediments, s ws (2-42) The attenuation of sound in the megahertz range in seawater at small (<30m) depths is primarily a function of water temperature. The formula by Fisher and Simmons [1977] gives the approximation to within 4% 22243355.92.374.77103.481010wfTTT&&& (2-43) where f is the sound frequency in megahertz, T is the water temperature in degrees centigrade.

PAGE 47

28 The sediment attenuation coefficient at a range r is given by 2t sbn (2-44) where bn is the number of particles per unit volume, t is the total scattering crosssection of the particle. Using the equations 2-9 and 2-31 and averaging over the sediment sizes, the sediment attenuation coefficient becomes 2 33 4s sa rMrrMr a (2-45) where is the sediment attenuation constant evaluated at range r 2 33 4sa a (2-46) The attenuation of the sound traveling within interval 12, rr due to scattering and absorption by suspended sediments can be approximated by 2 1211r s rrMrdr rr (2-47) If the sediment attenuation constant is not changing with range, then the attenuation is equal to s M (2-48) where M is the mean sediment concentration between ranges 12, rr Lognormal and Normal Distributions of Sediment Sizes Sediment sizes vary from very fine (clay) to coarse (sand, gravel, shell) [e.g., Sleath 1984]. Lognormal distribution can be used to approximate the actual distribution

PAGE 48

29 of sediment sizes. Its probability density function is given by 2 2ln 1 ,exp 2 2a Laa a aaM paMS S aS (2-49) To obtain the probability density function in terms of sediment radius a in meters the following substitutions were made 1ln23ln10 ln2a aM S (2-50) where and are the mean and standard deviation in phi units. The phi scale is defined as the logarithm of sediment size in millimeters. The sediment size, given in meters, can be written as 3 2log210a& (2-51) If the probability density function of a is given by equation 2-49 then the quantity lna is normally distributed. In terms of the probability density function is given by 2 21 ,exp 2 2Np (2-52) The parameters and can be approximated from the sieve analysis as [ Sleath 1984] 12 284.115.9logdd and 12 84.1 2 15.9logd d (2-53) where nd is the diameter in millimeters for which n % of the sample by weight is finer. The raw moments of the lognormal distribution are 2 1 0,exp 2a LaaaS aapaMSdaM (2-54)

PAGE 49

30 )*222 2 0,exp2LaaaaaapaMSdaMS (2-55) 2 33 3 09 ,exp3 2a LaaaS aapaMSdaM (2-56) Averaged Form Factor and Normalized Cross-Section Terms The averaged terms in the equations 2-11 and 2-46 can be calculated with the assumption that sediment sizes obey some known distribution. Assuming the lognormal distribution ( 2-49 ), the following simplification can be made. The term 2 2() afka, where denotes averaging over the sediment sizes, is equal to 22 22 0()()afkaafkapada (2-57) where the sediment size is distributed with the density function p a If p a is the lognormal distribution, the expression above simplifies to 22 22 0()(),LaaafkaafkapaMSda (2-58a) ln,da aydy a (2-58b) 2 2 2 21 ()exp 2 2a yy a ayM efkedy S S (2-58c) )* 2 2 2 2 22 1 ()exp2exp 2 2aa y aa a ayMS f keMSdy S S (2-58d) 2 2ˆˆ (),y N f kepydy (2-58e)

PAGE 50

31 2 2 ˆˆ ,()Ny pyfke (2-58f) Here p x is the expected value of the argument with the probability density function p x ˆˆ ,Npy is the normal distribution probability density function 2 2ˆ 1 ˆˆ ,exp ˆ 2 ˆ 2Ny py (2-59) with 2ˆ 2aa M S and ˆaS. With backward substitution the term 2 2() afka becomes 2 2()ln, da afkaaydy a (2-60a) 2 2 2 2 0ˆ ln 1 ()exp ˆ 2 ˆ 2 a f kada a (2-60b) 2 2 0ˆˆ (),L f kapada (2-60c) 2 2 ˆˆ ,()Lpafka (2-60d) 2 2 ˆˆ ,()Lpaafka (2-60e) The modified term 2 2 2 2 ˆˆ 3 3() ()Lpafka afka a (2-61a) ) )* 2 2 2 ˆˆ ,exp22 () exp392Laa pa aaMS fka MS (2-61b)

PAGE 51

32 22 ˆˆ 2 1()L apa Sfka e (2-61c) 22 ˆˆ 2()L apa Sfka e a (2-61d) The expressions for the averaged normalized total scattering cross-section can be obtained following similar derivations 22 ˆˆ ,Lpaakaaka (2-62) 2 2 ˆˆ 3 3Lpaka aka a (2-63a) 2ˆˆ 2L apa Ska e a (2-63b) Circular Transducer Directivity and Half Intensity Beam Width The directivity of the transducer can be defined as the ratio of sound pressure at point ,r, where is latitude angle measured from central axis of the transducer, to the axial pressure at the same range r ,0Pr D P r (2-64) Let ta be the radius of transducer and 2k be the sound wave number, where is the sound wavelength. At great distances from the transducer (1rk), the far-field directivity of a circular piston source is [Medwin and Clay, 1997, p.139; Morse and Ingard, 1968, p.381] 12sin sint tJka D ka (2-65)

PAGE 52

33 where 1 J x is the first order Bessel function of first kind. The curve of the squared directivity function, 2D, is shown in Figure 2-8 6 4 2 0 2 4 6 0 0.25 0.5 0.75 1 Directivity, D 2katsin 3.8317 1.6163 Figure 2-8. Circular piston directivity function, 2D. The time averaged acoustic intensity, I relates to the RMS pressure as [Medwin and Clay, 1997] 2 rms A P I c (2-66) where A is the acoustic density that relates to the bulk modulus of elasticity, E as 2 A E c. The half intensity beam width, can be defined as 2 where is the angle from the center axis to where the acoustic intensity is equal to the half of the axial value. For a circular piston source, the directivity function of which is given by equation 2-65 20.5tD at sin1.6163tka. Although most of the sound energy is transmitted within the cone defined by the half intensity beam width, as shown in Figure 2-1 the low frequency transducers can have wider main lobe and significant side lobes.

PAGE 53

34 Similar to the half intensity beam width, the width of the main lobe, main, can be determined from 20tD, which is satisfied at sin3.8317tka. The beam width at a given range, r, from the transducer, for example at r=1 m, can be found as ()2sin2Wrr and ()2sin2mainmainWrr The values of the beam widths and W for frequencies 1–5MHz are shown in Table 2-2 It can be seen, that for higher sound frequencies the beam widths are narrower. Table 2-2. Beam width for a circular piston transducer (ta= 5 mm and c= 1500 m/s). Frequency 1 MHz 2 MHz 3 MHz 4 MHz 5 MHz tka 21.0 41.9 62.8 83.8 104.7 deg 8.9 4.4 2.9 2.2 1.8 main, deg 21.1 10.5 7.0 5.2 4.2 W, m 0.15 0.08 0.05 0.04 0.03 mainW, m 0.37 0.18 0.12 0.09 0.07 Near-Field Correction Factor Close to the transducer, the sound waves from different parts of the transducer surface interact with each other, both destructively and constructively. The structure of sound pressure in the near-field is different from the form given by 2-1 Regardless of the complexity, it still can be calculated accurately [Lockwood and Willette, 1973]. For the application to remote measurements of suspended sediment concentration and sediment size, the detailed information about the sound field close to the sensor is usually [Downing et al., 1995; Thorne et al., 1993a] substituted by the near-field correction factor, r, which is given by [Downing et al., 1995] 3.2 3.211.322.5 1.322.5 z z z zz (2-67) where nf z rr is the normalized distance, and nfr is the near-field range as given by

PAGE 54

35 2 t nfa r (2-68) The value of nfr depends on the transducer radius and the sound wavelength. For a fixed transducer size, the near-field range increases linearly with the transducer frequency. For illustration, the values of the near-field range for frequencies 1–5 MHz are shown in Table 2-3 Table 2-3. Near-field range nr for a circular piston source (ta=5mm and c=1500m/s). Frequency 1 MHz 2 MHz 3 MHz 4 MHz 5 MHz nfr, m 0.05 0.10 0.16 0.21 0.26 0 0.1 0.2 0.3 0.4 0.5 0.6 0 0.2 0.4 0.6 0.8 1 Range, m1 / (r) 1 MHz 2 MHz 3 MHz 4 MHz 5 MHz Figure 2-9. Near-field correction factor, 1r, for ta=5mm and c=1500m/s. The near-field correction factor is applied to the far-field pressure in form of 1r. The corresponding curves for 1–5MHz frequency transducers are shown in Figure 2-9 For higher frequencies the effect of the proximity to the transducer on transmitted sound is larger, and the near-field region is farther.

PAGE 55

36 CHAPTER 3 CALIBRATION OF THE ACOUSTIC BACKSCATTER SYSTEM Calibration of the acoustic backscatter system is necessary to obtain system function given by equation 2-11 [Thorne and Hanes, 2002]. For this purpose, the measurements of profiles of backscattered sound intensity can be taken in an environment with known characteristics. Different scatterers, such as thin wire targets, glass spheres or sand, are used in the calibration in various experimental setups [e.g., Hay, 1991; Jansen, 1979; Sheng and Hay, 1993]. Calibration in a chamber with a homogeneous suspension of sand or glass spheres is the most common technique [e.g., Ludwig and Hanes, 1990; Schaafsma, 1989; Tamura and Hanes, 1986; Thorne et al., 1993a; Thosteson and Hanes, 1998]. It has several advantages over other methods. The calibration with similar scatterers to those found in the field requires no changes in backscattering parameters such as form function or normalized cross-section. The experiments with the thin wire require modification of the acoustic backscatter equation due to cylindrical spreading of backscattered sound compared to spherical spreading from the sand grains and spheres. Also, the definition of sample volume for the wire is different from the one done for the suspension of particles. Although a modified form of equation 2-10 needs to be applied [Sheng and Hay, 1993], the method with the thin wire targets provides the easiest calibration which requires only a simple setup (similar to shown in Figure 4-2 in Chapter 4 ). In experiments with a turbulent jet [Hay, 1991], a mixture of suspended particles and water is used. The suspension within the cross-section of the jet is not homogeneous and therefore modifications to the acoustic backscatter equation (equation 2-10 ) are

PAGE 56

37 needed. The turbulence provides proper mixing and makes the orientation of the particles random relative to the transducer. Freely falling particles [Jansen, 1979] can have predefined orientation relative to the positions of transmitter and receiver, although the effect of this is not significant for particles with approximately spherical shapes [Sheng and Hay, 1988]. For calibration and other tests, two acoustic backscatter systems were used. Both systems were developed by Centre for Environment, Fisheries and Aquaculture Science (CEFAS), Lowestoft, UK. The first system, hereafter referred as “ABS1”, has three 10mm transducers with operational frequencies of 1.08, 2.07 and 4.70MHz. The data collection was carried out by a specially developed acquisition system [Thosteson, 1997]. The acquisition system operates at maximum rate of 88 profiles per second, and records a maximum of 4 root-mean-square (RMS) profiles a second by averaging over 22 consecutive raw voltage profiles. The second system, hereafter referred as “ABS2”, also has three 10mm transducers, with frequencies 1.0, 2.5 and 5.0MHz. Aquatec Electronics Ltd, UK, developed the data logger, which allows obtaining raw ABS profiles at a maximum of 80 profiles per second for each frequency. The resolutions of the two systems are different. For the ABS1, the voltage is measured in a 12-bit range, while the ABS2 has 16-bit resolution. Spatial resolution is defined by the pulse duration. It is 10s for the ABS1 and 13s for the ABS2, which results in a bin size of approximately 7.5 and 10mm, respectively. Both systems use time varying gain (TVG). The form of the TVG was set to account for the spherical spreading, and is nominally given by a linear function of the range r, although the exact form of the TVG may deviate from linear.

PAGE 57

38 Calibration Chamber The recirculating chamber at the University of Florida Coastal Laboratory was used to perform the calibration of the acoustic backscatter systems. This chamber was previously used in a number of projects [Lee, 1994; Ludwig and Hanes, 1990; Tamura and Hanes, 1986; Thosteson, 1997]. The chamber is schematically shown in Figure 3-1 It consists of a vertical 2m acrylic tube with inside diameter of 0.19m. A rotary pump moves a water-sediment mixture through a 4cm PVC pipe upward, where it discharges into the tank through the four nozzles. The nozzles are at the same level, about 10cm below the free surface. The jets are directed towards the center of the tank from 90-degree sectors (see Figure 3-3 ), where they collide and generate a turbulent flow, mixing the suspension. The mixture flows down the chamber with a mean cross-section velocity of approximately 4cm/s. The bottom of the tank has a funnel, which leads the flow through a pipe back into the pump. The funnel prevents the sediments from accumulating at the bottom. The suction ports were closed with metal plugs, which protruded into the interior of the chamber for about 5mm. Before taking measurements, the chamber was filled with tap water and left for about 20 hours to allow air bubbles to leave the system. For additional 2–4 hours the system was left with the pump turned on. To establish the predefined level of suspended sediment concentration, portions of sand were added into the water. Every portion was initially dried, weighed and placed for several hours in container with a small amount of water and detergent. The detergent was used to detach air bubbles from the surface of the sand grains. Excess water with detergent was removed from the container before the sand was added into the chamber.

PAGE 58

39 40cm. 30cm. 30cm. 30cm. Transducer(s) Nozzles Levels 0, 15, and 30 cm Suction ports Pump Funnel 20cm. 200cm. Flow Flow Figure 3-1. Calibration chamber—side view.

PAGE 59

40 ABS Tests in the Chamber The chamber has a different environment for measurements than field experiments. Because of the laboratory setup, various artificial conditions can affect the measurements, which may lead to erroneous conclusions about estimated parameters. A series of tests were performed in the University of Florida Coastal Laboratory to investigate the effects of the chamber on calibration measurements. For every test, profiles of root-mean-square voltage were obtained. Each profile was calculated as the average of 120 individual RMS profiles recorded at 4Hz for 30 seconds. To obtain each individual RMS profile, the ABS was set to average 22 raw profiles. Tests were made with fewer averaged profiles to check if the setting has an effect on the measurements. The higher sampling rate theoretically can lead to the interaction between several consecutive pulses, when previous sound pulses were not fully attenuated. The tests did not reveal this possibility. Additionally, to the average profiles, the 10th and 90th percentile profiles were calculated to illustrate the scatter among the measured RMS profiles. Tests in Air and Still Water Two tests were performed to examine the lower bound for voltage profiles. For this purpose the measurements were taken with the transducers left in the air and placed in the chamber with still clear water. The resulting curves are shown in Figure 3-2 The curves, which correspond to the measurements in the air, take on the lowest values except for the 1.08MHz transducer, which has a peak near approximately 4cm range. None of the profiles start with zero values—on the contrary they start from the value around 50–100 in relative units and decay within a few bins. Within this range there is no difference between profiles measured in the air and in the water. Most likely this is an effect of resid-

PAGE 60

41 ual vibration of transducer surface after the sound pulse was emitted. The measurements in the water show that the scatter of the RMS profiles (in figures plotted with dashed lines) can be significant especially for high frequency transducers. The peaks around 9cm for 1.08 and 2.07MHz are caused by the reflection of emitted sound from the chamber walls, which will be illustrated further. 0 0.2 0.4 0.6 0.8 0 50 100 150 200 1.08MHz a).Bit air water/conc. 0 g/l water/conc. 0 g/l (center line) 0 0.2 0.4 0.6 0.8 0 50 100 150 200 2.07MHz b).BitRange,m 0 0.2 0.4 0.6 0.8 0 50 100 150 4.70MHz c). Range,mBit Figure 3-2. RMS voltage profiles (solid lines) with scatter intervals (dashed line) from the measurements in air and still clear water: a) 1.08MHz, b) 2.07MHz, and c) 4.70MHz. Homogeneity of Sediment Suspension and Effect of Chamber Walls The jet system and high flow rate serve to generate a homogeneous suspension. To investigate the homogeneity of the mixture in vertical and horizontal directions, samples were taken from the four openings in the side of the chamber wall (see Figure 3-1 ). The results of this investigation was previously published by Ludwig and Hanes [1990]. From their tests, it was concluded that the suspension in the chamber is homogeneous within the measurement error. Additional measurements were currently conducted to test effects of the proximity of a transducer to the jet system and sidewalls. The ABS1 system with 3 frequencies

PAGE 61

42 (1.08, 2.07, and 4.70MHz) was used for the tests. One transducer at a time was placed at different vertical levels relative to the nozzles (see Figure 3-1 ) and at different distances from the sidewall (see Figure 3-3 ). Mixture M i x t u r e Mixture M i x t u r e 18mm. Suction ports Nozzles Nozzles Transducer positions R 70mm. Figure 3-3. Calibration chamber—top view. The jet system is shown with transducer positions for the sidewall proximity tests. For the wall proximity tests, measurements were taken at the center of the chamber corresponding to a distance of 95mm, and also at 30 and 50mm distances from the sidewall. In the vertical, the transducer was placed at 3 levels: 0, 15 and 30cm. The 0cm level corresponds to the upper edge of the nozzles, therefore the jets of sand-water mixture were colliding just under the transducer surface. The profiles of RMS voltage were recorded for several different levels of suspended sand concentration (see Tables 3-1 3-2 and 3-3 ). Sand collected from Jacksonville Beach, FL with sieve fraction of size 0.212–0.250mm was used for the test. Table 3-1. Locations and concentration levels in tests for 1.08MHz transducer. Distance from wall (row) and level (column) 0cm 15cm 30cm 30mm 0, 0.2, 0.5g/l 0, 0.2, 0.5g/l 0g/l 50mm 0, 0.2, 0.5g/l 0, 0.2, 0.5g/l 0g/l 95mm (center) 0, 0.2, 0.5, 0.8g/l0, 0.2, 0.5, 0.8g/l0, 0.2, 0.5, 0.8g/l

PAGE 62

43 Table 3-2. Locations and concentration levels in tests for 2.07MHz transducer. Distance from wall (row) and level (column) 0cm 15cm 30cm 30mm 0.2, 0.5g/l 0.2, 0.5g/l – 50mm 0.2, 0.5g/l 0.2, 0.5g/l – 95mm (center) 0.2, 0.5, 0.8g/l0.2, 0.5, 0.8g/l0.2, 0.5, 0.8g/l Table 3-3. Locations and concentration levels in tests for 4.70MHz transducer. Distance from wall (row) and level (column) 0cm 15cm 30cm 95mm (center) 0.5, 0.8g/l0.5, 0.8g/l0.5, 0.8g/l 0 100 200 300 400 level 0 cm / conc. 0.0 g/l a).Bit 95 mm 50 mm 30 mm 0 50 100 150 200 1.08 MHz level 15 cm / conc. 0.0 g/l b). 95 mm 50 mm 30 mm 0 50 100 150 level 30 cm / conc. 0.0 g/l c). 95 mm 50 mm 30 mm 0 100 200 300 400 level 0 cm / conc. 0.2 g/l d).Bit 95 mm 50 mm 30 mm 0 100 200 300 400 level 15 cm / conc. 0.2 g/l e). 95 mm 50 mm 30 mm 0 100 200 300 400 level 30 cm / conc. 0.2 g/l f). 95 mm 0 200 400 600 level 0 cm / conc. 0.5 g/l g).Bit 95 mm 50 mm 30 mm 0 200 400 600 level 15 cm / conc. 0.5 g/l h). 95 mm 50 mm 30 mm 0 200 400 600 level 30 cm / conc. 0.5 g/l i). 95 mm 0 0.2 0.4 0.6 0.8 0 200 400 600 800 level 0 cm / conc. 0.8 g/l j). Range, mBit 95 mm 0 0.2 0.4 0.6 0.8 0 200 400 600 level 15 cm / conc. 0.8 g/l k). Range, m 95 mm 0 0.2 0.4 0.6 0.8 0 200 400 600 level 30 cm / conc. 0.8 g/l l). Range, m 95 mm Figure 3-4. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for 1.08MHz transducer for sidewall proximity tests.

PAGE 63

44 Figures 3-4 and 3-5 show the results of measurements with the 1.08 and 2.07MHz transducers for different concentrations of sand. For the 1.08MHz transducer the profiles have several peaks, which correspond to the positions of the suction port plugs. As shown earlier (Table 2-2 ), the 1.08MHz transducer has the widest directional beam. Therefore, a part of the sound can reflect from the plugs and be detected by the receiver. The profiles for 2.07MHz transducer do not have similar peaks. 0 200 400 600 800 level 0 cm / conc. 0.2 g/l a).Bit 95 mm 50 mm 30 mm 0 200 400 600 800 2.25 MHz level 15 cm / conc. 0.2 g/l b). 95 mm 50 mm 30 mm 0 200 400 600 800 level 30 cm / conc. 0.2 g/l c). 95 mm 0 500 1000 level 0 cm / conc. 0.5 g/l d).Bit 95 mm 50 mm 30 mm 0 500 1000 level 15 cm / conc. 0.5 g/l e). 95 mm 50 mm 30 mm 0 500 1000 level 30 cm / conc. 0.5 g/l f). 95 mm 0 0.2 0.4 0.6 0.8 0 500 1000 1500 level 0 cm / conc. 0.8 g/l g). Range, mBit 95 mm 0 0.2 0.4 0.6 0.8 0 500 1000 1500 level 15 cm / conc. 0.8 g/l h). Range, m 95 mm 0 0.2 0.4 0.6 0.8 0 500 1000 1500 level 30 cm / conc. 0.8 g/l i). Range, m 95 mm Figure 3-5. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for 2.07MHz transducer for sidewall proximity tests. The measurements along the chamber centerline at different vertical levels are shown in Figures 3-6 3-7 and 3-8 The curves for the 1.08MHz have peaks at approximately 9 and 19cm. Comparing to the profiles at other distances from the wall (see Figure 3-4 ), these peaks appear only when the transducer is placed in the center of the chamber. The ranges at which the peaks appear correspond to multiple chamber radiuses. They can

PAGE 64

45 result from multiple reflection of the sound wave from side lobes between the chamber wall and transducer surface. This also demonstrates that the side lobes of the 1.08MHz transducer are quite significant. This can affect the measurements in the field as well, because other transducers may not detect the scatterers located within the side lobes of the 1.08MHz transducer. 0 100 200 300 400 conc. 0.0 g/l a).Bit 0 cm 15 cm 30 cm 0 100 200 300 400 conc. 0.2 g/l b).Bit 0 cm 15 cm 30 cm 0 0.2 0.4 0.6 0.8 0 200 400 600 conc. 0.5 g/l c). Range, mBit 0 cm 15 cm 30 cm 0 0.2 0.4 0.6 0.8 0 200 400 600 800 conc. 0.8 g/l d). Range, mBit 0 cm 15 cm 30 cm Figure 3-6. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured along centerline with 1.08MHz transducer 0 200 400 600 800 conc. 0.2 g/l a).Bit 0 cm 15 cm 30 cm 0 0.2 0.4 0.6 0.8 0 500 1000 conc. 0.5 g/l b). Range, mBit 0 cm 15 cm 30 cm 0 0.2 0.4 0.6 0.8 0 500 1000 1500 conc. 0.8 g/l c). Range, mBit 0 cm 15 cm 30 cm Figure 3-7. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured along centerline with 2.07MHz transducer.

PAGE 65

46 The effect of proximity to the nozzles is noticeable only for the 2.07 and 4.70MHz transducers. Figures 3-7 and 3-8 show that the profiles at level 0cm are lower than at levels 15 and 30cm. The drop in the intensity can be due to either Doppler shift of the frequency of the backscattered sound from faster moving sand grains, which cannot be resolved by the receiver with fixed frequency, or due to inhomogeneity of suspended sediment concentration in the upper part of the chamber. Although separate measurements of the concentration in the jet part of the chamber were not performed, the later seems to have more effect, since the maximum speed from the jets is not very high (approximately 3m/s) and resulted Doppler shift would be about 0.2% of the transducer frequency. However, the measurements show that if the transducer is placed lower than the jets, the difference between profiles from different layers is very small. Hence, the vertical homogeneity of the concentration in the chamber is confirmed and by the current tests. 0 0.2 0.4 0.6 0.8 0 500 1000 1500 conc. 0.5 g/l a). Range, mBit 0 cm 0 cm 15 cm 15 cm 30 cm 30 cm 0 0.2 0.4 0.6 0.8 0 500 1000 1500 conc. 0.8 g/l b). Range, mBit 0 cm 15 cm 30 cm Figure 3-8. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured along center line with 4.70MHz transducer Figures 3-9 and 3-10 show the difference between the measurements at different levels along the chamber, but now the transducers are placed at 30 and 50mm from the chamber wall. Similar to the tests when the transducers were located at the center of the chamber, the tests for 2.07MHz transducer show a slight decrease in voltage at 0cm level.

PAGE 66

47 However, the profiles for 1.08MHz do not have peaks at ranges equal to multiples of the chamber radius. 0 50 100 150 at 30 mm / conc. 0.0 g/l a).Bit 0 cm 15 cm 30 cm 0 50 100 150 at 50 mm / conc. 0.0 g/l b).Bit 0 cm 15 cm 30 cm 0 100 200 300 at 30 mm / conc. 0.2 g/l c).Bit 0 cm 15 cm 0 100 200 300 at 50 mm / conc. 0.2 g/l d).Bit 0 cm 15 cm 0 0.2 0.4 0.6 0.8 0 200 400 600 at 30 mm / conc. 0.5 g/l e). Range, mBit 0 cm 15 cm 0 0.2 0.4 0.6 0.8 0 200 400 600 at 50 mm / conc. 0.5 g/l f). Range, mBit 0 cm 15 cm Figure 3-9. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured near the chamber wall with 1.08MHz transducer. 0 200 400 600 800 at 30 mm / conc. 0.2 g/l a).Bit 0 cm 15 cm 0 200 400 600 800 at 50 mm / conc. 0.2 g/l b).Bit 0 cm 15 cm 0 0.2 0.4 0.6 0.8 0 500 1000 at 30 mm / conc. 0.5 g/l c). Range, mBit 0 cm 15 cm 0 0.2 0.4 0.6 0.8 0 500 1000 at 50 mm / conc. 0.5 g/l d). Range, mBit 0 cm 15 cm Figure 3-10. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) measured near the chamber wall with 2.07MHz transducer.

PAGE 67

48 Results The tests show that the jet system generates homogeneous suspension in the region approximately 15cm below the nozzles. The chamber wall effects are not significant for 2.07 and 4.70MHz transducers, but for the measurements with the 1.08MHz transducer, it should not be placed in the center of the chamber to avoid multiple reflections from the side lobes of the beam. To obtain the smooth profile, the peaks corresponding to the positions of the suction port plugs can be filtered out by interpolating between the nearest data points. A few bins at the beginning of the profile do not contain useful information on the suspended sediment concentration, since it is contaminated by an electronic noise or residual vibration of the transducer membrane. Estimation of System Function The system functions of the acoustic backscatter, v K can be found from the calibration by using the expression 12 2 1216 0.963r t vrms ska r K Ve KMc (3-1) where s K is calculated by equation 2-11 M is the level of suspended sand concentration in the chamber, rmsV is the measured root-mean-square profile of voltage, is the nearfield correction factor ( 2-67 ), k and c are the sound wavenumber and the sound speed in water, ta is the transducer radius, is the pulse duration, and is the total attenuation coefficient ( 2-42 ) Since the transducer frequency is preset and changes in the speed of sound in water are relatively small, the system parameter t K given by equation 2-11a is usually obtained from the calibration instead of v K [ Thorne and Hanes 2002]. However, the effec-

PAGE 68

49 tive transducer radius, which determines the directivity of the sound beam and extent of the near-field region, is usually 5–20% less than the geometrical radius. By using the parameterv K the best-fit value for ta may be found from the calibration. The system function v K is constant if the time varying gain (TVG) is not applied, otherwise, v K is a function of range r. The system function describes the properties of the system and, therefore, does not vary with sediment size or concentration. However, the functions estimated from the measurements taken for different concentrations of sediments of different size distributions will most likely be not the same due to variations in sediment shapes, mineralogy, etc., which are not accounted for in the acoustic equation 2-10 The modified form function and normalized cross-section function with parameters f and given by equations 2-40 and 2-41 may be used to account for these differences, as well as to better approximate the system function. Let i v K be the ith estimation of the inherited system function, v K where each i v K is calculated by equation 3-1 at the concentration level i measM from the measured voltage profile i rmsV for a given sediment size distribution. The average function can be found as the mean of estimates as 11N i vv i K K N (3-2) where N is the number of estimations, to provide an estimation of the system function for a particular sediment size distribution. Approximation by 2nd Order Polynomial The system function calculated by equation 3-2 is not necessarily a smooth function. Therefore, v K was fitted with a 2nd order polynomial over the range of bins

PAGE 69

50 nfnjn -(fn is the near-field bin index and n is the total number of bins) to produce an estimation of system function v K % as 2 210 v K prprp % (3-3) where i p i=0, 1, or 2, are the coefficients of the best-fit polynomial. Calibration of the ABS1 The ABS1 was calibrated with quartz sand, collected at SISTEX’99 experiment [ Vincent et al. 2001], with mean size =2.17 and standard deviation =0.27 estimated from the sieve analysis. From the fall velocity measurements, the mean sediment size was approximated to be1.96 in phi units. Both measurements and corresponding normal distribution curves are shown in Figure 3-11 The transducer radius was taken to be 90% of the geometrical radius. The modified form function and normalized cross section were calculated by equations 2-34 and 2-35 with initial value of 0.5 for parameters and f 1 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 Sieve: = 2.17 = 0.27 !Fraction coarsera). 1 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 Fall velocity: = 1.96 = 0.27 !Fraction coarserb). data fit Figure 3-11. Estimation of sediment size distribution used for the calibration of ABS1 by: a) sieve analysis, b) fall velocity.

PAGE 70

51 The resulted system functions are shown in Figures 3-12 – 3-14 Individual estimations, i v K are shown together with the curves for v K % (marked as “fit”) on plots “b” in Figures 3-12 – 3-14 The errors on plot “b” were calculated as the average of relative rootmean-square deviations from the 2nd order fit and are given by the expression 2 111 1v nfi nN vjvj K jni nf vjKrKr nnN Kr% % (3-4) The near-field bin index nfn for current calculations was taken to be equal to 15, but can be adjusted for any particular system. The value was chosen from the sidewall proximity tests which showed that the measurements at bins <15 are not accurate. Calculated by 2-10 voltage profiles were compared to the measured profiles. Their ratios are plotted on plots “a.” Using 2-12 the inverted profiles of the suspended sediment concentration are shown on plots “c.” Relative errors for each of the concentration levels between inverted and measured concentrations are also shown with the mean error, which was calculated as the average of individual relative errors: 111 1nfii Nn calcjmeasj M i ijn nf measj M rMr Nnn Mr (3-5) where the subscript “meas” denotes measured or expected concentration in the chamber and subscript “calc” denotes estimated concentration profiles calculated with 2-12

PAGE 71

52 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 2.5 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 10 20 30 40 System coefficients b). Err = 11.7254% Kv = 33.8551 & r2 11.4308 & r+10.9485KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 5 Mean error = 20.46% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 41.0% 0.20 27.1% 0.30 13.2% 0.40 9.7% 0.50 3.1% 0.70 4.2% 1.00 15.7% 1.50 28.7% 2.00 25.6% 3.00 36.2% Conc,g/l Err,% Figure 3-12. Estimation of the system function for 1.08MHz transducer: a) ratio of measured and calculated voltage profiles, b) estimated system functions, and c) inverted sediment concentration. 0.2 0.4 0.6 0.8 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 20 40 60 80 100 System coefficients b). Err = 3.5695% Kv = 17.4659 & r2+45.4530 & r+21.1931KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 Mean error = 5.58% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 11.7% 0.20 6.2% 0.30 2.7% 0.40 4.2% 0.50 5.1% 0.70 5.0% 1.00 4.2% Conc,g/l Err,% Figure 3-13. Estimation of the system function for 2.07MHz transducer: a–c) same as in Figure 3-12

PAGE 72

53 0.2 0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 50 100 150 200 250 System coefficients b). Err = 4.8066% Kv = 209.7920 & r2+377.1953 & r+25.0449KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 Mean error = 18.90% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 18.2% 0.20 12.7% 0.30 3.1% 0.40 26.2% 0.50 34.3% Conc,g/l Err,% Figure 3-14. Estimation of the system function for 4.70MHz transducer: a–c) same as in Figure 3-12 The estimated system functions show that the expected linear dependence due to the TVG is present only for the 2.07MHz transducer, although the 2nd order coefficient is still significant. The curves of voltage ratios on plot “a” for the 4.70MHz transducer (see Figure 3-14 ) are not uniform with range compared to the curves for the other two transducers. This may indicate erroneously estimated attenuation at each level of concentration. The errors corresponding to the different concentration levels are not uniform, especially for 1.08 and 4.70MHz. Since the estimated system functions were applied, the errors grow for smaller and larger concentrations. Calibration of the ABS2 The calibration of the ABS2 was performed similarly to the ABS1. The sand from Jacksonville Beach, FL was used for the calibration. The estimations of sediment size

PAGE 73

54 distribution parameters by sieve analysis and based on fall velocity are shown in Figure 3-15 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 Sieve: = 2.41 = 0.13 !Fraction coarsera). 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 Fall velocity: = 2.28 = 0.23 !Fraction coarserb). data fit Figure 3-15. Estimation of sediment size distribution for sand used in calibration of ABS2 by: a) sieve analysis, b) fall velocity. 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 2 2.5 3 Ratio of voltages Range, mVmeas / Vcalca). f = 1.00MHz at = 4.50 mm = 2.41 = 0.13 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 1 0 200 400 600 800 1000 System coefficients b). Err = 4.2356% Kv = 246.7460 & r2+341.1658 & r+8.7842KvRange, m fit 0.2 0.4 0.6 0.8 1 0 0.5 1 1.5 2 2.5 Mean error = 7.60% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.02 15.0% 0.04 6.7% 0.07 5.2% 0.20 3.9% 0.50 4.3% 1.00 5.0% 2.00 13.1% Conc,g/l Err,% Figure 3-16. Estimation of the system function for 1MHz of ABS2: a–c) same as in Figure 3-12

PAGE 74

55 The resulting system functions are shown in Figure 3-16 Unlike the 1.08MHz transducer of the ABS1, the curves for the system functions are very close to each other and the correction of parameter was not needed. The 1MHz transducers of the both systems are very similar, and one would expect to obtain similar results in calibration of both systems. However, the current tests show that the results are very different. The calibration of the 1.08MHz transducer of the ABS1 does not result in a unique system function. The application of the average of estimated system functions can lead to large errors in inverted concentrations. For these cases, a semi-empirical model was developed to provide the solution for suspended sediment concentration. This model is described in Appendix B Sensitivity Analysis For Estimation of System Function A particular estimate of system function v K depends on several parameters, which can be initially approximated with some error or uncertainty. In order to investigate the sensitivity of the system function v K to changes in the parameters, such as form function, normalized cross-section, and sediment distribution, the following analysis was carried out. Using equation 3-1 the system function v K can be rewritten as a function of parameter transducer radius ta and sediment size distribution parameters and as ,,,vtKFa (3-6) The form function and normalized cross section were estimated by equations 2-40 and 2-41 The dependence of v K on listed parameters is not explicit, but through the form function, normalized cross-section, beam directivity, and near-field correction factor. Generally for calibration, these parameters need to be known and fixed. For current

PAGE 75

56 analysis, a subset of the listed parameters set to be free variables (i.e. by letting them vary within some intervals). This produces the new system functions, vfit K estimated by minimizing the error calculated by equation 3-4 The results of the calculations, based on the measurements for the ABS1, are gathered in tables in Figures 3-17 – 3-19 For each test, some of four parameters ta , and were fixed (marked with “dots”). Others were variable, for them the best-fit values are presented. Bounds for free parameters are shown in the last column of the table. For fixed parameters, their initial values were used. The errors of inverted concentrations together with the mean error (black line) were calculated and plotted for each test. The values of the mean errors are shown in the table in “Mean error, %” row. Values in row “Fit errors, %” were calculated by 3-4 The bounds of free parameters were chosen to cover a realistic, but still wide, range of values. In many cases, the minima were reached at boundaries. If unconstrained minimization would be performed, the minimum could be reached at unrealistic values, such as 0mm for transducer radius. It can be seen from the figures that not all the parameters have the same effect on calibration. The parameter and mean sediment size have the most significant effects, while transducer radius ta and sediment size sorting have almost no effect. Since both the mean sediment size and parameter are present in the expression for normalized cross-section, minimization on one of them can be compensated by an equivalent change in another. Because of this relationship, use of incorrect values for may lead to erroneous approximation of sediment size from the measurements.

PAGE 76

57 0 5 10 15 20 25 30 35 40 45 Error, % 0.10 0.20 0.30 0.40 0.50 0.70 1.00 1.50 2.00 3.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 11.7 20.5 0.50 4.50 2.17 0.27 1 10.9 18.5 & & & 0.05 2 10.2 16.6 & & 2.71 & 3 9.89 16 & & 2.71 0.05 4 11.7 20.5 & 5.00 & & 5 10.9 18.5 & 5.00 & 0.05 6 10.2 16.6 & 5.00 2.71 & 7 9.89 16 & 5.00 2.71 0.05 8 10.4 17.2 2.50 & & & 9 10.2 16.7 2.50 & & 0.05 10 9.86 15.9 2.50 & 2.71 & 11 9.74 15.6 2.50 & 2.71 0.05 12 10.4 17.2 2.50 5.00 & & 13 10.2 16.7 2.50 5.00 & 0.05 14 9.86 15.9 2.50 5.00 2.71 & 15 9.74 15.6 2.50 5.00 2.71 0.05 Bounds 2.50 2.50 3.00 5.00 1.63 2.71 0.05 1.00 Figure 3-17. Values of fitted parameters ta , and with corresponding errors in inverted concentrations for the 1.08MHz transducer of the ABS1. 2 4 6 8 10 12 Error, % 0.10 0.20 0.30 0.40 0.50 0.70 1.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 3.57 5.58 0.50 4.50 2.17 0.27 1 2.84 5.02 & & & 0.05 2 2.65 5.2 & & 2.47 & 3 2.65 5.2 & & 2.40 0.21 4 3.57 5.63 & 3.00 & & 5 2.84 5.06 & 3.00 & 0.05 6 2.65 5.24 & 3.00 2.47 & 7 2.65 5.24 & 3.00 2.40 0.21 8 2.65 5.2 0.77 & & & 9 2.65 5.2 0.18 & & 0.05 10 2.65 5.2 0.04 & 2.36 & 11 2.65 5.2 0.01 & 2.21 0.05 12 2.65 5.24 0.77 3.00 & & 13 2.65 5.24 0.19 3.00 & 0.07 14 2.65 5.24 0.04 3.00 2.36 & 15 2.65 5.24 0.01 3.00 2.22 0.09 Bounds 2.50 2.50 3.00 5.00 1.63 2.71 0.05 1.00 Figure 3-18. Same as on Figure 3-17 for the 2.07MHz transducer of the ABS1.

PAGE 77

58 0 5 10 15 20 25 30 35 Error, % 0.10 0.20 0.30 0.40 0.50 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 4.81 18.9 0.50 4.50 2.17 0.27 1 4.56 18.5 & & & 0.05 2 4.8 18.9 & & 2.13 & 3 4.4 18.3 & & 2.01 0.05 4 4.8 19 & 3.00 & & 5 4.55 18.7 & 3.00 & 0.05 6 4.8 19 & 3.00 2.13 & 7 4.39 18.4 & 3.00 2.01 0.05 8 2.45 14 1.20 & & & 9 2.45 14 1.18 & & 0.19 10 2.45 14 1.20 & 2.22 & 11 2.45 14 1.18 & 2.11 0.19 12 2.44 14.2 1.20 3.00 & & 13 2.44 14.2 1.18 3.00 & 0.19 14 2.44 14.2 1.20 3.00 2.22 & 15 2.44 14.2 1.18 3.00 2.12 0.20 Bounds 2.50 2.50 3.00 5.00 1.63 2.71 0.05 1.00 Figure 3-19. Same as on Figure 3-17 for the 4.70MHz transducer of the ABS1. The biggest decrease in mean error ( 3-5 )—from 18.9 to 14.0%—is shown in Figure 3-19 for the 4.70MHz transducer by varying parameter After the corrections to sediment attenuation, not only the mean error was reduced, but also the errors for most of concentration levels. The sediment attenuation for transducer 1.08MHz was insignificant; neither nor have any influence on errors. For the 2.07MHz transducer parameter was estimated to be –0.77, which results in decreasing of the originally estimated (by equation 2-35 ) normalized cross-section and therefore sediment attenuation. The corresponding value of for the 4.7MHz is 1.20, which implies that the sediment attenuation is larger than estimated by equation 2-35 The system functions for the 4.70MHz transducer with corrected are shown in Figure 3-20 The errors in inverted concentrations at low concentration levels (0.1 and 0.2g/l) were decreased more than 2 times, and also became more uniform with range. The

PAGE 78

59 spreading among vfit K curves was significantly reduced resulting in a mean relative error of 2.4%. 0.2 0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 1.20 0.2 0.4 0.6 0.8 0 50 100 150 200 250 System coefficients b). Err = 2.4457% Kv = 198.7014 & r2+405.7229 & r+22.0241KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 Mean error = 13.96% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 6.0% 0.20 5.9% 0.30 3.9% 0.40 25.1% 0.50 28.8% Conc,g/l Err,% Figure 3-20. Estimation of system function v K for 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12 Dependence on Sediment Sorting The dependence of v K on sorting is rather complicated, because of filtering the data (either theoretical or experimental) used in obtaining the form function and normalized cross-section. From equations 2-61 and 2-63 v K can be simplified to ) )*)*2 2 2 33 2 2 22 212 12 12 2 12 2 2 34exp exp expaa aafka aka v aa fka SS aa S aKCrCr CreCre CrSCre & & (3-7) where ln2aS, and functions iCr do not depend on but on other parameters.

PAGE 79

60 By approximating exponents to 2nd order, the following expression can be obtained ) 2 3441ln412v K CrCrCr" (3-8) where v K now depends on to second order. Therefore, small changes in will have small effect on minimized relative error. This is supported by Figures 3-17 – 3-19 where there is no consistency in calculated values for ; the errors sometime reach their minima at either limits of the constraint interval for Based on current analysis, it can be concluded that if the sediments are well sorted, then it becomes nearly impossible to successfully approximate sorting from measurements using this minimization technique. Dependence on Transducer Radius The dependence of v K on transducer radius ta can be explicitly found and derivative becomes )* )* 2 22 1.010.01t tv r t z a tt r z aK CraCrz aa Crz / / % // (3-9) where Cr is the coefficient not depending on ta For far-field ( z ) the relation between v K and ta is almost linear with a very small second order term, and therefore does not significantly affect the relative error ( 3-4 ). However, if the near-field region extends significantly, as in the case of 4.70MHz transducer, changes in transducer radius can be noticeable. Determination of Parameter f For a homogeneous suspension of sediments, coefficient s K (equation 2-11 a) is constant with range. Therefore, any uncertainties in determination of s K affect only the

PAGE 80

61 magnitude of the estimated system function v K Since the system function for a given transducer is unique, any deviation of the estimated system functions can be compensated by using the modified form functions ( 2-37 or 2-40 ) with the appropriate parameter f Because the coefficient s K appears as a multiplier in equation 3-1 the parameter f cannot be found from measurements with a homogeneous suspension of one sediment sample. Thus measurements with different sediment samples are required. From equations 2-11 and 2-40 it can be found that there exists a constant 22222 ,, 222 ,,hh f hafaf akaf (3-10) which for small standard deviation gives f ka (3-11) Similar to the estimation of the system functions, v K % which were obtained from the measurements with particular sediments (see equations 3-2 and 3-3 ), let the system function v K be written in the form of a 2nd order polynomial over the range nfnjn -2 210 v K prprp (3-12) with each coefficient k p k=0,1 or 2, found as 11m kkj j p p m (3-13) where kj p are the corresponding coefficients of the estimation for one sediment size system function v K % and m is the number of estimations.

PAGE 81

62 Therefore, can also be written as 2 210 2 210 v v K prprp K prprp% (3-14) which needs to be satisfied for all values or r. This may not be possible for all times, therefore the estimation of using mean quantities can be used and the following expression obtained )*02 221100 0 01 0rpprpprppdr r (3-15) with the solution for as 2 21 000 2 22 00032 32pp rrp pp rrp (3-16) where 0r is some reference range. After obtaining the solution for parameter f can be found by inverting equation 3-11 Estimation of System Function for ABS1 To obtain the system function for the ABS1, the tests were carried out with four different sediment samples. The sediments were predominantly quartz sand. The parameters of the size distributions were estimated from sieve analysis and are shown in Table 3-4 The sample from Case 3 is the same sample used for the calibration of the ABS1 and described in the previous section Its size distribution is shown in Figure 3-11 The sediments for Case 4 were the mix of two sieve fractions—0.125–0.149 and 0.297–0.250mm. The estimations of distribution parameters are shown in Figure 3-21 The system functions were calculated for all four series of tests. The results of the calculations are shown in Appendix A (see Figures A-13 – A-36 ). The estimated parame-

PAGE 82

63 ters and f the mean errors (calculated by equations 3-4 and 3-5 ), and parameter are gathered into Table 3-5 The parameters f were approximated in both cases—for initial values of =0.5 and for best-fit values, estimated from the minimization of error 3-4 It was previously noted that for a wide sediment size distribution, the value of sorting can have a noticeable effect on estimated quantities. For example, Figure A-2 shows that if =0.44, then the corresponding parameter is equal to –1.98. But if the sorting would be estimated as =0.06, which is the best-fit value when is taken as a free parameter, the parameter is equal to –0.19. In Table 3-5 the values of parameter correspond to the estimated from measurements values of (from the column for Test#8 in Figures A-1 – A-12 in Appendix A ). Table 3-4. Sediment sizes. Case 1 Case 2 Case 3Case 4 2.67 2.74 2.17 2.16 0.44 0.25 0.27 0.54 1 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 Sieve: = 2.16 = 0.54 !Fraction coarsera). 1 1.5 2 2.5 3 0 0.2 0.4 0.6 0.8 1 Fall velocity: = 1.99 = 0.60 !Fraction coarserb). data fit Figure 3-21. Estimation of sediment size distribution for Case 4 used for calibration of ABS1 by: a) sieve analysis, b) fall velocity.

PAGE 83

64 The results in Table 3-5 show that for the fitted values of all mean errors decrease up to 50%. Case 1 for the 4.70MHz transducer is the only one, where the mean error in inverted concentrations increases slightly. Although the changes in parameters and f are also small. Table 3-5. Estimation of parameters and f for different sediment sizes. Case 1 Case 2 Case 3 Case 4 Initial Fitted InitialFittedInitialFittedInitial Fitted 1.08MHz 0.5 –2.5 0.5 –2.5 0.5 –2.5 0.5 –2.5 vK % 6.9 5.3 3.6 3.3 11.7 10.4 17.1 14.1 M % 12.3 9.5 6.8 6.0 20.5 17.3 37.8 26.3 1.04 1.04 1.51 1.53 0.80 0.80 0.65 0.63 f 0.90 0.91 2.5 2.5 –0.94 –0.93 –1.99 –2.16 2.07MHz 0.5 –1.98 0.5 0.37 0.5 –0.77 0.5 –2.17 vK % 7.3 3.6 2.6 2.6 3.6 2.6 12.2 8.7 M % 11.4 6.8 5.9 5.9 5.6 5.2 inf 17.11 1.08 1.06 1.15 1.21 1.01 1.04 0.76 0.69 f 0.92 0.85 1.85 2.5 0.56 0.66 –0.59 –0.88 4.70MHz 0.5 0.33 0.5 1.47 0.5 1.20 0.5 –2.5 vK % 3.2 3.1 5.8 2.1 4.8 2.4 6.2 4.7 M % 16.8 17 21.9 11.8 18.9 14.0 42.5 18.8 1.16 1.13 1.01 1.09 1.03 1.08 0.80 0.71 f 1.10 1.04 0.55 0.91 0.72 1.00 –0.82 –1.43 Table 3-6. Values of x ka for test cases. Frequency, MHz Case 1 Case 2Case 3Case 4 1.08 0.36 0.34 0.51 0.51 2.07 0.69 0.66 0.98 0.98 4.70 1.57 1.49 2.22 2.23 It was previously noticed, that even large changes in parameter for the 1.08MHz transducer have small effects on estimated errors—the fitted values for

PAGE 84

65 drop to –2.5 (the lower bound), the value chosen to be the lower limit in the calculations. For the 1.08MHz frequency, the values of x ka (see Table 3-6 ) are less than 0.51 for the Cases 1–4. This gives the values for the normalized cross-section of 0.01 or less (see equation 2-35 and Figure 2-6 ), which results in very small total attenuation ( 2-42 ) at low concentrations. Since, the parameter is found from an expression like ,hka ka % (3-17) where normalized cross-sections— % is estimated from data and h is given by equation 2-35 —are small numbers, approximated from their ratio values of can have large errors. Therefore, the initial value for =0.5 can be used. 0 10 20 30 40 50 Kva). 1.08MHz Case 1: = 2.67, = 0.44 Case 2: = 2.74, = 0.25 Case 3: = 2.17, = 0.27 Case 4: = 2.16, = 0.54 Mean 0.2 0.4 0.6 0.8 0 20 40 60 80 Kvb). 2.07MHz Range, m 0.2 0.4 0.6 0.8 0 50 100 150 200 250 Kvc). 4.70MHz Range, m Figure 3-22. Estimated system functions and their means for Cases 1–4 with =0.5 (dashed) and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz transducers of the ABS1.

PAGE 85

66 The estimated system functions for all four cases are shown in Figure 3-22 together with their means calculated by equation 3-12 The dashed lines correspond to the functions estimated by assuming the initial value =0.5. The solid lines correspond to the system functions estimated with fitted values for (see Table 3-5 ). The parameter was estimated by equation 3-16 with 0r=0.6m. Corresponding values of f were found by inverting equation 3-11 Both parameters are shown in Table 3-5 0.6 0.8 1 1.2 1.4 1.6 Kv/Kv,meana). 1.08MHz Case 1: = 2.67, = 0.44 Case 2: = 2.74, = 0.25 Case 3: = 2.17, = 0.27 Case 4: = 2.16, = 0.54 0.2 0.4 0.6 0.8 0.6 0.8 1 1.2 1.4 Kv/Kv,meanb). 2.07MHz Range, m 0.2 0.4 0.6 0.8 0.6 0.8 1 1.2 1.4 Kv/Kv,meanc). 4.70MHz Range, m Figure 3-23. Ratios between individual system functions for Case 1–4 and mean system function with =0.5 (dashed) and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz transducers of the ABS1. Although the system function should be unique for a given system, the estimated curves have some scatter, which results from the different sediments used in each case. For a homogeneous suspension, the curves differ approximately by a constant (equation 3-14 ). Therefore, the ratios of individual estimations to the mean were calculated with the results shown in Figure 3-23 For the 1.08 and 2.07MHz transducers, the calculated ratios

PAGE 86

67 are close to constant for r>0.1m. In most cases for these two transducers, the ratios are more uniform for the system functions estimated with fitted values of compared to ones estimated by using =0.5. This means that the corrections to the normalized crosssection (equation 2-41 ) lead to a better approximation of sediment attenuation. Also it should be noted that for the 4.70MHz transducer, the ratios are less uniform with range than for other two frequencies. Analysis of Relationship between and f As can be seen from the results in Table 3-5 by choosing appropriate values of and f corrections to the form function and normalized cross-section can be made (equations 2-40 and 2-41 ), which improve the accuracy of suspended sediment concentrations. Since the parameters and f were found individually for each frequency and each sediment sample, the following questions may arise: 01Can a single value be found, which could be used for correction of both the form function and the normalized cross-section for every individual or for all used frequencies? 01Can a single value of f be found which would represent a sediment sample and be unique for an applicable range of sound frequencies? 01Can a single value of be found which would represent a sediment sample and be unique for an applicable range of sound frequencies? In Figure 3-24 the estimated parameters f are plotted versus corresponding fitted values of for frequencies 2.07 and 4.70MHz. It can be seen that in most cases f ', as it obtained by Thorne and Buckingham [2004]. Because of large scatter and an insufficient number of available data points, the functional relationship between f

PAGE 87

68 and can only be approximated with a large degree of uncertainty. Nevertheless, it can be concluded that f and are most likely different even for a single frequency. 3 2 1 0 1 2 3 3 2 1 0 1 2 3 = ff 2.07MHz 4.70MHz Case 1 Case 2 Case 3 Case 4 Figure 3-24. Fitted parameters and corresponding parameters f for Cases 1–4. Values of f and for Cases 1–4 are plotted in Figure 3-25 It can be seen, that even with the large scatter, there is an overall trend in the parameters. Case 1 Case 2 Case 3 Case 4 3 2 1 0 1 2 3 f, 1.08MHz 2.07MHz 4.70MHz f Figure 3-25. Values of parameters f and for Cases 1–4 for different frequencies.

PAGE 88

69 Figure 3-26 shows calculated values for together with ,ka Dashed lines correspond values of x ka (see Table 3-6 ). Note that values of ,ka for the 1.08MHz transducer were not plotted since all fitted = –2.5. Solid lines correspond to the function ,ka (equation 2-39 ) with estimated as ,2.07,4.701 2ff (3-18) where ,2.07 f and ,4.70 f are the values of f for the 2.07 and 4.70MHz transducers. The lines fit data points for relatively well, especially in Cases 1 and 4. Although in Case 3 the point corresponding to the 1.08MHz frequency is much lower than points for the other two frequencies and also lower than the approximated curve. Overall, equation 3-18 gives a good approximation for parameter f for a given sediment sample. 100 101 0.4 0.6 0.8 1 1.2 1.4 1.6 k & a (k & a, ) = 0.9 = 1.1 = 0.8 = 0.9 Case 1 Case 2 Case 3 Case 4 (k & a, ) (k & a, f) Figure 3-26. Values of and ,ka as functions of ka with approximated functions ,ka for Cases 1–4 (solid lines).

PAGE 89

70 A similar approximation for will not be possible. The values of ,ka are close to each other only for Case 4, while other three cases look random. “Fun-Pump” Tests The homogeneous sediment suspension, generated in the chamber (Figure 3-1 ) for calibration, will most likely not be found during field or lab experiments, when the measurements are taken above a sandy bottom, and the suspension is generated by waves and currents. The following tests were conducted in order to simulate the “real” environment, which have variable concentration and size distribution profiles of sediment suspension. A simple technique is used to generate such nonuniform profiles in the chamber. With the sediments added to the chamber, the pump system generates a homogeneous suspension throughout the water column. If at any time the pump is turned off, the sediments will settle with some terminal velocity. Since there are no other sources of sediment near the top of the chamber, a nonzero gradient in sediment concentration will be generated. Also, if there are several sediment sizes present, then the sediments with different sizes will settle at different velocities. This will also generate a nonuniform profile of sediment size distributions. The measurements were taken with the ABS1 system. Three transducers were placed at the top of the chamber at about 10cm below the free water surface pointed downwards (see Figure 3-1 ). The ABS was operating at 88 profiles per second. Every 22 profiles were averaged to obtain a single root-mean-square (RMS) profile of voltage. The resulting RMS voltage profiles were recorded by the data logger at 4Hz. The duration of every run was three minutes, including 30 seconds of data with homogeneous suspension with the pump turned on. Then the pump was turned off, and the measurements were re-

PAGE 90

71 corded for another two minutes, while the sediments were settling down. After that, the pump was turned back on and data were collected for another 30 seconds. Sample Sediment Size Distribution A mixture of two sediment samples was used for the tests. These are the same sediments as were used for the calibration (see Case 4 ). The first sample is the sieve fraction of size 0.125–0.149mm (2.47–2.75 phi units) of the sand from Jacksonville, FL. It will be referred as Sand 1. The other sample is the sieve fraction of size 0.297–0.250mm (1.25– 1.75 phi units) will be referred as Sand 2. Assuming a uniform distribution of sediment sizes within each fraction, one can approximate the mean and standard deviation of a mixture of two sediment samples. Let the mixture contains two fractions 1 and 2 of Sand 1 and Sand 2. Therefore, 121 (3-19) The size distributions are given by 1 ,, 0,,ii ii i iiab ba p ab 2 3 (3-20) where i corresponds to Sand 1 or Sand 2, and ,iiab are the size ranges. The mean and variance of sediment sizes in mixture becomes 1122 222222 111222 (3-21) where i and i are the mean and standard deviation of Sand 1 and Sand 2, determined as 2ii iab 212ii iab (3-22)

PAGE 91

72 Different concentration levels were acquired by adding proper amounts of Sand 1 or Sand 2 into the chamber. The fractions of each sample are shown in Table 3-7 together with the resulting size distributions. Based on the fall velocity analysis, a 1:1 mixture of two samples had the mean of 1.99 and standard deviation of 0.60 (see Figure 3-21 ). Note that the values of parameters of the distribution for the 1:1 mixture in Figure 3-21 from the sieve analysis were different from Table 3-7 It is because for the calibration in Case 4 average values of the parameters were used. Table 3-7. Concentration levels, mean and standard deviation of the mixture for “funpump” tests. Concentration, g/l 1 2 0.4 1.00 0.002.610.08 0.8 0.50 0.502.060.57 1.2 0.67 0.332.240.53 1.6 0.50 0.502.060.57 2.0 0.60 0.402.170.55 2.4 0.50 0.502.060.57 Sand 1: = 2.61 = 0.08 Sand 2: = 1.5 = 0.14PDF! 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 2 2.5 3 3.5 Sand 1 Sand 2 Mix 1:1 Figure 3-27. Sediment size distribution for Sand 1, Sand 2, and 1:1 mixture. The fall velocity analysis provided estimations of size distribution functions (PDF) for both sand samples and the 1:1 mixture. The results are shown in Figure 3-27 Note that the size distribution for Sand 1 is wider than that approximated from the sieve analysis.

PAGE 92

73 Prediction of Concentration and Sediment Size Distribution The profiles of sediment concentration and size distribution parameters at any given time after the pump was turned off can be predicted with a 1D advection-diffusion model (see Appendix C for a more detailed description of the model). The sediment diffusivity parameter (equation C-1 ) can be taken to be zero, because of the absence of turbulence production when the pump is off. When the pump has just been turned off, the turbulence is not zero. Therefore, a small constant value can be used for the sediment diffusivity to account for the residual turbulence. The sediment diffusivity was taken to be 0.0005m2/s. Other values within 50-200% interval were also tested; this did not result in significant difference in predicted profiles. 0 0.02 0.04 0.06 0.08 FractionSize distribution at Z = 0.30 m a). 0 0.02 0.04 0.06 0.08 FractionSize distribution at Z = 0.60 m b). 1 1.5 2 2.5 3 3.5 0 0.02 0.04 0.06 0.08 Sediment size, !FractionSize distribution at Z = 0.75 m c). 0 0.2 0.4 0.6 0.8 Sediment concentrations at given times, g/l d). 0 10 20 30 40 50 60 70 Time, s 2.2 2.4 2.6 2.8 e). Mean of sediment size distribution, 0 0.2 0.4 0.6 0.8 1 0.2 0.3 0.4 0.5 f). STD of sediment size distribution, !Distance from the top, m Figure 3-28. Predicted with 1D advection-diffusion model profiles of concentration and sediment size distributions for maximum concentration of 0.8g/l at times 0– 70s: size distributions at a) 30, b) 50, and c) 75cm below the free surface, d) sediment concentration profiles, e) profiles of mean sediment size in phi units, and f) profiles of standard deviations in phi units.

PAGE 93

74 The suspended sediment concentration profiles and profiles of size distribution parameters predicted with the 1D advection-diffusion model for the initial concentration in the chamber of 0.8g/l are shown in Figure 3-28 All the profiles are shown from 0 to 70 seconds at 10-second intervals. The initial size distribution was approximated from the fall velocity analysis (see Figure 3-27 ). The parameters of size distribution were estimated by fitting the normal distribution curve (equation 2-52 ) to the calculated by the model distribution curves. The model results show that most of the concentration profiles have a step, which originates from the difference in terminal settling velocity of two sediment samples; the coarse sediment, Sand 2, settles faster than the finer sediment, Sand 1. A similar irregularity can be seen in the profiles for the approximated mean sediment sizes and standard deviations. The mean sediment size at every range increases (in phi units) with time, while the sorting decreases. The difference in the sorting along the profile can be quite significant. It can be seen especially at times 10 and 20 seconds after the pump was turned off, when the STD changes from approximately 0.2 to 0.5 along a single profile. This can affect the accuracy of the inversion of the measurements taken by the ABS into concentration and sediment sizes, as already was previously discussed Parameters of Inversion For the inversion of RMS voltage profiles, the model defined by equations 2-10 and 2-11 was used. The modified high-pass model for backscattered intensity, defined by equation 2-40 for the form function, and by equation 2-41 for the normalized crosssection were applied with the parameters f and determined from the calibration (see parameters for Case 4 in Table 3-5 ). The system function v K was approximated by equation 3-12 based on all available calibration profiles (see section “ Determination of

PAGE 94

75 Parameter f ”). The resulting coefficients of the second order polynomial are given in Table 3-8 Table 3-8. Coefficients of 2nd order polynomial for system functions. Frequency, MHz 2 p 1 p 0 p 1.08 24.65 –1.16211.62 2.07 –4.118 55.37 18.56 4.70 –199.9 359.6 27.18 The form function and normalized cross-section are shown in Figure 3-29 Both the initial unmodified high-pass model curves for h f and h (equations 2-34 and 2-35 ) and the modified versions h f and h (equations 2-40 and 2-41 ) are plotted. With the provided parameters f and the modified functions have lower values than the initial ones. It should be pointed out that according to the plots the 1.08MHz frequency is the least sensitive to the sediment size represented by Sand 1. For this sand, the sediment attenuation, s (equation 2-47 ), is also close to zero. Therefore in the test with 0.4g/l initial concentration when only Sand 1 was used (see Table 3-7 ), the inverted concentration for the 1.08MHz transducer was much lower (with an average approximately equal to 0.07g/l) than for other two transducers (see Figure 3-30 ). The mean concentration profiles were also estimated to be lower. The profiles of mean sediment sizes were found by minimization of the error between estimated concentration profiles for each frequency at a given range (equation 2-25 ). The standard deviation of sediment sizes needs to be known in advance. Theoretically, the sediment sorting can be specified at each time for every profile. But since field measurements usually lack of such information, and the sorting is estimated from the

PAGE 95

76 sieve or fall velocity analyses. Therefore, a single value equal to 0.54 was used in all calculations. !f1.08 1.08 2.07 2.07 4.7 4.7a). 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 !4.7 4.7 2.07 2.07 1.08 1.08b). 1 1.5 2 2.5 3 3.5 0 0.5 1 1.5 f ,h h f ,h h Sand 1: = 2.61 = 0.08 Sand 2: = 1.50 = 0.14 Figure 3-29. High-pass model for backscattered intensity and its modified version used for “fun-pump” test: a) form functions, b) normalized cross-sections. The empirical model described in Appendix B was also used to invert the ABS measurements into concentration profiles. The required parameters of the model were estimated by using the same calibration data as for system functions and parameters f and Results of Inversion for Concentration The results of inverting the concentrations from the “fun-pump” test with a maximum concentration of 0.4g/l are shown in Figure 3-30 The profiles of suspended sediment concentration estimated from the 1.08MHz frequency transducer are shown together with the mean profiles. The mean profiles were calculated as an average of the estimations from the three frequencies. It was already noted in the previous section, that the resulting concentrations from 1.08MHz data are almost four times lower than expected. There may be several causes of this discrepancy. Some of them are:

PAGE 96

77 01Erroneous initial values for parameters of sediment size distribution, used for calibration, which can result in biased values for parameter f and system functions; 01Lower sensitivity of the 1.08MHz transducer to the used sediment sizes. The mean sediment size specified for the calibration was taken as an average of the sediment sizes for all used concentration levels. It was equal to 2.16 (see Table 3-4 ). But since a mixture of two sediment samples was used for the calibration, none of sample mean sizes was equal to it. As can be seen from Figure 3-29 the 1.08MHz transducer is less sensitive to the finer sample than to the coarse sediments. Since two sample sizes were not equally resolved, the use of their mean value can result in overestimation of the system function and in obtaining a biased value for f By exclusion of the 1.08MHz frequency, one can improve the accuracy in estimation of the mean concentration profiles. But the decision should be made based on ability of the transducer’s particular frequency to resolve sediment sizes present in the sample. 0 0.2 0.4 0.6 0.8 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 Range, mConcentration, g/l 1.081.081.08 1.081.081.08 1 .081.081.081.08 1 08 1.081.08 108 1.081.08 1081. 081.08 108 1.08 108108108108 Ini: Mmax = 0.4 g/l, = 2.61, = 0.50 pump off off 10 sec off 20 sec off 30 sec off 40 sec off 50 sec off 60 sec off 70 sec Figure 3-30. Inversion by unmodified high-pass model concentration profiles for “funpump” test with maximum concentration of 0.4g/l: mean profiles (thick lines) and 1.08MHz profiles (thin lines).

PAGE 97

78 Figures 3-31 and 3-32 show inverted concentration profiles for the “fun-pump” test with a maximum concentration of 0.8g/l. The profiles in Figure 3-31 were calculated with the initial unmodified high-pass model, which corresponds to the modified model with values of parameters f and equal to 0.5 for all frequencies. The profiles in Figure 3-32 were calculated with the modified model with f and estimated from calibration (see Table 3-5 ). 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 1 1.2 Range, mConcentration, g/l2.072.072.072.072.072.072.07 2 2.07 207 2.072.072.07 207 2.072.07 207 2.07 207 2.07 207207207 207 Ini: Mmax = 0.8 g/l, = 2.06, = 0.57 pump off off 10 sec off 20 sec off 30 sec off 40 sec off 50 sec off 60 sec off 70 sec Figure 3-31. Inversion by unmodified high-pass model concentration profiles for “funpump” test with maximum concentration of 0.8g/l: mean profiles (thick lines) and 2.07MHz profiles (thin lines). At the initial time, when the sediment suspension in the chamber was homogeneous (profiles labeled “pump off”) both models give similar results, since the system functions were calibrated by using these profiles, and the correction for ’s is not yet necessary. The high jumps in the concentration at the ranges near the transducer show that the inversion is not accurate there for either model. After the pump was turned off and the sediments started to settle, the difference in the inverted profiles between two models become more evident. The unmodified model overestimates the concentration over the range

PAGE 98

79 r>0.5m at time 10 seconds after the pump was turned off. The profiles also have “steps” which correspond to erroneously determined sediment sizes (see discussion in the following section ). 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 1 1.2 Range, mConcentration, g/l2.072.072.072.072.07 2.07 207 2.07 207 2.07 207 2.07 207 207 207207 Ini: Mmax = 0.8 g/l, = 2.06, = 0.57 pump off off 10 sec off 20 sec off 30 sec off 40 sec off 50 sec off 60 sec off 70 sec Figure 3-32. Inversion by modified model concentration profiles for “fun-pump” test with maximum concentration of 0.8g/l: mean profiles (thick lines) and 2.07MHz profiles (thin lines). The concentration profiles, both inverted using the modified model and calculated by the 1D advection-diffusion model, are shown in Figure 3-33 The mean profiles calculated with empirical model are shown as green lines. The mean concentration profiles were calculated by averaging of the profiles inverted from all three frequencies. The relative errors were calculated as 1calcjmodelj j modelj M rMr N Mr (3-23) where modelM is the concentration estimated by the 1D advection-diffusion model, calcM is the inverted concentration, with the summation carried over the indices at which modelj M r>0.05g/l, and N is the number of included indices.

PAGE 99

80 0 1 2 3 4 a). Errors: mean 10.9%; ABn 8.3% 0 secConc., g/l 1.08 MHz 2.07 MHz 4.70 MHz Mean Model Mean (ABn) 0 0.5 1 1.5 b). Errors: mean 8.4%; ABn 31.5% 10 secConc., g/l 0 0.5 1 c). Errors: mean 45.2%; ABn 17.9% 20 secConc., g/l 0 0.2 0.4 0.6 0.8 d). Errors: mean 59.9%; ABn 15.9% 30 secConc., g/l 0 0.1 0.2 0.3 0.4 e). Errors: mean 60.6%; ABn 15.0% 40 secConc., g/l 0 0.1 0.2 0.3 0.4 f). Errors: mean 56.6%; ABn 40.2% 50 secConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 g). Errors: mean 60.7%; ABn 49.5% 60 sec Range, mConc., g/l 0 0.2 0.4 0.6 0.8 0 0.05 0.1 0.15 0.2 h). Errors: mean 72.8%; ABn 39.0% 70 sec Range, mConc., g/l Figure 3-33. Comparison of concentration profiles inverted by modified high-pass model with profiles calculated by the 1D advection-diffusion and empirical models (labeled as “ABn”) for the “fun-pump” test with maximum concentration of 0.8g/l at 10 second intervals. The sequence of plots in Figure 3-33 shows the dynamics of the two sediment samples settling. The profiles corresponding to individual frequencies (blue lines) show increasing stratification of sediment suspension over time due to the different settling velocity. Again, when the content of coarse sediments gets low, the concentration estimated from the 1.08MHz frequency is much lower than obtained by the other frequencies. The 4.70MHz becomes the most accurate predictor of the concentration when only fine sedi-

PAGE 100

81 ments are present. The concentrations estimated from the 2.07MHz measurements are closest to the estimated mean concentration. The empirical model provides the best estimations for suspended sediment concentration. Although it does not use the information on sediment size, it still can be applied for the inversion. Results of Inversion for Sediment Size The sediment size distribution parameters were also estimated. The results for the test with the maximum concentration of 0.8g/l are shown in Figures 3-34 and 3-35 for unmodified and modified models; they correspond to the results presented in Figures 3-31 and 3-32 It can be seen from Figure 3-34 that the unmodified model provides estimations of mean sediment size which are closest to the expected values at very low concentrations (at time>30 seconds), but fails at higher concentrations, for example, profiles at 10 and 20 seconds. The model predicts sediment sizes which were not present in the chamber (see Figure 3-27 ), aliasing the estimations towards finer sediment sizes. The only approximation which is reasonable is at time t =0 seconds; and that profile was used for the calibration of the system, therefore the inversion is much more accurate. The estimations with the modified model are shown in Figures 3-35 and 3-36 Several definite improvements can be noticed. The model now captures correctly the dynamics of the changes in mean sediment sizes along the chamber with time. It was predicted by 1D advection-diffusion model (see Figure 3-28 e) that the mean sediment sizes decrease with time at any horizontal level after the pump was turned off. Figure 3-35 shows that the inverted profiles behave similarly to those predicted by the numerical (1D) model. Also, there is a limit for estimated sediment size at approximately 2.4 which is close to the mean size of Sand 1 sample, used in the test.

PAGE 101

82 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 1 1.5 2 2.5 3 3.5 4 Ini: Mmax = 0.8 g/l, = 2.06, = 0.57 !Range, m pump off off 10 sec off 20 sec off 30 sec off 40 sec off 50 sec off 60 sec off 70 sec Figure 3-34. Inverted profiles of mean sediment sizes for “fun-pump” test with maximum concentration of 0.8g/l by unmodified high-pass model. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.5 1 1.5 2 2.5 3 3.5 Ini: Mmax = 0.8 g/l, = 2.06, = 0.57 !Range, m pump off off 10 sec off 20 sec off 30 sec off 40 sec off 50 sec off 60 sec off 70 sec Figure 3-35. Inverted by modified high-pass model profiles of mean sediment sizes for “fun-pump” test with maximum concentration of 0.8g/l. In Figure 3-36 the comparison of estimated sediment sizes with the 1D advectiondiffusion model results are shown for each 10-second interval. Although the mean absolute errors are within the initially approximated standard deviation for the sediment samples (see Table 3-7 ), there is a steady shift of 0.4 This shift is most likely to be due to the underestimation of parameter f for the 1.08MHz frequency. Figure 3-26 shows that

PAGE 102

83 the value of f estimated for 1.08MHz results in lower values of x than for the other two frequencies. Also the 1.08MHz transducer has the largest uncertainty in determination of the system function (see Figure 3-12 ), which would result in a biased f estimation. a). Error 0.8 0 sec! 1 1.5 2 2.5 3 3.5 Invert Model Sand 1 Sand 2 b). Error 0.5 10 sec! 1 1.5 2 2.5 3 3.5 c). Error 0.4 20 sec! 1 1.5 2 2.5 3 3.5 d). Error 0.4 30 sec! 1 1.5 2 2.5 3 3.5 e). Error 0.4 40 sec! 1 1.5 2 2.5 3 3.5 f). Error 0.4 50 sec! 1 1.5 2 2.5 3 3.5 g). Error 0.4 60 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 h). Error 0.4 70 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 Figure 3-36. Comparison of inverted profiles of mean sediment sizes by modified highpass model with profiles calculated by the 1D advection-diffusion model for the “fun-pump” test with maximum concentration of 0.8g/l at 10-second intervals. Similar results were obtained for other tests with different concentration levels by using the same values of f and (see Appendix D ). Overall, the predictions of sedi-

PAGE 103

84 ment sizes and profiles of suspended sediment concentration are quite accurate. This is achieved by the applying the following improved techniques: 01 More accurate estimation of system functions by using 2nd order polynomials, which help to account for nonlinear effects resulting from using the time varying gain factor (TVG). 01 Modification of the high-pass model for backscattered intensity by using the correction function x with parameters f and estimated for a particular sediment sample. The developed empirical model also shows very good results in obtaining of profiles of sediment concentrations. Among the other applied methods (modified and unmodified high-pass models), it gives the best accuracy in inversion.

PAGE 104

85 CHAPTER 4 MEASUREMENTS OF TARGET POSITION AND BOTTOM ECHO REMOVAL ALHORITHM A Simplified Bottom Echo Model The acoustic backscatter system allows measurement of suspended sediment concentration based on the intensity of backscattered sound from the scatterers in the water column. At low concentrations the sound pulse emitted by the transmitter reaches the bottom and reflects back where it is recorded by the receiver. The recorded root-meansquare (RMS) voltage profiles contain information from the scatterers in the water and the bottom. The later can be used, for example, as a reference for correction of the sediment attenuation [ Thorne et al. 1995]. Because the acoustic properties of the sandy bottom are different from the properties of the suspended sediments [e.g., Medwin and Clay 1997, p.610], the acoustic backscatter equation with the parameters described in Chapter 2 cannot be applied to the measurements contaminated by the bottom echo; in vicinity of the bottom the signal becomes saturated since the intensity of the sound reflected from the bottom is higher than the intensity of the sound backscattered from the suspended sediment particles. Here an algorithm is developed which allows to separate these signals and provides a technique for accurate measurement of bottom location. After the signal is registered by the ABS, it passes through a lowpass filter, which removes high frequency noise. The resulting signal is a convolution of the registered signal with the response function of the filter

PAGE 105

86 0 t rmsenvVtVtthtdt %%% (4-1) where rmsV is the recorded RMS voltage, envV is the amplitude envelope of the amplified signal registered by the receiver, and h is the filter impulse response function. The registered signal is assumed to be a linear combination of the bottom echo, bV and the signal backscattered from suspended sediments, s V envbsVtVtVt (4-2) which with the substitution trc results in rmsbsVrVrVr (4-3) where “overbar” means filtered signals. To check the validity of the linear assumption a series of test were performed. The results of the tests are discussed in the following section The second term is given by equation 2-10 The RMS voltage from the bottom echo can be approximated as [ Medwin and Clay 1997, p.594] 2 0R g bbVtRVte (4-4) where b R is the bottom reflection coefficient, R g is the acoustical roughness parameter, and 0V is the recorded voltage from the perfect mirror-reflected surface. The acoustical roughness is a function of an incident angle and the physical roughness of the surface. By adopting a simplified expression for the bottom echo voltage 2brr b bbVrve (4-5) where bv is the magnitude of the bottom echo return, br is the bottom location relative to

PAGE 106

87 the transducer, and b is the spreading parameter, equation 4-3 can be used to extract the bottom echo return from the measured signal. A set of the parameters bv br and b required to the bottom echo model can be approximated from a series of measured RMS profiles. The magnitude of the bottom echo, bv depends on the total attenuation of the signal, which is mainly a function of sediment concentration in the water ( 2-42 ). When the measurements are obtained in the field, the suspended sediment concentration changes with time as waves pass over the measurement site [ Hanes 1991]. Therefore over a period of time a set of profiles can be found which have the same mean concentration level. For these profiles a single value for bv can be determined. From a voltage profile, the relative attenuation can be approximated by calculating the integral over the range from 0 to the approximate bottom location. Figure 4-1 shows an example of the measured RMS profile with the integral equal to the shaded area. 0 50 100 150 0 200 400 600 800 1000 1200 1400 1600 1800 Point "A" Level = 500 BinBit Figure 4-1. A sample of voltage profile with definition of “A”-point. The bottom location is usually assumed to correspond to the location of the maximum intensity, however, the position of the maximum can be shifted and its magnitude

PAGE 107

88 can be increased if other scatterers are located close to the bottom (from equation 4-3 ). Another approximation can be used—after the maximum is reached, the signal decays and the location where the profile reaches some level of intensity can be found. This location is labeled as “A” in Figure 4-1 An arbitrary level can be chosen to locate the position of the point “A”. The level 500 is shown in the figure and has been used for the 2.07MHz transducer (400 for the 1.08MHz and 100 for the 4.70Mhz). A set of the profiles with approximately equal attenuation is used to determine the profile for the bottom echo. Since the location of the point “A” changes with time, the relative shift, br needs to be calculated for each profile, so the “A” points for all profiles coincide. The lower envelope found as the lowest signal levels can now be used as an estimation of the bottom return. Assuming a constant total attenuation, higher levels of signal should relate to the presence of suspended sediments, since all profiles should have a similar bottom return. After the bottom echo profile is determined, the parameters in equation 4-5 can be found as the best-fit values. Finally, a correction for br needs to be done with the previously calculated shift, br The resulting RMS profiles for suspended sediments can be found from the initially measured RMS profiles minus the bottom return, as srmsbVrVrVr (4-6) Tests with a Steel Wire A series of the tests was conducted at the University of Florida Coastal Laboratory with the acoustic backscatter system. The purpose of the tests was to test the ability of the ABS to distinguish presence of a target such as a steel wire above the sand surface and to measure its elevation. The tests were carried out in a tank filled with tap water (see Fig-

PAGE 108

89 ure 4-2 ). A pan with quartz sand with the mean size of approximately 0.5mm was placed on the bottom of tank and the surface was made flat. The target was a steel wire with the diameter of 0.33mm, which was strained on a frame attached to the ruler. The ruler could move up or down with 1mm precision. ABS elevation Wire elevation ABS (3 frequencies) Ruler Wire Sand Figure 4-2. Setup for the tests with steel wire. A series of 31 tests were performed at different elevations of the wire above the sand surface (see Table 4-1 ). The measurements were taken with the ABS1 system (see Chapter 3 ). The RMS voltage profiles were collected at 2Hz for 15 seconds at every position of the wire. The resulting 30 profiles were averaged to obtain mean profiles, which are shown in Figures 4-3 and 4-4 The location of the sand surface corresponds to 452mm distance and on plots is shown with a solid line. The wire position is shown as a dashed line. The data from three transducers—1.08, 2.07, and 4.70MHz—are presented. As it can be seen from the figures, the position of the wire and the sand surface location coincide with corresponding maxima. A correction can be made to the estimation of the range vector, r, which originally was calculated by equation 2-13 to ensure that

PAGE 109

90 the measured locations correspond to the expected. A bin size is given as 2 c, where is the sound pulse duration and c is the speed of sound in the water. For the given system, =10–5s which with c =1480m/s results in the bin size of 7.4mm. Therefore one can expect that the wire will not be visible at an elevation of the wire lower than 7mm above the sand surface. This is supported by the data in Figures 4-3 and 4-4 Tests 11–23 show that the signal from the target is not yet separated from the signal from the sand bottom as it is in tests 23–31, but the recorded profile is not symmetrical and has a slightly higher intensity at the peak. This is support the assumptions discussed in the previous section The profiles in Figures 4-5 and 4-6 are calculated with equation 4-6 for which the bottom echo profile was taken from the measurements with no wire present (test 1). As it can be seen from the measurements with the 4.70MHz transducer, the ABS can detect the target at 4mm elevation above the sand surface, which is lower than 7.4mm bin size. Although the strength is much smaller than in cases of higher elevations, the used technique for separation of the signals reveal the difference between the cases with no target and with a target present close to the reflecting surface. Thee tests provide the basis for the bottom echo removal algorithm. In the next tests with a setup as shown in Figure 4-2 a wire with the 2mm diameter was used as a target. It was positioned at 5cm intervals from the transducer surface. The resulting RMS profiles are shown in Figure 4-7 The wire is located at 10, 15, and 20cm distance from the transducer. Figure shows the presence of second “ghost-wire” reflection, especially for the 2.07 and 4.70MHz transducers. The location of the second reflection exactly correspond to the twice the distance between the wire and the transducer. This means that the sound should reflect three times—two from the wire and ones from the transducer in order to generate this type of record. As it was previously discussed in

PAGE 110

91 Chapter 2 the multiple reflections are possible in presence of stronger scatterers. The current tests show the possibility of such event. Table 4-1. Tests with 0.33mm steel wire. Position number Time, min Ruler, mm Elevation above bottom, mm Distance from the ABS, mm 1 0.5 87.0 0.0 452.1 2 1.0 87.5 0.5 451.6 3 1.5 88.0 1.0 451.1 4 2.0 88.5 1.5 450.6 5 2.5 89.0 2.0 449.1 6 3.0 89.5 2.5 449.6 7 3.5 90.0 3.0 449.1 8 4.0 90.5 3.5 448.6 9 4.5 91.0 4.0 448.1 10 5.0 91.5 4.5 447.6 11 5.0 92.0 5.0 447.1 12 6.0 93.0 6.0 446.1 13 6.5 94.0 7.0 445.1 14 7.0 95.0 8.0 444.1 15 7.5 96.0 9.0 443.1 16 8.0 97.0 10.0 442.1 17 8.5 98.0 11.0 441.1 18 9.0 99.0 12.0 440.1 19 9.5 100.0 13.0 439.1 20 10.0 101.0 14.0 438.1 21 10.5 102.0 15.0 437.1 22 11.0 103.0 16.0 436.1 23 11.5 104.0 17.0 435.1 24 12.0 114.0 27.0 425.1 25 12.5 124.0 37.0 415.1 26 13.0 134.0 47.0 405.1 27 13.5 144.0 57.0 395.1 28 14.0 154.0 67.0 385.1 29 14.5 177.0 90.0 362.1 30 15.0 207.0 120.0 332.1 31 17.0 257.0 170.0 282.1

PAGE 111

92 0 1000 2000 Bit#1 @0mm 1.08MHz 0 2000 4000 2.07MHz 0 1000 2000 4.70MHz 0 1000 2000 Bit#2 @1mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#3 @1mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#4 @2mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#5 @2mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#6 @3mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#7 @3mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#8 @4mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#9 @4mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#10 @5mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#11 @5mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#12 @6mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#13 @7mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#14 @8mm 0 2000 4000 0 1000 2000 0 2000 4000 Bit#15 @9mm 0 2000 4000 0 2000 4000 300 350 400 450 0 1000 2000 Bit#16 @10mm Range, mm 300 350 400 450 0 2000 4000 Range, mm 300 350 400 450 0 1000 2000 Range, mm Figure 4-3. Root-mean-square voltage profiles for the tests with 0.33mm steel wire. Tests 1–16 with the wire location (dashed line) and the bottom location (solid line).

PAGE 112

93 0 2000 4000 Bit#17 @11mm 1.08MHz 0 2000 4000 2.07MHz 0 2000 4000 4.70MHz 0 2000 4000 Bit#18 @12mm 0 2000 4000 0 2000 4000 0 2000 4000 Bit#19 @13mm 0 2000 4000 0 2000 4000 0 2000 4000 Bit#20 @14mm 0 2000 4000 0 2000 4000 0 2000 4000 Bit#21 @15mm 0 2000 4000 0 2000 4000 0 2000 4000 Bit#22 @16mm 0 2000 4000 0 2000 4000 0 2000 4000 Bit#23 @17mm 0 2000 4000 0 2000 4000 0 1000 2000 Bit#24 @27mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#25 @37mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#26 @47mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#27 @57mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#28 @67mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#29 @90mm 0 2000 4000 0 1000 2000 0 1000 2000 Bit#30 @120mm 0 2000 4000 0 1000 2000 300 350 400 450 0 1000 2000 Bit#31 @170mm Range, mm 300 350 400 450 0 2000 4000 Range, mm 300 350 400 450 0 1000 2000 Range, mm Figure 4-4. Same as in Figure 4-3 for tests 17–31.

PAGE 113

94 0 100 200 Bit#1 @0mm 1.08MHz 0 100 200 2.07MHz 0 100 200 4.70MHz 0 100 200 Bit#2 @1mm 0 100 200 0 100 200 0 100 200 Bit#3 @1mm 0 100 200 0 100 200 0 100 200 Bit#4 @2mm 0 100 200 0 100 200 0 100 200 Bit#5 @2mm 0 100 200 0 100 200 0 100 200 Bit#6 @3mm 0 100 200 0 100 200 0 100 200 Bit#7 @3mm 0 100 200 0 100 200 0 100 200 Bit#8 @4mm 0 100 200 0 100 200 0 100 200 Bit#9 @4mm 0 100 200 0 100 200 0 100 200 Bit#10 @5mm 0 100 200 0 100 200 0 100 200 Bit#11 @5mm 0 100 200 0 200 400 0 100 200 Bit#12 @6mm 0 100 200 0 200 400 0 200 400 Bit#13 @7mm 0 200 400 0 500 0 100 200 Bit#14 @8mm 0 500 0 500 1000 0 200 400 Bit#15 @9mm 0 500 1000 0 500 1000 300 350 400 450 0 200 400 Bit#16 @10mm Range, mm 300 350 400 450 0 500 1000 Range, mm 300 350 400 450 0 500 1000 Range, mm Figure 4-5. Root-mean-square voltage profiles with subtracted bottom echo profile for the tests with 0.33mm steel wire. Tests 1–16 with the wire location (dashed line) and the bottom location (solid line).

PAGE 114

95 0 500 1000 Bit#17 @11mm 1.08MHz 0 500 1000 2.07MHz 0 500 1000 4.70MHz 0 500 1000 Bit#18 @12mm 0 500 1000 0 500 1000 0 500 1000 Bit#19 @13mm 0 500 1000 0 1000 2000 0 500 1000 Bit#20 @14mm 0 1000 2000 0 1000 2000 0 500 1000 Bit#21 @15mm 0 1000 2000 0 1000 2000 0 500 1000 Bit#22 @16mm 0 1000 2000 0 1000 2000 0 500 1000 Bit#23 @17mm 0 500 1000 0 1000 2000 0 500 1000 Bit#24 @27mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#25 @37mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#26 @47mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#27 @57mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#28 @67mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#29 @90mm 0 1000 2000 0 1000 2000 0 1000 2000 Bit#30 @120mm 0 1000 2000 0 1000 2000 300 350 400 450 0 1000 2000 Bit#31 @170mm Range, mm 300 350 400 450 0 1000 2000 Range, mm 300 350 400 450 0 1000 2000 Range, mm Figure 4-6. Same as in Figure 4-5 for tests 17–31.

PAGE 115

96 0 0.1 0.2 0.3 0.4 0.5 0.6 0 1000 2000 3000 a)1.08MHz Bit 0 0.1 0.2 0.3 0.4 0.5 0.6 0 500 1000 1500 2000 b)2.07MHz Bit 0 0.1 0.2 0.3 0.4 0.5 0.6 0 500 1000 1500 2000 c)4.70MHz BitRange, m Figure 4-7. Root-mean-square voltage profiles (solid) for tests with 2mm steel wire with the wire locations (dashed) and “ghost wire” reflection locations (dash-dotted) from a) 1.08, b) 2.07, and c) 4.70MHz transducers. Tests with a Water-Sediment Jet Measurements in a larger tank with the water-sediment mixture jet were performed. A setup which is similar to the experiments of Hay [1991] is schematically shown in Figure 4-8 The mixture was generated in the calibration chamber (Figure 3-1 ) and was moved by the pump system trough a flexible pipe connected to one of the chamber jets. The discharge of 2 liters per minute through a 3.2mm nozzle at the end of the pipe was large enough to prevent any accumulation of sediment in the pipe. The end of the pipe with the nozzle was attached to a ruler, which allowed positioning the jet with the 1mm precision. The sediment concentration in the chamber was 1g/l. Quartz sand with the

PAGE 116

97 mean size of 2.17 in phi units (approximately 0.22mm) and standard deviation of 0.27 was used. 50mm. ABS-to-jet distance ABS (3 frequencies) Jet Nozzle 3.2mm. Ruler Figure 4-8. Setup for the tests with the sediment jet. 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 50 100 150 a)1.08MHz Bit 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 200 400 600 b)2.07MHz Bit 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 500 1000 1500 c)4.70MHz BitRange, m Figure 4-9. Root-mean-square voltage profiles (solid) for tests for different watersediment jet locations (dashed) from a) 1.08, b) 2.07, and c) 4.70MHz transducers. The measurements with the ABS1 (see Chapter 3 ) were taken at the rate of 4Hz during 20 seconds at each jet location. The recorded RMS voltage profiles were averaged

PAGE 117

98 to provide with a single profile for each jet position. For the tests shown in Figure 4-9 the jet was located at 10cm intervals with the closest distance to the transducers of 30cm. As it can be seen from the measurements, the location of the jet exactly coincides with the peak of the profile. This is the same result as it was obtained in the tests with the wires. Therefore the assumption about the distance from the transducer to the bottom estimated from the ABS measurements as a location of the maximum of the peak is supported by these tests. Note that because of the low concentration the test with the jet showed no evidence of second reflection, which also was discussed in Chapter 2 Result of Bottom Echo Removal To illustrate the application of currently developed algorithm for the separation of signals from suspended scatterers and bottom echo data obtained in SandyDuck’97 field [ Mouraenko and Hanes 2003] experiment were used. A sample run with the data recorded on October 18th, 1997 (run “DAIH430”) at approximately 4m water depth over sandy bottom showed events of suspension of sand under the waves with the significant wave height of 1.9m. The measurements of suspended sediment concentration profiles were performed with the acoustic backscatter system. The profiles of RMS voltage were converted into the sediment concentration profiles with the model described in Chapter 2 for two cases—with the bottom echo present and removed. The results are shown in Figure 4-10 As it can be seen, the estimated bottom elevation fluctuates slightly with the waves passing over the instrument location. It especially noticeable for the time 17–20 seconds when a high wave passed over and the suspended sediment concentration near the bottom increased. The concentrations for that moment estimated from the initial (with bottom echo present) profiles increase up to 200g/l and higher. Theoretically, this concentration can be found near the bottom—the concentration

PAGE 118

99 for fully packed sand bottom can reach 1600g/l. But the attenuation of sound at that concentration level would be very high and the measured intensity should rapidly decrease after that concentration is reached. This is not true in the presented case. Therefore this estimation of the concentration is a result of the unbounded solution of inverse problem (equation 2-12 ) which is caused by very high intensity of the signal reflected from the bottom. The other plot shows the estimated concentration profiles with the bottom echo removed from the initial voltage profiles. The maximum concentration at the same time is now approximated as 50g/l with the signal decreasing afterwards. To obtain the total concentration, the bottom echo needs to be converted into concentration by a future inversion model. Concentraion, g/l 0.01 0.1 0.2 0.5 1 5 10 50 200 Range, m a) 0 5 10 15 20 0.6 0.7 0.8 0.9 Concentraion, g/l 0.01 0.1 0.2 0.5 1 5 10 50 200 Range, m b) 0 5 10 15 20 0.6 0.7 0.8 0.9 0 5 10 15 20 1 0 1 Time, s+ mc) Figure 4-10. Time series of suspended sediment concentrations estimated with a) bottom echo and b) bottom echo removed; c) free surface elevation.

PAGE 119

100 CHAPTER 5 CONCLUSIONS The accuracy and temporal resolution of measurements with the acoustic backscatter system depend strongly on the inversion model used for converting recorded data into profiles of suspended sediment concentrations and sediment sizes. The acoustic backscatter equation provides the basis for developing of the inversion model, but it is necessary to determine scattering parameters of sediment suspension before the inversion can be performed. The form function and normalized cross-section obtained from the analytical expressions for glass spheres or from the measurements with quartz sand cannot describe natural sediments of all possible mineralogical contents and shapes. In current work, a modified scattering correction function is proposed. It provides a simple way to account for the differences in sediment properties. The form of correction function depends on a single parameter Therefore, two different parameters can be defined—each for the form function and normalized crosssection. A series of laboratory tests with different sediment samples showed that the parameter varies in the range 32.5 ((. The values of parameter for correction of the form function, f and normalized cross-section, are most likely different. It was found that f is greater than The parameters f and obtained from different sound frequencies can be different, but single values for f and can be found.

PAGE 120

101 Application of the modified correction function leads to a significant increase in accuracy of estimating of sediment sizes. During the tests in the calibration chamber with a mixture of two sediment samples a series of measurements was collected with the acoustic backscatter system. The freely settling sediments were generating rapidly changing profiles with a gradient in sediment size and concentration. The results of inversion were compared to the 1D advection-diffusion model. The corrected inversion model predicts very well the overall dynamics of the profiles of sediment sizes with high temporal resolution (based on profiles averaged within every 3 seconds). The resulting errors lay within the sediment size standard deviation. The correct approximation of parameter provides the necessary adjustment to the sediment attenuation, what makes the inversion with the implicit method more stable. The calibration tests show that the system function is better approximated by a second order polynomial if the time varying gain (TVG) is applied. From the tests with different concentration levels the best-fit approximation can be found, and the corresponding parameters can be determined. Also, by testing over different sediment samples, the system function can be obtained. The system function should not be a function of a particular sediment sample. Although several approximations can exist, the general system function can be estimated by the mean of these approximations. From the ratios of each approximation and the general system function the parameters f can be determined. Therefore, a pair of parameters—f and —can describe every sediment sample, while the system functions become unique for each of the transducers. Further research needs to be done to evaluate the universality of the correction function.

PAGE 121

102 A sensitivity analysis was carried out for the system function, as a function of sediment size distribution parameters, parameter and the transducer radius. The effective acoustical radius of the transducer is usually 10-20% smaller than its geometrical radius, which was previously noticed by other investigators. Current analysis shows that the transducer radius has a very small (<1%) influence on inversion accuracy in far-field region. The standard deviation of sediment size distribution has a second order effect on accuracy of inversion; therefore, for well-sorted sediments ( <0.5) the ability of the instrument to successfully estimate the sorting from measurements becomes small. In case of wide or bimodal sediment size distribution the transducers unequally resolve presence of the sediments of all sizes. This can result in large variability of estimates for the sorting among the transducers and inability to resolve the sorting accurately. The predefined value of sorting, for example from sieve analysis, is recommended to use in the inversion. Multiple scattering of the sound waves can be ignored for the concentrations of natural sediments of <10g/l, but at higher concentrations or in presence of air bubbles it can become important. Therefore, a model was developed to examine the effect of multiple scattering by using a multiple layer approximation. The transmission and reflection coefficients were defined by analogy with the acoustic backscatter equation. A binary tree walk algorithm was implemented to account for all possible paths along which sound wave can travel. The resulting sound intensities can be found as the sum of all intensities of the sound waves arriving at the same time. The number of different paths increases rapidly with the number of layers, but the number of reflections can be limited to 3 if the

PAGE 122

103 number of layers is large (>15). The results show that, for the uniform concentration profiles, the difference can reach 12–14% at maximum concentration of 3g/l. With linearly changing concentration profiles, the difference does not exceed 1–3% at maximum concentration of 3g/l at range approximately 1m. “Bump” type profiles results in differences up to 20%, (at maximum concentration at “bump” of 5–7g/l) and produce the “ghost” picks, which can be seen from the wire tests. For further analysis, the model needs to be tested with data collected in presence of bubbles in order to investigate its validity for such applications. To improve the accuracy of the acoustic backscatter system in measuring suspended sediment concentration in the vicinity of the bottom, the bottom echo removal algorithm was developed. When the suspended sediment concentrations along the water column are low, the bed echo signal saturates the signal corresponding to the sound backscattered from suspended particles. The algorithm separates these two signals by assuming a bell-shaped form of bottom echo signal. This solution provides information on bottom location. It qualitatively well predicts the backscattered intensities from suspended sediments near the bottom, but requires future comparison with other types of measurements.

PAGE 123

104 APPENDIX A CALIBRATION TEST RESULTS The calibration of the ABS1 with transducers of frequencies 1.08, 2.07, and 4.70MHz was performed with four sediment samples. The parameters of size distributions of each sample is gathered into Table 3-4 In Figures A-1 – A-12 the best-fit parameters, calculated from minimization of error (equation 3-4 ) are shown. The results of estimated system functions are shown in Figures A-13 – A-36 0 5 10 15 20 25 30 Error, % 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 2.00 3.00 4.00 5.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 6.89 12.3 0.50 4.50 2.67 0.44 1 5.03 9.03 & & & 0.05 2 4.92 8.84 & & 3.34 & 3 4.76 8.41 & & 3.34 0.05 4 6.89 12.3 & 3.00 & & 5 5.03 9.03 & 4.50 & 0.05 6 4.92 8.84 & 4.50 3.34 & 7 4.76 8.41 & 4.50 3.34 0.05 8 5.33 9.51 2.50 & & & 9 4.86 8.72 2.50 & & 0.05 10 4.81 8.59 2.50 & 3.34 & 11 4.76 8.39 2.50 & 3.34 0.05 12 5.33 9.51 2.50 4.50 & & 13 4.86 8.72 2.50 4.50 & 0.05 14 4.81 8.59 2.50 4.50 3.34 & 15 4.76 8.39 2.50 4.42 3.34 0.05 Bounds 2.50 2.50 3.00 5.00 2.00 3.34 0.05 1.00 Figure A-1. Case 1: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer.

PAGE 124

105 0 5 10 15 20 25 30 Error, % 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00 2.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 7.27 11.4 0.50 4.50 2.67 0.44 1 3.7 6.78 & & & 0.05 2 3.62 6.76 & & 3.26 & 3 3.62 6.76 & & 2.76 0.05 4 7.27 11.4 & 3.00 & & 5 3.7 6.83 & 3.00 & 0.05 6 3.62 6.81 & 3.00 3.26 & 7 3.62 6.81 & 3.00 2.76 0.05 8 3.62 6.76 1.98 & & & 9 3.62 6.76 0.19 & & 0.06 10 3.62 6.76 0.35 & 3.13 & 11 3.62 6.76 1.57 & 2.39 0.08 12 3.62 6.81 1.98 3.00 & & 13 3.62 6.81 0.19 3.00 & 0.06 14 3.62 6.81 0.35 3.00 3.13 & 15 3.62 6.81 1.58 3.00 2.39 0.08 Bounds 2.50 2.50 3.00 5.00 2.00 3.34 0.05 1.00 Figure A-2. Case 1: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer. 0 10 20 30 40 50 Error, % 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 3.21 16.8 0.50 4.50 2.67 0.44 1 3.15 17 & & & 0.63 2 3.15 17 & & 2.78 & 3 3.15 17 & & 2.76 0.42 4 3.2 17.2 & 3.00 & & 5 3.14 17.4 & 3.00 & 0.30 6 3.14 17.4 & 3.00 2.78 & 7 3.14 17.4 & 3.00 2.79 0.45 8 3.15 17 0.33 & & & 9 3.15 17 0.31 & & 0.51 10 3.15 17 0.41 & 2.72 & 11 3.15 17 0.40 & 2.73 0.46 12 3.14 17.4 0.33 3.00 & & 13 3.14 17.4 0.31 3.00 & 0.50 14 3.14 17.4 0.38 3.00 2.70 & 15 3.14 17.4 0.41 3.00 2.73 0.45 Bounds 2.50 2.50 3.00 5.00 2.00 3.34 0.05 1.00 Figure A-3. Case 1: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer.

PAGE 125

106 3 4 5 6 7 8 9 10 Error, % 0.20 0.50 1.00 3.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 3.65 6.81 0.50 4.50 2.74 0.25 1 3.32 6.16 & & & 0.05 2 2.98 5.66 & & 3.42 & 3 2.95 5.61 & & 3.43 0.05 4 3.65 6.81 & 3.00 & & 5 3.32 6.16 & 4.50 & 0.05 6 2.98 5.66 & 4.50 3.43 & 7 2.95 5.61 & 4.50 3.43 0.05 8 3.26 6.05 2.50 & & & 9 3.13 5.85 2.50 & & 0.05 10 2.96 5.62 2.50 & 3.43 & 11 2.93 5.59 2.50 & 3.43 0.05 12 3.26 6.05 2.50 4.50 & & 13 3.13 5.85 2.50 4.50 & 0.05 14 2.96 5.62 2.50 4.49 3.43 & 15 2.93 5.59 2.50 4.49 3.43 0.05 Bounds 2.50 2.50 3.00 5.00 2.06 3.43 0.05 1.00 Figure A-4. Case 2: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer. 5 5.5 6 6.5 7 Error, % 0.20 0.50 1.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 2.57 5.85 0.50 4.50 2.74 0.25 1 2.57 5.9 & & & 0.24 2 2.57 5.9 & & 2.76 & 3 2.57 5.9 & & 2.75 0.24 4 2.57 5.89 & 3.00 & & 5 2.57 5.94 & 3.00 & 0.24 6 2.57 5.94 & 3.00 2.76 & 7 2.57 5.94 & 3.00 2.75 0.25 8 2.57 5.9 0.37 & & & 9 2.57 5.9 0.50 & & 0.24 10 2.57 5.9 0.50 & 2.76 & 11 2.57 5.9 0.50 & 2.75 0.24 12 2.57 5.94 0.37 3.00 & & 13 2.57 5.94 0.50 3.00 & 0.24 14 2.57 5.94 0.50 3.00 2.76 & 15 2.57 5.94 0.50 3.00 2.75 0.25 Bounds 2.50 2.50 3.00 5.00 2.06 3.43 0.05 1.00 Figure A-5. Case 2: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer.

PAGE 126

107 5 10 15 20 25 30 Error, % 0.20 0.50 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 5.81 21.9 0.50 4.50 2.74 0.25 1 4.76 19.8 & & & 0.51 2 3.55 17 & & 2.11 & 3 3.29 16.3 & & 2.06 0.05 4 5.81 21.6 & 5.00 & & 5 4.76 19.5 & 5.00 & 0.51 6 3.55 16.7 & 5.00 2.11 & 7 3.29 16 & 5.00 2.06 0.05 8 2.09 11.8 1.47 & & & 9 2.09 11.8 1.31 & & 0.63 10 2.09 11.8 1.64 & 2.83 & 11 2.09 11.8 2.47 & 2.06 0.99 12 2.08 11.2 1.47 5.00 & & 13 2.08 11.2 1.31 5.00 & 0.63 14 2.08 11.2 1.64 5.00 2.83 & 15 2.08 11.2 2.42 5.00 2.06 0.98 Bounds 2.50 2.50 3.00 5.00 2.06 3.43 0.05 1.00 Figure A-6. Case 2: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer. 0 5 10 15 20 25 30 35 40 45 Error, % 0.10 0.20 0.30 0.40 0.50 0.70 1.00 1.50 2.00 3.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 11.7 20.5 0.50 4.50 2.17 0.27 1 10.9 18.5 & & & 0.05 2 10.2 16.6 & & 2.71 & 3 9.89 16 & & 2.71 0.05 4 11.7 20.5 & 5.00 & & 5 10.9 18.5 & 5.00 & 0.05 6 10.2 16.6 & 5.00 2.71 & 7 9.89 16 & 5.00 2.71 0.05 8 10.4 17.2 2.50 & & & 9 10.2 16.7 2.50 & & 0.05 10 9.86 15.9 2.50 & 2.71 & 11 9.74 15.6 2.50 & 2.71 0.05 12 10.4 17.2 2.50 5.00 & & 13 10.2 16.7 2.50 5.00 & 0.05 14 9.86 15.9 2.50 5.00 2.71 & 15 9.74 15.6 2.50 5.00 2.71 0.05 Bounds 2.50 2.50 3.00 5.00 1.63 2.71 0.05 1.00 Figure A-7. Case 3: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer.

PAGE 127

108 2 4 6 8 10 12 Error, % 0.10 0.20 0.30 0.40 0.50 0.70 1.00 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 3.57 5.58 0.50 4.50 2.17 0.27 1 2.84 5.02 & & & 0.05 2 2.65 5.2 & & 2.47 & 3 2.65 5.2 & & 2.40 0.21 4 3.57 5.63 & 3.00 & & 5 2.84 5.06 & 3.00 & 0.05 6 2.65 5.24 & 3.00 2.47 & 7 2.65 5.24 & 3.00 2.40 0.21 8 2.65 5.2 0.77 & & & 9 2.65 5.2 0.18 & & 0.05 10 2.65 5.2 0.04 & 2.36 & 11 2.65 5.2 0.01 & 2.21 0.05 12 2.65 5.24 0.77 3.00 & & 13 2.65 5.24 0.19 3.00 & 0.07 14 2.65 5.24 0.04 3.00 2.36 & 15 2.65 5.24 0.01 3.00 2.22 0.09 Bounds 2.50 2.50 3.00 5.00 1.63 2.71 0.05 1.00 Figure A-8. Case 3: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer. 0 5 10 15 20 25 30 35 Error, % 0.10 0.20 0.30 0.40 0.50 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 4.81 18.9 0.50 4.50 2.17 0.27 1 4.56 18.5 & & & 0.05 2 4.8 18.9 & & 2.13 & 3 4.4 18.3 & & 2.01 0.05 4 4.8 19 & 3.00 & & 5 4.55 18.7 & 3.00 & 0.05 6 4.8 19 & 3.00 2.13 & 7 4.39 18.4 & 3.00 2.01 0.05 8 2.45 14 1.20 & & & 9 2.45 14 1.18 & & 0.19 10 2.45 14 1.20 & 2.22 & 11 2.45 14 1.18 & 2.11 0.19 12 2.44 14.2 1.20 3.00 & & 13 2.44 14.2 1.18 3.00 & 0.19 14 2.44 14.2 1.20 3.00 2.22 & 15 2.44 14.2 1.18 3.00 2.12 0.20 Bounds 2.50 2.50 3.00 5.00 1.63 2.71 0.05 1.00 Figure A-9. Case 3: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer.

PAGE 128

109 0 10 20 30 40 50 60 70 80 Error, % 0.50 0.60 0.70 0.80 1.00 1.20 1.40 1.60 2.00 2.40 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 17.1 37.8 0.50 4.50 2.16 0.54 1 13.9 25.6 & & & 0.05 2 14.7 27.9 & & 2.70 & 3 13 23 & & 2.70 0.05 4 17.1 37.8 & 5.00 & & 5 13.9 25.6 & 5.00 & 0.05 6 14.7 27.9 & 5.00 2.70 & 7 13 23 & 5.00 2.70 0.05 8 14.2 26.3 2.50 & & & 9 13.3 23.8 2.50 & & 0.05 10 13.5 24.3 2.50 & 2.70 & 11 12.9 22.6 2.50 & 2.70 0.05 12 14.2 26.3 2.50 5.00 & & 13 13.3 23.8 2.50 5.00 & 0.05 14 13.5 24.3 2.50 5.00 2.70 & 15 12.9 22.6 2.50 5.00 2.70 0.05 Bounds 2.50 2.50 3.00 5.00 1.62 2.70 0.05 1.00 Figure A-10. Case 4: Errors in inverted concentrations from varying of parameters ta , and for 1.08MHz transducer. 0 50 100 150 200 250 Error, % 0.40 0.50 0.60 0.70 0.80 1.00 1.20 1.40 1.60 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 12.2 Inf 0.50 4.50 2.16 0.54 1 8.97 18.1 & & & 0.05 2 9.3 19.6 & & 2.70 & 3 8.72 17.1 & & 2.34 0.05 4 12.2 Inf & 5.00 & & 5 8.97 18.2 & 5.00 & 0.05 6 9.3 19.6 & 5.00 2.70 & 7 8.72 17.1 & 5.00 2.34 0.05 8 8.72 17.1 2.17 & & & 9 8.72 17.1 0.45 & & 0.07 10 8.72 17.1 1.10 & 2.62 & 11 8.72 17.1 0.16 & 2.22 0.05 12 8.72 17.1 2.17 5.00 & & 13 8.72 17.1 0.45 5.00 & 0.07 14 8.72 17.1 1.10 5.00 2.62 & 15 8.72 17.1 0.16 5.00 2.22 0.05 Bounds 2.50 2.50 3.00 5.00 1.62 2.70 0.05 1.00 Figure A-11. Case 4: Errors in inverted concentrations from varying of parameters ta , and for 2.07MHz transducer.

PAGE 129

110 0 10 20 30 40 50 60 70 Error, % 0.40 0.50 0.60 mean Conc, g/l Test # Fit error, % Mean error, % at, mm !!Initial 6.25 42.5 0.50 4.50 2.16 0.54 1 4.44 13.3 & & & 1.00 2 5.33 26.5 & & 1.62 & 3 4.43 12.9 & & 1.69 0.90 4 6.24 42.7 & 5.00 & & 5 4.44 12.7 & 5.00 & 1.00 6 5.33 26.2 & 5.00 1.62 & 7 4.43 12.4 & 5.00 1.68 0.90 8 4.75 18.8 2.50 & & & 9 4.43 12.9 0.77 & & 1.00 10 4.71 18.3 2.50 & 1.62 & 11 4.43 12.9 0.27 & 1.62 0.87 12 4.74 18.4 2.50 5.00 & & 13 4.43 12.4 2.11 5.00 & 0.97 14 4.71 17.9 2.50 5.00 1.62 & 15 4.43 12.4 0.26 5.00 1.66 0.88 Bounds 2.50 2.50 3.00 5.00 1.62 2.70 0.05 1.00 Figure A-12. Case 4: Errors in inverted concentrations from varying of parameters ta , and for 4.70MHz transducer. 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.67 = 0.44 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 10 20 30 40 50 System coefficients b). Err = 6.8873% Kv = 30.7307 & r2 4.5384 & r+12.7797KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 Mean error = 12.29% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 16.0% 0.30 18.8% 0.40 9.0% 0.50 6.8% 0.60 6.8% 0.70 3.8% 0.80 7.0% 0.90 6.6% 1.00 8.2% 2.00 25.4% 3.00 13.8% 4.00 24.5% 5.00 12.9% Conc,g/l Err,% Figure A-13. Case 1: Estimation of the system function for the 1.08MHz transducer initial value of parameter : a–c) same as in Figure 3-12

PAGE 130

111 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.67 = 0.44 f = 0.50 = 2.50 0.2 0.4 0.6 0.8 0 10 20 30 40 System coefficients b). Err = 5.3265% Kv = 28.6465 & r2 4.1686 & r+12.6589KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 5 6 Mean error = 9.51% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 12.4% 0.30 15.4% 0.40 6.3% 0.50 5.3% 0.60 6.8% 0.70 4.5% 0.80 6.6% 0.90 6.2% 1.00 6.6% 2.00 22.2% 3.00 8.4% 4.00 14.6% 5.00 8.3% Conc,g/l Err,% Figure A-14. Case 1: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 2.5 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.67 = 0.44 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 50 100 150 System coefficients b). Err = 7.2707% Kv = 3.2122 & r2+64.9692 & r+19.3012KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 Mean error = 11.38% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 24.4% 0.20 15.2% 0.30 9.5% 0.40 5.9% 0.50 3.5% 0.60 5.2% 0.70 5.2% 0.80 7.5% 0.90 7.7% 1.00 12.9% 2.00 28.1% Conc,g/l Err,% Figure A-15. Case 1: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 131

112 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 2.5 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.67 = 0.44 f = 0.50 = 1.98 0.2 0.4 0.6 0.8 0 20 40 60 80 System coefficients b). Err = 3.6196% Kv = 7.7387 & r2+61.5636 & r+19.3143KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 2 Mean error = 6.76% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 13.2% 0.20 6.3% 0.30 3.2% 0.40 3.4% 0.50 4.1% 0.60 7.8% 0.70 5.6% 0.80 5.0% 0.90 3.0% 1.00 4.3% 2.00 18.5% Conc,g/l Err,% Figure A-16. Case 1: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.67 = 0.44 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 50 100 150 200 250 300 System coefficients b). Err = 3.2057% Kv = 230.6803 & r2+464.3218 & r+14.8929KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.2 0.4 0.6 0.8 Mean error = 16.78% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 9.0% 0.20 5.3% 0.30 4.3% 0.40 4.5% 0.50 9.4% 0.60 21.0% 0.70 32.9% 0.80 47.9% Conc,g/l Err,% Figure A-17. Case 1: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 132

113 0.2 0.4 0.6 0.8 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.67 = 0.44 f = 0.50 = 0.33 0.2 0.4 0.6 0.8 0 50 100 150 200 250 300 System coefficients b). Err = 3.1461% Kv = 233.4131 & r2+459.3537 & r+15.4429KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.2 0.4 0.6 0.8 Mean error = 17.00% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 7.9% 0.20 6.4% 0.30 4.8% 0.40 4.5% 0.50 9.3% 0.60 21.3% 0.70 33.4% 0.80 48.3% Conc,g/l Err,% Figure A-18. Case 1: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.74 = 0.25 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 10 20 30 40 50 System coefficients b). Err = 3.6493% Kv = 19.5280 & r2+12.5779 & r+15.8376KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 Mean error = 6.81% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 7.9% 0.50 4.8% 1.00 5.3% 3.00 9.2% Conc,g/l Err,% Figure A-19. Case 2: Estimation of the system function for the 1.08MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 133

114 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.74 = 0.25 f = 0.50 = 2.50 0.2 0.4 0.6 0.8 0 10 20 30 40 50 System coefficients b). Err = 3.2610% Kv = 19.0202 & r2+12.5043 & r+15.8217KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 Mean error = 6.05% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 7.3% 0.50 4.2% 1.00 5.5% 3.00 7.2% Conc,g/l Err,% Figure A-20. Case 2: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.74 = 0.25 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 20 40 60 80 System coefficients b). Err = 2.5705% Kv = 19.2774 & r2+76.4350 & r+21.2817KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 Mean error = 5.85% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 5.3% 0.50 6.8% 1.00 5.4% Conc,g/l Err,% Figure A-21. Case 2: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 134

115 0.2 0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.74 = 0.25 f = 0.50 = 0.37 0.2 0.4 0.6 0.8 0 20 40 60 80 System coefficients b). Err = 2.5674% Kv = 19.4754 & r2+76.2805 & r+21.2894KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 Mean error = 5.90% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 5.2% 0.50 6.9% 1.00 5.7% Conc,g/l Err,% Figure A-22. Case 2: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.7 0.8 0.9 1 1.1 1.2 1.3 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.74 = 0.25 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 50 100 150 200 System coefficients b). Err = 5.8109% Kv = 245.9268 & r2+340.1723 & r+37.8448KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.2 0.4 0.6 0.8 Mean error = 21.88% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 18.3% 0.50 25.5% Conc,g/l Err,% Figure A-23. Case 2: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 135

116 0.2 0.4 0.6 0.8 0.7 0.8 0.9 1 1.1 1.2 1.3 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.74 = 0.25 f = 0.50 = 1.47 0.2 0.4 0.6 0.8 0 50 100 150 200 250 System coefficients b). Err = 2.0882% Kv = 251.3325 & r2+396.7159 & r+32.3872KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.2 0.4 0.6 0.8 Mean error = 11.81% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.20 6.4% 0.50 17.3% Conc,g/l Err,% Figure A-24. Case 2: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 2.5 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 10 20 30 40 System coefficients b). Err = 11.7254% Kv = 33.8551 & r2 11.4308 & r+10.9485KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 5 Mean error = 20.46% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 41.0% 0.20 27.1% 0.30 13.2% 0.40 9.7% 0.50 3.1% 0.70 4.2% 1.00 15.7% 1.50 28.7% 2.00 25.6% 3.00 36.2% Conc,g/l Err,% Figure A-25. Case 3: Estimation of the system function for the 1.08MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 136

117 0.2 0.4 0.6 0.8 0 0.5 1 1.5 2 2.5 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 2.50 0.2 0.4 0.6 0.8 0 5 10 15 20 25 30 System coefficients b). Err = 10.4178% Kv = 32.1612 & r2 11.0111 & r+10.8400KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 5 Mean error = 17.25% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 38.5% 0.20 24.3% 0.30 10.2% 0.40 11.1% 0.50 4.0% 0.70 5.5% 1.00 15.4% 1.50 24.6% 2.00 18.5% 3.00 20.5% Conc,g/l Err,% Figure A-26. Case 3: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 20 40 60 80 100 System coefficients b). Err = 3.5695% Kv = 17.4659 & r2+45.4530 & r+21.1931KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 Mean error = 5.58% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 11.7% 0.20 6.2% 0.30 2.7% 0.40 4.2% 0.50 5.1% 0.70 5.0% 1.00 4.2% Conc,g/l Err,% Figure A-27. Case 3: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 137

118 0.2 0.4 0.6 0.8 0.5 1 1.5 2 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.77 0.2 0.4 0.6 0.8 0 20 40 60 80 System coefficients b). Err = 2.6485% Kv = 12.3413 & r2+44.3597 & r+21.1283KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 Mean error = 5.20% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 7.0% 0.20 3.9% 0.30 3.1% 0.40 5.3% 0.50 4.6% 0.70 3.3% 1.00 9.2% Conc,g/l Err,% Figure A-28. Case 3: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 50 100 150 200 250 System coefficients b). Err = 4.8066% Kv = 209.7920 & r2+377.1953 & r+25.0449KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 Mean error = 18.90% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 18.2% 0.20 12.7% 0.30 3.1% 0.40 26.2% 0.50 34.3% Conc,g/l Err,% Figure A-29. Case 3: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 138

119 0.2 0.4 0.6 0.8 0.8 1 1.2 1.4 1.6 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.17 = 0.27 f = 0.50 = 1.20 0.2 0.4 0.6 0.8 0 50 100 150 200 250 System coefficients b). Err = 2.4457% Kv = 198.7014 & r2+405.7229 & r+22.0241KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 Mean error = 13.96% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.10 6.0% 0.20 5.9% 0.30 3.9% 0.40 25.1% 0.50 28.8% Conc,g/l Err,% Figure A-30. Case 3: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.7 0.8 0.9 1 1.1 1.2 1.3 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.16 = 0.54 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 5 10 15 20 25 30 System coefficients b). Err = 17.0961% Kv = 24.0417 & r2 2.9426 & r+7.4190KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 Mean error = 37.80% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.50 45.6% 0.60 26.6% 0.70 26.6% 0.80 40.7% 1.00 15.2% 1.20 9.9% 1.40 29.3% 1.60 67.9% 2.00 40.3% 2.40 75.7% Conc,g/l Err,% Figure A-31. Case 4: Estimation of the system function for the 1.08MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 139

120 0.2 0.4 0.6 0.8 0.7 0.8 0.9 1 1.1 1.2 1.3 Ratio of voltages Range, mVmeas / Vcalca). f = 1.08MHz at = 4.50 mm = 2.16 = 0.54 f = 0.50 = 2.50 0.2 0.4 0.6 0.8 0 5 10 15 20 25 System coefficients b). Err = 14.1598% Kv = 19.4051 & r2 2.0512 & r+7.1964KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 1 2 3 4 Mean error = 26.29% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.50 38.9% 0.60 22.0% 0.70 19.6% 0.80 34.7% 1.00 17.1% 1.20 10.1% 1.40 23.8% 1.60 44.2% 2.00 22.6% 2.40 30.0% Conc,g/l Err,% Figure A-32. Case 4: Estimation of the system function for the 1.08MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0.4 0.6 0.8 1 1.2 1.4 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.16 = 0.54 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 20 40 60 80 100 System coefficients b). Err = 12.1824% Kv = 18.0542 & r2+43.3926 & r+12.5010KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 2 4 6 8 10 x 1067 Mean error = Inf% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.40 48.2% 0.50 17.5% 0.60 11.7% 0.70 8.6% 0.80 25.8% 1.00 40.0% 1.20 78.7% 1.40 242.5% 1.60 Inf% Conc,g/l Err,% Figure A-33. Case 4: Estimation of the system function for the 2.07MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 140

121 0.2 0.4 0.6 0.8 0.4 0.6 0.8 1 1.2 1.4 Ratio of voltages Range, mVmeas / Vcalca). f = 2.07MHz at = 4.50 mm = 2.16 = 0.54 f = 0.50 = 2.17 0.2 0.4 0.6 0.8 0 10 20 30 40 50 60 System coefficients b). Err = 8.7214% Kv = 1.8421 & r2+39.5602 & r+12.4132KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 2 Mean error = 17.11% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.40 30.9% 0.50 11.5% 0.60 18.5% 0.70 17.9% 0.80 13.2% 1.00 19.0% 1.20 15.8% 1.40 9.9% 1.60 17.2% Conc,g/l Err,% Figure A-34. Case 4: Estimation of the system function for the 2.07MHz transducer with corrected parameter : a–c) same as in Figure 3-12 0.2 0.4 0.6 0.8 0 0.5 1 1.5 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.16 = 0.54 f = 0.50 = 0.50 0.2 0.4 0.6 0.8 0 50 100 150 200 System coefficients b). Err = 6.2478% Kv = 110.8176 & r2+224.4238 & r+33.2869KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 2 2.5 Mean error = 42.52% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.40 32.6% 0.50 32.1% 0.60 62.9% Conc,g/l Err,% Figure A-35. Case 4: Estimation of the system function for the 4.70MHz transducer with initial value of parameter : a–c) same as in Figure 3-12

PAGE 141

122 0.2 0.4 0.6 0.8 0 0.5 1 1.5 Ratio of voltages Range, mVmeas / Vcalca). f = 4.70MHz at = 4.50 mm = 2.16 = 0.54 f = 0.50 = 2.50 0.2 0.4 0.6 0.8 0 50 100 150 System coefficients b). Err = 4.7454% Kv = 122.7166 & r2+191.2057 & r+35.7818KvRange, m fit 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.5 1 1.5 Mean error = 18.78% c). Measured (dashed) and calculated (solid) concentrations Concentration, g/lRange, m 0.40 20.6% 0.50 19.0% 0.60 16.7% Conc,g/l Err,% Figure A-36. Case 4: Estimation of the system function for the 4.70MHz transducer with corrected parameter : a–c) same as in Figure 3-12

PAGE 142

123 APPENDIX B EMPIRICAL MODEL FOR ACOUSTIC BACKSCATTER The empirical models were previously used to convert measured by the acoustic backscatter system sound intensities into suspended sediment concentration profiles [e.g., Tamura and Hanes 1986]. The currently developed model provides an accurate method to invert ABS measurements by using a set of coefficients. The coefficients for the empirical model can be obtained from the calibration in the chamber (Figure 3-1 ) with the sediments taken from the field site. Therefore this limits the application of the model to the used sediment. The solution of the acoustic backscatter equation described in Chapter 2 allows using the same system coefficients for the sediment of wide range sizes and properties, as long as the system parameters remain unchanged. The problem arises when the unique system coefficients cannot be found. This can lead to increasing errors in estimation of suspended sediment concentrations. From the solution for the acoustic backscatter system given by equations 2-10 and 2-11 recorded voltage is determined as a function of suspended sediment concentration, sediment size, and system parameters. The total sound attenuation is given by 2-42 The attenuation due to the water is small compared to the sediment attenuation, thus can be neglected. The sediment attenuation at a given point is described by the sediment concentration and the sediment attenuation constant ( 2-47 ). In order to obtain the system constants calibration of the system is performed. Measurements are obtained on homogeneous suspension at known concentration levels with the sediments taken from field site. With system and sediment parameters assumed

PAGE 143

124 to be constant during the measurements, voltage curves can be written as a function of distance from the transducer. The approximation can be made as follows 12 Br rmsM VAe r (B-1) with two constants A and B The parameter A depends on sediment size and system constant only, but not on concentration. The parameter B is defined by the attenuation and relates to the concentration as 01 222r s B MdrM r (B-2) when and M are constants for homogeneous suspension. The tests in the calibration chamber show that it is not always possible to reproduce concentration profile by varying only two parameters, such as and concentration. The system parameter ( 2-11 ) is not always constant, especially when the time varying gain (TVG) is used. The semi-empirical model for conversion of the ABS measurements into concentrations only is introduced. The voltage curves from the calibration are approximated with the tree parameter model as 12MMnBr fitMVMAre (B-3) where constantsMA M B and Mn are chosen to minimize the error with measured data for each of the concentration level M From the original backscatter equation for farfield the parameter MA is a constant, Mn is equal to –1 if the TVG is not used, and M B depends on concentration linearly. In present model all three parameters are free with dependence on concentration only.

PAGE 144

125 Since the measurements for some discrete concentration levels can be obtained, the triple of parameters ,,MMMABn are approximated for other concentrations by extrapolation from known values. The power law curves were used to extrapolate the values of parameters for concentrations up twice maximum concentration measured in form byaxc (B-4) where y is one of the parameters MA M B or Mn and a b, and c are the coefficients. Therefore, each of MA M B or Mn have corresponded set of parameters ,, abc which are fully specify the model. Comparing to the initial solution, where only two parameters (mean sediment size and concentration) are variable and other parameters can be found from the theoretical or empirical expressions, the proposed model requires 9 parameters—coefficients ,, abc for each of ,,MMMABn —obtained from the calibration. The sand used in the calibration of the model needs to be from the field site, since the information on sediment size is not directly included into model and cannot be extracted. To invert the measured voltage into concentration iterative solution method is applied. Follow the algorithm described elsewhere (e.g. [ Thorne and Hanes 2002]), the concentration near the transducer needs to be determined first. Since the measurements at the ranges just near the transducer surface usually contain electronic noise, the concentration can be taken to be zero. The concentrations at further ranges are calculated by iteration using formula 12 ,1 1 1kkkrmsk k nBr kkV M Are (B-5)

PAGE 145

126 where k M is the suspended sediment concentration at range kr and ,,kkkABn are the parameters calculated using the corresponding power law curves (equation B-4 ). The concentration used in obtaining of ,,kkkABn is equal to the mean concentration at previously calculated ranges. Figures B-1 – B-3 show the calibration of the transducers with frequencies 1.08, 2.07, and 4.70MHz with calculated coefficients ,,MMMABn Measured voltages for different concentration levels (on plots “a”) are shown with approximating curves calculated by fitting equation B-3 The corresponding coefficients ,,MMMABn are gathered into the table on each figure. The values of the coefficients are plotted as circles on plots “b”–“d”. The extrapolated by the power law curves with coefficients ,, abc which are also gathered into the table, show the approximations for ,,MMMABn at higher concentrations. According to equation B-2 coefficients M B depend linearly on concentration. The power coefficient b for the transducers with frequencies 2.07 and 4.70MHz are 0.82 and 1.17 correspondingly, which are closer to 1 than the power coefficient for 1.08MHz transducer, which is equal to 0.53. The parameter M B corresponds to the attenuation of the sound, and therefore has to be negative. Despite, it has positive values for all concentrations for 1.08MHz transducer, and up to 0.5g/l for 2.07MHz. This comes from the coupling of two parameters—M B and Mn —when the changes in one of the parameter can lead to the changes in another. Although, if one of the parameters is set to be constant, ability of the model to provide better fit to the voltage curves becomes limited, which therefore results in larger errors in inversion into the sediment concentration.

PAGE 146

127 0 20 40 60 80 100 120 0 200 400 600 800 1000 1200 Fit curves: Vfit = sqrt(M) & A & rn& eB*rBinCountsa). M | A | B | n 0.10 | 11.13 | 2.3641 | 1.5118 0.20 | 15.99 | 2.0960 | 1.3626 0.30 | 16.82 | 2.1847 | 1.3493 0.40 | 15.67 | 2.4398 | 1.3780 0.50 | 21.43 | 1.9793 | 1.2496 0.70 | 24.86 | 1.7846 | 1.2009 1.00 | 27.02 | 1.7256 | 1.1824 1.50 | 28.94 | 1.6908 | 1.1490 2.00 | 37.25 | 1.2816 | 1.0444 3.00 | 39.65 | 1.1783 | 1.0122 Coefficients [a b c]: A = [ 26.535906 0.380370 0.191433]; B = [ 0.828189 0.529510 2.606169]; n = [ 2.856120 0.051853 4.032534]; 0 20 40 60 Fit curve: y = a & xb+cAb). 0 1 2 3 Bc). 0 1 2 3 4 5 6 1.6 1.4 1.2 1 0.8 nM, mean concentration d). Figure B-1. Calibration coefficients for 1.08 MHz transducer: a) measured voltage (dots) and fitted by curves B-3 and calculated coefficients b) A, c) B and d) n 0 20 40 60 80 100 120 0 200 400 600 800 1000 1200 Fit curves: Vfit = sqrt(M) & A & rn& eB*rBinCountsa). M | A | B | n 0.10 | 172.50 | 0.5905 | 0.8134 0.20 | 204.42 | 0.3838 | 0.7435 0.30 | 230.66 | 0.2201 | 0.6823 0.40 | 257.35 | 0.0624 | 0.6308 0.50 | 238.02 | 0.1373 | 0.6500 0.70 | 299.86 | 0.2650 | 0.5405 1.00 | 355.77 | 0.6645 | 0.4658 1.50 | 394.97 | 1.0320 | 0.3886 2.00 | 492.35 | 1.6548 | 0.2847 Coefficients [a b c]: A = [198.614575 0.766785 144.602059]; B = [ 1.355532 0.816231 0.767856]; n = [ 0.534683 0.428192 1.010497]; 0 200 400 600 800 Fit curve: y = a & xb+cAb). 4 2 0 2 Bc). 0 1 2 3 4 1 0.5 0 nM, mean concentration d). Figure B-2. Calibration coefficients for 2.07MHz transducer: a–d) same as in Figure B-1

PAGE 147

128 0 20 40 60 80 100 120 0 500 1000 1500 Fit curves: Vfit = sqrt(M) & A & rn& eB*rBinCountsa). M | A | B | n 0.10 | 2016.54 | 1.8475 | 0.1132 0.20 | 2322.68 | 2.3017 | 0.0379 0.30 | 2566.73 | 2.7131 | 0.0272 0.40 | 3062.75 | 3.2112 | 0.1401 0.50 | 3566.10 | 3.7140 | 0.1926 0.70 | 4599.33 | 4.6715 | 0.3080 1.00 | 7132.18 | 6.3905 | 0.4842 Coefficients [a b c]: A = [5120.106312 1.756009 1979.662202]; B = [ 4.830602 1.169421 1.541309]; n = [ 0.737061 0.748176 0.251822]; 0 0.5 1 1.5 2 x 104 Fit curve: y = a & xb+cAb). 15 10 5 0 Bc). 0 0.5 1 1.5 2 0.5 0 0.5 1 nM, mean concentration d). Figure B-3. Calibration coefficients for 4.70MHz transducer: a–d) same as in Figure B-1 The method was applied to the measurements used for calibration of the ABS1 and the results are shown on Figure B-4 Though the initial concentration near the transducer should be approximated, it was taken to be 0.001g/l for all 3 transducers. The method shows very good accuracy for the 1.08 and 2.07MHz transducers compared to the results in Figures 3-12 and 3-13 For the 2.07 and 4.70MHz transducers the method works very well for low concentrations and underestimates the concentrations at higher levels. This would not happen if the initial concentration were determined properly (for example, known concentrations were used). For the 4.70MHz transducer reasonable results can be achieved at low (less than 0.5g/l) concentration levels.

PAGE 148

129 0 1 2 3 4 Concentration, g/la).Mean error 6.49% 0.1 6.5% 0.2 6.4% 0.3 5.7% 0.4 12.0% 0.5 5.1% 0.7 6.6% 1.0 5.8% 1.5 6.0% 2.0 5.5% 3.0 5.4% Conc,g/l Err,% 0 1 2 3 Concentration, g/lb).Mean error 5.23% 0.1 5.4% 0.2 4.2% 0.3 3.3% 0.4 4.2% 0.5 4.1% 0.7 2.5% 1.0 3.1% 1.5 4.5% 2.0 15.8% 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0 0.2 0.4 0.6 0.8 1 Distance, mConcentration, g/lc).Mean error 21.64% 0.1 6.5% 0.2 14.2% 0.3 19.8% 0.4 11.3% 0.5 10.3% 0.7 22.0% 1.0 67.3% Figure B-4. Concentration profiles converted with the empirical model from the measurements with a) 1.08, b) 2.07, and c) 4.70MHz transducer. The disadvantage of the method is that the parameters such as sediment of the calibration system should be similar to the ones found at experiment site. To obtain better conversion the range of concentration levels for calibration should be wide.

PAGE 149

130 APPENDIX C 1D ADVECTION-DIFFUSION MODEL The distribution of suspended sediments in the chamber is modeled with the advection-diffusion equation [ Nielsen 1992, p.241] ss M MM w trrr//////// (C-1) where M is the suspended sediment concentration, s is the sediment diffusivity, and s w is the settling velocity. With the vertical coordinate, r, directed downward the advection term is taken with the positive sign. The settling velocity is a function of the sediment size which varies with the range r. In general the sediment diffusivity is a function of the range as well. To simplify a solution of the equation, sediment sizes are assumed to obey some distribution, p or can be divided into fractions. The total concentration can be found as the superposition of mass concentrations for each fraction (,), M trMtrd or ,, M trMtr (C-2) where M is the suspended sediment concentration of the fraction of sediments of size The sediment size distribution can be calculated as ,, ,tr M tr p M tr (C-3)

PAGE 150

131 Now the equation can be rewritten for each sediment size fraction as ssMMM w trrr //// //// (C-4) where s w is the settling velocity for the sediment size The settling velocity for each fraction is now can be assumed constant, and the sediment diffusivity is assumed to be a function of r only. If the sediment diffusivity is zero than the equation is a pure advection. The equation is solved for 0, rD 2 where D is the depth of a domain, with the initial condition for the suspended sediment concentration given by 0 0,t M trMr (C-5) and the boundary condition at the free surface 0,0rMtr (C-6) A form of the sediment diffusivity defines properties of the model. The following functions can be applied: 1. No diffusion 0sr 2. Constant diffusivity s sr 3. Linear diffusivity sruDr4 4. Parabolic diffusivity sruDrrD4 where 4 is von Karman constant which is equal to 0.4, and *u is the friction velocity. The distribution of the sediment sizes at time 0 t needs to be specified. The normal distribution ( 2-52 ) with mean and standard deviation may be used as the initial.

PAGE 151

132 The settling velocity for each sediment size was approximated by Gibbs formula [ Gibbs et al. 1971] 2239(1)(0.0038690.02480) 0.0116070.07440s g dsd w d,, (C-7) where d is the sediment size in centimeters, s w is the settling velocity in cm/s, is the kinematic viscosity of fluid in cm2/s, g is the acceleration of gravity in cm/s2, and s is the specific gravity of sediments. Numerical Solution A numerical solution of equation C-4 was obtained on a uniform grid with CrankNicolson 2nd order scheme [ Fletcher 1991, p.228] (indices “ ” were omitted) 111 1111 111 1111 221 222 22 1 0 2nnnnnn iisiiii s i nnnnnn iiiiii s iMMdMMMM w tdrrr MMMMMM rr (C-8) The implicit approximations lead to the tridiagonal matrix form, which is solved with efficient Thomas algorithm, 111 11,2,nnn iiiiiiiaMbMcMfiN (C-9) with the coefficients 00 a 01 b 00 c 00 f (C-10a) 21 42s ifls i id tt aC rdrr & (C-10b) 21is it b r (C-10c) 21 42s ifls i id tt cC rdrr & (C-10d)

PAGE 152

133 11 11 21 4 2 2nnn s iiflii i nnn siii id t fMCMM rdr t MMM r & (C-10e) 12fl NC a 11 2fl NC b 10Nc 1112fl nnn NNNNC fMMM (C-10f) where flC is Courant number equal to flsCwtr N is the number of grid cells, and n and i are the running indexes. There are no constrains for the value of flC required for stability of the numerical scheme. But since the settling velocity depends on the sediment size, steps in time and vertical direction were chosen to satisfy flC=1 for the coarsest fraction. For other sediment sizes Courant number are smaller. Vertical Sediment Size Distribution Four test cases were calculated to illustrate how sediment size distribution parameters change if different types of diffusion applied. The same initial parameters were used in all cases. A domain size, D was set to be equal 1.5m. Other parameters are gathered into Table C-1 Table C-1. Advection-diffusion model test parameters. Diffusion type 0 M r ,g/l Case 1 no diffusion s = 0m2/s 1.0 2.260.24 Case 2 constant s = 0.001m2/s 1.0 2.260.24 Case 3 linear *u = 0.01m/s 1.0 2.260.24 Case 4 parabolic *u = 0.01m/s 1.0 2.260.24 The results of the calculations are shown in Figures C-1 – C-4 Although the sediment sizes were no longer normally distributed at any time t >0, their distribution can be approximated by the normal distribution. The parameters of the distribution can be estimated by fitting the normal probability curve to one calculated by C-3 The approximated

PAGE 153

134 parameters and are shown on plots “e” and “f” in Figures C-1 – C-4 as functions of a distance from the free surface. The plots show that the mean sediment size (in mm) decreases with time and distance, as well as the standard deviation. In the case of zero diffusivity, the rate of that change is higher than for other types. For the linear diffusivity case, the rate of diffusion is higher towards the free surface. This reduces settling speed of the sediments and lowers the rate of change in sediment sorting. All the test cases show that the sorting is not uniform with vertical elevation, but if the diffusivity is higher (as in linear case) than more mixing occurs and the sorting become more uniform. For the linear diffusivity case, the sorting decreases to approximately 80% of the initial value, while for the zero diffusivity, it decreases down to approximately 27% of the initial value in 90 seconds. The sorting is a key parameter for the correct inversion of sediment size and concentration from the ABS measurements. It is usually (see Chapter 2 ) assumed to be constant, though it is possible to use an average vertical profile of the standard deviations of sediment sizes to account for the changes is sorting. The linear and parabolic diffusivities approximate the mixing in the wave boundary layer more accurately [e.g., Puleo et al. 2004]. Therefore, by applying these types of diffusivity a character of changes in the sediment sorting with time and distance in a wave boundary layer can be approximated.

PAGE 154

135 0 0.05 0.1 FractionSize distribution at Z = 0.25 m a). 0 0.05 0.1 0.15 0.2 FractionSize distribution at Z = 0.50 m b). 1 1.5 2 2.5 3 3.5 0 0.05 0.1 0.15 0.2 Sediment size, FractionSize distribution at Z = 1.00 m c). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Sediment concentrations at given times, g/l d). 0 10 20 30 40 50 60 70 80 90 Time, s 2 2.5 3 e). Mean of sediment size distribution, 0 0.5 1 1.5 0.1 0.2 0.3 0.4 f). STD of sediment size distribution, !Distance from the top, m Figure C-1. Case 1—zero diffusivity: sediment size distribution at distance a) 0.25, b) 0.5, and c) 1m; d) mass concentration; e) approximate mean, and f) standard deviation. 0 0.02 0.04 0.06 FractionSize distribution at Z = 0.25 m a). 0 0.02 0.04 0.06 FractionSize distribution at Z = 0.50 m b). 1 1.5 2 2.5 3 3.5 0 0.02 0.04 0.06 Sediment size, !FractionSize distribution at Z = 1.00 m c). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Sediment concentrations at given times, g/l d). 0 10 20 30 40 50 60 70 80 90 Time, s 2 2.5 3 e). Mean of sediment size distribution, 0 0.5 1 1.5 0.16 0.18 0.2 0.22 0.24 f). STD of sediment size distribution, !Distance from the top, m Figure C-2. Case 2—constant diffusivity: a–f) same as in Figure C-1

PAGE 155

136 0 0.02 0.04 FractionSize distribution at Z = 0.25 m a). 0 0.02 0.04 FractionSize distribution at Z = 0.50 m b). 1 1.5 2 2.5 3 3.5 0 0.02 0.04 Sediment size, FractionSize distribution at Z = 1.00 m c). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Sediment concentrations at given times, g/l d). 0 10 20 30 40 50 60 70 80 90 Time, s 2.2 2.4 2.6 e). Mean of sediment size distribution, 0 0.5 1 1.5 0.2 0.22 0.24 0.26 f). STD of sediment size distribution, !Distance from the top, m Figure C-3. Case 3—linear diffusivity: a–f) same as in Figure C-1 0 0.02 0.04 0.06 FractionSize distribution at Z = 0.25 m a). 0 0.02 0.04 0.06 FractionSize distribution at Z = 0.50 m b). 1 1.5 2 2.5 3 3.5 0 0.02 0.04 0.06 Sediment size, !FractionSize distribution at Z = 1.00 m c). 0 0.2 0.4 0.6 0.8 1 1.2 1.4 Sediment concentrations at given times, g/l d). 0 10 20 30 40 50 60 70 80 90 Time, s 2 2.5 3 e). Mean of sediment size distribution, 0 0.5 1 1.5 0.16 0.18 0.2 0.22 0.24 f). STD of sediment size distribution, !Distance from the top, m Figure C-4. Case 4—parabolic diffusivity: a–f) same as in Figure C-1

PAGE 156

137 APPENDIX D RESULTS FOR “FUN-PUMP” TESTS The following plots show the results of inversion of suspended sediment concentration and mean size profiles for “fun-pump” (see Chapter 3 ) tests for initial maximum concentration levels of 1.2, 1.6, 2.0, and 2.4g/l. 0 1 2 3 a). Errors: mean 41.0%; ABn 28.8% 0 secConc., g/l 1.08 MHz 2.07 MHz 4.70 MHz Mean Model Mean (ABn) 0 0.5 1 1.5 b). Errors: mean 38.3%; ABn 11.5% 10 secConc., g/l 0 0.5 1 c). Errors: mean 45.7%; ABn 17.0% 20 secConc., g/l 0 0.2 0.4 0.6 0.8 d). Errors: mean 49.0%; ABn 27.8% 30 secConc., g/l 0 0.2 0.4 0.6 0.8 e). Errors: mean 58.4%; ABn 29.5% 40 secConc., g/l 0 0.2 0.4 0.6 0.8 f). Errors: mean 72.3%; ABn 40.0% 50 secConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 g). Errors: mean 79.9%; ABn 46.7% 60 sec Range, mConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 h). Errors: mean 74.6%; ABn 64.4% 70 sec Range, mConc., g/l Figure D-1. “Fun-pump” test results for concentration profiles with maximum concentration of 1.2g/l: a–h) same as for 3-33

PAGE 157

138 a). Error 1.1 0 sec! 1 1.5 2 2.5 3 3.5 Invert Model Sand 1 Sand 2 b). Error 0.8 10 sec! 1 1.5 2 2.5 3 3.5 c). Error 0.7 20 sec! 1 1.5 2 2.5 3 3.5 d). Error 0.7 30 sec! 1 1.5 2 2.5 3 3.5 e). Error 0.6 40 sec! 1 1.5 2 2.5 3 3.5 f). Error 0.5 50 sec! 1 1.5 2 2.5 3 3.5 g). Error 0.6 60 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 h). Error 0.7 70 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 Figure D-2. “Fun-pump” test results for mean sediment size profiles with maximum concentration of 1.2g/l: a–h) same as for 3-36

PAGE 158

139 0 1 2 3 4 a). Errors: mean 34.9%; ABn 37.3% 0 secConc., g/l 1.08 MHz 2.07 MHz 4.70 MHz Mean Model Mean (ABn) 0 0.5 1 1.5 2 b). Errors: mean 27.2%; ABn 6.7% 10 secConc., g/l 0 0.5 1 1.5 c). Errors: mean 54.2%; ABn 20.5% 20 secConc., g/l 0 0.5 1 d). Errors: mean 62.9%; ABn 27.8% 30 secConc., g/l 0 0.2 0.4 0.6 0.8 e). Errors: mean 66.3%; ABn 26.7% 40 secConc., g/l 0 0.2 0.4 0.6 0.8 f). Errors: mean 74.2%; ABn 35.1% 50 secConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 g). Errors: mean 81.5%; ABn 46.0% 60 sec Range, mConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 h). Errors: mean 91.8%; ABn 58.1% 70 sec Range, mConc., g/l Figure D-3. “Fun-pump” test results for concentration profiles with maximum concentration of 1.6g/l: a–h) same as for 3-33

PAGE 159

140 a). Error 1.2 0 sec! 1 1.5 2 2.5 3 3.5 Invert Model Sand 1 Sand 2 b). Error 0.8 10 sec! 1 1.5 2 2.5 3 3.5 c). Error 0.6 20 sec! 1 1.5 2 2.5 3 3.5 d). Error 0.5 30 sec! 1 1.5 2 2.5 3 3.5 e). Error 0.4 40 sec! 1 1.5 2 2.5 3 3.5 f). Error 0.4 50 sec! 1 1.5 2 2.5 3 3.5 g). Error 0.4 60 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 h). Error 0.4 70 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 Figure D-4. “Fun-pump” test results for mean sediment size profiles with maximum concentration of 1.6g/l: a–h) same as for 3-36

PAGE 160

141 0 2 4 6 a). Errors: mean 54.2%; ABn 51.5% 0 secConc., g/l 1.08 MHz 2.07 MHz 4.70 MHz Mean Model Mean (ABn) 0 0.5 1 1.5 2 b). Errors: mean 39.4%; ABn 21.8% 10 secConc., g/l 0 0.5 1 1.5 2 c). Errors: mean 53.3%; ABn 19.5% 20 secConc., g/l 0 0.5 1 1.5 d). Errors: mean 61.4%; ABn 29.4% 30 secConc., g/l 0 0.5 1 e). Errors: mean 62.9%; ABn 28.7% 40 secConc., g/l 0 0.2 0.4 0.6 0.8 f). Errors: mean 72.0%; ABn 32.9% 50 secConc., g/l 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 g). Errors: mean 78.6%; ABn 49.6% 60 sec Range, mConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 h). Errors: mean 85.5%; ABn 52.2% 70 sec Range, mConc., g/l Figure D-5. “Fun-pump” test results for concentration profiles with maximum concentration of 2.0g/l: a–h) same as for 3-33

PAGE 161

142 a). Error 1.3 0 sec! 1 1.5 2 2.5 3 3.5 Invert Model Sand 1 Sand 2 b). Error 0.9 10 sec! 1 1.5 2 2.5 3 3.5 c). Error 0.8 20 sec! 1 1.5 2 2.5 3 3.5 d). Error 0.7 30 sec! 1 1.5 2 2.5 3 3.5 e). Error 0.6 40 sec! 1 1.5 2 2.5 3 3.5 f). Error 0.5 50 sec! 1 1.5 2 2.5 3 3.5 g). Error 0.4 60 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 h). Error 0.4 70 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 Figure D-6. “Fun-pump” test results for mean sediment size profiles with maximum concentration of 2.0g/l: a–h) same as for 3-36

PAGE 162

143 0 2 4 6 a). Errors: mean 56.6%; ABn 57.8% 0 secConc., g/l 1.08 MHz 2.07 MHz 4.70 MHz Mean Model Mean (ABn) 0 1 2 3 b). Errors: mean 34.5%; ABn 18.9% 10 secConc., g/l 0 0.5 1 1.5 2 c). Errors: mean 59.6%; ABn 29.8% 20 secConc., g/l 0 0.5 1 1.5 d). Errors: mean 66.5%; ABn 36.4% 30 secConc., g/l 0 0.5 1 1.5 e). Errors: mean 69.0%; ABn 28.9% 40 secConc., g/l 0 0.2 0.4 0.6 0.8 f). Errors: mean 76.3%; ABn 33.4% 50 secConc., g/l 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 g). Errors: mean 83.0%; ABn 41.5% 60 sec Range, mConc., g/l 0 0.2 0.4 0.6 0.8 0 0.1 0.2 0.3 0.4 h). Errors: mean 91.1%; ABn 55.4% 70 sec Range, mConc., g/l Figure D-7. “Fun-pump” test results for concentration profiles with maximum concentration of 2.4g/l: a–h) same as for 3-33

PAGE 163

144 a). Error 1.5 0 sec! 1 1.5 2 2.5 3 3.5 Invert Model Sand 1 Sand 2 b). Error 1.0 10 sec! 1 1.5 2 2.5 3 3.5 c). Error 0.8 20 sec! 1 1.5 2 2.5 3 3.5 d). Error 0.7 30 sec! 1 1.5 2 2.5 3 3.5 e). Error 0.5 40 sec! 1 1.5 2 2.5 3 3.5 f). Error 0.5 50 sec! 1 1.5 2 2.5 3 3.5 g). Error 0.4 60 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 h). Error 0.4 70 sec Range, m! 0 0.2 0.4 0.6 0.8 1 1.5 2 2.5 3 3.5 Figure D-8. “Fun-pump” test results for mean sediment size profiles with maximum concentration of 2.4g/l: a–h) same as for 3-36

PAGE 164

145 LIST OF REFERENCES Abramowitz, M., and I.A. Stegun, Handbook of mathematical functions: with formulas, graphs, and mathematical tables xiv, 1046 pp., Dover Publications, New York, 1965. Crawford, A.M., and A.E. Hay, Determining suspended sand size and concentration from multifrequency acoustic backscatter, Journal of the Acoustical Society of America 94 (6), 3312-3324, 1993. Dohmen-Janssen, C.M., and D.M. Hanes, Sheet flow dynamics under monochromatic nonbreaking waves, Journal of Geophysical Research-Oceans 107 (C10), 13-1-1321, 2002. Downing, A., P.D. Thorne, and C.E. Vincent, Backscattering from a suspension in the near field of a piston transducer, Journal of the Acoustical Society of America 97 (3), 1614-1620, 1995. Faran, J.J., Sound scattering by solid cylinders and spheres, Journal of the Acoustical Society of America 23 (4), 405-418, 1951. Fisher, F.H., and V.P. Simmons, Sound-absorption in sea-water, Journal of the Acoustical Society of America 62 (3), 558-564, 1977. Fletcher, C.A.J., Computational techniques for fluid dynamics Vol. 1, 402 pp., SpringerVerlag, Berlin; New York, 1991. Gibbs, R.J., M.D. Matthews, and D.A. Link, Relationship between sphere size and settling velocity, Journal of Sedimentary Petrology 41 (1), 7-18, 1971. Hanes, D.M., Suspension of sand due to wave groups, Journal of Geophysical ResearchOceans 96 (C5), 8911-8915, 1991. Hanes, D.M., V. Alymov, Y.S. Chang, and C. Jette, Wave-formed sand ripples at Duck, North Carolina, Journal of Geophysical Research-Oceans 106 (C10), 2257522592, 2001. Hanes, D.M., and C.E. Vincent, Detailed dynamics of neashore suspended sediment, in Proceedings Coastal Sediments '87 pp. 285-299, New Orleans, LA, 1987. Hanes, D.M., C.E. Vincent, D.A. Huntley, and T.L. Clarke, Acoustic measurements of suspended sand concentration in the C2S2 experiment at Stanhope-Lane, PrinceEdward-Island, Marine Geology 81 (1-4), 185-196, 1988.

PAGE 165

146 Hay, A.E., Sound scattering from a particle-laden, turbulent jet, Journal of the Acoustical Society of America 90 (4), 2055-2074, 1991. Hay, A.E., and D.G. Mercer, On the theory of sound scattering and viscous absorption in aqueous suspensions at medium and short wavelengths, Journal of the Acoustical Society of America 78 (5), 1761-1771, 1985. Hay, A.E., and A.S. Schaafsma, Resonance scattering in suspensions, Journal of the Acoustical Society of America 85 (3), 1124-1138, 1989. Hay, A.E., and J.Y. Sheng, Vertical profiles of suspended sand concentration and size from multifrequency acoustic backscatter, Journal of Geophysical ResearchOceans 97 (C10), 15661-15677, 1992. Jansen, R.H., An ultrasonic doppler scatterometer for measuring suspended sand transport, in Proceedings Ultrasonic International 79 (UI 79) pp. 366-369, Graz, Austria, 1979. Johnson, R.K., Sound scattering from a fluid sphere revisited, Journal of the Acoustical Society of America 61 (2), 375-377, 1977. Knuth, D.E., The art of computer programming. Fundamental Algorithms Vol. 1, 650 pp., Addison-Wesley, Reading, Mass., 1997. Lee, T.H., Acoustic measurement and modeling of the vertical distribution of suspended sediment driven by waves and currents, Ph.D. dissertation, University of Florida, Gainesville, FL, 1994. Lee, T.H., and D.M. Hanes, Direct inversion method to measure the concentration profile of suspended particles using backscattered sound, Journal of Geophysical Research-Oceans 100 (C2), 2649-2657, 1995. Libicki, C., K.W. Bedford, and J.F. Lynch, The interpretation and evaluation of a 3-Mhz acoustic backscatter device for measuring benthic boundary-layer sediment dynamics, Journal of the Acoustical Society of America 85 (4), 1501-1511, 1989. Lockwood, J.C., and J.G. Willette, High-speed method for computing exact solution for pressure variations in nearfield of a baffled piston, Journal of the Acoustical Society of America 53 (3), 735-741, 1973. Ludwig, K.A., and D.M. Hanes, A laboratory evaluation of optical backscatterance suspended-solids sensors exposed to sand-mud mixtures, Marine Geology 94 (1-2), 173-179, 1990. Medwin, H., and C.S. Clay, Fundamentals of acoustical oceanography 712 pp., Academic Press, Boston, 1997. Morse, P.M., and K.U. Ingard, Theoretical acoustics xix, 927 pp., McGraw-Hill, New York, 1968.

PAGE 166

147 Mouraenko, O.A., and D.M. Hanes, SandyDuck'97: University of Florida data set on small-scale sediment processes, from FRF Data Server http://dksrv.usace.army.mil/SD97/UFDH retrieved November 3, 2004. Neubauer, W.G., R.H. Vogt, and Dragonet.Lr, Acoustic reflection from elastic spheres. 1. Steady-state signals, Journal of the Acoustical Society of America 55 (6), 11231129, 1974. Nielsen, P., Coastal bottom boundary layers and sediment transport xxiii, 324 pp., World Scientific, Singapore; River Edge, N.J., 1992. Puleo, J.A., O. Mouraenko, and D.M. Hanes, One-dimensional wave bottom boundary layer model comparison: specific eddy viscosity and turbulence closure models, Journal of Waterway, Port, Coastal, and Ocean Engineering 130 (6), 322-325, 2004. Richards, S.D., The effect of temperature, pressure, and salinity on sound attenuation in turbid seawater, Journal of the Acoustical Society of America 103 (1), 205-211, 1998. Schaafsma, A.S., Acoustic measurement of concentration and size of sediment particle suspensions, in Proceedings Ultrasonics International'89 pp. 338-393, Madrid, 1989. Schat, J., Multifrequency acoustic measurement of concentration and grain size of suspended sand in water, Journal of the Acoustical Society of America 101 (1), 209217, 1997. Sheng, J.Y., and A.E. Hay, An examination of the spherical scatterer approximation in aqueous suspensions of sand, Journal of the Acoustical Society of America 83 (2), 598-610, 1988. Sheng, J.Y., and A.E. Hay, Spherical wave backscatter from straight cylinders---thin-wire standard targets, Journal of the Acoustical Society of America 94 (5), 2756-2765, 1993. Sleath, J.F.A., Sea bed mechanics 335 pp., Wiley, New York, 1984. Smerdon, A.M., J.M. Rees, and C.E. Vincent, An acoustic backscatter instrument to measure near-bed sediment processes, Technical report Aquatec Electronics Ltd, Hampshire, UK, 1998. Tamura, T., and D.M. Hanes, Laboratory calibration of a 3 megahertz acoustic concentration meter to measure suspended sand concentration, Technical report TR 86-004 Rosenftiel School of Marine and Atmospheric science, Miami, FL, 1986.

PAGE 167

148 Thorne, P.D., and M.J. Buckingham, Measurements of scattering by suspensions of irregularly shaped sand particles and comparison with a single parameter modified sphere model, Journal of the Acoustical Society of America 116 (5), 2876-2889, 2004. Thorne, P.D., and S.C. Campbell, Backscattering by a suspension of spheres, Journal of the Acoustical Society of America 92 (2), 978-986, 1992. Thorne, P.D., and D.M. Hanes, A review of acoustic measurement of small-scale sediment processes, Continental Shelf Research 22 (4), 603-632, 2002. Thorne, P.D., and P.J. Hardcastle, Acoustic measurements of suspended sediments in turbulent currents and comparison with in-situ samples, Journal of the Acoustical Society of America 101 (5), 2603-2614, 1997. Thorne, P.D., P.J. Hardcastle, and R.L. Soulsby, Analysis of acoustic measurements of suspended sediments, Journal of Geophysical Research-Oceans 98 (C1), 899-910, 1993a. Thorne, P.D., G.P. Holdaway, and P.J. Hardcastle, Constraining acoustic backscatter estimates of suspended sediment concentration profiles using the bed echo, Journal of the Acoustical Society of America 98 (4), 2280-2288, 1995. Thorne, P.D., C. Manley, and J. Brimelow, Measurements of the form function and total scattering cross-section for a suspension of spheres, Journal of the Acoustical Society of America 93 (1), 243-248, 1993b. Thosteson, E.D., Developmental and field application of a littoral processes monitoring system for examination of the relevant time scales of sediment suspension processes, Ph.D. dissertation, University of Florida, Gainesville, FL, 1997. Thosteson, E.D., and D.M. Hanes, A simplified method for determining sediment size and concentration from multiple frequency acoustic backscatter measurements, Journal of the Acoustical Society of America 104 (2), 820-830, 1998. Traykovski, P., A.E. Hay, J.D. Irish, and J.F. Lynch, Geometry, migration, and evolution of wave orbital ripples at LEO-15, Journal of Geophysical Research-Oceans 104 (C1), 1505-1524, 1999. Vincent, C.E., and M.O. Green, Field-measurements of the suspended sand concentration profiles and fluxes and of the resuspension coefficient0 over a rippled bed, Journal of Geophysical Research-Oceans 95 (C7), 11591-11601, 1990. Vincent, C.E., D.M. Hanes, C.M. Dohmen-Janssen, G. Klopman, S.R. McLean, and J.S. Ribberink, Suspension by regular and group waves over bedforms in a large wave flume (SISTEX 99), in, Coastal Dynamics’01 pp. 303–312, Lund, Sweden, 2001. Weisstein, E.W., Catalan Number, from MathWorld--A Wolfram Web Resource http://mathworld.wolfram.com/CatalanNumber.html retrieved October 10, 2004.

PAGE 168

149 Wren, D.G., B.D. Barkdoll, R.A. Kuhnle, and R.W. Derrow, Field techniques for suspended-sediment measurement, Journal of Hydraulic Engineering-Asce 126 (2), 97-104, 2000. Young, R.A., J.T. Merrill, T.L. Clarke, and J.R. Proni, Acoustic profiling of suspended sediments in the marine bottom boundary layer, Geophysical Research Letters 9 (3), 175-178, 1982.

PAGE 169

150 BIOGRAPHICAL SKETCH The author was born the second son of Nadezhda and Alexander Mouraenko in 1974 in the small town of Myaundzha in the Magadan region of northeast of Russia. From early on he spent a great deal of time with his father, a professional driver, from whom he inherited a passion for mechanics and learned to respect perfection at work. At the age of thirteen he and his family moved to Barnaul, Russia, where in 1991 he finished high school with a silver medal and entered the Mathematical Department of the Altai State University. In 1996 he graduated with honors and chose to continue his education in applied sciences by entering the graduate school at the Institute for Water and Environmental Problems. In 1999 he continued his study at the University of Florida, and in 2001 he received his master’s degree in coastal and oceanographic engineering. In his spare time, he enjoys long distance running and competing in various races including marathons.


Permanent Link: http://ufdc.ufl.edu/UFE0008389/00001

Material Information

Title: Acoustic measurement techniques for suspended sediments and bedforms
Physical Description: Mixed Material
Language: English
Creator: Mouraenko, Oleg A. ( Dissertant )
Kennedy, Andrew B. ( Thesis advisor )
Dean, Robert G. ( Reviewer )
Thieke, Robert J. ( Reviewer )
Slinn, Donald N. ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2004
Copyright Date: 2004

Subjects

Subjects / Keywords: Civil and Coastal Engineering thesis, Ph.D
Dissertations, Academic -- UF -- Civil and Coastal Engineering

Notes

Abstract: The acoustic backscatter system provides a nonintrusive method for measuring of profiles of suspended sediment size and concentrations. Therefore, an inversion model is required to convert the recorded series of intensities into parameters of sediment suspension. The measurement precision highly depends on the choice of the inversion model. In this work the high-pass backscattering model is modified to allow accurate inversion for both sediment sizes and concentrations. The modified correction function with a single parameter is introduced for the form function and normalized cross-section. The parameters can be found from the calibration of the system in the chamber with generated homogeneous suspension. To account for the difference in shape and mineralogy, a set of two parameters can be determined for a particular sediment. Their use greatly improves the accuracy of mean sediment size measurements obtained with the three-frequency acoustic backscatter system. The inversion model was applied to data obtained in the chamber with settling mixture of two sediment samples. The results were compared to the predictions by the 1D advection-diffusion model. Good agreement was shown for concentrations and sediment size measurements, with the error for sediment size approximation lying within the standard deviation of size distribution. The accuracy of the concentration measurements was different among the transducer frequencies, and was found to depend on the sensitivity of the transducer to the particular sediment size. The sediment sizes were estimated by minimizing the variance between concentrations, determined from different frequencies, and found to depend on the accuracy of estimated concentrations. For correction of recorded voltage data with applied time-varying gain, the system functions were found to be better approximated by second order polynomials. A multiple reflection model was developed based on the multiple layer approximation. The numerical simulation with the initial uniform concentration profile of 3g/l shows that the effect in intensities from the multiple reflections can be up to 14%. Also, a bottom echo removal algorithm was developed in order to eliminate the effect of high intensity scattering on the concentration measurements close to the bottom.
Subject: acoustic, backscatter, bedform, bottom, concentration, flux, measurement, profile, sediment, size, suspended, system, technique, transport
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 169 pages.
General Note: Includes vita.
Thesis: Thesis (Ph.D.)--University of Florida, 2004.
Bibliography: Includes bibliographical references.
General Note: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0008389:00001

Permanent Link: http://ufdc.ufl.edu/UFE0008389/00001

Material Information

Title: Acoustic measurement techniques for suspended sediments and bedforms
Physical Description: Mixed Material
Language: English
Creator: Mouraenko, Oleg A. ( Dissertant )
Kennedy, Andrew B. ( Thesis advisor )
Dean, Robert G. ( Reviewer )
Thieke, Robert J. ( Reviewer )
Slinn, Donald N. ( Reviewer )
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2004
Copyright Date: 2004

Subjects

Subjects / Keywords: Civil and Coastal Engineering thesis, Ph.D
Dissertations, Academic -- UF -- Civil and Coastal Engineering

Notes

Abstract: The acoustic backscatter system provides a nonintrusive method for measuring of profiles of suspended sediment size and concentrations. Therefore, an inversion model is required to convert the recorded series of intensities into parameters of sediment suspension. The measurement precision highly depends on the choice of the inversion model. In this work the high-pass backscattering model is modified to allow accurate inversion for both sediment sizes and concentrations. The modified correction function with a single parameter is introduced for the form function and normalized cross-section. The parameters can be found from the calibration of the system in the chamber with generated homogeneous suspension. To account for the difference in shape and mineralogy, a set of two parameters can be determined for a particular sediment. Their use greatly improves the accuracy of mean sediment size measurements obtained with the three-frequency acoustic backscatter system. The inversion model was applied to data obtained in the chamber with settling mixture of two sediment samples. The results were compared to the predictions by the 1D advection-diffusion model. Good agreement was shown for concentrations and sediment size measurements, with the error for sediment size approximation lying within the standard deviation of size distribution. The accuracy of the concentration measurements was different among the transducer frequencies, and was found to depend on the sensitivity of the transducer to the particular sediment size. The sediment sizes were estimated by minimizing the variance between concentrations, determined from different frequencies, and found to depend on the accuracy of estimated concentrations. For correction of recorded voltage data with applied time-varying gain, the system functions were found to be better approximated by second order polynomials. A multiple reflection model was developed based on the multiple layer approximation. The numerical simulation with the initial uniform concentration profile of 3g/l shows that the effect in intensities from the multiple reflections can be up to 14%. Also, a bottom echo removal algorithm was developed in order to eliminate the effect of high intensity scattering on the concentration measurements close to the bottom.
Subject: acoustic, backscatter, bedform, bottom, concentration, flux, measurement, profile, sediment, size, suspended, system, technique, transport
General Note: Title from title page of source document.
General Note: Document formatted into pages; contains 169 pages.
General Note: Includes vita.
Thesis: Thesis (Ph.D.)--University of Florida, 2004.
Bibliography: Includes bibliographical references.
General Note: Text (Electronic thesis) in PDF format.

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0008389:00001


This item has the following downloads:


Full Text











ACOUSTIC MEASUREMENT TECHNIQUES
FOR SUSPENDED SEDIMENTS AND BEDFORMS















By

OLEG A. MOURAENKO


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2004































To my family.















ACKNOWLEDGMENTS


I wish to express my appreciation to my chairman, Dr. Andrew B. Kennedy, for the

freedom granted to me while pursuing my research interests, and for the guidance and

support. I also wish to thank the members of my graduate committee, Dr. Robert G.

Dean, Dr. Robert J. Thieke, Dr. Donald N. Slinn, for their intelligence and comprehen-

sive supervision. My special thanks go to Dr. Renwei Mei, the external member of the

committee, for his interest and involvement. I thank Dr. Peter D. Thorne for the guidance

and great discussions on various aspects of acoustic measurements. I also thank Dr.

Christopher R. Sherwood for providing the acoustic backscatter system and technical

help.

I thank my father, Alexander Mouraenko, for his patience and love.

My special thanks go to the staff at the Coastal Engineering Lab, especially to Sid-

ney Schofield, Jim Joiner, Victor Adams, and Chuck Broward, for their help, support,

and technical guidance, without which this research would not have been possible.

I thank Jamie MacMahan and Justin Davis for their help and endless support, with

special thanks to Brian Barr for his exceptional knowledge, great vision of life and nu-

merous discussions.

This work was supported by the ONR Coastal Geosciences program and the Na-

tional Science Foundation through the National Ocean Partnership Program (NOPP).















TABLE OF CONTENTS

Page

A C K N O W L E D G M E N T S ................................................................................................. iii

LIST OF TABLES ............................................ .................................... vi

LIST OF FIGURES ......................................... .................................... vii

K E Y T O SY M B O L S ....................................................................................... xv

A B S T R A C T ................................................................................................................... x v iii

CHAPTER

1 IN T R O D U C T IO N .................................................. .. ....................................1.. .. ... 1

2 MULTIPLE REFLECTION SOLUTION FOR ACOUSTIC BACKSCATTER ........7

A acoustic B ackscatter E quation ................................... ..................... 7
Inverse Problem for Sedim ent Concentration........................................ .................. 11
M multiple L ayer A pproxim ation ................................................................................ 11
Numerical Realization of M multiple Reflections.............................. ............... 15
Determination of Sediment Size and Concentration ......................................17
Number of Paths ......................................................... .......... .. 18
Effect of M multiple Reflections .............. ............. .... ................ 19
Special Parameters for Acoustic Backscatter Equation.........................................23
Form Function and Normalized Cross-Section for Elastic Spheres.................23
H igh-Pass M odel for Quartz Sand.................................................. ................ 24
M odified M odel for Elastic Spheres ..................................................... 25
Modified High-Pass Model for Natural Sediments........................................26
Sound A ttenuation C oefficient....................................................... ................ 27
Lognormal and Normal Distributions of Sediment Sizes............................... 28
Averaged Form Factor and Normalized Cross-Section Terms ........................30
Circular Transducer Directivity and Half Intensity Beam Width ....................32
N ear-F ield C orrection F actor ..............................................................................34

3 CALIBRATION OF THE ACOUSTIC BACKSCATTER SYSTEM....................36

C alib ratio n C h am b er ................................................................................................... 3 8
A B S T ests in the C ham ber ......................................... ........................ ................ 40
T ests in A ir and Still W ater ............................................................ ................ 40









Homogeneity of Sediment Suspension and Effect of Chamber Walls.............41
R e su lts .............. .. ........................................................................................... . 4 8
Estimation of System Function..................... ................ 48
Approxim ation by 2nd Order Polynomial...................................... ................ 49
C alibration of the A B S 1 ................................................................. ................ 50
C alibration of the A B S2 ................................... ..... ................ 53
Sensitivity Analysis For Estimation of System Function......................................55
D ependence on Sedim ent Sorting .................................................. ................ 59
D ependence on Transducer R adius ................................................ ................ 60
D term nation of P aram eter f f .................................................................................60
Estim ation of System Function for ABS1 ...................................... ................ 62
Analysis of Relationship between z and f ............. ................................... 67
"F u n -P u m p T e sts ............................................................... .................. .................... 7 0
Sam ple Sedim ent Size D istribution.................................................................. 71
Prediction of Concentration and Sediment Size Distribution...........................73
P aram eters of Inversion .................................................................. ................ 74
Results of Inversion for Concentration........................................... ................ 76
Results of Inversion for Sedim ent Size ............... .................................... 81

4 MEASUREMENTS OF TARGET POSITION AND BOTTOM ECHO REMOVAL
A L H O R IT H M ....................................................... .. .............................................. 8 5

A Sim plified B ottom Echo M odel ......................................................... ................ 85
Tests w ith a Steel W ire .. ................................................................................ 88
T ests w ith a W ater-Sedim ent Jet ........................................................... ............... 96
R esult of B ottom E cho R em oval ........................................................... ................ 98

5 CONCLUSIONS ...................................... ........ ................... 100

APPENDIX

A CA LIBR A TION TE ST RE SU LTS ..........................................................................104

B EMPIRICAL MODEL FOR ACOUSTIC BACKSCATTER................................123

C ID ADVECTION-DIFFUSION MODEL ............... .............. ..................... 130

N um erical Solution ............................................................................... 132
Vertical Sedim ent Size Distribution ...... ........ .. ...... ..................... 133

D RESULTS FOR "FUN-PUMP" TESTS........................................137

LIST O F REFEREN CE S .. .................................................................... ............... 145

BIOGRAPH ICAL SKETCH .................. .............................................................. 150















LIST OF TABLES


Table Page

2-1: F irst 24 C atalan num bers......................................... ......................... ............... 19

2-2: Beam width for a circular piston transducer (a,= 5 mm and c= 1500 m/s). ..........34

2-3: Near-field range r for a circular piston source (a,=5mm and c =1500m/s). .........35

3-1: Locations and concentration levels in tests for 1.08MHz transducer....................42

3-2: Locations and concentration levels in tests for 2.07MHz transducer....................43

3-3: Locations and concentration levels in tests for 4.70MHz transducer....................43

3 -4 : S ed im en t sizes .......................................................................................................... 6 3

3-5: Estimation of parameters Ifz and ff for different sediment sizes.......................64

3-6: V alues of x = ka for test cases ........................................................... ................ 64

3-7: Concentration levels, mean and standard deviation of the mixture for "fun-pump"
te sts ............................................................... 7 2

3-8: Coefficients of 2nd order polynomial for system functions................................75

4-1: T ests w ith 0.33m m steel w ire .............................................................. ................ 9 1

C- 1: Advection-diffusion model test parameters. .......................................133















LIST OF FIGURES


Figure Page

2-1: Example of the scattered sound within the narrow beam....................................12

2-2: The multilayer approximation for a possible sound pulse traveling path. The time
interval corresponds to rc/2 ............................................................... ............... 13

2-3: Generation of a binary tree for numerical realization of multiple reflections.......... 16

2-4: Difference between intensities modeled with one and three reflections for selected
profiles of concentrations: a) constant profiles, e) linear profiles, b-d) and
f-h) percent difference. ......... ..... .......................................... ...............2 1

2-5: Difference between intensities modeled with one and three reflections for selected
profiles of concentrations: a) profiles of constant concentration with "bump",
e) profiles of constant concentrations with "spike", b-d) and f-h) same as in
F ig u re 2 -4 ............................................................................................................... 2 2

2-6: Form function and normalized cross section for quartz sand (blue) and spheres
(red): a) form function, b) normalized cross-section...........................................25

2-7: Modified y function for selected values of parameter fl ............... ................26

2-8: Circular piston directivity function, D 2 (0) ...................................... ................ 33

2-9: Near-field correction factor, 1/1V(r), for a,=5mm and c=1500m/s....................35

3-1: C alibration cham ber- side view ........................................................ ................ 39

3-2: RMS voltage profiles (solid lines) with scatter intervals (dashed line) from the
measurements in air and still clear water: a) 1.08MHz, b) 2.07MHz, and
c) 4 .7 0 M H z .............................................................................................................. 4 1

3-3: Calibration chamber-top view. The jet system is shown with transducer positions
for the sidew all proxim ity tests ........................................................... ................ 42

3-4: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for
1.08MHz transducer for sidewall proximity tests ..............................................43









3-5: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for
2.07M Hz transducer for sidewall proximity tests. .............................. ................ 44

3-6: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured along centerline with 1.08MHz transducer.........................................45

3-7: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured along centerline with 2.07MHz transducer.........................................45

3-8: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured along center line with 4.70MHz transducer........................................46

3-9: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured near the chamber wall with 1.08MHz transducer ...............................47

3-10: Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured near the chamber wall with 2.07MHz transducer ...............................47

3-11: Estimation of sediment size distribution used for the calibration of ABS1 by:
a) sieve analysis, b) fall velocity ......................................................... ................ 50

3-12: Estimation of the system function for 1.08MHz transducer: a) ratio of measured
and calculated voltage profiles, b) estimated system functions, and c) inverted
sedim ent concentration ........................................................................ ................ 52

3-13: Estimation of the system function for 2.07MHz transducer: a-c) same as in
Figure 3-12. ............. ........................................................................ ....... ..... 52

3-14: Estimation of the system function for 4.70MHz transducer: a-c) same as in
F figure 3-12 ............. ............................................................................... ..... 53

3-15: Estimation of sediment size distribution for sand used in calibration of ABS2 by:
a) sieve analysis, b) fall velocity ......................................................... ................ 54

3-16: Estimation of the system function for IMHz of ABS2: a-c) same as in
F figure 3-12 ............. ............................................................................... ..... 54

3-17: Values of fitted parameters /fi, a, p and q,, with corresponding errors in
inverted concentrations for the 1.08MHz transducer of the ABS1.......................57

3-18: Same as on Figure 3-17 for the 2.07MHz transducer of the ABS1. ......................57

3-19: Same as on Figure 3-17 for the 4.70MHz transducer of the ABS1. ......................58

3-20: Estimation of system function K, for 4.70MHz transducer with corrected
param eter fi : a-c) sam e as in Figure 3-12. ....................................... ................ 59









3-21: Estimation of sediment size distribution for Case 4 used for calibration of ABSI
by: a) sieve analysis, b) fall velocity ................................................... ................ 63

3-22: Estimated system functions and their means for Cases 1-4 with 83 =0.5 (dashed)
and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz transducers of the ABS1. .......65

3-23: Ratios between individual system functions for Case 1-4 and mean system function
with 8z=0.5 (dashed) and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz
transducers of the A B S ...................................................................... ................ 66

3-24: Fitted parameters f#3 and corresponding parameters f, for Cases 1-4 ..............68

3-25: Values of parameters /f, and/ ,, for Cases 1-4 for different frequencies ............68

3-26: Values of j and y(ka,137) as functions of ka with approximated functions
y(ka, f ) for C ases 1-4 (solid lines) .................................................. ................ 69

3-27: Sediment size distribution for Sand 1, Sand 2, and 1:1 mixture .............................72

3-28: Predicted with ID advection-diffusion model profiles of concentration and
sediment size distributions for maximum concentration of 0.8g/l at times 0-70s:
size distributions at a) 30, b) 50, and c) 75cm below the free surface, d) sediment
concentration profiles, e) profiles of mean sediment size in phi units, and f) profiles
of standard deviations in phi units....................................................... ................ 73

3-29: High-pass model for backscattered intensity and its modified version used for "fun-
pump" test: a) form functions, b) normalized cross-sections...............................76

3-30: Inversion by unmodified high-pass model concentration profiles for "fun-pump"
test with maximum concentration of 0.4g/l: mean profiles (thick lines) and
1.08M H z profiles (thin lines) ................................... ...................... ................ 77

3-31: Inversion by unmodified high-pass model concentration profiles for "fun-pump"
test with maximum concentration of 0.8g/l: mean profiles (thick lines) and
2.07M H z profiles (thin lines) ................................... ...................... ................ 78

3-32: Inversion by modified model concentration profiles for "fun-pump" test with
maximum concentration of 0.8g/l: mean profiles (thick lines) and 2.07MHz profiles
(th in lin e s). .............................................................................................................. 7 9

3-33: Comparison of concentration profiles inverted by modified high-pass model with
profiles calculated by the ID advection-diffusion and empirical models (labeled as
"ABn") for the "fun-pump" test with maximum concentration of 0.8g/l at 10
seco n d in terv als. ....................................................................................................... 8 0









3-34: Inverted profiles of mean sediment sizes for "fun-pump" test with maximum
concentration of 0.8g/l by unmodified high-pass model.....................................82

3-35: Inverted by modified high-pass model profiles of mean sediment sizes for "fun-
pump" test with maximum concentration of 0.8g/1.............................................82

3-36: Comparison of inverted profiles of mean sediment sizes by modified high-pass
model with profiles calculated by the ID advection-diffusion model for the "fun-
pump" test with maximum concentration of 0.8g/l at 10-second intervals.............. 83

4-1: A sample of voltage profile with definition of"A"-point ...................................... 87

4-2: Setup for the tests w ith steel w ire........................................................ ............... 89

4-3: Root-mean-square voltage profiles for the tests with 0.33mm steel wire. Tests 1-16
with the wire location (dashed line) and the bottom location (solid line) .............92

4-4: Sam e as in Figure 4-3 for tests 17-31 ................................................ ................ 93

4-5: Root-mean-square voltage profiles with subtracted bottom echo profile for the tests
with 0.33mm steel wire. Tests 1-16 with the wire location (dashed line) and the
bottom location (solid line) ...................................... ....................... ................ 94

4-6: Sam e as in Figure 4-5 for tests 17-31 ................................................ ................ 95

4-7: Root-mean-square voltage profiles (solid) for tests with 2mm steel wire with the
wire locations (dashed) and "ghost wire" reflection locations (dash-dotted) from a)
1.08, b) 2.07, and c) 4.70M H z transducers. ........................................ ................ 96

4-8: Setup for the tests w ith the sedim ent jet............................................. ............... 97

4-9: Root-mean-square voltage profiles (solid) for tests for different water-sediment jet
locations (dashed) from a) 1.08, b) 2.07, and c) 4.70MHz transducers ................97

4-10: Time series of suspended sediment concentrations estimated with a) bottom echo
and b) bottom echo removed; c) free surface elevation. .....................................99

A-1: Case 1: Errors in inverted concentrations from varying of parameters /fz, at, ,,
and ,, for 1.08MHz transducer. ................................ 104

A-2: Case 1: Errors in inverted concentrations from varying of parameters /fz, a, ,,
and ,, for 2.07M Hz transducer. ...... ...... ...... ...................... 105

A-3: Case 1: Errors in inverted concentrations from varying of parameters /fz, a,, ,,
and o, for 4.70M Hz transducer. ...... ...... ...... ...................... 105









A-4: Case 2: Errors in inverted concentrations from varying of parameters /fz, a, p, ,
and ,, for 1.08M H z transducer. ...... .................................. .. ...................... 106

A-5: Case 2: Errors in inverted concentrations from varying of parameters /fz, a, p, ,
and ,, for 2.07M H z transducer. ...... .................................. .. ...................... 106

A-6: Case 2: Errors in inverted concentrations from varying of parameters /fz, a, p, ,
and a,, for 4.70M H z transducer. ...... .................................. .. ...................... 107

A-7: Case 3: Errors in inverted concentrations from varying of parameters f6z, a,, p ,
and a for 1.08M H z transducer. ...... .................................. .. ...................... 107

A-8: Case 3: Errors in inverted concentrations from varying of parameters f6z, a,, p ,
and a,, for 2.07M H z transducer. ...... .................................. .. ...................... 108

A-9: Case 3: Errors in inverted concentrations from varying of parameters f6z, a,, p ,
and ,, for 4.70M H z transducer. ...... .................................. .. ...................... 108

A-10: Case 4: Errors in inverted concentrations from varying of parameters /fz, a, p, ,
and ,, for 1.08M H z transducer. ...... .................................. .. ...................... 109

A-11: Case 4: Errors in inverted concentrations from varying of parameters /fz, a,, ,,
and o, for 2.07M H z transducer. ...... .................................. .. ...................... 109

A-12: Case 4: Errors in inverted concentrations from varying of parameters /fz, a, p, ,
and a,, for 4.70M H z transducer. ...... .................................. .. ...................... 110

A-13: Case 1: Estimation of the system function for the 1.08MHz transducer initial value
of parameter f6z: a-c) same as in Figure 3-12.......................... .................... 110

A-14: Case 1: Estimation of the system function for the 1.08MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12. ........................................111

A-15:Case 1: Estimation of the system function for the 2.07MHz transducer with initial
value of parameter fz: a-c) same as in Figure 3-12 .......................................... 111

A-16: Case 1: Estimation of the system function for the 2.07MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12. .............. ................112









A-17: Case 1: Estimation of the system function for the 4.70MHz transducer with initial
value of parameter fiz: a-c) same as in Figure 3-12. .................... ................ 112

A-18: Case 1: Estimation of the system function for the 4.70MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12 .............. ................ 113

A-19: Case 2: Estimation of the system function for the 1.08MHz transducer with initial
value of parameter fiz: a-c) same as in Figure 3-12. .................... ................ 113

A-20: Case 2: Estimation of the system function for the 1.08MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12 .............. ................ 114

A-21:Case 2: Estimation of the system function for the 2.07MHz transducer with initial
value of parameter fiR: a-c) same as in Figure 3-12. .................... ................ 114

A-22: Case 2: Estimation of the system function for the 2.07MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12 .............. ................ 115

A-23:Case 2: Estimation of the system function for the 4.70MHz transducer with initial
value of parameter fiz: a-c) same as in Figure 3-12. .................... ................ 115

A-24: Case 2: Estimation of the system function for the 4.70MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12 .............. ................ 116

A-25:Case 3: Estimation of the system function for the 1.08MHz transducer with initial
value of parameter fiz: a-c) same as in Figure 3-12. .................... ................ 116

A-26: Case 3: Estimation of the system function for the 1.08MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12 .............. ................ 117

A-27: Case 3: Estimation of the system function for the 2.07MHz transducer with initial
value of parameter fiR: a-c) same as in Figure 3-12. .................... ................ 117

A-28: Case 3: Estimation of the system function for the 2.07MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12 .............. ................ 118

A-29: Case 3: Estimation of the system function for the 4.70MHz transducer with initial
value of parameter fiz: a-c) same as in Figure 3-12. .................... ................ 118

A-30: Case 3: Estimation of the system function for the 4.70MHz transducer with
corrected parameter Jf : a-c) same as in Figure 3-12 .............. ................ 119









A-31:Case 4: Estimation of the system function for the 1.08MHz transducer with initial
value of parameter fi : a-c) same as in Figure 3-12. ...................... ................ 119

A-32: Case 4: Estimation of the system function for the 1.08MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12. ............... ....... ............ 120

A-33:Case 4: Estimation of the system function for the 2.07MHz transducer with initial
value of parameter fi : a-c) same as in Figure 3-12. ...................... ................ 120

A-34:Case 4: Estimation of the system function for the 2.07MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12. ............... ....... ............ 121

A-35:Case 4: Estimation of the system function for the 4.70MHz transducer with initial
value of parameter fi : a-c) same as in Figure 3-12. ...................... ................ 121

A-36: Case 4: Estimation of the system function for the 4.70MHz transducer with
corrected parameter If : a-c) same as in Figure 3-12. ............... ....... ............ 122

B-1: Calibration coefficients for 1.08 MHz transducer: a) measured voltage (dots) and
fitted by curves B-3 and calculated coefficients b) A, c) B, and d) n...............127

B-2: Calibration coefficients for 2.07MHz transducer: a-d) same as in Figure B-1 ....127

B-3: Calibration coefficients for 4.70MHz transducer: a-d) same as in Figure B-1 ....128

B-4: Concentration profiles converted with the empirical model from the measurements
with a) 1.08, b) 2.07, and c) 4.70M Hz transducer...... ................... ............... 129

C-1: Case 1-zero diffusivity: sediment size distribution at distance a) 0.25, b) 0.5, and
c) Im; d) mass concentration; e) approximate mean, and f) standard deviation.... 135

C-2: Case 2-constant diffusivity: a-f) same as in Figure C-1. ...............................135

C-3: Case 3-linear diffusivity: a-f) same as in Figure C-1............... ...................136

C-4: Case 4-parabolic diffusivity: a-f) same as in Figure C-1..............................136

D-1: "Fun-pump" test results for concentration profiles with maximum concentration of
1.2g/l: a-h) sam e as for 3-33. ................................................................ 137

D-2: "Fun-pump" test results for mean sediment size profiles with maximum
concentration of 1.2g/l: a-h) sam e as for 3-36....... .................... .................. 138

D-3: "Fun-pump" test results for concentration profiles with maximum concentration of
1.6g/l: a-h) sam e as for 3-33. ................................................................ 139









D-4: "Fun-pump" test results for mean sediment size profiles with maximum
concentration of 1.6g/l: a-h) sam e as for 3-36....... .................... ................... 140

D-5: "Fun-pump" test results for concentration profiles with maximum concentration of
2.0g/l: a-h) sam e as for 3-33. ................................................................ 141

D-6: "Fun-pump" test results for mean sediment size profiles with maximum
concentration of 2.0g/l: a-h) sam e as for 3-36....... .................... ................... 142

D-7: "Fun-pump" test results for concentration profiles with maximum concentration of
2.4g/l: a-h) sam e as for 3-33. ................................................................ 143

D-8: "Fun-pump" test results for mean sediment size profiles with maximum
concentration of 2.4g/l: a-h) sam e as for 3-36....... .................... ................... 144














KEY TO SYMBOLS


a [m] sediment radius
[a,b] [phi] sediment size range
(a, b, c) coefficients of power-law curve
(A, B, n) parameters of empirical model
at [m] transducer radius
b [m] root-mean-square width of bottom echo return
B = a-rc, pressure correction factor
c [m/s] speed of sound in water
Cfl Courant number
c, [m/s] speed of compressional wave in spherical particle
C" n th Catalan number
ct [m/s] speed of shear wave in spherical particle
d [mm] sediment diameter
Dt transmitter directivity function
E modulus of elasticity
f [MHz] sound frequency
f- nondimensional form function
f- (0, x) nondimensional form function
h filter impulse response function
h(1) spherical Hankel function of first kind
I sound intensity
J, first order Bessel function of first kind
jl spherical Bessel function of first kind
k [rad/m] sound wave number
K, scattering parameter
Kt system parameter
K, system function
LI complex acoustic scattering length
M [kg/m3] concentration of suspended sediments
MA [m] mean sediment size
N number of scatters
nb [1/m3] density of scatters









P0 [Pa] reference sound pressure at r = ro
p, coefficients of best-fit polynomial for K,
PL lognormal probability density function
PN normal probability density function
PI. [Pa] root-mean-squared backscattered pressure
P, [Pa] far-field sound pressure
Px [Pa] total pressure an receiver
r [m] range from transducer
91 transducer receive sensitivity
ro [m] reference range
rb [m] range to the bottom location
Rb bottom reflection coefficient
R, reflection coefficient
rI. [m] near-field range
S, [m] sediment sorting, standard deviation of sediment sizes
t [s] time
T [C] water temperature
T transmission coefficient
Voltage transfer function
V [m3] gated volume
VO [volt] signal from mirror-reflected surface
vb [volt] magnitude of bottom echo signal
Vb [volt] recorded bottom echo return
VeII [volt] amplitude envelope of recorded voltage
Vrm [volt] recorded root-mean-squared voltage
V [volt] recorded signal from suspended scatterers
W [m] reference beam diameter at r =lm
Wman [m] reference main lobe diameter at r =lm
wS [m/s] sediment settling velocity
x =ka
z normalized distance
a [Np/m] total sound attenuation coefficient
as [Np/m] attenuation coefficient due to suspended sediments
a. [Np/m] water attenuation coefficient
fl form function correction parameter
f3z normalized cross-section correction parameter
y modified scattering correction function
70 scattering correction function









Ar [m] thickness of insonified layer
c relative error
e2 variance of estimated concentrations
e [m 2/s] sediment diffusivity
sediment attenuation constant
0 [rad] latitude angle from central axis of transducer
0 [rad] half intensity beam width
Oma.n [rad] width of main lobe
2 [1/m] sound wave length
/ [phi] mean sediment size
correction factor, calculated from data
p [kg/m3] water density
PA acoustic density
p, [kg/m3] sediment density
[7i [m2] total scattering cross-section
oa [phi] sediment sorting, standard deviation of sediment sizes
T [s] sound pulse duration
(p phi scale
" normalized scattering cross-section
Vf near-field correction function
T D integrated beam pattern


xvii















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

ACOUSTIC MEASUREMENT TECHNIQUES
FOR SUSPENDED SEDIMENTS AND BEDFORMS

By

Oleg A. Mouraenko

December, 2004

Chair: Andrew B. Kennedy
Major Department: Civil and Coastal Engineering

The acoustic backscatter system provides a nonintrusive method for measuring of

profiles of suspended sediment size and concentrations. Therefore, an inversion model is

required to convert the recorded series of intensities into parameters of sediment suspen-

sion. The measurement precision highly depends on the choice of the inversion model. In

this work the high-pass backscattering model is modified to allow accurate inversion for

both sediment sizes and concentrations. The modified correction function with a single

parameter is introduced for the form function and normalized cross-section. The parame-

ters can be found from the calibration of the system in the chamber with generated ho-

mogeneous suspension. To account for the difference in shape and mineralogy, a set of

two parameters can be determined for a particular sediment. Their use greatly improves

the accuracy of mean sediment size measurements obtained with the three-frequency

acoustic backscatter system.


xviii









The inversion model was applied to data obtained in the chamber with settling mix-

ture of two sediment samples. The results were compared to the predictions by the ID

advection-diffusion model. Good agreement was shown for concentrations and sediment

size measurements, with the error for sediment size approximation lying within the stan-

dard deviation of size distribution. The accuracy of the concentration measurements was

different among the transducer frequencies, and was found to depend on the sensitivity of

the transducer to the particular sediment size. The sediment sizes were estimated by

minimizing the variance between concentrations, determined from different frequencies,

and found to depend on the accuracy of estimated concentrations. For correction of re-

corded voltage data with applied time-varying gain, the system functions were found to

be better approximated by second order polynomials.

A multiple reflection model was developed based on the multiple layer approxima-

tion. The numerical simulation with the initial uniform concentration profile of 3g/l

shows that the effect in intensities from the multiple reflections can be up to 14%. Also, a

bottom echo removal algorithm was developed in order to eliminate the effect of high

intensity scattering on the concentration measurements close to the bottom.















CHAPTER 1
INTRODUCTION

The application of acoustics for the study of sediment transport and development of

prediction models for morphological changes of coastlines and bathymetry has been

widely recognized and developed during the last few decades [Thorne andHanes, 2002].

By providing remote, no intrusive methods of measurements of flow velocities, bottom

topography, and profiles of suspended sediment concentration, acoustic instruments have

proven a reliable way of obtaining high resolution data during field and laboratory ex-

periments [Hanes, 1991; Hanes et al., 2001; Hanes and Vincent, 1987; Hanes et al.,

1988; Jansen, 1979; Libicki et al., 1989; Traykovski et al., 1999; Vincent and Green,

1990; Young et al., 1982].

The ultrasound at megahertz frequency range is used for the measurements of sedi-

ment suspension parameters within a meter range above the bottom [e.g., Smerdon et al.,

1998]. The system consists of a sound transmitter and receiver. The properties of water-

sediment mixture along an insonified path can then be estimated from the changes in

sound attenuation and from the intensity of scattered sound [Morse andlngard, 1968].

Two configurations of such systems are usually applied: bistatic, when the transmitter

and receiver are spatially separated, and monostatic, when the same transducer serves as

a transmitter and a receiver. The later configuration measures the backscattered intensity

from the scatters in suspension sound.

The acoustic backscatter system allows measurement of both suspended sediment

concentration profiles and distribution of sediment sizes [e.g., Craw ford and Hay, 1993;









Hay and M.Niig, 1992; Schat, 1997; Thorne andHardcastle, 1997; Thosteson andHanes,

1998]. Compared to other techniques, acoustics are a nonintrusive way of measuring sus-

pended sediments with a high temporal (in the order of a second) and spatial (in order of

a centimeter) resolutions [Wren et al., 2000], with comparable accuracy [Hanes et al.,

1988; Thorne andHardcastle, 1997; Thorne et al., 1995].

Since the acoustic backscatter system is an indirect method of measurement, an in-

version algorithm is required for converting measured backscattered sound intensities

into parameters of suspension. The acoustic backscatter equation [e.g., Medwin and Clay,

1997, p.353] provides the basis for the development of such an algorithm.

The difficulties arise from the dependence of a backscattered signal from the prop-

erties of scatterers, their concentration, directivity of the sound beam, and also particular

electronic implementation. Semi-empirical approximations of these parameters [e.g., Ta-

mura andHanes, 1986] can be used to convert the data into concentration profiles. Al-

though a better description of the parameters can be found from theoretical approxima-

tion of sound scattering properties of natural sediments.

The acoustic properties of scatterers depend on many parameters, such as size,

shape, mineralogy, distribution of those parameters in a sample, speed and direction of

motion, their concentration, relative position to the sound transmitter and receiver, and

others. Only a limited set of properties can be considered for the inversion. The simpli-

fied model for acoustic backscattering allows a user to obtain the most important proper-

ties for the study of sediment transport, while others can be approximated by general ex-

pressions. Only suspended sediment concentration, mean sediment size and sorting are

considered to provide the necessary description of the water-sediment suspension, while

the lognormal distribution of sediment sizes in a sample is assumed. The scattering pa-









rameters then are approximated by two functions-the form function and normalized

cross-section. The choice of these two functions determines the accuracy and reliability

of inversion model. The following works define the necessary framework for develop-

ment of the inversion model: [Downing et al., 1995; Faran, 1951; Hay, 1991; Hay and

Mercer, 1985; Hay and Schaafsma, 1989; Neubauer etal., 1974; .iheig and Hay, 1988;

Thorne and Campbell, 1992; Thorne et al., 1993a].

A multiple frequency setup needs to be used in order to estimate sediment sizes.

Schaafsma [1989] used a bistatic configuration of variable frequency transducers, which

covered the frequency range between 1-10MHz, to measure concentration and sediment

sizes of a mixture of two sediment samples. He showed that concentrations could be ob-

tained with 20% error, while the sediment sizes can be resolved within a factor of 2. For

field measurements monostatic configurations are usually used. A system consisting of a

set of three or more transducers with frequencies 1-5MHz is able to estimate the mean

sediment sizes along the insonified column [Crawford andHay, 1993; Hay and.\heig,

1992; Schat, 1997; Thosteson andHanes, 1998]. The accuracy of the estimation highly

depends on correct approximation of scattering properties of the sediments, as well as on

the accuracy of estimations of suspended sediment concentrations from each of the fre-

quencies. It has been shown that the accuracy and stability of estimations of sediment

sizes can be achieved by averaging among a series of consecutive profiles, which leads to

significant loss in temporal resolution [e.g., Thorne andHardcastle, 1997].

Several problems regarding measurement accuracy were addressed by the current

dissertation. Their resolution improves both the accuracy and the temporal resolution of

the measurements, and provides a basis for future improvements.









Usually, only measurements at low suspended sediment concentrations (<10g/l) are

performed. Therefore, the multiple scattering of the sound waves can be ignored [e.g.,

\/Vieg andHay, 1988]. Although it is true for natural sand, in the presence of stronger

scatterers, like air bubbles, or at higher concentrations, as in the region near the sand bot-

tom, multiple scattering can become important. Here a multiple reflection model is de-

veloped and the effects of multiple scattering on resulted errors are shown.

The acoustic properties of different natural sediments vary from sample to sample.

By using the approximated parameters, such as form function and cross-section, obtained

for a particular type of scatters like glass spheres or quartz sand, the inversion algorithm

can lead to erroneous results if the actual sediment differs from the approximation in

mineralogy and shape [Thorne and Buckingham, 2004]. Therefore, acoustical properties

of the actual sediments need to be well defined before the inversion process can be done.

The detailed study of each sediment sample, obtained from the experimental site, for

mineralogical content, distribution of shapes, and overall acoustical performance is an

even more difficult problem, which even in a case of positive solution cannot guarantee

correct results, since the sediment content can change during the period of measurements.

The current work describes a new scattering model, which modifies a correction function

originally presented by Thorne and Buckingham [2004], with a technique to determine

necessary parameters. The description of the acoustic backscatter equation, an implemen-

tation of the multiple layer solution, and the newly developed model are discussed in

Chapter 2.

The calibration of an acoustical backscatter system is performed in order to obtain

system transfer function, which describes the dependence of sound intensity and recorded

voltage levels. Different methods of calibration can be applied, such as by using thin-wire









targets [./heng andHay, 1993], calibration with a particle-laden turbulent jet [Hay, 1991],

or in a calibration chamber with a homogeneous suspension [Tamura andHanes, 1986;

Thorne et al., 1993a]. The time varying gain (TVG) is usually applied for the amplifica-

tion of low level sound signals, arriving from the farther distances from the transducer.

With the proper correction for the TVG, the system function should be uniform with the

range and therefore, can be determined from the calibration. In fact the system functions

found from the calibrations are not always uniform with the range and may depend of

particular sediment sample used for the calibration [Thorne and Buckingham, 2004]. The

variations in system constant can reach 25%, which result in a corresponding error in es-

timation of suspended sediment concentrations. By providing proper corrections to the

system functions developed here, these errors can be lowered.

The solution of the acoustic backscatter equation for suspended sediment concen-

tration can be obtained explicitly with some limitations [Lee andHanes, 1995]. Further

development of the method [Thosteson andHanes, 1998] allowed us to estimate the pro-

files of sediment sizes. The implicit method [Thorne et al., 1993a] provides a more flexi-

ble way of solution, although it can become unstable; when the concentrations become

large and the sediment attenuation becomes significant, even small 5-10% errors can re-

sult in unbounded profiles of concentrations from the iterative solution [Thorne et al.,

1995]. If the concentration profile is approximately known, for example, from the pump

sampler measurements [e.g., Thorne et al., 1993a], the correction for sediment attenua-

tion can be performed more accurately.

In Chapter 3 calibration of the acoustic backscatter system together with a tech-

nique of evaluation of the system parameters and the parameters for the correction func-

tions is described. Also, the results of a series of laboratory tests are shown and compared









to the ID advection-diffusion model predictions for sediment size and concentration pro-

files.

The bed echo contains important information, which can be used to improve accu-

racy of the acoustic backscatter system [Thorne et al., 1995]. Also, the intensity of the

bed echo at low concentrations is much greater than the intensity of the sound backscat-

tered from suspended particles. Because the acoustic backscatter system requires averag-

ing and filtering of the input signal, signals from sand bottom interchange with the sig-

nals from suspended sediments. The algorithm for separation of these signals can provide

one with the information on bed location and allow to determine concentrations of sus-

pended sediments in the vicinity of the bottom. The algorithm is discussed in Chapter 4.















CHAPTER 2
MULTIPLE REFLECTION SOLUTION FOR ACOUSTIC BACKSCATTER

Acoustic Backscatter Equation

The acoustic backscatter equation is derived from the time integral of squared pres-

sure backscattered from the objects in an insonified volume with following assumptions

[Medwin and Clay, 1997]:

1. The objects in the insonified volume are alike, which implies that they are of a
similar shape, size and mineral structure, and can be substituted with a population
of objects with average properties.

2. The distribution of the objects within a gated volume is random and uniform. The
gated volume is a volume of fluid, insonified within the short time interval-time
gate.

3. The time gate is assumed to be large compared to the sound pulse duration. In prac-
tice, however, the time gate is often chosen to be equal to half of the pulse duration
[e.g., Smerdon et al., 1998].

4. The attenuation of sound within the gated volume is ignored.

The far-field sound pressure from a circular piston source in iso-speed medium is

then equal to [Medwin and Clay, 1997]


P,(t)= D (0) P t-r ro e-ar (2-1)


where D, (0) is the transmitter directivity function given by 2-65, 1P (t) is the sound

pressure at the range ro along the central axis, a is the total sound attenuation coefficient

given by 2-42, c is the speed of sound in water, and r is the range from the transducer.

After the transducer emits a sound pulse, it travels through the volume of water with sus-

pended sediments. A portion of the sound is scattered by any particle in the direction of









the receiver. The pressure at the receiver is equal to


P (t) = D (0)P (t) L e-ar (2-2)
r

where Dr (0) is the directivity of the receiver, L, is the complex acoustic scattering

length, and the factor 1/r accounts for the spherical spreading of the sound wave. The

complex acoustic scattering length describes the scattering properties of the particle. It is

a function of angles between incident and reflected waves relative to the particle, and the

sound wave number k. For spheres it can be rewritten as [e.g., .\heig andHay, 1988]

af (0,x) (2-3)
2

where a is the particle radius, and f (0, x) is the nondimensional form function with

x = ka .

In general, the form function is a complex valued function. For scatterers of simple

shapes, like cylinders and spheres, it can be found analytically [Morse andIngard, 1968].

But for particles of irregular shapes, such as sand grains, approximations are used to de-

scribe the amplitude of backscattered form function [Craw ford and Hay, 1993; Neubauer

et al., 1974; .ihelg andHay, 1988; Thorne and Buckingham, 2004; Thorne et al., 1993b].

It relates to the form function as

f = f (0, x) (2-4)

The total pressure at the receiver, P., is a sum of all backscattered pressures from

individual particles. The squared pressure amplitude can be calculated as









N N N
PX (tr =i (t),* (t)+fii (t),(t) (2-5)
1=1 1=1 J=1
J1#

where N is the total number of particles, and asterisk denotes complex conjugate value.

The second term is the sum of the cross-products of the pressures. In the case of a

homogeneous suspension it can be shown that this term approaches zero either with in-

creasing the number of particles or by ensemble averaging of the total backscattered pres-

sure realizations [Libicki et al., 1989; Smerdon et al., 1998; Thorne et al., 1993a]. Previ-

ously mentioned assumptions 1 and 2 define the homogeneity of the water-sediment sus-

pension. In practice, both methods of averaging are used to obtain the root-mean-squared

(RMS) pressure profiles [e.g., Smerdon et al., 1998].

The RMS backscattered pressure for a given range r is calculated as is the integral

over the gated volume of the total backscattered pressure squared, which for the gated

time equal to sound pulse duration, r, can be written as [Hay, 1991]


P 2 1 sinh B-4r (2-6)
0 P0 4 bArTDr 4 B e (2-6)


p2 2 difcs t e p
where P0 = [ 0] (t-) dt is the reference pressure, nb = N/V is the density of the


particles in gated volume, Ar = rc/2 is the extent of the sample volume in r, and (...)

implies expected value over the sediment size distribution.

The term sinh B/B with B = arc is the correction factor if the attenuation over the

gated volume is large [Hay, 1991; Thosteson andHanes, 1998].

The integrated beam pattern is calculated as

2)r;r/2
PD = fJ D,2 (, 0)D (), 0)sin d0dd (2-7)
0 0









which for the monostatic system, when transmitter and receiver are located in the same

point and their directivities are the same, can be approximated as [Thorne andHardcas-

tie, 1997]


YD 2 096 ka, > 10 (2-8)


where a, is the transducer radius.

The density of the particles can be calculated by approximation of the particle

shape by spheres with radius a. Therefore, it can be written as a function of particle mass

concentration, M, as


nb (2-9)
4 (a3)ps

where ps is the density of sediments.

The root-mean-square voltage, recorded by the acoustic backscatter system can be

written in the form [Thorne and Hanes, 2002; Thorne andHardcastle, 1997]

1 M1/2
S= 9 = K K -- e-2r (2-10)


a 1 2f2/2
Ks= Ia (2-1 la)



K, = (K 3rc 1/2 0.96 (2-1 lb)
S16 ) ka,

K = 9T?0ro (2-1 1c)

where 9% is the transducer receive sensitivity, T is the voltage transfer function of the

system, y is the near-field correction factor given by equation 2-67. K, is the system









parameter, which is a constant if all other parameters of the system are fixed, or a func-

tion, if the time varying gain (TVG) is applied. It can be determined by the calibration

with known concentrations and sediment properties [Hay, 1991; Thorne andHanes,

2002; Thorne et al., 1993a].

Inverse Problem for Sediment Concentration

The parameters of the suspension such as sediment concentration and sediment size

distribution parameters can be evaluated by solving the inverse to equation 2-10 problem

[Thorne andHanes, 2002]. For the concentration, the solution can be obtained from the

equation

2
M(r) = \ Vyrr e4ra (2-12)


It cannot be solved explicitly, since the reflection parameter Ks and total attenuation co-

efficient a are functions of sediment size distribution and concentration. The solution by

iterations is usually applied. As can be seen from the equation, the solution for the con-

centration can be unstable, since it has positive exponential dependence on attenuation.

Multiple Layer Approximation

In the derivation of the equation 2-6 for RMS pressure, the multiple scattering be-

tween the particles was ignored. Multiple scattering for the suspension of sand can be ne-

glected for suspended sediment concentrations of less than 10 g/1 [.\/eng and Hay, 1988].

At higher concentrations, which can be found at the ranges close to the bottom [e.g.,

Dohmen-Janssen andHanes, 2002], or in the presence of stronger scatterers in the fluid,

such as a wire or bubbles, multiple reflections can affect the measurements.

After the transducer emits a sound pulse, it travels within a narrow directed beam.

Part of the sound wave is scattered by the suspended particles, and it may be rescattered









by other particles or reflected from the transducer surface multiple times before the back-

scattered sound is recorded by the receiver. The receiver detects only the sound arriving

from the direction within the narrow beam, which is defined by the directivity of the

transducer (2-65). Therefore, the sound scattered backwards in the direction of the inci-

dent sound will most likely remain within the beam and be detected by the receiver (see

Figure 2-1). Table 2-2 shows the approximate beam widths for transducers of different

frequencies.






D (0)


-3dB














Figure 2-1. Example of the scattered sound within the narrow beam.

The RMS pressure (2-6) is calculated as the integral of the backscattered pressure

over the time gate interval, which is again assumed to be equal to half of the pulse dura-

tion. During each time interval, the sound wave travels the distance Tc/2. Therefore, the

whole range from the transducer surface can be represented as a set of layers with thick-

nesses rc/2 (see Figure 2-2). The center of the nth layer is located at distance










r=n (2-13)
2

where n =0 corresponds to the transducer surface.


Time
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
TRANSDUCER
0



1-




--J


3-


-- leading edge
4-
*- trailing edge

Figure 2-2. The multilayer approximation for a possible sound pulse traveling path. The
time interval corresponds to rc/2.

The sound waves can travel along many possible paths within the beam and scatter

from the layers of suspended particles. In the multiple layer approximation all the paths

begin at the layer 0 and pass between the layers with possible changes in direction. The

mean squared pressure at the transducer at time t,, = 1r/c is equal to the sum of mean

squared pressures of the sound followed all possible paths

Cn 1
p2 pZn (2-14)
wheree the C is the total number of the paths given by equation 2-26. Obviously, the

where the Cn, is the total number of the paths given by equation 2-26. Obviously, the


length of every path in the summation should be equal to r.









Let the sound pressure corresponding to an individual path at level i be ]P. With

spherical spreading and attenuation of the sound, the backscattered pressure from level j

becomes

R
Pj = P T (2-15)


where ; is the distance between layers, T is the transmission coefficient, and Rj is the

backscattering coefficient. The backscattering coefficient defines a part of sound which is

scattered from the layer of suspended particles in the direction of the incident sound

wave. It is proportional to the number of scatterers within the layer and their scattering

length (2-3). Using the same assumptions as for the acoustic backscatter equation, the

coefficient can be written as

Sa2f 3Mi rc
R 4=)P 2 Dj (2-16)


where the term (a f2) /4 approximates the scattering length of the scatterers in the

layer j, and the term 'DrD2 Tc/2 defines the insonified volume at layer j. The transmit-

ting coefficient T is equal to

T= e e-a (2-17)

where the attenuation coefficient is calculated using the trapezoidal rule as


a,- = +a +---+a a- + (2-18)
2 2

Therefore, the RMS pressure corresponding to a path of length r is equal to


=P 2,ljl e 2ayr (2-19)
Ur ^









where P,,, is initially transmitted pressure (2-1) with the near-field correction factor ap-

plied (2-67). The multiplication is made over all sub paths, whose total length is




Numerical Realization of Multiple Reflections

The main difficulty in numerical realization is to account for all possible paths that

sound waves can travel along. In the newly developed algorithm the binary tree is used to

solve this problem. The generation of the binary tree is possible, because from any layer a

sound wave is assumed to be traveling only along the directional beam in two direc-

tions-up or down. For the given number of layers, the binary tree will look like it is

shown in Figure 2-3. The nodes of the tree correspond to the centers of the layers, and

their numeration was done in a way that simplifies the generation of the binary tree. The

root of the tree is located at the layer corresponding to the transducer surface. There is a

set of parameters associated with every node. They include the indices of the nodes for

possible up and down directions, as well as calculated reflection (2-16) and attenuation

(2-42) coefficients. The same reflection and attenuation coefficients are applied to all

nodes located at the same layer.

After the binary tree is generated, a fast algorithm of tree-walk is used [e.g., Knuth,

1997]. The walk over the binary tree starts with the node at the transducer surface, where

the reflection coefficient is assumed to be 1 and the attenuation coefficient is equal the

value for plain water. This implies complete reflection of the sound wave from the trans-

ducer surface and zero suspended sediment concentration. After the first step "down" is

made (from the node "1" to "2" in Figure 2-3), neither reflection nor attenuation coeffi-

cients are known at layer 1 (node "2"); therefore they are initially assigned with the val-






16


ues 1 and 0 correspondingly. The intensity of the sound at layer 1 becomes

2
I' = I, e 'Ar (2-20)


where I, is the initial intensity of sound emitted by the transducer with the index corre-

sponding to the node number, and V is the near-field correction factor calculated at

r Ar.


Bin numbers
0 1 2 3 4 5 6


u1 4 4 9 3 6 f2!?
S5 4 27

2'


CZ 4 X1 X7 X22

4-
X5 2 18

5-

6-
-- :: possible direction
S i- possible path

Figure 2-3. Generation of a binary tree for numerical realization of multiple reflections.

A step "up" is made next (from the node "2" to "8" in Figure 2-3), and the calcu-

lated intensity of sound arriving back to layer 0 becomes


I' I e- Ar (2-21)


where "primes" are used for the intensities calculated with initial values R' = 1 and

a= 0. The calculated intensity I, should be equal to the measured intensity correspond-









ing to bin one and recorded at time r, I, This provides us with an expression from

which the concentration at leyer 1 can be found, and hence, corresponding reflection and

attenuation coefficients

Iml, = R2e -21A, (2-22)
'8,

After the reflection and attenuation coefficients for layer 1 are determined, the in-

tensities along all the paths within the layers 0 and 1 can now be calculated and accumu-

lated according to the equation 2-14. The next step "down" is made (from the node "2" to

"3") and the procedure is repeated until the last layer is reached. The expression to find

the concentration in general form becomes

I, -I R e Ar (2-23)
I'

where IJ, is the sum of intensities calculated for the paths which arrive at the transducer

surface at the time corresponding to bin i. The node number is found as


n,= i n- 1 i+1 (2-24)


with n number of layers.

Determination of Sediment Size and Concentration

Three parameters need to be known at every layer to calculate reflection and at-

tenuation coefficients: suspended sediment concentration M, and parameters of sediment

size distribution, MA and S,.

If suspended sediment concentration is the only unknown parameter and mean

sediment size, MA, and standard deviation, So, are known somehow (for example, from

additional measurements or assumed values are used), then equation 2-23 can be solved









via iterations [Thorne et al., 1993a]. To determine the sediment size parameters and the

suspended sediment concentration more than one measurement with different sound fre-

quencies are needed. The techniques were developed to obtain concentration, M, and

mean sediment size, MA, by using 3 transducers, while standard deviation, So, is as-

sumed to be known [Hay and.\heng, 1992], [Crawford and Hay, 1993], [Thosteson and

Hanes, 1998]. To obtain all the parameters--M, MA, and S,-a system with six trans-

ducers can be used [Schat, 1997].

More often the 3-transducer setup is used. It usually requires sorting, So, of the

sediments to be known. The mean sediment size can be estimated from the minimization

of the variance of sediment concentration approximated by each of the transducers


c2 (MO)= (Mk -M) (2-25)
k=1 M3

where M = Mk /3.

Number of Paths

The number of possible paths grows very quickly with the number of layers n. For

a specified n the total number of paths is equal to (n-1)th Catalan number [Weisstein].

The n th Catalan number is equal to


C- (2n)! (2-26)
(n+1)!n!

For illustration the first 24 numbers are given in Table 2-1. The asymptotic form of n th

Catalan number is









4"
C ~4- (2-27)


which for the number of layers n = 120 gives C120 ~ 7.58.1068.

Table 2-1. First 24 Catalan numbers
n C,, n n C,,
1 1 9 1,430 17 35,357,670
2 1 10 4,862 18 129,644,790
3 2 11 16,796 19 477,638,700
4 5 12 58,786 20 1,767,263,190
5 14 13 208,012 21 6,564,120,420
6 42 14 742,900 22 24,466,267,020
7 132 15 2,674,440 23 91,482,563,640
8 429 16 9,694,845 24 343,059,613,650

Because of the large number of the paths, the execution time becomes very long

even for a small number of layers. Therefore, not all the paths can be included in calcula-

tion. The number of reflections which a sound wave is exposed can be chosen as a crite-

ria for the path to be included in the calculation. The number of reflections for each path

varies between 1 and n(n 1)/2, where n is the number of layers. The reflection coeffi-

cient for the layer of scatterers is determined by equation 2-16. To approximate the order

of magnitude of the coefficient, the following values can be used: r =lm, r =10lOs,

c=1500m/s, T'D=0.015, p,=2650kg/m3, M=3g/1l, a=0.2mm, for which R=0.01. Two

reflections result in attenuation of incident pressure on the order of 10-4, three-10-6. Al-

though, the number of the paths can be very large, the sound waves of small amplitude

will be absorbed by the suspension. Therefore, the number of reflections can be limited to

3 if the number of layers is large (n > 15).

Effect of Multiple Reflections

Four tests were performed to analyze the relative effect of multiple reflections on

measurement of sediment concentration. The ratios of intensities calculated with one and









three reflections using the model described above are plotted in Figures 2-4 and 2-5. For

uniform concentration profiles (see Figure 2-4 a-d), the effect of multiple reflections is

almost uniform with distance. The difference reaches its maximum of 12-14% at 3g/l,

but at lg/1 it is only around 5%. For the concentration profiles, which linearly change

from Og/1 to maximum values of 0.1-3g/1 with distance (see Figure 2-4 e-h), the differ-

ence is even smaller-on the order of 1-3%. This is probably the most common situation

for field measurements, when the concentration near the transducer is much smaller than

it is near the bottom.

However, it is possible for a cloud of sediments to be injected by a turbulent eddy

into the water column with the concentration at higher elevations to become greater than

the concentration near the bottom. Figure 2-5 a-d shows the results when uniform con-

centration profiles were modified by a jet-type profile with maximum concentration of

2g/l, located at around 30cm distance from the transducer. The difference becomes more

visible, especially at the ranges just after the jet location. Compared to uniform concen-

tration profiles, the intensity increases around 1.2 times.

In the next test, a wire-type response was modeled with results shown in Figure

2-5 e-h. For the profiles at 0.1 and 0.5g/l there are "ghost" peaks present, which corre-

spond to the second reflection of sound between the transducer surface and "wire". This

situation can be modeled in the lab with a strong reflector, like steel wire (see Chapter 4).

Although the second reflection is present in Figure 2-5 b-d, there are no such "ghost"

peaks in the tests with the sediment mixture jet (see Chapter 4). In the field, air bubbles

can be present in the water column, and since the bubbles are strong reflectors the second

reflection from them can affect the measurements at longer ranges.










Concentration profiles


1.08MHz


1.08MHz


2.25MHz 2.25MHz


4.75MHz


4.75MHz


0 0.2 0.4 0.6
Distance


0.2 0.4 0.6 0.8
Distance


Figure 2-4. Difference between intensities modeled with one and three reflections for se-
lected profiles of concentrations: a) constant profiles, e) linear profiles, b-d)
and f-h) percent difference.


Concentration profiles











6

-4


Concentration profiles

a).
ia */


2 10M


1.08MHz


2.25MHz
20










10
0



0 0.2 0.4 0.6 0.8
Distance


0

20


o- 10


Concentration profiles
e).





1.08MHz


2.25MHz
30

20 7 5

10

0
4.75MHz


0.2 0.4 0.6 0.8
Distance


Figure 2-5. Difference between intensities modeled with one and three reflections for se-
lected profiles of concentrations: a) profiles of constant concentration with
"bump", e) profiles of constant concentrations with "spike", b-d) and f-h)
same as in Figure 2-4.


f) A
~ZZ~f\Ezzzzr


A I








Special Parameters for Acoustic Backscatter Equation
Form Function and Normalized Cross-Section for Elastic Spheres
The theoretical far-field form function for elastic spheres can be expressed as

[Thorne and Campbell, 1992]

f (x= (-1)" (2n+1)b (2-28)
Ix n 0

where

A a12 a13 a a12 a13
b =2 a22 a23 a21 a22 a23 (2-29a)
0 a32 a33 0 a32 a33
4, = x,2 (P/Psji(,), nXi (


= x2 p ) (x),2 =-xh (x),

12 = [2n(n+1)-x ] ( )-4xj (x,),

22= X (X ) 32 =2 [j( x ) xj (x ) (2-29b)

a3 =2n(n+1) [x j: (x )-j (x )i ,

23 =n(n +l)j (x,),

a33 = 2xj (x,)+[x2 -2n(n+1)+2] j (x),

where x, = x(c/c), x, = x(c/c,), c is the sound speed in the water, c, and c, are the

speeds of shear and compressional waves in the spherical particle, p and p, are the fluid

and particle densities, ji is the spherical Bessel function, and h~1) is the spherical Hankel

function of first kind. The derivatives of the functions can be found by the recurrence re-

lation [Abramowitz and Stegun, 1965]










S= (-)+ ng (x) (2-30)
x

where g (x) is either a spherical Bessel or Hankel function.

The normalized scattering cross section is defined as


X= t (2-31)
2ra2

where <7, = L 2 is the total scattering cross section of a single particle of radius a. For

the spherical particles it can be found to be


S(x) = 2 (2n+1) b 2 (2-32)
X n=0

Equations 2-28 and 2-32 involve infinite series, but only a finite number of terms

can be calculated. The following approximate expression can be used to calculate the

number of terms sufficient for the resulting accuracy of 0 (10-6) for x < 50

N10-6 = 1.2246x + 6.4232 (2-33)

High-Pass Model for Quartz Sand

The modified high-pass model for backscattered intensity [Crawford andHay,

1993; Johnson, 1977; .\heig andHay, 1988; Thorne et al., 1993b] provides simpler ex-

pressions for the form function and normalized scattering cross section for quartz sand


fh (x)CO Kfx42 (2-34a)


C =-vexp 2exp )212 (2-34b)


4 K, x4
3Kx4 (2-35)
(')C) I+ x2 4 K C4
3 K,?










where v, = 0.25, x, =1.4, i7 =0.5 and V2 = 0.37, x2 =2.8, 172 =2.2, Kf =1.1,

K, = 0.18, and x=ka.

The functions given by equations 2-34 and 2-35 are shown in Figure 2-6 together

with corresponding curves for quartz spheres (2-28 and 2-32). The values of the veloci-

ties of shear and compressional waves in quartz spheres were taken to be c, =3760m/s

and c, =5980m/s.


2 1.2
a). b).
1.5
0.8
1 0.6
0.4
0.5 0.2

0 0
100 102 100 102
k-a k-a
Figure 2-6. Form function and normalized cross section for quartz sand (blue) and
spheres (red): a) form function, b) normalized cross-section.

Modified Model for Elastic Spheres

Based on the spherical model, modified expressions were developed by Thorne and

Buckingham [2004] to better describe the scattering properties of natural sediments. The

following empirical function was introduced


3 +0.5x+3.5 (2-36)
x3 +3.5

and applied to the low-pass filtered spherical form function and normalized cross section

in form









f ( X)= (X,) f= (2-37)


(x)= o(x,= Jz (2-38)

The f3 parameters describe the deviation of the modified model from the spherical

model due to irregularities in shapes and sizes, which occur in natural sediments. While

the parameters /~. and fz are generally independent, their values are close and usually

change between 1.3 and 2.2. The value of / =1.9 can be taken for both of the parameters

to provide a reasonable fit with the data [Thorne and Buckingham, 2004]. It can be seen,

that the function 70 (x, 8) is greater than 1 for all values of x. It implies that the values

of modified form function and normalized cross section are greater than the original ones

calculated by spherical model.








1.5
2 | -7:: :.







0.5
10-1 100 101 102
k.a
Figure 2-7. Modified y function for selected values of parameter 3 .

Modified High-Pass Model for Natural Sediments

Similarly, the high-pass model can be modified to better model the properties of

particular sediments. Therefore, the original function 70 (x,f8) (2-36) was currently








modified as

{ 0(x,?3), /3>1
;(xO)=Lx'+(8-0.5)x+3.5 (2-39)
1 x' +3.5 <8<

The function y(x,/3) is equal to 1 for /3=0.5. The lower limit 8/ implies

y(x, 3) >0 and is equal to 0.5-5.25/i.75 =-3.85. The graphs of the function y(x,/3)

for selected 83 parameters equal to -1, 0.5, 1, and 1.9 are shown in Figure 2-7.

The modified high-pass model becomes

fh ( x) = y(x, /3)f ,, (2-40)

X (x))= y(x,]3z)X, (2-41)

Sound Attenuation Coefficient

The sound attenuation coefficient defines the exponential rate of decay of sound in-

tensity with distance. It depends on temperature, salinity, pressure, presence of suspended

particles and microbubbles, and varies with sound frequency [Richards, 1998]. The total

attenuation coefficient can be written as a sum of the attenuation coefficient in clear wa-

ter, a,, and the attenuation due to scattering by suspended sediments, as,

a = a, + a (2-42)

The attenuation of sound in the megahertz range in seawater at small (<30m)

depths is primarily a function of water temperature. The formula by Fisher and Simmons

[1977] gives the approximation to within 4%

c= f2 (55.9-2.37T+4.77.10-2T2-3.48.10-4T3).10-3 (2-43)

where f is the sound frequency in megahertz, T is the water temperature in degrees

centigrade.








The sediment attenuation coefficient at a range r is given by

a, =nb (2-44)
2

where nb is the number of particles per unit volume, <7, is the total scattering cross-

section of the particle. Using the equations 2-9 and 2-31 and averaging over the sediment

sizes, the sediment attenuation coefficient becomes

3Ka2
a (r) = M (r) = (r)M (r) (2-45)
4p, a )

where 4 is the sediment attenuation constant evaluated at range r

3Ka2)
( = (2-46)
4p, a3

The attenuation of the sound traveling within interval [r, r2 ] due to scattering and

absorption by suspended sediments can be approximated by

a, = -1 fJ (r)M(r)dr (2-47)
r2 ~ r2 r,

If the sediment attenuation constant is not changing with range, then the attenuation is

equal to

a, = "M (2-48)

where M is the mean sediment concentration between ranges [r', r2].

Lognormal and Normal Distributions of Sediment Sizes

Sediment sizes vary from very fine (clay) to coarse (sand, gravel, shell) [e.g.,

.Se'ith, 1984]. Lognormal distribution can be used to approximate the actual distribution








of sediment sizes. Its probability density function is given by

pL S 1 (In a -Ma)2
a aMSS = exp 2 (2-49)
a S, 2S

To obtain the probability density function in terms of sediment radius a in meters

the following substitutions were made

MA =-(l+,u)ln2-31nl0 (2-50)
S = = o ln2

where /u and a, are the mean and standard deviation in phi units. The phi scale is de-

fined as the logarithm of sediment size in millimeters. The sediment size, given in meters,

can be written as

(p -log2 (2a-103) (2-51)

If the probability density function of a is given by equation 2-49, then the quantity In a

is normally distributed. In terms of (p the probability density function is given by


P, (p a,) =exp (2-52)


The parameters u,, and a,, can be approximated from the sieve analysis as [Sleath,

1984]

/* N1/2
// =-log2 (d841d1591/2 and 07 = log92 d841 (2-53)
d,159

where da is the diameter in millimeters for which n % of the sample by weight is finer.

The raw moments of the lognormal distribution are


=(a) = ap, (a M ,S)da=exp MA + (2-54)









,2 a2)=Ja2 pL (aMa,S)da= exp{2(Ma,+S )} (2-55)
0

3 =(a' =japL,(aMSa,S)da=exp 3MA,+9a (2-56)

Averaged Form Factor and Normalized Cross-Section Terms
The averaged terms in the equations 2-11 and 2-46 can be calculated with the as-

sumption that sediment sizes obey some known distribution. Assuming the lognormal

distribution (2-49), the following simplification can be made.

The term Ka2 f2(ka) 2), where (...) denotes averaging over the sediment sizes, is

equal to

a2 f (ka)2 f (ka) p(a)da (2-57)
0

where the sediment size is distributed with the density function p (a). If p (a) is the log-

normal distribution, the expression above simplifies to

af (ka)=a 2(ka) pL(aMa,SO)da (2-58a)
0

l= na=y,da= dy (2-58b)
_a

2F ezyf(e2 1 (y-M)2
f e (key) exp 2 dy (2-58c)


2 S 2(k 2 xp 21 y (M + 2S 2 )]2
(= (key) exp{2(M+ S ) } exp 2Sdy (2-58d)



= J2 f(key)2 pN,(y/it,& )dy (2-58e)





31


=2Kf (keY)) (2-58f)

Here (... p(x) is the expected value of the argument with the probability density function

p (x). PN, (y |, 6() is the normal distribution probability density function


PN (Y ) )=- exp (2-59)

with J = M + 2S2 and S& = S.

With backward substitution the term Ka2 f2(ka)2) becomes

a2 f (ka)2 = In a = y, = dy (2-60a)


= /2 f(ka)2 1 exp (lna- da (2-60b)


= f (ka)2pL (ap, (.) da (2-60c)
0

= /2 ( (ka)2) (2-60d)

(a2) f (ka)2 ) (2-60e)

The modified term

(a2 ((ka) 2 \(ka)f)
S a)- PI(a (ka) ) (2-61a)
Ka3) /3

exp{2M< + 2S }
exp{ 2M +92S} f (ka)2) (2-61b)
exp{3M0+9 2SJ} IL(a^)









=e-2s'i fka)p(, (2-61c)
/A

f(ka)2)
=e-2s f--a) PL4) (2-61d)


The expressions for the averaged normalized total scattering cross-section can be

obtained following similar derivations

a 2X(ka)) = a2 )(ka))P(a,a) (2-62)

a 2X(ka)) )2 (K(ka ) P(2-63a)
-= --- (2-63a)


= e-2si (a) (2-63b)

(a)

Circular Transducer Directivity and Half Intensity Beam Width
The directivity of the transducer can be defined as the ratio of sound pressure at

point (r, 0), where 0 is latitude angle measured from central axis of the transducer, to

the axial pressure at the same range r

D ( r,0) ) (2-64)
P(r,0)

Let at be the radius of transducer and k = 2r/2A be the sound wave number, where

A is the sound wavelength. At great distances from the transducer (rk > 1), the far-field

directivity of a circular piston source is [Medwin and Clay, 1997, p. 139; Morse and n-

gard, 1968, p.381]

D() 2J1 (ka, sin 0)
Dka sin (2-65)
kat sin 6










where J, (x) is the first order Bessel function of first kind. The curve of the squared di-


rectivity function, D2 (0), is shown in Figure 2-8.


1-


0.75-


'5 0.5 -.
\

0.25 I
1.6163 3.8317


-6 -4 -2 0 2 4 6
katsinO

Figure 2-8. Circular piston directivity function, D2 (0).

The time averaged acoustic intensity, I, relates to the RMS pressure as [Medwin

and Clay, 1997]

P2
jI = ]m (2-66)
PAC

where PA is the acoustic density that relates to the bulk modulus of elasticity, E, as


PA = E/c2.

The half intensity beam width, 0, can be defined as 0 = 20 where 0 is the angle

from the center axis to where the acoustic intensity is equal to the half of the axial value.

For a circular piston source, the directivity function of which is given by equation 2-65,

D2 (0) = 0.5 at ka, sin = 1.6163 Although most of the sound energy is transmitted

within the cone defined by the half intensity beam width, as shown in Figure 2-1, the low

frequency transducers can have wider main lobe and significant side lobes.









Similar to the half intensity beam width, the width of the main lobe, .ma.n, can be

determined from D,2 (0) = 0, which is satisfied at kat sin = 3.8317.

The beam width at a given range, r, from the transducer, for example at r =1 m,

can be found as W(r) = 2r sin (0/2) and Wea, (r) = 2r sin (On /2). The values of the

beam widths 0 and W for frequencies 1-5MHz are shown in Table 2-2. It can be seen,

that for higher sound frequencies the beam widths are narrower.

Table 2-2. Beam width for a circular piston transducer (a,= 5 mm and c= 1500 m/s).
Frequency 1 MHz 2 MHz 3 MHz 4 MHz 5 MHz
ka, 21.0 41.9 62.8 83.8 104.7
0, deg 8.9 4.4 2.9 2.2 1.8
.ma.n, deg 21.1 10.5 7.0 5.2 4.2
W, m 0.15 0.08 0.05 0.04 0.03
Wma I. m 0.37 0.18 0.12 0.09 0.07

Near-Field Correction Factor

Close to the transducer, the sound waves from different parts of the transducer sur-

face interact with each other, both destructively and constructively. The structure of

sound pressure in the near-field is different from the form given by 2-1. Regardless of the

complexity, it still can be calculated accurately [Lockwood and Willette, 1973]. For the

application to remote measurements of suspended sediment concentration and sediment

size, the detailed information about the sound field close to the sensor is usually

[Downing et al., 1995; Thorne et al., 1993a] substituted by the near-field correction fac-

tor, y~(r), which is given by [Downing et al., 1995]


1+1.32z + (2.5z)32
( z = 32 (2-67)
1.32z+(2.5z)

where z = r/rl is the normalized distance, and r1f is the near-field range as given by










r.,= (2-68)


The value of r1f depends on the transducer radius and the sound wavelength. For a

fixed transducer size, the near-field range increases linearly with the transducer fre-

quency. For illustration, the values of the near-field range for frequencies 1-5 MHz are

shown in Table 2-3.

Table 2-3. Near-field range r, for a circular piston source (a, =5mm and c =1500m/s).
Frequency 1 MHz 2 MHz 3 MHz 4 MHz 5 MHz
rf, m 0.05 0.10 0.16 0.21 0.26






0.8 -


-56:. /


0.4-
04- 1 MHz
-- 2 MHz
0.2 -- 3 MHz
/ 4 MHz
S-- 5 MHz
0
0 0.1 0.2 0.3 0.4 0.5 0.6
Range, m

Figure 2-9. Near-field correction factor, 1/ly(r), for a,=5mm and c=1500m/s.

The near-field correction factor is applied to the far-field pressure in form of

i/I (r). The corresponding curves for 1-5MHz frequency transducers are shown in

Figure 2-9. For higher frequencies the effect of the proximity to the transducer on trans-

mitted sound is larger, and the near-field region is farther.















CHAPTER 3
CALIBRATION OF THE ACOUSTIC BACKSCATTER SYSTEM

Calibration of the acoustic backscatter system is necessary to obtain system func-

tion given by equation 2-11 [Thorne andHanes, 2002]. For this purpose, the measure-

ments of profiles of backscattered sound intensity can be taken in an environment with

known characteristics. Different scatterers, such as thin wire targets, glass spheres or

sand, are used in the calibration in various experimental setups [e.g., Hay, 1991; Jansen,

1979; .he/ug and Hay, 1993]. Calibration in a chamber with a homogeneous suspension

of sand or glass spheres is the most common technique [e.g., Ludwig andHanes, 1990;

Schaafsma, 1989; Tamura and Hanes, 1986; Thorne et al., 1993a; Thosteson and Hanes,

1998]. It has several advantages over other methods. The calibration with similar scatter-

ers to those found in the field requires no changes in backscattering parameters such as

form function or normalized cross-section. The experiments with the thin wire require

modification of the acoustic backscatter equation due to cylindrical spreading of back-

scattered sound compared to spherical spreading from the sand grains and spheres. Also,

the definition of sample volume for the wire is different from the one done for the sus-

pension of particles. Although a modified form of equation 2-10 needs to be applied

[.\heig and Hay, 1993], the method with the thin wire targets provides the easiest calibra-

tion which requires only a simple setup (similar to shown in Figure 4-2 in Chapter 4).

In experiments with a turbulent jet [Hay, 1991], a mixture of suspended particles

and water is used. The suspension within the cross-section of the jet is not homogeneous

and therefore modifications to the acoustic backscatter equation (equation 2-10) are









needed. The turbulence provides proper mixing and makes the orientation of the particles

random relative to the transducer. Freely falling particles [Jansen, 1979] can have prede-

fined orientation relative to the positions of transmitter and receiver, although the effect

of this is not significant for particles with approximately spherical shapes [.\//,g and

Hay, 1988].

For calibration and other tests, two acoustic backscatter systems were used. Both

systems were developed by Centre for Environment, Fisheries and Aquaculture Science

(CEFAS), Lowestoft, UK. The first system, hereafter referred as "ABS 1", has three

10mm transducers with operational frequencies of 1.08, 2.07 and 4.70MHz. The data col-

lection was carried out by a specially developed acquisition system [Thosteson, 1997].

The acquisition system operates at maximum rate of 88 profiles per second, and records a

maximum of 4 root-mean-square (RMS) profiles a second by averaging over 22 consecu-

tive raw voltage profiles. The second system, hereafter referred as "ABS2", also has three

10mm transducers, with frequencies 1.0, 2.5 and 5.0MHz. Aquatec Electronics Ltd, UK,

developed the data logger, which allows obtaining raw ABS profiles at a maximum of 80

profiles per second for each frequency.

The resolutions of the two systems are different. For the ABS1, the voltage is

measured in a 12-bit range, while the ABS2 has 16-bit resolution. Spatial resolution is

defined by the pulse duration. It is 10ps for the ABS1 and 13jps for the ABS2, which re-

sults in a bin size of approximately 7.5 and 10mm, respectively. Both systems use time

varying gain (TVG). The form of the TVG was set to account for the spherical spreading,

and is nominally given by a linear function of the range r, although the exact form of the

TVG may deviate from linear.









Calibration Chamber

The recirculating chamber at the University of Florida Coastal Laboratory was used

to perform the calibration of the acoustic backscatter systems. This chamber was previ-

ously used in a number of projects [Lee, 1994; Ludwig andHanes, 1990; Tamura and

Hanes, 1986; Thosteson, 1997]. The chamber is schematically shown in Figure 3-1. It

consists of a vertical 2m acrylic tube with inside diameter of 0.19m. A rotary pump

moves a water-sediment mixture through a 4cm PVC pipe upward, where it discharges

into the tank through the four nozzles. The nozzles are at the same level, about 10cm be-

low the free surface. The jets are directed towards the center of the tank from 90-degree

sectors (see Figure 3-3), where they collide and generate a turbulent flow, mixing the

suspension. The mixture flows down the chamber with a mean cross-section velocity of

approximately 4cm/s. The bottom of the tank has a funnel, which leads the flow through

a pipe back into the pump. The funnel prevents the sediments from accumulating at the

bottom. The suction ports were closed with metal plugs, which protruded into the interior

of the chamber for about 5mm.

Before taking measurements, the chamber was filled with tap water and left for

about 20 hours to allow air bubbles to leave the system. For additional 2-4 hours the sys-

tem was left with the pump turned on.

To establish the predefined level of suspended sediment concentration, portions of

sand were added into the water. Every portion was initially dried, weighed and placed for

several hours in container with a small amount of water and detergent. The detergent was

used to detach air bubbles from the surface of the sand grains. Excess water with deter-

gent was removed from the container before the sand was added into the chamber.













Transducer(s)



Nozzles




Levels 0, 15, and
30 cm


Suction ports





















A I







Funnel






Pump l


20cm.

Figure 3-1. Calibration chamber-side view.









ABS Tests in the Chamber

The chamber has a different environment for measurements than field experiments.

Because of the laboratory setup, various artificial conditions can affect the measurements,

which may lead to erroneous conclusions about estimated parameters. A series of tests

were performed in the University of Florida Coastal Laboratory to investigate the effects

of the chamber on calibration measurements.

For every test, profiles of root-mean-square voltage were obtained. Each profile

was calculated as the average of 120 individual RMS profiles recorded at 4Hz for 30 sec-

onds. To obtain each individual RMS profile, the ABS was set to average 22 raw profiles.

Tests were made with fewer averaged profiles to check if the setting has an effect on the

measurements. The higher sampling rate theoretically can lead to the interaction between

several consecutive pulses, when previous sound pulses were not fully attenuated. The

tests did not reveal this possibility. Additionally, to the average profiles, the 10th and

90th percentile profiles were calculated to illustrate the scatter among the measured RMS

profiles.

Tests in Air and Still Water

Two tests were performed to examine the lower bound for voltage profiles. For this

purpose the measurements were taken with the transducers left in the air and placed in the

chamber with still clear water. The resulting curves are shown in Figure 3-2. The curves,

which correspond to the measurements in the air, take on the lowest values except for the

1.08MHz transducer, which has a peak near approximately 4cm range. None of the pro-

files start with zero values-on the contrary they start from the value around 50-100 in

relative units and decay within a few bins. Within this range there is no difference be-

tween profiles measured in the air and in the water. Most likely this is an effect of resid-










ual vibration of transducer surface after the sound pulse was emitted. The measurements

in the water show that the scatter of the RMS profiles (in figures plotted with dashed

lines) can be significant especially for high frequency transducers. The peaks around 9cm

for 1.08 and 2.07MHz are caused by the reflection of emitted sound from the chamber

walls, which will be illustrated further.


1.08MHz 2.07MHz
200 200
a). b).
150 150 /
s 100 n 100
50 50 -
0o 0o
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
4.70MHz Range,m
150
c).
100 -- air
water/conc. 0 g/I
I \ -- water/conc. 0 g/1I
50 J (center line)

0
0 0.2 0.4 0.6 0.8
Range,m

Figure 3-2. RMS voltage profiles (solid lines) with scatter intervals (dashed line) from the
measurements in air and still clear water: a) 1.08MHz, b) 2.07MHz, and c)
4.70MHz.

Homogeneity of Sediment Suspension and Effect of Chamber Walls

The jet system and high flow rate serve to generate a homogeneous suspension. To

investigate the homogeneity of the mixture in vertical and horizontal directions, samples

were taken from the four openings in the side of the chamber wall (see Figure 3-1). The

results of this investigation was previously published by Ludwig and Hanes [1990]. From

their tests, it was concluded that the suspension in the chamber is homogeneous within

the measurement error.

Additional measurements were currently conducted to test effects of the proximity

of a transducer to the jet system and sidewalls. The ABS1 system with 3 frequencies









(1.08, 2.07, and 4.70MHz) was used for the tests. One transducer at a time was placed at

different vertical levels relative to the nozzles (see Figure 3-1) and at different distances

from the sidewall (see Figure 3-3).





Nozzles R
R 70mm.







Suction ports Nozzles




Figure 3-3. Calibration chamber-top view. The jet system is shown with transducer po-
sitions for the sidewall proximity tests.

For the wall proximity tests, measurements were taken at the center of the chamber

corresponding to a distance of 95mm, and also at 30 and 50mm distances from the side-

wall. In the vertical, the transducer was placed at 3 levels: 0, 15 and 30cm. The Ocm level

corresponds to the upper edge of the nozzles, therefore the jets of sand-water mixture

were colliding just under the transducer surface.

The profiles of RMS voltage were recorded for several different levels of sus-

pended sand concentration (see Tables 3-1, 3-2, and 3-3). Sand collected from Jackson-

ville Beach, FL with sieve fraction of size 0.212-0.250mm was used for the test.

Table 3-1. Locations and concentration levels in tests for 1.08MHz transducer.
Distance from wall (row) 0cm 15cm 30cm
and level (column)
30mm 0, 0.2, 0.5g/l 0, 0.2, 0.5g/l Og/1
50mm 0, 0.2, 0.5g/l 0, 0.2, 0.5g/l Og/1
95mm (center) 0, 0.2, 0.5, 0.8g/l 0, 0.2, 0.5, 0.8g/l 0, 0.2, 0.5, 0.8g/l







43



Table 3-2. Locations and concentration levels in tests for 2.07MHz transducer.
Distance from wall (row) 0cm 15cm 30cm
and level (column)
30mm 0.2, 0.5g/l 0.2, 0.5g/l
50mm 0.2, 0.5g/l 0.2, 0.5g/l
95mm (center) 0.2, 0.5, 0.8g/l 0.2, 0.5, 0.8g/l 0.2, 0.5, 0.8g/l


Table 3-3. Locations and concentration levels in tests for 4.70MHz transducer.
Distance from wall (row) 0cm 15cm 30cm
and level (column)
95mm (center) 0.5, 0.8g/l 0.5, 0.8g/l 0.5, 0.8g/l


level 0 cm / conc. 0.0 g/l


level 0 cm / conc. 0.2 g/l


1.08 MHz
level 15 cm / conc. 0.0 g/l


level 30 cm / conc. 0.0 g/l


level 15 cm / conc. 0.2 g/l level 30 cm / conc. 0.2 g/l


level 0 cm / conc. 0.5 g/l


level 0 cm / conc. 0.8 g/l


level 15 cm / conc. 0.5 g/l
600
h). 95 mm
400 50mm
400- 30mm

200

0
level 15 cm / conc. 0.8 g/l


level 30 cm / conc. 0.5 g/l


level 30 cm / conc. 0.8 g/l


0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Range, m Range, m Range, m

Figure 3-4. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for
1.08MHz transducer for sidewall proximity tests.


400

300

a 200

100

0

400

300

a 200

100

0


800

600

a 400

200











Figures 3-4 and 3-5 show the results of measurements with the 1.08 and 2.07MHz

transducers for different concentrations of sand. For the 1.08MHz transducer the profiles

have several peaks, which correspond to the positions of the suction port plugs. As shown

earlier (Table 2-2), the 1.08MHz transducer has the widest directional beam. Therefore, a

part of the sound can reflect from the plugs and be detected by the receiver. The profiles

for 2.07MHz transducer do not have similar peaks.


2.25 MHz
level 0 cm / conc. 0.2 g/l level 15 cm / conc. 0.2 g/l level 30 cm / conc. 0.2 g/l
800 800 800
a). 95 mm b)' 95 mm c)' 95 mm
600 50mm 600 50mm 600
30 mm : 30mm : :
400 400 400 : :

200 200 200
0 0 0
level 0 cm / conc. 0.5 g/l level 15 cm / conc. 0.5 g/l level 30 cm / conc. 0.5 g/l
1000 1000 1000
d). 95 mm 95 mm m. 95 mm
50 mm 50 mm
30 mm 30 mm
E 500 500 500


0 0 0
level 0 cm / conc. 0.8 g/l level 15 cm / conc. 0.8 g/l level 30 cm / conc. 0.8 g/l
1500 1500 1500
g). 95mm h). 95mm i). 95mm
1000 1000 1000

500 500 500 : .

0 0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Range, m Range, m Range, m

Figure 3-5. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines) for
2.07MHz transducer for sidewall proximity tests.

The measurements along the chamber centerline at different vertical levels are

shown in Figures 3-6, 3-7, and 3-8. The curves for the 1.08MHz have peaks at approxi-

mately 9 and 19cm. Comparing to the profiles at other distances from the wall (see Figure

3-4), these peaks appear only when the transducer is placed in the center of the chamber.

The ranges at which the peaks appear correspond to multiple chamber radiuses. They can











result from multiple reflection of the sound wave from side lobes between the chamber


wall and transducer surface. This also demonstrates that the side lobes of the 1.08MHz


transducer are quite significant. This can affect the measurements in the field as well, be-


cause other transducers may not detect the scatterers located within the side lobes of the


1.08MHz transducer.


conc. 0.0 g/1I conc. 0.2 g/1I
400 400
a)' -- Ocm cm
300 -- 15cm 300 15 cm
30 cm 30 cm
S200 30200cm

100 100



conc. 0.5 g/I conc. 0.8 g/I
600 800
9). -- 0 cm d). : 0 cm
-- 15cm 600 15cm
400 -- 30 cm 30 cm
S400 4


0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Range, m Range, m

Figure 3-6. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured along centerline with 1.08MHz transducer

conc. 0.2 g/1I conc. 0.5 g/1I
800 1000
a -- Ocm 0cm
600 15 cm 15 cm
., : 30 cm '^,- 30 cm
2 400 500 m



00
0 0.2 0.4 0.6 0.8
conc. 0.8 g/1I Range, m
1500
c). 0 cm
30 cm


500 ...

0
0 0.2 0.4 0.6 0.8
Range, m

Figure 3-7. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured along centerline with 2.07MHz transducer.










The effect of proximity to the nozzles is noticeable only for the 2.07 and 4.70MHz

transducers. Figures 3-7 and 3-8 show that the profiles at level Ocm are lower than at lev-

els 15 and 30cm. The drop in the intensity can be due to either Doppler shift of the fre-

quency of the backscattered sound from faster moving sand grains, which cannot be re-

solved by the receiver with fixed frequency, or due to inhomogeneity of suspended sedi-

ment concentration in the upper part of the chamber. Although separate measurements of

the concentration in the jet part of the chamber were not performed, the later seems to

have more effect, since the maximum speed from the jets is not very high (approximately

3m/s) and resulted Doppler shift would be about 0.2% of the transducer frequency. How-

ever, the measurements show that if the transducer is placed lower than the jets, the dif-

ference between profiles from different layers is very small. Hence, the vertical homoge-

neity of the concentration in the chamber is confirmed and by the current tests.


conc. 0.5 g/I conc. 0.8 g/1I
1500 a 0 cm 1500
10 a). m- :'n : -0 Ocm
100-- .,,-n, :- 15cm
1000 -- :n', 1000 30cm
ED -- n. E n'
500 1 :, :.n',n 500

0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Range, m Range, m
Figure 3-8. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured along center line with 4.70MHz transducer

Figures 3-9 and 3-10 show the difference between the measurements at different

levels along the chamber, but now the transducers are placed at 30 and 50mm from the

chamber wall. Similar to the tests when the transducers were located at the center of the

chamber, the tests for 2.07MHz transducer show a slight decrease in voltage at Ocm level.








47



However, the profiles for 1.08MHz do not have peaks at ranges equal to multiples of the


chamber radius.


at 30 mm / conc. 0.0 g/1I

a). 0 cm"
S- 15cm m
30 cm




at 30 mm / conc. 0.2 g/I

0cm
15 cm


m 'cs^ /


0 0.2 0.4 0.6 0.8
Range, m


at 50 mm / conc. 0.0 g/I
150
b). 0 cm
100 15 cm -
30 cm




at 50 mm / conc. 0.2 g/I
300 c-
F : 0 cm
200o 15 cm

100 I

0
at 50 mm /conc. 0.5 g/I
600
f) o cm
400 : 15cm

200

0-
0 0.2 0.4 0.6 0.8
Range, m


Figure 3-9. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured near the chamber wall with 1.08MHz transducer.


at 30 mm / conc. 0.2 g/1I


at 30 mm / conc. 0.5 g/1I

0- Ocm
15 cm

N --
v- ,,


0 0.2 0.4 0.6 0.8
Range, m


800

600

2 400

200

0


1000



2 500



0


at 50 mm /conc. 0.2 g/1I

b). : -- Ocm
\ 15 cm

.- :-


at 50 mm / conc. 0.5 g/1I


0 0.2 0.4 0.6 0.8
Range, m


Figure 3-10. Profiles of RMS voltage (solid lines) with scattering ranges (dashed lines)
measured near the chamber wall with 2.07MHz transducer.


at 30 mm / conc. 0.5 g/I


800

600

2 400

200

0


1000



m 500









Results

The tests show that the jet system generates homogeneous suspension in the region

approximately 15cm below the nozzles. The chamber wall effects are not significant for

2.07 and 4.70MHz transducers, but for the measurements with the 1.08MHz transducer, it

should not be placed in the center of the chamber to avoid multiple reflections from the

side lobes of the beam. To obtain the smooth profile, the peaks corresponding to the posi-

tions of the suction port plugs can be filtered out by interpolating between the nearest

data points. A few bins at the beginning of the profile do not contain useful information

on the suspended sediment concentration, since it is contaminated by an electronic noise

or residual vibration of the transducer membrane.

Estimation of System Function

The system functions of the acoustic backscatter, K, can be found from the cali-

bration by using the expression


K =V Vr kat 1 1/6 2ar (3-1)
KM1/2 0.96 3rc


where K5 is calculated by equation 2-11, M is the level of suspended sand concentration

in the chamber, Vrm, is the measured root-mean-square profile of voltage, y is the near-

field correction factor (2-67), k and c are the sound wavenumber and the sound speed in

water, a, is the transducer radius, r is the pulse duration, and a is the total attenuation

coefficient (2-42)

Since the transducer frequency is preset and changes in the speed of sound in water

are relatively small, the system parameter Kt, given by equation 2-1 la, is usually ob-

tained from the calibration instead of K [Thorne andHanes, 2002]. However, the effec-









tive transducer radius, which determines the directivity of the sound beam and extent of

the near-field region, is usually 5-20% less than the geometrical radius. By using the pa-

rameter KA, the best-fit value for a, may be found from the calibration.

The system function K, is constant if the time varying gain (TVG) is not applied,

otherwise, K is a function of range r. The system function describes the properties of

the system and, therefore, does not vary with sediment size or concentration. However,

the functions estimated from the measurements taken for different concentrations of

sediments of different size distributions will most likely be not the same due to variations

in sediment shapes, mineralogy, etc., which are not accounted for in the acoustic equation

2-10. The modified form function and normalized cross-section function with parameters

fl, and 8z, given by equations 2-40. and 2-41, may be used to account for these differ-

ences, as well as to better approximate the system function.

Let K' be the i th estimation of the inherited system function, K, where each K'

is calculated by equation 3-1 at the concentration level MA',eas from the measured voltage

profile V, for a given sediment size distribution. The average function can be found as

the mean of estimates as


K =- Y K (3-2)
v V


where N is the number of estimations, to provide an estimation of the system function

for a particular sediment size distribution.

Approximation by 2nd Order Polynomial

The system function calculated by equation 3-2 is not necessarily a smooth func-

tion. Therefore, K, was fitted with a 2nd order polynomial over the range of bins










nfnf < j < n (nf is the near-field bin index and n is the total number of bins) to produce


an estimation of system function K' as


K = P2r2 +pr+pO (3-3)


where p,, i =0, 1, or 2, are the coefficients of the best-fit polynomial.


Calibration of the ABS1

The ABS1 was calibrated with quartz sand, collected at SISTEX'99 experiment

[Vincent et al., 2001], with mean size /u =2.17 and standard deviation a7 =0.27 esti-


mated from the sieve analysis. From the fall velocity measurements, the mean sediment

size was approximated to bel.96 in phi units. Both measurements and corresponding

normal distribution curves are shown in Figure 3-11. The transducer radius was taken to

be 90% of the geometrical radius. The modified form function and normalized cross sec-

tion were calculated by equations 2-34 and 2-35 with initial value of 0.5 for parameters

Iz and ff


1 ,--- 1-b). ---- 0 _[
a). b). 0 0 data
fit
0.8 0.8

o 0.6 0.6
o 0o
o o

0.4 0.4

j Sieve: Fall velocity:
0.2 / = 2.17 0.2- = 1.96
o =0.27 / =0.27

1 1.5 2 2.5 3 1 1.5 2 2.5 3
9 9
Figure 3-11. Estimation of sediment size distribution used for the calibration of ABS1 by:
a) sieve analysis, b) fall velocity.









The resulted system functions are shown in Figures 3-12-3-14. Individual estima-

tions, K', are shown together with the curves for K' (marked as "fit") on plots "b" in

Figures 3-12-3-14. The errors on plot "b" were calculated as the average of relative root-

mean-square deviations from the 2nd order fit and are given by the expression


1 n 1 N K_ ( r) K r ( t ) 2
C-K =nn +1 K (3-4)
nf j =

The near-field bin index nnf for current calculations was taken to be equal to 15, but can

be adjusted for any particular system. The value was chosen from the sidewall proximity

tests which showed that the measurements at bins <15 are not accurate.

Calculated by 2-10, voltage profiles were compared to the measured profiles. Their

ratios are plotted on plots "a." Using 2-12 the inverted profiles of the suspended sediment

concentration are shown on plots "c." Relative errors for each of the concentration levels

between inverted and measured concentrations are also shown with the mean error, which

was calculated as the average of individual relative errors:

1 N 1 .. MA, c (r, ) -M'as (r) (35)
N n-nnf +1 =nn Me', a(r,)

where the subscript "meas" denotes measured or expected concentration in the chamber

and subscript "calc" denotes estimated concentration profiles calculated with 2-12.













2.5

2

>1.5

>

0.5

0



5

-4



0
2
0
0 1

0


System coefficients


'>20


10


0.2 0.4 0.6 0.8 0.2
Range, m
Measured (dashed) and calculated (solid) concentrations


0.4 0.6 0.8
Range, m
Conc,g/l Err,%

0.10-41.0%
-- 0.20-27.1%
0.30- 13.2%
0.40- 9.7%
-- 0.50- 3.1%
0.70- 4.2%
1.00- 15.7%
1.50- 28.7%
2.00 25.6%
3.00 36.2%


0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Range, m

Figure 3-12. Estimation of the system function for 1.08MHz transducer: a) ratio of meas-
ured and calculated voltage profiles, b) estimated system functions, and c) in-
verted sediment concentration.


System coefficients


0.5



1.5

0)
S1



o 0.5
0

0


0.2 0.4 0.6 0.8 0.2
Range, m
Measured (dashed) and calculated (solid) concentrations


0.4 0.6 0.8
Range, m
Conc,g/l Err,%


0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Range, m

Figure 3-13. Estimation of the system function for 2.07MHz transducer: a-c) same as in
Figure 3-12.


Ratio of voltages












1.6

1.4

>1.2



0.8


0.5

0.4
0)

- 0.3

o 0.2
0
0 0.1

0


Ratio of voltages
250
f =:4.70MHzat = 4.50 mm b).
PO = 2.170 = 0.27 200
Sf = 0.50 P = 0.50
.... :150

100

50
K =-20
0
0.2 0.4 0.6 0.8 0.2
Range, m
Measured (dashed) and calculated (solid) concentrations


System coefficients

fit






Err = 4.8066%
19.7920 r2+377.1953 r+25.0449

0.4 0.6 0.8
Range, m
Conc,g/l Err,%

90% 0.10- 18.2%
0.20- 12.7%
0.30- 3.1%
0.40 26.2%
0.50 34.3%


0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Range, m

Figure 3-14. Estimation of the system function for 4.70MHz transducer: a-c) same as in
Figure 3-12.


The estimated system functions show that the expected linear dependence due to


the TVG is present only for the 2.07MHz transducer, although the 2nd order coefficient is


still significant. The curves of voltage ratios on plot "a" for the 4.70MHz transducer (see


Figure 3-14) are not uniform with range compared to the curves for the other two trans-


ducers. This may indicate erroneously estimated attenuation at each level of concentra-


tion. The errors corresponding to the different concentration levels are not uniform, espe-


cially for 1.08 and 4.70MHz. Since the estimated system functions were applied, the er-


rors grow for smaller and larger concentrations.


Calibration of the ABS2


The calibration of the ABS2 was performed similarly to the ABS1. The sand from


Jacksonville Beach, FL was used for the calibration. The estimations of sediment size








54



distribution parameters by sieve analysis and based on fall velocity are shown in Figure


3-15.


1-
-


0.8


Z 0.6


S0.4


0.2


0
1.5


0.8 -


z^--' 0
2 2.5 3 1.


Figure 3-15. Estimation of sediment size distribution for
ABS2 by: a) sieve analysis, b) fall velocity.


Ratio of voltages
3 1000
a). f = 1.00MHz a = 4.50 mm b).
2.5 : : 2.41 = 0.13 800
.52 J\ Pf= 0.50 3 =:0,50
>o x 600

1 __ ...,. 400

0.5 200
0.5
0 0
0.2 0.4 0.6 0.8 1
Range, m
Measured (dashed) and calculated (solid) concentra


1.5


0.2 0.4 0.6
Range, m


0 data
- fit


2 2.5 3


sand used in calibration of


System coefficients


0.2 0.4

tions


0.6 0.8 1
Range, m
Conc,g/l Err,%

0.02- 15.0%
0.04- 6.7%
0.07- 5.2%
0.20- 3.9%
-- 0.50- 4.3%
1.00- 5.0%
-- 2.00- 13.1%


0.8 1


Figure 3-16. Estimation of the system function for 1MHz of ABS2: a-c) same as in Fig-
ure 3-12.


b).










Fall velocity:
P = 2.28

S= 0.23


c). Mean error = 7.60%





/7

n.[^ -


5









The resulting system functions are shown in Figure 3-16. Unlike the 1.08MHz

transducer of the ABS 1, the curves for the system functions are very close to each other

and the correction of /f parameter was not needed.

The 1MHz transducers of the both systems are very similar, and one would expect

to obtain similar results in calibration of both systems. However, the current tests show

that the results are very different. The calibration of the 1.08MHz transducer of the ABS1

does not result in a unique system function. The application of the average of estimated

system functions can lead to large errors in inverted concentrations. For these cases, a

semi-empirical model was developed to provide the solution for suspended sediment

concentration. This model is described in Appendix B.

Sensitivity Analysis For Estimation of System Function

A particular estimate of system function K depends on several parameters, which

can be initially approximated with some error or uncertainty. In order to investigate the

sensitivity of the system function K, to changes in the parameters, such as form function,

normalized cross-section, and sediment distribution, the following analysis was carried

out. Using equation 3-1, the system function K, can be rewritten as a function of pa-

rameter ,fi, transducer radius at, and sediment size distribution parameters jl, and a.

as

K = F( lz,a,, p ) (3-6)

The form function and normalized cross section were estimated by equations 2-40 and

2-41. The dependence of K on listed parameters is not explicit, but through the form

function, normalized cross-section, beam directivity, and near-field correction factor.

Generally for calibration, these parameters need to be known and fixed. For current









analysis, a subset of the listed parameters set to be free variables (i.e. by letting them vary

within some intervals). This produces the new system functions, K ,,fi estimated by

minimizing the error calculated by equation 3-4.

The results of the calculations, based on the measurements for the ABS 1, are gath-

ered in tables in Figures 3-17-3-19. For each test, some of four parameters Jfz, a,, uV ,

and a,, were fixed (marked with "dots"). Others were variable, for them the best-fit val-

ues are presented. Bounds for free parameters are shown in the last column of the table.

For fixed parameters, their initial values were used. The errors of inverted concentrations

together with the mean error (black line) were calculated and plotted for each test. The

values of the mean errors are shown in the table in "Mean error, %" row. Values in row

"Fit errors, %" were calculated by 3-4.

The bounds of free parameters were chosen to cover a realistic, but still wide, range

of values. In many cases, the minima were reached at boundaries. If unconstrained mini-

mization would be performed, the minimum could be reached at unrealistic values, such

as Omm for transducer radius.

It can be seen from the figures that not all the parameters have the same effect on

calibration. The parameter fiz and mean sediment size p/ have the most significant ef-

fects, while transducer radius a, and sediment size sorting a, have almost no effect.

Since both the mean sediment size and parameter Jfz are present in the expression

for normalized cross-section, minimization on one of them can be compensated by an

equivalent change in another. Because of this relationship, use of incorrect values for 8fz

may lead to erroneous approximation of sediment size from the measurements.








57


45 Conc, g/I

40 -0- 0.10
-e- 0.20
35 -- 0.30
0.40
30 0.50
0.70
25 1.00
20 ~ 1.50
-W- 2.00
15 -0- 3.00
-- mean
10 -

5 -

0
Test # Initial 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Fit error, % 11.7 10.9 10.2 9.89 11.7 10.9 10.2 9.89 10.4 10.2 9.86 9.74 10.4 10.2 9.86 9.74 Bounds
Mean error, % 20.5 18.5 16.6 16 20.5 18.5 16.6 16 17.2 16.7 15.9 15.6 17.2 16.7 15.9 15.6
P 0.50 -. 2.50-2.50-2.50-2.50-2.50-2.50-2.50-2.50 -2.50+ 2.50
at, mm 4.50 5.00 5.00 5.00 5.00 5.00 5.00 5.00 5.00 3.00- 5.00
1 2.17 2.71 2.71 2.71 2.71 2.71 2.71 2.71 2.71 1.63- 2.71
0 0.27 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05 0.05- 1.00

Figure 3-17. Values of fitted parameters fz, at, p and a. with corresponding errors

in inverted concentrations for the 1.08MHz transducer of the ABS1.

12 T F Conc, g/1

-0- 0.10
10 -0- 0.20
-0- 0.30
e e e 0e e e -e 0.40
-e- 0.50
8o 0.70
0E-- 1.00
20 o o e-e e e e -e-omean








Test # Initial 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Fit error, % 3.57 2.84 2.65 2.65 3.57 2.84 2.65 2.65 2.65 2.65 2.65 2.65 2.65 2.65 2.65 2.65 Bounds
Mean error, % 5.58 5.02 5.2 5.2 5.63 5.06 5.24 5.24 5.2 5.2 5.2 5.2 5.24 5.24 5.24 5.24
P3 0.50 -0.77-0.18-0.040.01-0.77-0.19-0.04-0.01 -2.50-2.50
at, mm 4.50 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00 3.00- 5.00
t 2.17 2.47 2.40 2.47 2.40 2.36 2.21 2.36 2.22 1.63- 2.71
0 0.27 0.05 0.21 0.05 0.21 0.05 0.05 0.07 0.09 0.05- 1.00

Figure 3-18. Same as on Figure 3-17 for the 2.07MHz transducer of the ABS1.










35 r T I T0 Conc, g/I

30\ --- 0.10
30 E E o 0 -e- 0.20
-e- 0.30
25 0.40
-e- 0.50
o20 -0 mean

S15 -

10



0 1 1 1 1 1 1 1 1 1 1 1 1 1
Test # Initial 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Fit error, % 4.81 4.56 4.8 4.4 4.8 4.55 4.8 4.39 2.45 2.45 2.45 2.45 2.44 2.44 2.44 2.44 Bounds
Mean error, % 18.9 18.5 18.9 18.3 19 18.7 19 18.4 14 14 14 14 14.2 14.2 14.2 14.2
P 0.50 .- .. 1.20 1.18 1.20 1.18 1.20 1.18 1.20 1.18 -2.50-2.50
at, mm 4.50 3.00 3.00 3.00 3.00 .. 3.00 3.00 3.00 3.00 3.00- 5.00
S 2.17 2.13 2.01 2.13 2.01 2.22 2.11 2.22 2.12 1.63- 2.71
G 0.27 0.05 0.05 0.05 0.05 0.19 0.19 0.19 0.20 0.05+ 1.00

Figure 3-19. Same as on Figure 3-17 for the 4.70MHz transducer of the ABS1.

The biggest decrease in mean error (3-5)-from 18.9 to 14.0%-is shown in Figure


3-19 for the 4.70MHz transducer by varying parameter z After the corrections to


sediment attenuation, not only the mean error was reduced, but also the errors for most of

concentration levels. The sediment attenuation for transducer 1.08MHz was insignificant;


neither p/ nor fz have any influence on errors. For the 2.07MHz transducer parameter


#z was estimated to be -0.77, which results in decreasing of the originally estimated (by


equation 2-35) normalized cross-section and therefore sediment attenuation. The corre-


sponding value of f8z for the 4.7MHz is 1.20, which implies that the sediment attenua-


tion is larger than estimated by equation 2-35.


The system functions for the 4.70MHz transducer with corrected f3z are shown in


Figure 3-20. The errors in inverted concentrations at low concentration levels (0.1 and

0.2g/l) were decreased more than 2 times, and also became more uniform with range. The










spreading among K,it curves was significantly reduced resulting in a mean relative error


of 2.4%.


Ratio of voltages System coefficients
1.6 250
a). f = 4.70MHz at = 4.50 mm b).
1.4 = 2.17z = 0.27 200 -
> 1.2Pf = 0.50 P = 1.20 1
>"1.2 150

1 100
0.8 : 50 Err = 2.4457%
0 5 K = -198.7014. r2+405.7229 r+22.0241
0
0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8
Range, m Range, m
Measured (dashed) and calculated (solid) concentrations Conc,g/l Err,%
0 .5 |--------
c). Mean error = 13.96% -- 0.10- 6.0%
._--4_ -- 0.20- 5.9%
0.4 0.30- 3.9%
S.. ... ......... .. 0.40- 25.1%

@0.2 ----- -- ----
0
0
0.1 .. ..-.- -
0


0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Range, m

Figure 3-20. Estimation of system function KV for 4.70MHz transducer with corrected
parameter /fz: a-c) same as in Figure 3-12.


Dependence on Sediment Sorting

The dependence of K on sorting is rather complicated, because of filtering the


data (either theoretical or experimental) used in obtaining the form function and normal-

ized cross-section. From equations 2-61 and 2-63, K can be simplified to


r(a2 ,--1/2(
K- = C (r (3 -.exp C2(r)



= C (r)[es- .exp -C2(r)e-2S = (3-7)


=C3 (r) exp {S-C4(r)e -2S}


where S, = 7 In 2, and functions C, (r) do not depend on 0a, but on other parameters.






60

By approximating exponents to 2nd order, the following expression can be obtained

K C3 (r){l-C4 (r)+o2 Iln4(l+ 2C4 (r))} (3-8)

where K now depends on a,, to second order. Therefore, small changes in a,, will have

small effect on minimized relative error. This is supported by Figures 3-17-3-19, where

there is no consistency in calculated values for a, ; the errors sometime reach their min-

ima at either limits of the constraint interval for o',. Based on current analysis, it can be

concluded that if the sediments are well sorted, then it becomes nearly impossible to suc-

cessfully approximate sorting from measurements using this minimization technique.

Dependence on Transducer Radius

The dependence of K, on transducer radius a, can be explicitly found and deriva-

tive becomes


Ka = C(r) a, +y = C(r){-2zVi+V,} rA
Za'2 z(3-9)
=C(r){1.01-0.01l(z-fr)} rA


where C(r) is the coefficient not depending on at. For far-field (z > )T) the relation be-

tween K, and at is almost linear with a very small second order term, and therefore does

not significantly affect the relative error (3-4). However, if the near-field region extends

significantly, as in the case of 4.70MHz transducer, changes in transducer radius can be

noticeable.

Determination of Parameter fL

For a homogeneous suspension of sediments, coefficient K, (equation 2-1 la) is

constant with range. Therefore, any uncertainties in determination of K affect only the









magnitude of the estimated system function K. Since the system function for a given

transducer is unique, any deviation of the estimated system functions can be compensated

by using the modified form functions (2-37 or 2-40) with the appropriate parameter ,ff.

Because the coefficient K, appears as a multiplier in equation 3-1, the parameter flf

cannot be found from measurements with a homogeneous suspension of one sediment

sample. Thus measurements with different sediment samples are required.

From equations 2-11 and 2-40 it can be found that there exists a constant

2 a2f 2h) Ka2 fh2)(-
2 (ka, 1f ) (3-10)
={aY(ka,/3 ,)fh

which for small standard deviation a, gives

j r(ka,f/) (3-11)

Similar to the estimation of the system functions, K', which were obtained from

the measurements with particular sediments (see equations 3-2 and 3-3), let the system

function K, be written in the form of a 2nd order polynomial over the range n nf < j < n

K,=pr + pr+p (3-12)

with each coefficient pk, k =0,1 or 2, found as

1 m
k Pk, (3-13)


where pk,j are the corresponding coefficients of the estimation for one sediment size sys-

tem function K', and m is the number of estimations.









Therefore, j can also be written as

K$ P2r2 +Plr+po (3-14)
Kv p2r + pr + po

which needs to be satisfied for all values or r. This may not be possible for all times,

therefore the estimation of using mean quantities can be used and the following ex-

pression obtained


1 p -p, )r 2{+(p( -p)r+)(po4 -p)} dr=0 (3-15)


with the solution for j as

P2 r02 +P r +P0
= 2 (3-16)
P2 r2 +P2r + po
3 2

where r0 is some reference range. After obtaining the solution for J, parameter flf, can

be found by inverting equation 3-11.

Estimation of System Function for ABS1

To obtain the system function for the ABS1, the tests were carried out with four

different sediment samples. The sediments were predominantly quartz sand. The parame-

ters of the size distributions were estimated from sieve analysis and are shown in Table

3-4. The sample from Case 3 is the same sample used for the calibration of the ABS1 and

described in the previous section. Its size distribution is shown in Figure 3-11. The sedi-

ments for Case 4 were the mix of two sieve fractions-0.125-0.149 and 0.297-0.250mm.

The estimations of distribution parameters are shown in Figure 3-21.

The system functions were calculated for all four series of tests. The results of the

calculations are shown in Appendix A (see Figures A-13-A-36). The estimated parame-










ters f6z and ,ff, the mean errors (calculated by equations 3-4 and 3-5), and parameter 4


are gathered into Table 3-5. The parameters /f, were approximated in both cases-for


initial values of fz =0.5 and for best-fit values, estimated from the minimization of error


3-4. It was previously noted that for a wide sediment size distribution, the value of sort-

ing a, can have a noticeable effect on estimated quantities. For example, Figure A-2


shows that if o, =0.44, then the corresponding parameter fz is equal to -1.98. But if the


sorting would be estimated as o. =0.06, which is the best-fit value when a,, is taken as a


free parameter, the parameter 8fz is equal to -0.19. In Table 3-5 the values of parameter


I#z correspond to the estimated from measurements values of a (from the column for


Test#8 in Figures A-l-A-12 in Appendix A).

Table 3-4. Sediment sizes.
Case 1 Case 2 Case 3 Case 4
/, 2.67 2.74 2.17 2.16
o' 0.44 0.25 0.27 0.54


1 ---O 1 O
a). b). 0 0 data
fit
0.8 0.8

0.6 0.6 0
o o
C 0 0 0 00 0
o o o
t 0.4 0.4 0

0.2 Sieve: Fall velocity:
0.2 t = 2.16 0.2 = 1.99
G = 0.54 G =0.60
0 0=0 0000
1 1.5 2 2.5 3 1 1.5 2 2.5 3
9 9
Figure 3-21. Estimation of sediment size distribution for Case 4 used for calibration of
ABS1 by: a) sieve analysis, b) fall velocity.










The results in Table 3-5 show that for the fitted values of /, all mean errors de-

crease up to 50%. Case 1 for the 4.70MHz transducer is the only one, where the mean

error in inverted concentrations increases slightly. Although the changes in parameters

I# and /fi are also small.

Table 3-5. Estimation of parameters /6 and fL for different sediment sizes.
Case 1 Case 2 Case 3 Case 4
Initial Fitted Initial Fitted Initial Fitted Initial Fitted


1.08MHz
1# 0.5
K % 6.9
CM, % 12.3
1.04
'yf 0.90
2.07MHz
1# 0.5
K, % 7.3
M, % 11.4
1.08
'if 0.92
4.70MHz
1# 0.5
K, % 3.2
,M % 16.8
1.16
'/f 1.10

Table 3-6. Values
Frequency, MHz
1.08
2.07


-2.5
5.3
9.5
1.04
0.91

-1.98
3.6
6.8
1.06
0.85


0.33
3.1
17
1.13
1.04

of x=
Case 1
0.36
0.69


0.5
3.6
6.8
1.51
2.5

0.5
2.6
5.9
1.15
1.85


0.5
5.8
21.9
1.01
0.55

for test
Case 2
0.34
0.66


-2.5
3.3
6.0
1.53
2.5

0.37
2.6
5.9
1.21
2.5


1.47
2.1
11.8
1.09
0.91

cases.
Case
0.51
0.98


0.5
11.7
20.5
0.80
-0.94

0.5
3.6
5.6
1.01
0.56


0.5
4.8
18.9
1.03
0.72


-2.5
10.4
17.3
0.80
-0.93

-0.77
2.6
5.2
1.04
0.66


1.20
2.4
14.0
1.08
1.00


0.5
17.1
37.8
0.65
-1.99

0.5
12.2
inf
0.76
-0.59


0.5
6.2
42.5
0.80
-0.82


-2.5
14.1
26.3
0.63
-2.16

-2.17
8.7
17.11
0.69
-0.88


-2.5
4.7
18.8
0.71
-1.43


3 Case 4
0.51
0.98


4.70 1.57 1.49 2.22 2.23


It was previously noticed, that even large changes in parameter 8, for the


1.08MHz transducer have small effects on estimated errors-the fitted values for )6,


ka











drop to -2.5 (the lower bound), the value chosen to be the lower limit in the calculations.


For the 1.08MHz frequency, the values of x = ka (see Table 3-6) are less than 0.51 for


the Cases 1-4. This gives the values for the normalized cross-section of 0.01 or less (see


equation 2-35 and Figure 2-6), which results in very small total attenuation (2-42) at low


concentrations. Since, the parameter /z is found from an expression like


;I(kaJZ) h (ka)


(3-17)


where normalized cross-sections- is estimated from data and -h is given by equation


2-35-are small numbers, approximated from their ratio values of y can have large er-


rors. Therefore, the initial value for z =0.5 can be used.


1.08MHz


2.07MHz


4.70MHz


0.2 0.4 0.6 0.8
Range, m



- Case 1: ta =2.67, o =0.44
- Case 2: pt = 2.74, o = 0.25
- Case 3: pt =2.17, o =0.27
Case 4: at = 2.16, o = 0.54
- Mean


0.2 0.4 0.6 0.8
Range, m

Figure 3-22. Estimated system functions and their means for Cases 1-4 with fz =0.5

(dashed) and fitted (solid): a) 1.08, b) 2.07, and c) 4.70MHz transducers of the
ABS1.


50

40

30

20

10

0


250

200

150

100

50

0







66


The estimated system functions for all four cases are shown in Figure 3-22 together

with their means calculated by equation 3-12. The dashed lines correspond to the func-

tions estimated by assuming the initial value fz =0.5. The solid lines correspond to the


system functions estimated with fitted values for fz (see Table 3-5). The parameter j


was estimated by equation 3-16 with ro =0.6m. Corresponding values of Jfl, were found


by inverting equation 3-11. Both parameters are shown in Table 3-5.


1.08MHz 2.07MHz


1.6t .


E1.2


0.8
0.6 .


4.70MHz
1.4 )'
'IV


.0.8 --- .- ...... ... ..--.-.--

0.6 .
0.2 0.4 0.6 0.8
Range, m


.1.2 I -- -- Case 1: po = 2.67, o = 0.44
I --. --- -- Case 2: pt = 2.74, o = 0.25
S1 Case 3: ta = 2.17, o = 0.27
0.8 Case 4: It = 2.16, G = 0.54

0.6
0.2 0.4 0.6 0.8
Range, m
Figure 3-23. Ratios between individual system functions for Case 1-4 and mean system
function with f8z=0.5 (dashed) and fitted (solid): a) 1.08, b) 2.07, and
c) 4.70MHz transducers of the ABS 1.

Although the system function should be unique for a given system, the estimated

curves have some scatter, which results from the different sediments used in each case.

For a homogeneous suspension, the curves differ approximately by a constant (equation

3-14). Therefore, the ratios of individual estimations to the mean were calculated with the

results shown in Figure 3-23. For the 1.08 and 2.07MHz transducers, the calculated ratios









are close to constant for r >0. Im. In most cases for these two transducers, the ratios are

more uniform for the system functions estimated with fitted values of /z, compared to

ones estimated by using /z =0.5. This means that the corrections to the normalized cross-

section (equation 2-41) lead to a better approximation of sediment attenuation. Also it

should be noted that for the 4.70MHz transducer, the ratios are less uniform with range

than for other two frequencies.

Analysis of Relationship between )6 and fi

As can be seen from the results in Table 3-5, by choosing appropriate values of fz

and f/, corrections to the form function and normalized cross-section can be made

(equations 2-40 and 2-41), which improve the accuracy of suspended sediment concentra-

tions. Since the parameters Ifz and fl3. were found individually for each frequency and

each sediment sample, the following questions may arise:

* Can a single value 8? be found, which could be used for correction of both the form
function and the normalized cross-section for every individual or for all used fre-
quencies?

* Can a single value of fL be found which would represent a sediment sample and
be unique for an applicable range of sound frequencies?

* Can a single value of /z be found which would represent a sediment sample and
be unique for an applicable range of sound frequencies?

In Figure 3-24, the estimated parameters f/L are plotted versus corresponding fitted

values of /z for frequencies 2.07 and 4.70MHz. It can be seen that in most cases

J8f > fz, as it obtained by Thorne and Buckingham [2004]. Because of large scatter and

an insufficient number of available data points, the functional relationship between Jfl3







68



and f can only be approximated with a large degree of uncertainty. Nevertheless, it can


be concluded that Jfl. and I3 are most likely different even for a single frequency.


3-


2





cc- 0


-1


-2


-3


- 2.07MHz
- 4.70MHz
O Case 1
A Case 2
O Case 3
0 Case 4


-3 -2 -1 0 1 2 3


Figure 3-24. Fitted parameters I3# and corresponding parameters f. for Cases 1-4.


Values of fL and 8', for Cases 1-4 are plotted in Figure 3-25. It can be seen, that


even with the large scatter, there is an overall trend in the parameters.


3


2




1

--

C-1


Case 1 Case 2 Case 3 Case 4

Figure 3-25. Values of parameters flf and 8 for Cases 1-4 for different frequencies.


-- 1.08MHz
2.07MHz
4.70MHz -
A A 1
a OOf
0 0
A A



0

0
A

A









Figure 3-26 shows calculated values for j together with y(ka, fz). Dashed lines


correspond values of x = ka (see Table 3-6). Note that values of 7(ka, f ) for the


1.08MHz transducer were not plotted since all fitted z = -2.5. Solid lines correspond to


the function y(ka, 3) (equation 2-39) with f3 estimated as


I= ( f,2o07+ f,470) (3-18)


where f3f,207 and ff ,470 are the values of Jf3 for the 2.07 and 4.70MHz transducers. The

lines fit data points for j relatively well, especially in Cases 1 and 4. Although in Case 3

the point corresponding to the 1.08MHz frequency is much lower than points for the

other two frequencies and also lower than the approximated curve. Overall, equation 3-18

gives a good approximation for parameter f/f for a given sediment sample.


1.6 -
.... -- Case 1
Case 2
: : \'. I 0 A: /\ case 3k '







100 101
1.8- CC 4






k.a

Figure 3-26. Values of and "(ka,1 3) as functions of ka with approximated functions
y(ka, 3) for Cases 1-4 (solid lines).









A similar approximation for fz will not be possible. The values of y(ka, ,8) are

close to each other only for Case 4, while other three cases look random.

"Fun-Pump" Tests

The homogeneous sediment suspension, generated in the chamber (Figure 3-1) for

calibration, will most likely not be found during field or lab experiments, when the meas-

urements are taken above a sandy bottom, and the suspension is generated by waves and

currents. The following tests were conducted in order to simulate the "real" environment,

which have variable concentration and size distribution profiles of sediment suspension.

A simple technique is used to generate such nonuniform profiles in the chamber. With the

sediments added to the chamber, the pump system generates a homogeneous suspension

throughout the water column. If at any time the pump is turned off, the sediments will

settle with some terminal velocity. Since there are no other sources of sediment near the

top of the chamber, a nonzero gradient in sediment concentration will be generated. Also,

if there are several sediment sizes present, then the sediments with different sizes will

settle at different velocities. This will also generate a nonuniform profile of sediment size

distributions.

The measurements were taken with the ABS1 system. Three transducers were

placed at the top of the chamber at about 10cm below the free water surface pointed

downwards (see Figure 3-1). The ABS was operating at 88 profiles per second. Every 22

profiles were averaged to obtain a single root-mean-square (RMS) profile of voltage. The

resulting RMS voltage profiles were recorded by the data logger at 4Hz. The duration of

every run was three minutes, including 30 seconds of data with homogeneous suspension

with the pump turned on. Then the pump was turned off, and the measurements were re-









corded for another two minutes, while the sediments were settling down. After that, the

pump was turned back on and data were collected for another 30 seconds.

Sample Sediment Size Distribution

A mixture of two sediment samples was used for the tests. These are the same sedi-

ments as were used for the calibration (see Case 4). The first sample is the sieve fraction

of size 0.125-0.149mm (2.47-2.75 phi units) of the sand from Jacksonville, FL. It will be

referred as Sand 1. The other sample is the sieve fraction of size 0.297-0.250mm (1.25-

1.75 phi units) will be referred as Sand 2.

Assuming a uniform distribution of sediment sizes within each fraction, one can

approximate the mean and standard deviation of a mixture of two sediment samples. Let

the mixture contains two fractions a and a, of Sand 1 and Sand 2. Therefore,

a + 2 = 1 (3-19)

The size distributions are given by


p, ((p) = b-a (3-20)
0, [ ',b]

where i corresponds to Sand 1 or Sand 2, and [a,b, b] are the size ranges. The mean and

variance of sediment sizes in mixture becomes

// = + a2, = a+ a2 (2 +U)_ -) 2 (3-21)

where / and a, are the mean and standard deviation of Sand 1 and Sand 2, determined

as


A = ") = (3-22)
2 12







72


Different concentration levels were acquired by adding proper amounts of Sand 1

or Sand 2 into the chamber. The fractions of each sample are shown in Table 3-7 together

with the resulting size distributions. Based on the fall velocity analysis, a 1:1 mixture of

two samples had the mean of 1.99 and standard deviation of 0.60 (see Figure 3-21). Note

that the values of parameters of the distribution for the 1:1 mixture in Figure 3-21 from

the sieve analysis were different from Table 3-7. It is because for the calibration in

Case 4 average values of the parameters were used.

Table 3-7. Concentration levels, mean and standard deviation of the mixture for "fun-
pump" tests.
Concentration, g/1 a, a2 P aUV,

0.4 1.00 0.00 2.61 0.08
0.8 0.50 0.50 2.06 0.57
1.2 0.67 0.33 2.24 0.53
1.6 0.50 0.50 2.06 0.57
2.0 0.60 0.40 2.17 0.55
2.4 0.50 0.50 2.06 0.57


3.5 I
Sand 2: Sand 1: Sand 1
3 = 1.5 = 2.61 Sand 2
S-Mix 1:1
2.5- = 0.14 = 0.08
2-
IL 2
0
1.5



0.5 -

1 1.5 2 2.5 3 3.5


Figure 3-27. Sediment size distribution for Sand 1, Sand 2, and 1:1 mixture.

The fall velocity analysis provided estimations of size distribution functions (PDF)

for both sand samples and the 1:1 mixture. The results are shown in Figure 3-27. Note

that the size distribution for Sand 1 is wider than that approximated from the sieve analy-

sis.











Prediction of Concentration and Sediment Size Distribution

The profiles of sediment concentration and size distribution parameters at any


given time after the pump was turned off can be predicted with a ID advection-diffusion


model (see Appendix C for a more detailed description of the model). The sediment dif-


fusivity parameter (equation C-1) can be taken to be zero, because of the absence of tur-


bulence production when the pump is off. When the pump has just been turned off, the


turbulence is not zero. Therefore, a small constant value can be used for the sediment dif-


fusivity to account for the residual turbulence. The sediment diffusivity was taken to be


0.0005m2/s. Other values within 50-200% interval were also tested; this did not result in


significant difference in predicted profiles.


Size distribution at Z= 0.30 m Time, s Sediment concentrations at given times, g/1I

0.06 :10
| -/ 2. 30 0.6
0.04 -40
L 50
0.02 -- 60 0.4 -
070
Size distribution at Z= 0.60 m 0.2
0.08

0.06 0 -
0 \Mean of sediment size distribution, i
0.04- 2.8 e).
U-
0.02 2.6
2.2

Size distribution at Z= 0.75 m ___
0.08
c). STD of sediment size distribution, o
0 0.6 f)0
0.04 / 0.4
02.02 0.23

0
1 1.5 2 2.5 3 3.5 0 0.2 0.4 0.6 0.8 1
Sediment size, 0 Distance from the top, m

Figure 3-28. Predicted with ID advection-diffusion model profiles of concentration and
sediment size distributions for maximum concentration of 0.8g/l at times 0-
70s: size distributions at a) 30, b) 50, and c) 75cm below the free surface,
d) sediment concentration profiles, e) profiles of mean sediment size in phi
units, and f) profiles of standard deviations in phi units.









The suspended sediment concentration profiles and profiles of size distribution pa-

rameters predicted with the ID advection-diffusion model for the initial concentration in

the chamber of 0.8g/l are shown in Figure 3-28. All the profiles are shown from 0 to 70

seconds at 10-second intervals. The initial size distribution was approximated from the

fall velocity analysis (see Figure 3-27). The parameters of size distribution were esti-

mated by fitting the normal distribution curve (equation 2-52) to the calculated by the

model distribution curves.

The model results show that most of the concentration profiles have a step, which

originates from the difference in terminal settling velocity of two sediment samples; the

coarse sediment, Sand 2, settles faster than the finer sediment, Sand 1. A similar irregu-

larity can be seen in the profiles for the approximated mean sediment sizes and standard

deviations. The mean sediment size at every range increases (in phi units) with time,

while the sorting decreases. The difference in the sorting along the profile can be quite

significant. It can be seen especially at times 10 and 20 seconds after the pump was

turned off, when the STD changes from approximately 0.2 to 0.5 along a single profile.

This can affect the accuracy of the inversion of the measurements taken by the ABS into

concentration and sediment sizes, as already was previously discussed.

Parameters of Inversion

For the inversion of RMS voltage profiles, the model defined by equations 2-10

and 2-11, was used. The modified high-pass model for backscattered intensity, defined by

equation 2-40 for the form function, and by equation 2-41 for the normalized cross-

section were applied with the parameters /f, and fz determined from the calibration

(see parameters for Case 4 in Table 3-5). The system function K was approximated by

equation 3-12 based on all available calibration profiles (see section "Determination of









Parameter Jf "). The resulting coefficients of the second order polynomial are given in

Table 3-8.

Table 3-8. Coefficients of 2nd order polynomial for system functions.
Frequency, MHz p2 PA p0
1.08 24.65 -1.162 11.62
2.07 -4.118 55.37 18.56
4.70 -199.9 359.6 27.18

The form function and normalized cross-section are shown in Figure 3-29. Both the

initial unmodified high-pass model curves for f ,h and X ,h (equations 2-34 and 2-35)

and the modified versions f:h and %,h (equations 2-40 and 2-41) are plotted. With the

provided parameters flf and 83 the modified functions have lower values than the ini-

tial ones. It should be pointed out that according to the plots the 1.08MHz frequency is

the least sensitive to the sediment size represented by Sand 1. For this sand, the sediment

attenuation, a, (equation 2-47), is also close to zero. Therefore in the test with 0.4g/l ini-

tial concentration when only Sand 1 was used (see Table 3-7), the inverted concentration

for the 1.08MHz transducer was much lower (with an average approximately equal to

0.07g/1) than for other two transducers (see Figure 3-30). The mean concentration profiles

were also estimated to be lower.

The profiles of mean sediment sizes were found by minimization of the error be-

tween estimated concentration profiles for each frequency at a given range (equation

2-25). The standard deviation of sediment sizes needs to be known in advance. Theoreti-

cally, the sediment sorting can be specified at each time for every profile. But since field

measurements usually lack of such information, and the sorting is estimated from the










sieve or fall velocity analyses. Therefore, a single value equal to 0.54 (p was used in all

calculations.


1.5 1.5
a). b). -- ,
',h' X h
S-. Sand 1:
1 \ 1 1 25 3=.2.61
G =0.08
S \.and 2:
01. F 1.50
0.5 \ 0.5 NN




1 1.5 2 2.5 3 3.5 1 1.5 2 2.5 3 3.5


Figure 3-29. High-pass model for backscattered intensity and its modified version used
for "fun-pump" test: a) form functions, b) normalized cross-sections.

The empirical model described in Appendix B was also used to invert the ABS

measurements into concentration profiles. The required parameters of the model were

estimated by using the same calibration data as for system functions and parameters /f .


and I6.


Results of Inversion for Concentration

The results of inverting the concentrations from the "fun-pump" test with a maxi-

mum concentration of 0.4g/l are shown in Figure 3-30. The profiles of suspended sedi-

ment concentration estimated from the 1.08MHz frequency transducer are shown to-

gether with the mean profiles. The mean profiles were calculated as an average of the es-

timations from the three frequencies. It was already noted in the previous section, that the

resulting concentrations from 1.08MHz data are almost four times lower than expected.

There may be several causes of this discrepancy. Some of them are:










* Erroneous initial values for parameters of sediment size distribution, used for cali-
bration, which can result in biased values for parameter fi, and system functions;
* Lower sensitivity of the 1.08MHz transducer to the used sediment sizes.

The mean sediment size specified for the calibration was taken as an average of the

sediment sizes for all used concentration levels. It was equal to 2.16 (p (see Table 3-4).


But since a mixture of two sediment samples was used for the calibration, none of sample

mean sizes was equal to it. As can be seen from Figure 3-29, the 1.08MHz transducer is

less sensitive to the finer sample than to the coarse sediments. Since two sample sizes

were not equally resolved, the use of their mean value can result in overestimation of the

system function and in obtaining a biased value for f/i. By exclusion of the 1.08MHz


frequency, one can improve the accuracy in estimation of the mean concentration pro-

files. But the decision should be made based on ability of the transducer's particular fre-

quency to resolve sediment sizes present in the sample.


0.4
Ini: M = 0.4 g/l, p = 2.61, % = 0.50 -- pump off
0.35 max off 10 sec
off 20 sec
off 30 sec
0.3 off 40 sec
off 50 sec
0.25 off 60 sec
off 70 sec
0.2

0 0.15

0.1

0.05 -


0 0.2 0.4 0.6 0.8
Range, m

Figure 3-30. Inversion by unmodified high-pass model concentration profiles for "fun-
pump" test with maximum concentration of 0.4g/l: mean profiles (thick lines)
and 1.08MHz profiles (thin lines).







78


Figures 3-31 and 3-32 show inverted concentration profiles for the "fun-pump" test

with a maximum concentration of 0.8g/l. The profiles in Figure 3-31 were calculated with

the initial unmodified high-pass model, which corresponds to the modified model with

values of parameters ifl~ and fz equal to 0.5 for all frequencies. The profiles in Figure


3-32 were calculated with the modified model with ifl~ and fz estimated from calibra-


tion (see Table 3-5).



.2 Ini: M = 0.8 g/1, P = 2.06 .5 pump off
max off 10 sec
1 off 20 sec
off 30 sec
off 40 sec
-0.8 off 50 sec
0 -- off 60 sec
.0 0 off 70 sec
0.6




0.2

0
0 0.2 0.4 0.6 0.8
Range, m
Figure 3-31. Inversion by unmodified high-pass model concentration profiles for "fun-
pump" test with maximum concentration of 0.8g/l: mean profiles (thick lines)
and 2.07MHz profiles (thin lines).

At the initial time, when the sediment suspension in the chamber was homogeneous

(profiles labeled "pump off") both models give similar results, since the system functions

were calibrated by using these profiles, and the correction for fly's is not yet necessary.

The high jumps in the concentration at the ranges near the transducer show that the inver-

sion is not accurate there for either model. After the pump was turned off and the sedi-

ments started to settle, the difference in the inverted profiles between two models become

more evident. The unmodified model overestimates the concentration over the range










r >0.5m at time 10 seconds after the pump was turned off. The profiles also have "steps"

which correspond to erroneously determined sediment sizes (see discussion in the

following section).



lni: M = 0.8 g/1, = 2.06, = 0.57 pump off
max- off 10 sec
1 off 20 sec
off 30 sec
off 40 sec
0.8 _off 50 sec
-- off 60 sec
0 / 0' off 70 sec
.'0
0.6
0 0.4 -


0.2 0

0
0 0.2 0.4 0.6 0.8
Range, m

Figure 3-32. Inversion by modified model concentration profiles for "fun-pump" test with
maximum concentration of 0.8g/l: mean profiles (thick lines) and 2.07MHz
profiles (thin lines).

The concentration profiles, both inverted using the modified model and calculated

by the ID advection-diffusion model, are shown in Figure 3-33. The mean profiles calcu-

lated with empirical model are shown as green lines. The mean concentration profiles

were calculated by averaging of the profiles inverted from all three frequencies. The rela-

tive errors were calculated as


C = () mdl ( ) (3-23)
N M/i dl (r, )


where Mmodel is the concentration estimated by the ID advection-diffusion model, Mac,


is the inverted concentration, with the summation carried over the indices at which


Model (r~) >0.05g/l, and N is the number of included indices.











Errors: mean 10.9%; ABn 8.3% Errors: mean 8.4%; ABn 31.5%
4 1.5
0 sec b). 10 sec 1.08 MHz
3 2.07 MHz
S1- 4.70 MHz
2 -- Mean
5 8 0.5 / Model
Error: .- Mean (ABn)
0 0
Errors: mean 45.2%; ABn 17.9% Errors: mean 59.9%; ABn 15.9%
1 ----0.8

S0.6

0.5 0.4
o 0
0 0.5 -


0 0
Errors: mean 60.6%; ABn 15.0% Errors: mean 56.6%; ABn 40.2%
0.4 0.4
e). 40 s c f). 50 sec
0.3 0.3

6 0.2 02 -
o o.1- o. 1



Errors: mean 60.7%; ABn 49.5% Errors: mean 72.8%; ABn 39.0%
0.4 g-) 60-ee 0.2
g). 60 sec h). 70 sec
0.3 0.15

d 0.2 0.1-

0

0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Range, m Range, m

Figure 3-33. Comparison of concentration profiles inverted by modified high-pass model
with profiles calculated by the ID advection-diffusion and empirical models
(labeled as "ABn") for the "fun-pump" test with maximum concentration of
0.8g/l at 10 second intervals.

The sequence of plots in Figure 3-33 shows the dynamics of the two sediment sam-


ples settling. The profiles corresponding to individual frequencies (blue lines) show in-


creasing stratification of sediment suspension over time due to the different settling ve-


locity. Again, when the content of coarse sediments gets low, the concentration estimated


from the 1.08MHz frequency is much lower than obtained by the other frequencies. The


4.70MHz becomes the most accurate predictor of the concentration when only fine sedi-









ments are present. The concentrations estimated from the 2.07MHz measurements are

closest to the estimated mean concentration.

The empirical model provides the best estimations for suspended sediment concen-

tration. Although it does not use the information on sediment size, it still can be applied

for the inversion.

Results of Inversion for Sediment Size

The sediment size distribution parameters were also estimated. The results for the

test with the maximum concentration of 0.8g/l are shown in Figures 3-34 and 3-35 for

unmodified and modified models; they correspond to the results presented in Figures

3-31 and 3-32. It can be seen from Figure 3-34, that the unmodified model provides esti-

mations of mean sediment size which are closest to the expected values at very low con-

centrations (at time>30 seconds), but fails at higher concentrations, for example, profiles

at 10 and 20 seconds. The model predicts sediment sizes which were not present in the

chamber (see Figure 3-27), aliasing the estimations towards finer sediment sizes. The

only approximation which is reasonable is at time t=0 seconds; and that profile was used

for the calibration of the system, therefore the inversion is much more accurate.

The estimations with the modified model are shown in Figures 3-35 and 3-36. Sev-

eral definite improvements can be noticed. The model now captures correctly the dynam-

ics of the changes in mean sediment sizes along the chamber with time. It was predicted

by ID advection-diffusion model (see Figure 3-28e) that the mean sediment sizes de-

crease with time at any horizontal level after the pump was turned off. Figure 3-35 shows

that the inverted profiles behave similarly to those predicted by the numerical (ID)

model. Also, there is a limit for estimated sediment size at approximately 2.4 (p, which is

close to the mean size of Sand 1 sample, used in the test.