<%BANNER%>

Impacts of Diffuse Recharge on Transmissivity and Water Budget Calculations in the Unconfined Karst Aquifer of the Santa...

Permanent Link: http://ufdc.ufl.edu/UFE0021715/00001

Material Information

Title: Impacts of Diffuse Recharge on Transmissivity and Water Budget Calculations in the Unconfined Karst Aquifer of the Santa Fe River Basin
Physical Description: 1 online resource (145 p.)
Language: english
Creator: Ritorto, Michael J
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: fe, floridan, karst, oleno, river, santa, unconfined
Geological Sciences -- Dissertations, Academic -- UF
Genre: Geology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: In karst aquifers, high transmissivities make it difficult to conduct traditional aquifer tests. An alternative method to evaluate formation transmissivity on a large scale is passive monitoring of observation well response to changes in stream stage or conduit hydraulic head during high discharge events. During these events, increases in water levels at observation wells are dependent on the stream or conduit boundary condition, but are also influenced by the amount of diffuse recharge arriving at the water table. This study used data from the karstic Floridan aquifer in northern Florida to evaluate the effect of neglecting diffuse recharge on the transmissivity estimate. At the Santa Fe River Sink/Rise system, the Santa Fe River enters the Floridan aquifer at the River Sink, and re-emerges downstream at the River Rise. Between these locations, cave divers have mapped over 7 km of conduits. Diffuse recharge was estimated as rainfall minus evapotranspiration calculated using the Penman-Monteith method. The transit time of recharge through the unsaturated zone is not known; given the relatively thin unsaturated zone and its expected high hydraulic conductivity, the simplifying assumption was made that recharge reaches the aquifer on the same day as rainfall occurs. Hydraulic heads were monitored through time at locations within the conduit system and in wells monitoring the surrounding aquifer. Four storm events were recorded by the monitoring network. Aquifer responses to storm events were first predicted using an analytical solution created by Pinder et al. (1969) that does not include diffuse recharge. For comparison, aquifer response was then simulated using a one-dimensional model implemented in MODFLOW incorporating the calculated diffuse recharge during the storm event. In the model, the time-varying hydraulic head at the conduit was simulated as a specified head boundary condition. Calibration of the model for individual monitoring well hydrographs yields estimated transmissivity values for the aquifer between the monitoring well and conduit. Results suggest that neglecting to include diffuse recharge in the solution results in overpredicted transmissivity values. Results also suggest that the magnitude of diffuse recharge in comparison to head change in the conduit determines the relative effect of the diffuse component on the observed well hydrograph after a storm event. In terms of total water volumes, results show that diffuse recharge can be more than twice the input received from the conduit boundary during small storm events, and less than half the input received from the conduit boundary during large storm events. These results suggest that in future calculations it will be important to consider diffuse recharge as an influence on dissolution across the entire Santa Fe River basin.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Michael J Ritorto.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Screaton, Elizabeth J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021715:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021715/00001

Material Information

Title: Impacts of Diffuse Recharge on Transmissivity and Water Budget Calculations in the Unconfined Karst Aquifer of the Santa Fe River Basin
Physical Description: 1 online resource (145 p.)
Language: english
Creator: Ritorto, Michael J
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: fe, floridan, karst, oleno, river, santa, unconfined
Geological Sciences -- Dissertations, Academic -- UF
Genre: Geology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: In karst aquifers, high transmissivities make it difficult to conduct traditional aquifer tests. An alternative method to evaluate formation transmissivity on a large scale is passive monitoring of observation well response to changes in stream stage or conduit hydraulic head during high discharge events. During these events, increases in water levels at observation wells are dependent on the stream or conduit boundary condition, but are also influenced by the amount of diffuse recharge arriving at the water table. This study used data from the karstic Floridan aquifer in northern Florida to evaluate the effect of neglecting diffuse recharge on the transmissivity estimate. At the Santa Fe River Sink/Rise system, the Santa Fe River enters the Floridan aquifer at the River Sink, and re-emerges downstream at the River Rise. Between these locations, cave divers have mapped over 7 km of conduits. Diffuse recharge was estimated as rainfall minus evapotranspiration calculated using the Penman-Monteith method. The transit time of recharge through the unsaturated zone is not known; given the relatively thin unsaturated zone and its expected high hydraulic conductivity, the simplifying assumption was made that recharge reaches the aquifer on the same day as rainfall occurs. Hydraulic heads were monitored through time at locations within the conduit system and in wells monitoring the surrounding aquifer. Four storm events were recorded by the monitoring network. Aquifer responses to storm events were first predicted using an analytical solution created by Pinder et al. (1969) that does not include diffuse recharge. For comparison, aquifer response was then simulated using a one-dimensional model implemented in MODFLOW incorporating the calculated diffuse recharge during the storm event. In the model, the time-varying hydraulic head at the conduit was simulated as a specified head boundary condition. Calibration of the model for individual monitoring well hydrographs yields estimated transmissivity values for the aquifer between the monitoring well and conduit. Results suggest that neglecting to include diffuse recharge in the solution results in overpredicted transmissivity values. Results also suggest that the magnitude of diffuse recharge in comparison to head change in the conduit determines the relative effect of the diffuse component on the observed well hydrograph after a storm event. In terms of total water volumes, results show that diffuse recharge can be more than twice the input received from the conduit boundary during small storm events, and less than half the input received from the conduit boundary during large storm events. These results suggest that in future calculations it will be important to consider diffuse recharge as an influence on dissolution across the entire Santa Fe River basin.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Michael J Ritorto.
Thesis: Thesis (M.S.)--University of Florida, 2007.
Local: Adviser: Screaton, Elizabeth J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021715:00001


This item has the following downloads:


Full Text
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 E20101108_AAAABZ INGEST_TIME 2010-11-08T21:47:18Z PACKAGE UFE0021715_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 93909 DFID F20101108_AABRPS ORIGIN DEPOSITOR PATH ritorto_m_Page_039.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
1dd8f49f1311bb3742188319a2b8e09d
SHA-1
3911cae7b6a1dec29507efe4bf654ab5ac15c298
41880 F20101108_AABRQH ritorto_m_Page_057.jpg
7fe4552e4bfb56b8bae84d5a84958739
f3d313debc31fe19b6a7eaa7084f53e7efc4122b
81642 F20101108_AABRQI ritorto_m_Page_058.jpg
070701a917ce183d494a5981e8de50cb
05c25e0e04d183abb4afeded0006264af989021c
89667 F20101108_AABRPT ritorto_m_Page_040.jpg
ef2597b271881ca694634ad43ee2926e
d014fb3ae02136a0605b55a11f53f232d0e6dba9
86296 F20101108_AABRQJ ritorto_m_Page_059.jpg
0427651b317fd92ac67690febd52f576
55ef20db5a553a1d43186e7dd1230a67d34bcd57
89275 F20101108_AABRPU ritorto_m_Page_042.jpg
731d0638bf0092141123f08799250e74
fb266f22895993c9cb1c0797631f2942387211fa
92481 F20101108_AABRQK ritorto_m_Page_060.jpg
a959aa6dfe527cdc228b534a30a2f245
c58d5316b98f41aba0b1011cd689addf173188dd
68030 F20101108_AABRPV ritorto_m_Page_044.jpg
13ad8db32194d313b48d1e724993cb67
624559743e948bf5f40dc085118909488d660d1b
88289 F20101108_AABRQL ritorto_m_Page_061.jpg
30dad58d09aacde6530ff5a8e0b45b02
39057d39c22c7a12fe6db904a667add05cef1485
84524 F20101108_AABRPW ritorto_m_Page_045.jpg
e1708b302677d8074414a5fe6650fdae
d1d27ebcccc709c807dd335a3b278f302b8687c8
24311 F20101108_AABRRA ritorto_m_Page_083.jpg
4c3c48aa9824eaeca37bf4766db97b6b
1ea9ce5ed99b9d4f1df06afe7f0baf7022230fe7
93388 F20101108_AABRPX ritorto_m_Page_046.jpg
f351fae67e3c50c5e4a5326e0b046de3
dddc6c7eb547ee83623c1c8c50bd72757da93e95
23442 F20101108_AABRRB ritorto_m_Page_084.jpg
e0056fa47f69bce67326357692c3ab06
d4aa6d6fe04349fcce1ba8faec1f4482dc2b4fad
86947 F20101108_AABRQM ritorto_m_Page_062.jpg
0c34e681f282a6a819a4f64b7754b679
108d0037ec6324776677f32edca332e6a5495cfc
85412 F20101108_AABRPY ritorto_m_Page_047.jpg
536afd6410512978df13050f11b1e6c0
f0754c29a1f562d1ff254a51c1681438fd8f6856
20342 F20101108_AABRRC ritorto_m_Page_085.jpg
1dccb9fa3ded62bd96d790af8ee3765a
0342f09d733eb78923da15935072a2daf3ee8c77
93063 F20101108_AABRQN ritorto_m_Page_063.jpg
d78915982ef0d89c35333b6ddb043aa5
41deaea825bf878052ed9dfe765e524e2c2401f0
73309 F20101108_AABRPZ ritorto_m_Page_048.jpg
66f71be0166cab9a0465a92dc340c5c0
ac2a2af7add3cec1c292f2ea148aa51418b3b6d4
19391 F20101108_AABRRD ritorto_m_Page_086.jpg
c5f7785bf5caf0bbad25591e2de23153
ec389b07c6a1744ba712d5cc7f048e1702722623
95197 F20101108_AABRQO ritorto_m_Page_064.jpg
5a04a7eee1cd174fe81ec96bbc756bd4
16544f1c9c7cfc3a880f63d2ba152a132186753b
25773 F20101108_AABRRE ritorto_m_Page_087.jpg
66a0775e975b7a8ec17bf0e9f1e54bb0
5697003f6cb1117ff601455fcc437a40bdb74d71
94766 F20101108_AABRQP ritorto_m_Page_065.jpg
7256d15e96fdcf32cd6fbb667f1c2ab8
6c9c9ba117dfe93584a9f3dc38d80dc929f9f8d2
31847 F20101108_AABRRF ritorto_m_Page_088.jpg
323ff36ef216a3974b1c98efe283b9ee
cfb874f13aad2c75a610067b9790fbe5b32e6c74
86171 F20101108_AABRQQ ritorto_m_Page_067.jpg
3f847c6d0f435345e1556e4df4273bd2
42ee069dbf14286109b24f3207647421523c3ad5
32549 F20101108_AABRRG ritorto_m_Page_089.jpg
3df9db7bc5d303810477dbce309f53b3
f2b09c8b3589026b05b9aa5ce578264bba0fed5f
90418 F20101108_AABRQR ritorto_m_Page_068.jpg
a5ab95fcaca9626aae9d69f723a60b06
b3626cdd21391c3bdb9398325534e063606bc077
32817 F20101108_AABRRH ritorto_m_Page_090.jpg
a32e557774d49ac85cd3d4998fcce058
6c0bd30b566170b98ccf2fcd709a70ec2bcbcce8
92620 F20101108_AABRQS ritorto_m_Page_072.jpg
fbf2526b09088ddf4171cfe02bb752ff
85a48c14e7c7fa2c7b283eb1d39b2048adbee8a1
32111 F20101108_AABRRI ritorto_m_Page_091.jpg
17b18b15b5ec05be713248aa08c096cf
6b9b64ac4492615c8d350e3f49f1b640024d348a
92315 F20101108_AABRQT ritorto_m_Page_073.jpg
909116112d9f2481f4157638a44ce119
54e53e085aaa9974deb0e481c2a0aa9d92c7a4e7
33535 F20101108_AABRRJ ritorto_m_Page_092.jpg
4476969f398006c4f3e1406c02453ade
632bf1ac6019abcfff635e6020c136bf4023db56
88524 F20101108_AABRQU ritorto_m_Page_074.jpg
2d915e1409c17e47fa4e5acd99c513b7
6242d4a7106f83ffd91782383cdacab172a61124
48492 F20101108_AABRRK ritorto_m_Page_095.jpg
338e9d12c681491300901cbd745a7108
0d603774109bdcca10a551975486c35983bbfe4f
91491 F20101108_AABRQV ritorto_m_Page_075.jpg
1111783f799acf4ffbf922a027c3407e
51e232df5054ab80915802f8efa9a149821b2d6f
45415 F20101108_AABRRL ritorto_m_Page_097.jpg
77acab58a3f43a36096973c3b18b829e
1eb3978ae79e25c3f14ba59b547fc7ed3cedfc9f
81006 F20101108_AABRQW ritorto_m_Page_076.jpg
d7b6f37932ebe80e890ae6eab6af1124
091d3623c9e4e095e6b7f96e763e11a0e5f1ceb9
23854 F20101108_AABRRM ritorto_m_Page_098.jpg
743b208ebb023cb7fbac372910d771c8
999f26fa52cda66747cfdf5f4c94635d1e82736a
25270 F20101108_AABRQX ritorto_m_Page_078.jpg
f7171e8c349ab3b2e0cb1e887817ea91
5a3a9ca5e279dc041fbfba71cefdb59954312d79
22818 F20101108_AABRSA ritorto_m_Page_120.jpg
f430b83db9550922bc4a4654488bb4b1
28c89a59bc298f7809c2b9590d433bc4b1c555d3
30468 F20101108_AABRQY ritorto_m_Page_080.jpg
8c62d3bb7823e67d2d7426d8f8ca864c
99ea13e182ee8cebb47809acb494175150738216
35423 F20101108_AABRSB ritorto_m_Page_121.jpg
920d005dfa02fd3eb1f9f379df3e9608
f34e348e15578cf9f06271d2c792a6da5a0126f0
26305 F20101108_AABRRN ritorto_m_Page_100.jpg
d415a08a381e1a7a8e09cecc41e5d5ff
cbdb3c47e846f98f7044c19d22628aad4fcdf437
24900 F20101108_AABRQZ ritorto_m_Page_082.jpg
e9f534d72e0bae11e640cfdaac4f77ea
a698d5503f7abfc2281b457b4f4af92c8524a027
83843 F20101108_AABRSC ritorto_m_Page_122.jpg
d50d241134308f9c04c3164e5c72a2a5
193f20b4093cb456a2ee0cd9a17dac50c171ef87
42108 F20101108_AABRRO ritorto_m_Page_103.jpg
5903011d5902923fee5f6cc42d434903
83fdd56cc9c5478b731142012ff2a61b8c55693e
90838 F20101108_AABRSD ritorto_m_Page_123.jpg
dad29ec1b4fdaf24f21ad587b2d18eaa
eef5f3a045bc48bf3cfdf2926dd86819f469e773
42449 F20101108_AABRRP ritorto_m_Page_104.jpg
92061802de8255f282a7b4a2f67ab830
86b84754eca25b7d6f905dea480f06c9fe5d14bf
93825 F20101108_AABRSE ritorto_m_Page_124.jpg
af3abc89d81a3d0a95a2dc1b002bc691
ccb911e5219d4f6c8483fa3cc5f745d1e97b2ee2
46504 F20101108_AABRRQ ritorto_m_Page_105.jpg
c1e8a41d1971f06636b32b20be54756f
4d36b1136366eea8288e1ab7ecda14c419cded21
91542 F20101108_AABRSF ritorto_m_Page_125.jpg
5006db84eca5f29262b66fe05c5f0afd
7b303e641d143ee1f4af5f15457188f20154b9cd
47255 F20101108_AABRRR ritorto_m_Page_106.jpg
cced8d1cb0768d98ce4059bd073c5331
895817f61c1e4bbb66105bd850f671de734126f9
87756 F20101108_AABRSG ritorto_m_Page_126.jpg
c8eba4a433d1332047a84e7522def957
f9d0b922dcc2b55864df77862b56e2c602351b43
47476 F20101108_AABRRS ritorto_m_Page_107.jpg
6788f0ef735793511c957e83f1870805
3d2430de58d8ce510bdd177a8026a1aec454b0c6
87564 F20101108_AABRSH ritorto_m_Page_127.jpg
b9faad14392181ec4068d433511fd976
d6d1f59e5db7198cc63c3736a476ff6edecbaa10
42782 F20101108_AABRRT ritorto_m_Page_109.jpg
c65c34d3bb560c99d82bcd52974d7a4b
24c2ce2b54e3ddb409801b9f5b66bbfa47aeacf3
92330 F20101108_AABRSI ritorto_m_Page_128.jpg
913de458e1b81c8ddd8acfec13e40ed9
1d9ff8526c2936ee9962c7e25df21ae735f11d60
43003 F20101108_AABRRU ritorto_m_Page_111.jpg
b1143092d4812a864c636322b0966b78
0cdec4d95643f90124118bd64e797e1795d35abd
96603 F20101108_AABRSJ ritorto_m_Page_129.jpg
5c5e939bad7e89247f2f2abe9793cd36
07180e73c98ed9ea7b02e31eda9c0164259b99bb
43275 F20101108_AABRRV ritorto_m_Page_112.jpg
a0d91ef1bfbb574519794fee83fc8f0d
f4c12bc118288bcb461f70c7c115c14f864b4dc8
89238 F20101108_AABRSK ritorto_m_Page_132.jpg
9f397be9d9429ed2063c94939d95dbde
8aff04e3adab854a27e31a7f16ccceec15a63a9c
43508 F20101108_AABRRW ritorto_m_Page_113.jpg
c6a7d88c9a2a022752d303c9305c6d59
00399d05dfd44a54eba338bb8af96d60532fdc66
87075 F20101108_AABRSL ritorto_m_Page_133.jpg
7c531b2b2d2c83e6f68f8874cf2c4209
bed7e05ec6d6ec2b4e95905f9102f2c18d27b24c
36153 F20101108_AABRRX ritorto_m_Page_115.jpg
23faecbaf3e7dfeb8e5b54e8812fb62c
9a4a3419467980f56c1c91b034a9c4ecd34733d2
1051977 F20101108_AABRTA ritorto_m_Page_007.jp2
6494b9ab59a1919a8a2925c017a7d501
d48c811ab03272999d932e40021155df8946ad3b
93588 F20101108_AABRSM ritorto_m_Page_134.jpg
fe382a149f5ae9a7cd5d21bebf12e60d
c26fe805831b6782747c293c5701292aab30e068
24222 F20101108_AABRRY ritorto_m_Page_116.jpg
5279626c429140188349a5a42c670a9a
d752f5dcf3130063dc9436374203e17e755001c5
1051967 F20101108_AABRTB ritorto_m_Page_008.jp2
2c463a485a8ba6e5bf66de6c15f35577
34ff9c1402cf46c2049cb1832589fa8fe0257314
88642 F20101108_AABRSN ritorto_m_Page_135.jpg
b36440df20662d8404c2cb516e8906e6
c108925f59286b594673d6d8aa0b23d4ab29f1bc
34285 F20101108_AABRRZ ritorto_m_Page_119.jpg
5f05ea3ca9768b684f68c6c9d3bfc1e6
7a85266a0fc229980328dc42c80c5c4bb3d980ab
1051985 F20101108_AABRTC ritorto_m_Page_009.jp2
7b1fca676becb5915bffb8e481f43f32
d8ab7beecb5bbeb41d2bed1558715908ccaa2819
88221 F20101108_AABRTD ritorto_m_Page_013.jp2
e63413e1c878fc0bc398f6bba9f89bed
42b95225c864e887364297a07b86a4b204639f38
91620 F20101108_AABRSO ritorto_m_Page_136.jpg
d2c4e968d23d6ea10a60d2588a3c95f0
0721194e8afb85d09cf909c4491d2c91e1006bda
113634 F20101108_AABRTE ritorto_m_Page_015.jp2
3ffbabfc14686231d50383ec5e7f9ea8
d5e26dd58069ef0618002f29b6fdb430cef528ea
36464 F20101108_AABRSP ritorto_m_Page_137.jpg
33adc9ea630197a03293f3567387cdd9
6acc5f347100877990b43c9bb992c73cc45b6721
111961 F20101108_AABRTF ritorto_m_Page_017.jp2
760b2c2b47f06fab541b8b7a56f17cba
e17329d3875056d4d8ceeacc18f3c761c0a7f841
24679 F20101108_AABRSQ ritorto_m_Page_138.jpg
1549cbdec3436ee16f4bd85b3583a059
e3c45e6b83bf4337e18cbb4a1bad9a4388ddad5f
30821 F20101108_AABRTG ritorto_m_Page_018.jp2
4956b67df6a26225dd83f91f038acb70
38dd8c62f57a5b18df38333f8e7ec9682c42211a
82862 F20101108_AABRSR ritorto_m_Page_139.jpg
5dfecdb9c8bc27edca0ad0487226e943
427f5071f223131a8e3bd835bf54c735dfe31b02
103765 F20101108_AABRTH ritorto_m_Page_019.jp2
4d6a5cb3ceaf0ffef7e1927e52662277
78d308cfe5a68b5d13e5d0037fbc24a1636ec632
64402 F20101108_AABRSS ritorto_m_Page_140.jpg
f89a47c6170501c0073e452d816fcb30
f123406cf154c11803091f251cc60320adb6f905
119173 F20101108_AABRTI ritorto_m_Page_020.jp2
ba386184b9bc1f3c5ac595eba18c386d
df68d6048cf0ce0c8d4413237c0f77400b1dc0de
90236 F20101108_AABRST ritorto_m_Page_142.jpg
078df4afe76ef6d6e8d0aeaa8866fdc6
7c0f39b5187d21d18c7ae6f09e338b0a90574431
114324 F20101108_AABRTJ ritorto_m_Page_022.jp2
b6e426b3de1caf8e1463ce04bd9e2053
deafd79212275e0600332707858e86742d51dbcd
55717 F20101108_AABRSU ritorto_m_Page_144.jpg
502ccb9974c80d6002ed7e8a7dbfb390
502e7efa03b70b9ddbddeaea2c29b82c5c7db6ac
119513 F20101108_AABRTK ritorto_m_Page_024.jp2
5011c7f8d4ebcab279616c505c485e92
08e64294d27c7a6d3c654ef2ab22640fade075c9
32084 F20101108_AABRSV ritorto_m_Page_145.jpg
9dcefacfd54c1331d98f3ffef9acd920
96410c65b2c0b3607d81830177597820bfda3df8
117945 F20101108_AABRTL ritorto_m_Page_025.jp2
dc23e91d63ed0cacb77427864554b6fe
b0f9f46dec4a10ff3c12422da6ac566fc903b91c
27738 F20101108_AABRSW ritorto_m_Page_001.jp2
c57b139ecc7683d2d5347c9a02f270af
28962050dbddb5035315bcbe98beb28ff0b82639
86789 F20101108_AABRUA ritorto_m_Page_044.jp2
e48fc4807638e49bfdcd155b9fe39937
bba2c2fc08d204d6966120bf65ac7abf295db3d6
117901 F20101108_AABRTM ritorto_m_Page_028.jp2
c4f6d4f769f63a0cbce317a47fceaed7
8b51892adb3b2b02150401b5725cca6ee374bab5
5013 F20101108_AABRSX ritorto_m_Page_002.jp2
305811d0d1aeeda0ea2475b4a26061ea
dc98b9bee0f1573e676f6ac1588dc90244afa5c1
109191 F20101108_AABRUB ritorto_m_Page_045.jp2
16c1b7879f9fcc9e9844fe43a8b7c4d2
c8b666cc8a1d4b8143181af8c240cdb0838f9677
1051918 F20101108_AABRTN ritorto_m_Page_029.jp2
be009c06a2ed92c460a72be61807a736
28152930fbc18c50a1b14d4d13c729ba1126f720
1051961 F20101108_AABRSY ritorto_m_Page_005.jp2
b3890a2fa467d5fa939afbd4ed0766c5
c25b7fbd23c387dcadc7d0e8b52af89a282f6a88
94972 F20101108_AABRUC ritorto_m_Page_048.jp2
7cfad38b2eae4f5c698cd2a671b2090d
9ae965b910a7a7c7330e44460e8a1ac304b14e9d
121561 F20101108_AABRTO ritorto_m_Page_030.jp2
d04fa607aaef5e08b1ccecd097ef47d3
b4f74c8b52ffeb30b29dc6578a64ae312050a121
1051982 F20101108_AABRSZ ritorto_m_Page_006.jp2
5717902b6658a9ed34e8fa5d7e5a64cf
1ffbf340a5768145716a5bef184d319b36b83f20
105647 F20101108_AABRUD ritorto_m_Page_049.jp2
1b694493d874d626d903b0ba7ba823b7
bd4bb66f254b9120e4b2049112c42a1a05fc9de5
40703 F20101108_AABRUE ritorto_m_Page_053.jp2
9d28ce92d992eddb89d9d7e82b6eac15
78134daa5b19469a18b52a5c5d085c70fc8737f4
108821 F20101108_AABRTP ritorto_m_Page_031.jp2
fb18eeaf9b4f92f623a4d064e73ad3fb
b19896be1b8130316d8bf7eacc3d53fb9d7f7775
366095 F20101108_AABRUF ritorto_m_Page_055.jp2
cf389a56685283280d0709a83d22c2bf
eb61a7ab439c493eaab26adf934116e19311b603
116915 F20101108_AABRTQ ritorto_m_Page_032.jp2
f7f36f59f7e013a2aad704c6a0d61db6
89a6ba0a9910757cf0d0339ed6d0f484a5a7c8be
42598 F20101108_AABRUG ritorto_m_Page_057.jp2
5de3b5f2f8bf9fe988a43f1c3040dbef
4687812994e2d0a526a0e026b5085f7eaaca80ea
121451 F20101108_AABRTR ritorto_m_Page_033.jp2
900751fbbc20c109dbf480b536a98179
00db11fba3ca4d4784cce59cb2d136b3aa427dae
1053954 F20101108_AABSAA ritorto_m_Page_123.tif
b6f0e87ddde5d62587ee5e9ca2ba5800
02b3034f5fddef29bdda653762a5d04ba68c7846
70703 F20101108_AABRUH ritorto_m_Page_058.jp2
8737b12c539d83fee8bd681495c8af4e
48d92dee4955c7b2b4b69507506caacfdf695d89
94538 F20101108_AABRTS ritorto_m_Page_034.jp2
8d67842982ea7baca0b6c98197b2f747
3c10e6cc60c2703ff28279084a0e20e3b67524ef
F20101108_AABSAB ritorto_m_Page_124.tif
4f4953a56743e5d4005ae9e7fda4d5c3
27cb4a859ae66ab3d684475849be5bdb846b0080
111634 F20101108_AABRUI ritorto_m_Page_059.jp2
66f68452891fcda7e100472bb672f5fa
17ea023909e491fec1d5b3bb2baf31d6ff5a6875
478391 F20101108_AABRTT ritorto_m_Page_035.jp2
a6a99fcb79049ed747d83271ee6ed871
654fd6422de46650915d4b67ce9d6b3f1da6900f
F20101108_AABSAC ritorto_m_Page_125.tif
966d6c3c34644864e18d50401ca5ce3c
23c547ba7e4a9298fe44e43e239d7835abef465c
118328 F20101108_AABRUJ ritorto_m_Page_060.jp2
2f0d5eeecca0a6d13556328bfe731157
f9ec78074ab773f3856c21ea003408ee82c93a15
1051965 F20101108_AABRTU ritorto_m_Page_036.jp2
4740cd34a8e138e7a276dd6ccee1b1e0
18964442de53b90d088abf287b23471bdea088b3
F20101108_AABSAD ritorto_m_Page_126.tif
86e9cd5e80a2ae26ef4ed79317a2a85b
d6b5b71726f70027f0976764777e52e1ad546371
114499 F20101108_AABRUK ritorto_m_Page_061.jp2
3310dbced20e249c8870b3cfe8c4af8b
540177faaeb9ba3b3b7d5ae390e7a9ed96271f06
64902 F20101108_AABRTV ritorto_m_Page_037.jp2
c50da70243af8b7454bacbc209dc564a
9736e75ae77f17e7ef5e744cb95c0845a78e8f57
F20101108_AABSAE ritorto_m_Page_129.tif
267a92d191b2dc690ac5c4a18938c989
cf411b4ff077a53501a19ca55a481848fee3c5cf
120301 F20101108_AABRUL ritorto_m_Page_063.jp2
d0b218461603e7f3071d3e9e7b80ad01
5db386854024247a390e5cf6bf14c2607eea0075
113613 F20101108_AABRTW ritorto_m_Page_038.jp2
5a4d4baf9269c3be6a1bc5d51260fe89
644da245a323611f59ec836525a2cca5ec1a4f14
F20101108_AABSAF ritorto_m_Page_132.tif
27699b2bacad93875f77fdc8aff3fdc7
937e6a050292ec2a90e0a48e25b72f3b79de741f
123852 F20101108_AABRUM ritorto_m_Page_064.jp2
eba7c11ba760bfb1f6c56b6bbaba8eeb
bbd1282694657769afb2b4ad983b5ab56cce97cf
123190 F20101108_AABRTX ritorto_m_Page_039.jp2
a9f885ca14cb2c410c61826a9f498b20
a2f6b17235b830e76d4b45ce61ad3a2f6ed9b234
F20101108_AABSAG ritorto_m_Page_133.tif
bc4429f8ed8752154798c07d892c4da5
5a268c871e7fffb2b7392eccb6b2ab560e65e277
581780 F20101108_AABRVA ritorto_m_Page_083.jp2
b1241673a9c261675ef62d7008bb440c
f91b893da0dec5c697059aca34fd96da7c307dd7
120588 F20101108_AABRUN ritorto_m_Page_066.jp2
ca3abf5cd3d764df7fbadac120948cce
48bea477c8b49a3a9db7832ef7afe6116281610d
1051979 F20101108_AABRTY ritorto_m_Page_041.jp2
10f6bca8400d0fc850b876f6967a2a69
eb0d4120674c5a7e983ba5318fe7887f327a6639
F20101108_AABSAH ritorto_m_Page_134.tif
eeff492b115b3fc1d6c93668d4cac5e2
199f2d17407d61fc01ae9f7fc63876f0a9e101c3
393139 F20101108_AABRVB ritorto_m_Page_085.jp2
6f9322f7d9befd7d94e4df20ae8cb17d
b533ffa55da9fbb95025a56d6cad45ec2efc1c8d
118226 F20101108_AABRUO ritorto_m_Page_068.jp2
d26e7f0535c9e2daf43de73d1ed2a1fe
d308f392381009267b3dfcc42e6c1839beff55b6
116947 F20101108_AABRTZ ritorto_m_Page_042.jp2
98e0eb49573f15e51f7d1f7d77367a91
c3f44c085f4b4f037db28a0b6afa11e06deb8bfe
F20101108_AABSAI ritorto_m_Page_135.tif
57d37c261a7fe040772f66ca6e0251bd
919769613a10df62208696d6d2e1fccfdf48277b
292031 F20101108_AABRVC ritorto_m_Page_087.jp2
dba6ea873bea576ad9136cd4c08212a8
44bf2f728b56381c995b3c7c921a9edc67803cea
116668 F20101108_AABRUP ritorto_m_Page_069.jp2
64a24f3fd408261a80629c14bf0199a0
95590b9be0adbef717f79eacb58637ef5556ba88
F20101108_AABSAJ ritorto_m_Page_136.tif
3e039edabff7ef8546b853124395c16a
98f7c5ba7ca98d2b134d7ce46741ada45a1fcf14
625611 F20101108_AABRVD ritorto_m_Page_088.jp2
9a234c915218b279aab38aaece7901ac
1a1aee7fc498178690546feadb15cbd4f25f93fd
F20101108_AABSAK ritorto_m_Page_137.tif
696881076ddb26d179a848801094fd78
09a4d0c43870c1f2149aa225269f19d31a0202b1
642316 F20101108_AABRVE ritorto_m_Page_089.jp2
f6e786f0caf91a7f21748204ad8b316c
2639c2b477374b1b912537748c3bd99100c7d2bd
1054428 F20101108_AABSAL ritorto_m_Page_138.tif
c448b37811fe118f7617dd336d55cc4d
19d5c3033d868c527ea415c9d4678c177e5149ec
666361 F20101108_AABRVF ritorto_m_Page_090.jp2
9c123774652ac8c2154d423025334fc0
e5203a9fc69dbf1ef6a44d7d4988daf5925f1318
112780 F20101108_AABRUQ ritorto_m_Page_070.jp2
104fccd1e24a1f19d4ff41d3d6d62d57
7f6e8aa047d79882f64deca4b2cb6c879a75cef1
53563 F20101108_AABSBA ritorto_m_Page_021.pro
faebf2e06d9ac43dacf05e5d4936b113
ede8a4ca25b55d80bbf37e4e4fe9c8e6e3a7fa63
F20101108_AABSAM ritorto_m_Page_139.tif
b73c44c5876533d01df8f14430437b08
adb80b764dd6c19d74dcedeb89cace9f9a9caa76
681150 F20101108_AABRVG ritorto_m_Page_091.jp2
6608f10e706c47737bb123958dd14233
1089d7a116e245bccf428508fd2dd0aaf945e760
114922 F20101108_AABRUR ritorto_m_Page_071.jp2
29918da1b4712621eff530114d3cf826
168d919e4890bcc8fd44163e4a1399fb5c3f28be
56076 F20101108_AABSBB ritorto_m_Page_023.pro
6a0cff392f4667e447d3a121a3b0c510
6cdeeb25b42815f3035b662dddfb806dccd3f58e
F20101108_AABSAN ritorto_m_Page_141.tif
d78e854d4480235a9e92095914500b68
5bf97a73104a8285c2cf13aba262ba778e3c2070
709020 F20101108_AABRVH ritorto_m_Page_092.jp2
3bef69fbdda7971dca7192f2f3e0e7fb
e0a89a3b11789df08025a16278e3caca059c9928
120363 F20101108_AABRUS ritorto_m_Page_072.jp2
8e31f9b5fe25332f46a41b71c27b73c3
bf69fc3f7f846152d63eadff58722badbb859588
54626 F20101108_AABSBC ritorto_m_Page_025.pro
ea60192596691b579164e49fa9359194
519a7fae9de608dd9138fd95dc91ba2aba87fe4d
F20101108_AABSAO ritorto_m_Page_142.tif
ab261232e82171db22178421c5ca49ce
9e6f62084d10f23e0a62867ad4804e0c4dc794d9
760628 F20101108_AABRVI ritorto_m_Page_093.jp2
07306fb20767125a4984350c2551aca8
f463e791d462029793f102ace488ced82d2dc4c2
119515 F20101108_AABRUT ritorto_m_Page_073.jp2
470db43a26385b0ad3ac0f37877aec30
f415275181b4889a31b705d36095d91d1f9a75bf
57126 F20101108_AABSBD ritorto_m_Page_026.pro
a09b15c6704e6361c58b726f217b126d
37b28fd5842afdbd9224057393002137d6f514c4
F20101108_AABSAP ritorto_m_Page_144.tif
01b6d5f175f121fd4c00aabc291ae9c9
c82b42d3f07fe1542cf1c003cad0bb3854a66c20
731296 F20101108_AABRVJ ritorto_m_Page_095.jp2
a0ca9d676248e09967d972b6ca5b19c6
5067eb4b932fcbb23cc2fdef44d697efd2adf73b
116272 F20101108_AABRUU ritorto_m_Page_074.jp2
917c3e4dbc74e85cb43e8fb9db9f77a8
9709516798759d172fefa1d41176ee3fdcc0bac9
67409 F20101108_AABSBE ritorto_m_Page_027.pro
99c2cf3d3c08818333bdc568ab91c91d
fba4ce5989f0772d526cf3d0638410ddbd903c66
857 F20101108_AABSAQ ritorto_m_Page_002.pro
4ee21803392ea84eb9a5ce46f1e7457c
dc7301c3683a0c39d196ff4022d405cbe1d69d05
668830 F20101108_AABRVK ritorto_m_Page_096.jp2
bcccc01300a934db57bce754076fd354
4426ee1c623198f41c1c1f406587d9cc5300a3dd
119613 F20101108_AABRUV ritorto_m_Page_075.jp2
4f626a132dce037068b7e8ff08e7431e
a633269e4609d63214f36e9b0378a488de4bc81d
54528 F20101108_AABSBF ritorto_m_Page_028.pro
a3b783157f41448f7c25776c5d60c1a2
b590f0befcb76a7275c470ab9fc08922603a8a62
11825 F20101108_AABSAR ritorto_m_Page_004.pro
b18c33e183da0c0f06de31949ad9bbfb
fd65802381ff231c5cbe66e1f03ba9e44dafef79
336258 F20101108_AABRVL ritorto_m_Page_098.jp2
4d1d8bc9489f1f9cda015277a64dfeaa
c7f51a79375f7ea1d788d33115deab9e041d8e78
118902 F20101108_AABRUW ritorto_m_Page_077.jp2
051f36a30210fc679d5e17b2e4da1836
9151754092b2453c7b1cb2e221781d784d638cdb
55596 F20101108_AABSBG ritorto_m_Page_029.pro
e1490ae1aacd35c57d51f93ece412a6b
2d81f3bf3a2f603167adba1bc9243cc8bdeba2fe
47895 F20101108_AABRWA ritorto_m_Page_120.jp2
c81dbb1a7b8afcf3b4605799d4d9a430
487fc797cadf47386726a928a93e62f885beea28
76318 F20101108_AABSAS ritorto_m_Page_005.pro
31ca239cf794b4c608ab206ec8d4967c
333c0730a91a0008702cd4f0357c479d7aa9bcb0
20311 F20101108_AABRVM ritorto_m_Page_099.jp2
0b5fd98c98c6ff2459e7c3b616b7ff16
9a34e1507270388f9e613f16b655dad5f94f06cf
558744 F20101108_AABRUX ritorto_m_Page_078.jp2
42213e960f440631c6fcb42d96c641c5
e084492f87dea3c0183c03e2de56bd59bbe84cce
57245 F20101108_AABSBH ritorto_m_Page_030.pro
c024ddcbb4f067d9f69e8342212d6aec
b74ddefbc5595a38f2b81d16772a30001b91518d
75938 F20101108_AABRWB ritorto_m_Page_121.jp2
3314723f4681f8cd7a119476f2541d36
0914c7f8de77a4dc67f5973f8469a8c648956603
44098 F20101108_AABSAT ritorto_m_Page_006.pro
54b2f14bea53b778752ad38438c081a4
483b3990f486e4bad86ab597723b64e1abcb2148
379241 F20101108_AABRVN ritorto_m_Page_100.jp2
c39a05c5eaeef9f2f37b3737e371a556
8938f04cd55622a10d89ea8aecb92eba90eca7ee
773154 F20101108_AABRUY ritorto_m_Page_079.jp2
a0943cdd4c1e37ae2f23806f395f2b1a
08cd781437977d17b3322223dff3d252e3ed7eb9
51759 F20101108_AABSBI ritorto_m_Page_031.pro
fafc26da6148ee4707c6c68ba862dd83
db05b91f7557e61f31ac48cce110d4950d74c211
118756 F20101108_AABRWC ritorto_m_Page_123.jp2
a76d84d276943d434036985284003641
8c8d46bbb4b463e6da3df59f04640eeb3f1e9994
68188 F20101108_AABSAU ritorto_m_Page_008.pro
467ec156635f19236680dd179ae17bd1
2abcd0221ca4c6c28be3c6b044c335f234196e40
657700 F20101108_AABRVO ritorto_m_Page_101.jp2
22b879734d1c8f2949a904e05c0ab3c0
f283a54f28d2762736591dfd1420f768cf086587
559907 F20101108_AABRUZ ritorto_m_Page_081.jp2
9eafc0315f7ab7faee5d9ec77764708a
b96a510a48d827d5582c6b824740af0b645ca0fc
54794 F20101108_AABSBJ ritorto_m_Page_032.pro
c4c1f10f9159d4db976a42f7b10bbb7d
7a3797241134534c99b5ac027dfba54f8e43dbbc
122652 F20101108_AABRWD ritorto_m_Page_124.jp2
27a265a022508403fde0ce6c18700133
03291484b506d9418938b4884bd47edd151ca057
74181 F20101108_AABSAV ritorto_m_Page_010.pro
5b0fcf47d1a1efc0ff9c5a28f0089651
81f5d54c3f889de576ab942b623f475ef3ae40d1
613945 F20101108_AABRVP ritorto_m_Page_102.jp2
2ad2a06c00cf4466dbc631a028b92f30
a0999a0248e4b521cee80bb31d39506f82c7dda2
56943 F20101108_AABSBK ritorto_m_Page_033.pro
be49273a8f816d417276e1d90aac8aea
7d3095b3e7451a950eac26fd2e3484e856d8166b
119083 F20101108_AABRWE ritorto_m_Page_125.jp2
2fdbe9ce7a9bed7e4359f5b6ed7208c9
84b1680ffc611869f79b9631c46cb5a2f1252a71
45270 F20101108_AABSAW ritorto_m_Page_012.pro
915057f8531ffb5ab411bf712ee00763
04af6e49fedf053d3e56010abe40746b85e0f4bc
657565 F20101108_AABRVQ ritorto_m_Page_103.jp2
5eb2ec1ed91615da4e8602c02a9f21fa
c9bdafd8f0033e001181c938cd62ae3c3e35ce20
1403 F20101108_AABSBL ritorto_m_Page_035.pro
16f893e5edd84869715dfcc51acf1c37
fce31ad34512254678aa0f23b43c5d438a66d372
114325 F20101108_AABRWF ritorto_m_Page_126.jp2
4c8aacdb9fb7ea943b72db4f39bb09df
08a3f5b174b6797ef71e9304ef29c218a848b563
40124 F20101108_AABSAX ritorto_m_Page_013.pro
e3dd670e564d2f870b2198e1ed8b5639
457dc9516ee9ac173b13e6fe96e816f9f7c275bd
25806 F20101108_AABSBM ritorto_m_Page_037.pro
af3495f428fee4b43d49b4db2468d2d5
692c886ce393f20ab68b866de92454c0dc433555
120296 F20101108_AABRWG ritorto_m_Page_128.jp2
258beef16dfdb6c7fba713cea465c1db
1dcac6e02415ac86d67c353381081a2ae2e8818d
53367 F20101108_AABSAY ritorto_m_Page_014.pro
187086d84fe9900ce540efeff7977e3f
bac2eede2b7bfe8c1d438a835ca78d95c401bd41
656062 F20101108_AABRVR ritorto_m_Page_104.jp2
135e67fdbfe0bc530e8d9b490bd6e8a2
8bb07e38ef31a593c4e02681338d38488b471ecf
41569 F20101108_AABSCA ritorto_m_Page_058.pro
e84996be95fc1fc6c49fbcbce8bb4b77
42e92b9ea52353af34fb9cd9fe0a1a2b52f53790
52203 F20101108_AABSBN ritorto_m_Page_038.pro
0abfa5b85c315632bb9c6f9f79e86bec
f0a42a9ccb3c9d99e752bfa4d23c101cf01d4fee
124469 F20101108_AABRWH ritorto_m_Page_129.jp2
cfcb4b7ff0ad4f9126108cc3f6983c3c
844e9ca7c8bc17588c0558f5ef3c1c77244dd5d7
52228 F20101108_AABSAZ ritorto_m_Page_015.pro
773626c5ab5047d4aad8b1828242aadf
88495094c7b39d00d2be018d7a72008bb88af86d
792390 F20101108_AABRVS ritorto_m_Page_105.jp2
daa8451fe495bcd14d79b8596c2579d1
fccd7f754b9c3df29db9bd8a152fd30b1d903d8d
57976 F20101108_AABSCB ritorto_m_Page_065.pro
4de69260bdbdab660fbd0ccf09adbcff
22adfad4f00cee9178d1da173a47cb56007ddfbd
120890 F20101108_AABRWI ritorto_m_Page_130.jp2
565f2c47b86083ba39cc23cea953821f
75681d38bd743e1461b55d39e93a24da14d207a0
683902 F20101108_AABRVT ritorto_m_Page_109.jp2
f98f08cfa4433c79c7131db25c72ff07
b1c8d3e0af2499cd41bd24f6edc2424a011ef8cf
56520 F20101108_AABSCC ritorto_m_Page_066.pro
01d9a8774ebc09c80f1cd601a89bf077
3a6a9578a2ffe3a2f329de8404917c37dff40c53
57682 F20101108_AABSBO ritorto_m_Page_039.pro
d835e62dd8f1df3a1ebcc42e21bf198f
3027071241499e88e6253ec639b4f66c48ae8738
118972 F20101108_AABRWJ ritorto_m_Page_131.jp2
e37e28b2f7ec10490ee60fc3f2aea273
6217d1cca45847883b4394002b39c0014ba8b981
665175 F20101108_AABRVU ritorto_m_Page_110.jp2
2fce14cd1793ed66814d9211166ed77c
8142941bdd77da27f333103bff2342408c16b50b
51030 F20101108_AABSCD ritorto_m_Page_067.pro
81c0dcbce2203ad66b93f9c40cec18be
16cf400dbb27ebed5f71b169884b3c7784055516
54786 F20101108_AABSBP ritorto_m_Page_040.pro
26e60c630bf354edbb5826b6d4909dea
83347930cbc2adf1dbd9a8a71be7a21180584686
115528 F20101108_AABRWK ritorto_m_Page_132.jp2
2a14f10961f95add99ed6033faa6e400
db51fa2d5c72dfb3061acb4ae4620d0a6be8986b
732944 F20101108_AABRVV ritorto_m_Page_111.jp2
90a6140d1cece8e0fedc9e7748d160fd
06675342b4a428c2ae877a85b185f45ae31ee94f
53220 F20101108_AABSCE ritorto_m_Page_070.pro
c5acd3508945b5df17b3aed096223471
30540e8fd00aa91b0491ca7b35b5a9ec645b44b6
55811 F20101108_AABSBQ ritorto_m_Page_041.pro
9073ac8e980e35b36b84463e82a370bf
1b45ffd88d30f0ca8a6943e9dd91d56b06e38f12
122213 F20101108_AABRWL ritorto_m_Page_134.jp2
6e5686cccaf92d89e79ad29323e04f54
12ee0f81b39a01225511990e4092a86928662514
697173 F20101108_AABRVW ritorto_m_Page_113.jp2
d89c078f56ad63c4824b411622e6bdbb
55a59b7fb11a8c0b596bb6636484c507d34303cd
56593 F20101108_AABSCF ritorto_m_Page_073.pro
2ed1ab798c4a890ba5f94bb7ecf40d1e
b98ed3d543e7058819ff0bbac1d8a6b3e5b5e31b
54549 F20101108_AABSBR ritorto_m_Page_042.pro
329dfc62134fa61ba9ca527fab4e487e
e494804abf0496c8099ff5a2b912be5891bdbe7f
114496 F20101108_AABRWM ritorto_m_Page_135.jp2
19efe7358d9253a934a29bc56a99f752
7aef68ac603e80ca5adc938b342597492a9b6b07
78193 F20101108_AABRVX ritorto_m_Page_117.jp2
707f45061872364384e7837a11819aab
b8743e4e2d1de1137b1b2930bc5f112fd3afca3c
56770 F20101108_AABSCG ritorto_m_Page_075.pro
b625b411d79f232dffea04412733c64c
7eb34c15e992c30107156e765112079c776a30bb
F20101108_AABRXA ritorto_m_Page_016.tif
f2347be0cefb85f9f214e30a58a83334
53ad4d409bba4d68ffd8d4b00898bf2e453694ea
39161 F20101108_AABSBS ritorto_m_Page_044.pro
6f36a09c4c93f66fc5fdf277129eeb3b
ae3bf7ea67fdd3e56ca777e62870b0e5d81bd88c
43637 F20101108_AABRWN ritorto_m_Page_138.jp2
824944f0048d6ccd48369ca9c61e2da5
9ca077bc6e4113b0bd8a501a1a9339f470b811de
35368 F20101108_AABRVY ritorto_m_Page_118.jp2
251ca7658cc331c9a7655187ba8f36c2
3f35763fda0b227db3b10539b4e073a4059ff9c7
49064 F20101108_AABSCH ritorto_m_Page_076.pro
d7a68011f1d2e3184f175e9da32346c2
031c7a42019fa843df00f7fe5954afb3af6b1f96
F20101108_AABRXB ritorto_m_Page_019.tif
216b79b2cba6696af850bb9d6fa8538d
12822e804c1198e25442b996b14e165f6d61f219
51857 F20101108_AABSBT ritorto_m_Page_047.pro
c70728d1fe55ad6f306da7d39a5e8ff1
5255c8bf491cc2e0c53184b54491c1b7033bb444
71906 F20101108_AABRWO ritorto_m_Page_144.jp2
87913436df46fe360855256930f3048d
892676ca1bf810876db105892810be1da4ecaaf5
77945 F20101108_AABRVZ ritorto_m_Page_119.jp2
c3c8b7fe4244a68eed080750aa7c9af6
088b2a4d5f4122523136c3c77ecc6f734d4474c7
7054 F20101108_AABSCI ritorto_m_Page_080.pro
bb66fb765e9b980a833b0273c0cd9188
e58cfdfccd3aae60cc7f8a403bf5180556dfba65
F20101108_AABRXC ritorto_m_Page_020.tif
cffe0121b342dd54c0500d6c884d436c
f9c7dc5c863c49346f89d44aebfb5e62f04ed4c0
48332 F20101108_AABSBU ritorto_m_Page_049.pro
6ad09cd40ea6578c365ab5c53e1ff073
5a179ff3c4b2cc177bb8962143ee1841fbeb5e62
F20101108_AABRWP ritorto_m_Page_001.tif
47ad2f5d6ff0c6904fdb73a46dc64245
697d8d8bf7596e32cca6cd5e538176550f8042c7
7976 F20101108_AABSCJ ritorto_m_Page_081.pro
8a0111b15a04ee1f69a9a101b5bc2fc7
a37078689601cfc349876e553fc34757afeb93e2
F20101108_AABRXD ritorto_m_Page_021.tif
0ecf297ae65f7d04f71903c789e09abf
d5d854571d7bdb00919e1b901c280cb4458173c1
53469 F20101108_AABSBV ritorto_m_Page_050.pro
b04f42c8e109645f02a8657e6e1fb103
78e3663e38ac4d9f9e42cdc4cc8b5ca0065d7350
F20101108_AABRWQ ritorto_m_Page_003.tif
0985e77592263ada2efbefce70cbced4
0b5e57580f69d97743368843a01b00acd3ec2456
6952 F20101108_AABSCK ritorto_m_Page_082.pro
9f560c9d3e8bb9c8c9c1766c4f10a639
951c77eab2841d23cf6c269ff7ddc08f04845961
F20101108_AABRXE ritorto_m_Page_022.tif
8736ea5c99c49eded0ee465ec02e3858
da60f6d0e5bc448b2441e1a6b93b199b4bdfaedc
44637 F20101108_AABSBW ritorto_m_Page_051.pro
bb1495d0b815e6ac290cc4e0827eab52
07642307793375d32305231454927b1939889590
F20101108_AABRWR ritorto_m_Page_004.tif
c2fa02ac9ac87f0f3933572cfbaec550
33bbb1742a200fd41f6b5bad59438fc224c1888a
9649 F20101108_AABSCL ritorto_m_Page_083.pro
1b27e25ff47588927b5a733f003a87b9
6a9d2f563a0f9054c12c106353788d9d587d36da
25271604 F20101108_AABRXF ritorto_m_Page_023.tif
2353951a9970da8c4a9c29384f6e1b47
3a9ffc085f7df51385fc4a64c599a717599e5062
42204 F20101108_AABSBX ritorto_m_Page_052.pro
123346b5753d359c07558ac1d95d94b4
bb0b11fa5f6d0ba46c731c7b3cdc5bb82a27929f
10011 F20101108_AABSDA ritorto_m_Page_104.pro
eb5f53730ade760e7ac1e507e834ed0c
8b6249507c549f63657140cbe5675ade24f38af2
4047 F20101108_AABSCM ritorto_m_Page_084.pro
3369ec1494d7497ad489db943798e32d
b0c0ec489f201bb14152ee46a4bda7fd5aa35335
F20101108_AABRXG ritorto_m_Page_024.tif
71bcbb95f6ca93c42f814a89159f40e0
c302c224a4f5adc4b3fefb18a98405c48694d3be
3164 F20101108_AABSBY ritorto_m_Page_053.pro
9607e7f76ced5ef41409f8ddcacf30bf
e460509859d052d7f0836726db2e88870a971256
F20101108_AABRWS ritorto_m_Page_005.tif
9a8a23144fc2b4d65d440328dfab5e56
3490d8d1ae707dc8b29e4e134c84cedde9b71c19
8850 F20101108_AABSDB ritorto_m_Page_105.pro
93f11ee3d35873dde88b4c42d3ab212d
7033e9ba06184756185298eecd678eb80ad6fe81
4561 F20101108_AABSCN ritorto_m_Page_085.pro
0ae4ba4c4bdcc5ef46670096c0205d20
1e538d97df4339121d1710cdf05650d80f901b33
F20101108_AABRXH ritorto_m_Page_025.tif
b0542844205104c7d907dddb66d7fa91
85244a407cf11133538a5feb0c29241cfc72a347
6705 F20101108_AABSBZ ritorto_m_Page_056.pro
5564f49ea82addf7f45532e5215c93a3
9a2b40522b634459cdcb53a580ab8b997c0e3f33
F20101108_AABRWT ritorto_m_Page_006.tif
6094b05fe35bee67f83e21d7d5d5ca03
f97c414d6a2c016794aad33e94a0330e6d6fe623
9294 F20101108_AABSDC ritorto_m_Page_107.pro
f3c32d2a5e38dec8a87d2bbbc83d74c1
c948b7d1e2f501e74cbcb4f1357625860c2e2ec9
4783 F20101108_AABSCO ritorto_m_Page_086.pro
231aac7399bea187dafc21b6901e0173
62823df6b884dd16f6f37f25d0825370c5fc0cdc
F20101108_AABRXI ritorto_m_Page_026.tif
cd792c3a438fb868e192fc1c228694f4
b700a9125f87c63f4cec5451a016bfc47de660e8
F20101108_AABRWU ritorto_m_Page_007.tif
03ca14e192ec1edb91d10f7da2702ce5
7a916d7d2d81783234b26f4ed204ed4f20ea43fc
8644 F20101108_AABSDD ritorto_m_Page_108.pro
c5a5a9bbf1359a39e16ea674379953d2
e07d295df6d3deb8011adfa8288cd32b8fd0aba9
5361 F20101108_AABSCP ritorto_m_Page_087.pro
bf5dc1dad3847631d794d0ae4fd437b9
c3f82201dbb7754e97c93f0d90c9eff829d692a8
F20101108_AABRXJ ritorto_m_Page_028.tif
c77e73d164d96a733cd7a1c578c3a840
5c0b7d0d05dcfe0841aa7258c5292c67b403e170
F20101108_AABRWV ritorto_m_Page_008.tif
4cf9a18c6215e4a362326467cf7c7a91
80b5d819b2fafeabb84d5da1bf3c099e5301561c
8587 F20101108_AABSDE ritorto_m_Page_110.pro
5c01c2747a47bd8c4ed655e502a4c6e4
73c6758ef28001fb61b48615278ed04261837df3
6548 F20101108_AABSCQ ritorto_m_Page_089.pro
376ceae2cc22543b8da6fff63324f593
4fab8495fa82c16f49bf4c61756582bfba6a2d1f
F20101108_AABRXK ritorto_m_Page_029.tif
eb9e85bb9d28de9d0f1d3460a4fb55a5
c96c2dc320b975f241cce0a3a3ccc8ecaad8a1b3
F20101108_AABRWW ritorto_m_Page_009.tif
9cdc968f98dc601494eff9fbedbf9d34
33a1e5821290c20b1d4b74b2099b61c6794014ab
8768 F20101108_AABSDF ritorto_m_Page_112.pro
ad57a64eb6ffb555520405adb4e73482
80bc7d41f311c589174c86ea5b0561f29105ffe1
7564 F20101108_AABSCR ritorto_m_Page_090.pro
1b68cae644d730df6bee174230fb20fc
ad396b239d0f8882668eaafe040b86c682fcb176
F20101108_AABRXL ritorto_m_Page_030.tif
f4c5443f4fbbd4699e7ac570c8ffb542
17f1dda6125c986843623e6d0f5a7b72fe55427d
F20101108_AABRWX ritorto_m_Page_010.tif
9e9fd2e5885826904e181c37141cc8de
178a133cf6eaab08d6860339c8f0cf8a5f7f72d5
11235 F20101108_AABSDG ritorto_m_Page_113.pro
0b050e6a7ca90662a05667f25b548361
fd8bcd17f12ce4dc7d633055d7de1e0af953a61c
F20101108_AABRYA ritorto_m_Page_049.tif
c3b8554f24b362d727a2fa1c5fe84c12
971f38c87903a55c376d016f180acac90ca0917a
6717 F20101108_AABSCS ritorto_m_Page_092.pro
f71530ce86f1bc8d781e9fe2340b6a51
3c30ba72909eb525b96bf221007f6475e8e4c253
F20101108_AABRXM ritorto_m_Page_033.tif
09fb7be54b6bde06437e2c8193022bba
d8d3fe61c02d93a9bf16c84a8e91749e9905bd6b
F20101108_AABRWY ritorto_m_Page_011.tif
f7a0c46510575b12c8e5ae7ca494fd75
54aad5b7ab167df4d16a80d9e4127f30e0888d36
11246 F20101108_AABSDH ritorto_m_Page_114.pro
b4c39a5b7ab7ba240fbd5d591dea41ca
28a7766608d5246b333ea53c7dc9bcc2c65eb297
F20101108_AABRYB ritorto_m_Page_050.tif
3a62ae2679b854a6370cc3de002a184f
ebbaedb2514653d65aa97f2d6c9e322264050038
5974 F20101108_AABSCT ritorto_m_Page_093.pro
d2d7b4923981db83c1c1bbce8f1342c2
ddaeef68f923071caf2ab61ea13fa693b92e632a
F20101108_AABRXN ritorto_m_Page_034.tif
ef8fb57a2fe499a389d8650242fbbd0f
120cefc496e9ff219619228cec54e18d7a7ca582
F20101108_AABRWZ ritorto_m_Page_015.tif
7247acdcbbf78e5d0dad5836cfb39262
8879cc1b23bf2982874e2b4706fafef118e900ee
10648 F20101108_AABSDI ritorto_m_Page_115.pro
533454a9e8d6af7100d5be2fd46ae3b4
ae9a409fe378dbc81bd3b6fe4a92f07d434d9156
F20101108_AABRYC ritorto_m_Page_051.tif
9188133fc7f629a8c7c2cfd53c583849
a0d5ae784b219337c2e0dd47b0bd7a84adb1b521
12258 F20101108_AABSCU ritorto_m_Page_094.pro
7f8d244c1236c05d05f768caaf4e0490
15abb45b58116e0b7021b7abaae1a800616c2aaf
F20101108_AABRXO ritorto_m_Page_035.tif
e509c3b84736a1730c6f4aa57cff15dd
5dd47d0b87fc08625dcd21fde44400d121a759f9
16517 F20101108_AABSDJ ritorto_m_Page_118.pro
2cb807e63b514a467d55b693585026f9
3991396764151e780a571e4c6742233f333ba237
F20101108_AABRYD ritorto_m_Page_053.tif
c5a30f3a8385ac28d2b32a6ce7956164
ebc63039b77de2f93a5672dea449944178c70de9
12462 F20101108_AABSCV ritorto_m_Page_096.pro
91b6e72f1946a082e1689b7d1b1b9cfe
f8761121c4b1813e12852d3ab318914f05987431
F20101108_AABRXP ritorto_m_Page_036.tif
abb97ef2766d265988ea31dfdc732b85
261edb382c4a332a9000fdab01baeb00efd63c47
40929 F20101108_AABSDK ritorto_m_Page_119.pro
2fdd621c6ef22c3c3b2c1cfd2e52132c
3e2b1df672e4f8015b3fabd89174ca9bf194848c
F20101108_AABRYE ritorto_m_Page_054.tif
2925818749ff75eae2978ae872501cd4
06a6a89923fa4628960980ad870947d6b406b72b
7211 F20101108_AABSCW ritorto_m_Page_100.pro
c5dc040a3de38c3de7e052e9dbc5c7d2
ffaeec3b561b6788723f02d09c82df9df4f25122
F20101108_AABRXQ ritorto_m_Page_038.tif
1c9dea3328d982c0d34327eca1b6179b
33adc2af6ca69dbb9dbd1405793edf9f4a1be196
41665 F20101108_AABSDL ritorto_m_Page_121.pro
1d96dcd654326eb492f739dcbd13c40c
b1b2e7dc2b88059065e0129623fbdcf47828b215
F20101108_AABRYF ritorto_m_Page_055.tif
e7305463983bce003723ad78959227bb
efae1d6dea1bcfffd46e0e5ae591323f65bda2d5
10184 F20101108_AABSCX ritorto_m_Page_101.pro
81d249f5bb2651b91282c95287c5e819
4565061a36491b7d6303608673026fe20ab6c92c
F20101108_AABRXR ritorto_m_Page_039.tif
a54f5908ebfc4d4fa0e50b907e7abc82
9dc4480cc00b212050a6cf2bd494a2e11e99cb51
51506 F20101108_AABSDM ritorto_m_Page_122.pro
7a8a821f501ef6b6729e24dec72bcec1
6f28a8fd4c7b35fc8699d4e6a7e00e64480c3d3c
8423998 F20101108_AABRYG ritorto_m_Page_056.tif
e3f51aa39774fbf36c13400edf536bcb
5e23c4a5f2080c48ee8e5e5babd0b697e28af7e4
10104 F20101108_AABSCY ritorto_m_Page_102.pro
f3fc4cd7269e8f22dad6719355e7331d
b5fb7e8aa6b2b662bf782616d6dfba6f70c7646c
F20101108_AABRXS ritorto_m_Page_040.tif
b0dc8e04a14fdf8a0c8b336d0bc38b04
3d9b23d1ea19ce0cff8b4957e80c29aab50a26bf
50454 F20101108_AABSEA ritorto_m_Page_139.pro
9f2e68d1b72d92e49f316e6c0688b706
bd65efea9bad5e16f1f70ad541c7e910a8ad1703
55903 F20101108_AABSDN ritorto_m_Page_123.pro
83bfe48bba55dbf87d6c0ff5424799bf
e51bde7f55eb0275c70516c49e156cc5d96e2500
F20101108_AABRYH ritorto_m_Page_057.tif
a91dcce4589acef38be4410468e6573e
a256d886632a79927da5e3203fbd96f70cf99265
10286 F20101108_AABSCZ ritorto_m_Page_103.pro
3f40c0880da45d8b5e43565aaabd7277
97c785e1e471513db1727acf57cdb13db7f579d1
38199 F20101108_AABSEB ritorto_m_Page_140.pro
cc9131cef8c523ad4590d3e51aab67d1
c520125110045c4e062b88e570648d969707b413
58298 F20101108_AABSDO ritorto_m_Page_124.pro
fc38d0014498c86669c9fb722221377b
81014cb95f82012bfea1b5d76e0df927044d709c
F20101108_AABRYI ritorto_m_Page_058.tif
c0977583bdf93f981a4b734df9a53e01
729e7d9fbb4528a067d014e68e1744126bd1deb3
F20101108_AABRXT ritorto_m_Page_041.tif
011951d220d793ace56146f6c6e1cd9c
f463c6be3b14dada64a5f942f37f63dbe53287db
55944 F20101108_AABSEC ritorto_m_Page_141.pro
255573cd398341643f33cb4e26071a3a
3cbfadc8d9e9fc6eae55466504177584384945da
57062 F20101108_AABSDP ritorto_m_Page_125.pro
5df33b14e22773deef5f3645b2a61738
d3f41a496c624c16db2fa9a4e9be4e9d48f817f1
F20101108_AABRYJ ritorto_m_Page_059.tif
aec5bf7feba1cdcfbc9c769ab749cee8
fc474f6c28b69c9baef9e0cb67c24202f638d549
F20101108_AABRXU ritorto_m_Page_042.tif
12d01d88bea941a565d65171f65dc053
931434f2717fa2a81af4a5b06884fba9ae75d6d3
55042 F20101108_AABSED ritorto_m_Page_142.pro
34cd7971071917f1bb0f178d0ceb735c
eac8ee7e8bc146e610569c759d54e53894efb6b8
53107 F20101108_AABSDQ ritorto_m_Page_126.pro
0717422a1fc7903157daa6ba87fc42b4
630065615cb0b25843e6ba258d47e05f9c573872
F20101108_AABRYK ritorto_m_Page_062.tif
25b36ab0f33de221199e39276d4d41da
94d71dee1fda40c877f832e19f1159791e4323e0
F20101108_AABRXV ritorto_m_Page_043.tif
0218df4f89112ed2efb5654e8d6b0b05
5791d4417b5b739b473f79e5c8e7b3a4047937a0
61349 F20101108_AABSEE ritorto_m_Page_143.pro
da52e02fdfc32f908ab21e26cdd42faa
79f8976e05e84f83551fbbe2f70fe487773b9ad6
52829 F20101108_AABSDR ritorto_m_Page_127.pro
4aaff03e08d1576348e2bb7a3a385790
c4e86bd5fbe3e8d0da235878917ee81473282616
F20101108_AABRYL ritorto_m_Page_064.tif
7d26ffca42e4c305c00d43e4c5194bb7
a92e6bf372fafe7d60bf1b620ad3275a8459279c
F20101108_AABRXW ritorto_m_Page_044.tif
641a1617ed274681b6c021d65a644538
5fc7ccd14ecdea66b866f06e7629530797b7fc77
32380 F20101108_AABSEF ritorto_m_Page_144.pro
38075f77d10e220d66f91150e0ba5264
46d4c2e81d0edca0e0dbec3ff150710b8b321c9d
25265604 F20101108_AABRZA ritorto_m_Page_084.tif
3237fe1aca04a8b55a22d380f93c42f9
69e58c7e27fec74f1ff0d70f184c0c9623be8e7a
56257 F20101108_AABSDS ritorto_m_Page_128.pro
c9c84457f433b9fe45d8cade13966e5b
10433e0f7469f5646585db154a0e6f3fb9a53e5c
F20101108_AABRYM ritorto_m_Page_067.tif
35e39024b00ef4f439cef02732938f86
54156cfb5c8a28b652d05f09356ef8a847a7c5bc
F20101108_AABRXX ritorto_m_Page_045.tif
b882de4b33982ff26cab07923e29ba7c
73730026f547fc4a0619b4f927bf15eb018dbb4e
16655 F20101108_AABSEG ritorto_m_Page_145.pro
c78787ed35f0fd9c753c2455cfa57322
906d13fc5228fdf30d2618274a3b37a834be95f5
F20101108_AABRZB ritorto_m_Page_085.tif
89ee25eae5d5d3a1807cfced6d3328b6
9f5909049842639001f5551435186a2fad1bf369
57528 F20101108_AABSDT ritorto_m_Page_130.pro
6624754bd7f43b574fdf369f7b06001f
20fe9b4f3b81a537d0e5104f49f4aaf8e11e36a1
F20101108_AABRYN ritorto_m_Page_068.tif
0b5146ee050972ca6819b0989bed8007
02c5f831df6786912bb49e22de37b3c7cb7ab774
F20101108_AABRXY ritorto_m_Page_047.tif
d5efb9baa05efa0c99101affe69fc912
867d5e5f5854d60a280232555965990079b53cc1
550 F20101108_AABSEH ritorto_m_Page_001.txt
108951405faccc22814c2af4b936d216
9799ebb386785e8e6c3fe38d7e2165d295278d16
F20101108_AABRZC ritorto_m_Page_089.tif
072ad7006d160bf025389b57c9358408
a8872818aecc8a768a74bac1b03aa2a91c431951
53638 F20101108_AABSDU ritorto_m_Page_133.pro
b16cf66da2df57aa07ce9916f36e8393
1b7eb945c535f8109840fb9fe89593c4b9d41e63
F20101108_AABRYO ritorto_m_Page_069.tif
16e0f7ecfe8181af96bade335f153ab4
d9f244dfafa3300707c9de744287710915f2d36a
F20101108_AABRXZ ritorto_m_Page_048.tif
780e27b9c10fd552cb1ed8ccab47aca0
1e471963e85bee7902f191d08884c14f318bd82e
107 F20101108_AABSEI ritorto_m_Page_002.txt
19f73c9efaa6aec6a88255b71b47e64b
068fa0b2257de8179efa93ed5b39924843baa2f1
F20101108_AABRZD ritorto_m_Page_090.tif
f91fd8034d564d2436ef08dcf4b29466
a0c489fd94f013e7c77bd18d3a414880030a8500
57886 F20101108_AABSDV ritorto_m_Page_134.pro
952c9dda9431fc6be03729b56b834f9b
29bc0a141857d211b9337f8d8e4a4e85d696e409
F20101108_AABRYP ritorto_m_Page_070.tif
8270c12e746f82aa57e2898448ef1ce1
eae58a75780c44da45acb04e17d8f05c2bf523f3
94 F20101108_AABSEJ ritorto_m_Page_003.txt
7130b6c0310fcf3df4f692794a166799
7eb9bbd35f47ab18d86c4e0b8de7a610e16fb1fe
F20101108_AABRZE ritorto_m_Page_093.tif
8d2a9371fe231f74676ef784db643e09
21f4ced9e6e5ed4619e59062049bd68ca81b5876
53816 F20101108_AABSDW ritorto_m_Page_135.pro
49f8f8f0235778b0bc2cec963c169fb2
5d6a5a15ad604d06e1995d708b90455f45903a4d
F20101108_AABRYQ ritorto_m_Page_071.tif
b89f848e83df96bab84f2d5de012885f
94fe54dc21d155e163d15b718dacbbf2aca38ce3
3410 F20101108_AABSEK ritorto_m_Page_005.txt
d02ce080d0accff74a0c549ffddac3f4
5576d7c7e93ff0fc064c3e7cc10ee38c04e20b5b
F20101108_AABRZF ritorto_m_Page_096.tif
061faa39e90ddfa529c9cd07b0aca5c4
3be32354087f83a0b0c026627ec8e36e01a2d996
56282 F20101108_AABSDX ritorto_m_Page_136.pro
1eba2f8d2d7080e2609a03a67c9a9e05
f2de0d562782b819e4ac7ef0a3ac43d8b6f0cd9d
F20101108_AABRYR ritorto_m_Page_072.tif
1521286c4486cd1ec2d15da56d8ca793
bf931a471abe07b3156c00159637729b8d21fe74
1900 F20101108_AABSEL ritorto_m_Page_006.txt
f6534ac81c0130c2edc35f1d420fe4d8
0bc2eb77ebccc37f18350e6af4cb421647b6a79e
F20101108_AABRZG ritorto_m_Page_097.tif
f2605917846c508963e97b9df7342564
8f2d585d11a707a3245ecb3b99e6cd6a60ab1b7e
20641 F20101108_AABSDY ritorto_m_Page_137.pro
c2e14802233a163c70d398a428f74f1b
28e0b91febb41839c0d0f9fb7f7aad421d043395
F20101108_AABRYS ritorto_m_Page_073.tif
6063d2136f202023017eee9a65b55bc5
94b9630e3a2d8dd53fe550f23cc60f72b35f789a
1723 F20101108_AABSFA ritorto_m_Page_034.txt
be12baddbf8be60bc73390aa2548e149
fe577e84337b5764fb4c1a84c477ffe48b3d6b76
2826 F20101108_AABSEM ritorto_m_Page_008.txt
f2ddee9ec04fdd4c4045f076b6b8d092
bd0268155005bc2a163c0b92ea9c8c2ad75c0cba
F20101108_AABRZH ritorto_m_Page_098.tif
22e5921b99f9454c3526b9639cf20fb7
fa350324c694a85cf46a64165c612a1b88b458ed
24161 F20101108_AABSDZ ritorto_m_Page_138.pro
101097732788aa7400c170b4e4c69377
74004c6411f6568e335fd36800d2b59dfc396ae9
F20101108_AABRYT ritorto_m_Page_074.tif
36104ad6ffb6ab509d850b916bc44f4d
9b7eecac4b6f0a98a507f5e67b54effa26db565c
432 F20101108_AABSFB ritorto_m_Page_036.txt
5fd9bc862ffa52570c0a30fa309deb59
e28c01ff9344d737bb358908b77c694b3ffe3dc6
564 F20101108_AABSEN ritorto_m_Page_011.txt
59a1b7fd92daf1472034fa2c3cee2d29
87ae8672f17cab815ada2377411ac9a74d5df9af
F20101108_AABRZI ritorto_m_Page_099.tif
ea107ed5fa4b7a87600791d8ccb28aa2
63f256aa8884e757392942143ddc5a4f7e60cdf8
1921 F20101108_AABSFC ritorto_m_Page_037.txt
f983dcd070cc42446f140893b7e0a6f2
cc4e5a79b45eb4c43bb0fa57da9f554a1f332114
2063 F20101108_AABSEO ritorto_m_Page_015.txt
1473c68b9bbce5cad9203d2280f68586
86a2015cf0d8e18026bbc4b0bc553bd18fd5c219
F20101108_AABRZJ ritorto_m_Page_100.tif
afd1874ac31c7e1290d0a0c68318afb1
5becf41e43ddb74d78ace894cef512817a436826
F20101108_AABRYU ritorto_m_Page_075.tif
bf9958ee1b58cd6783d27254e816487b
6a4fadb99897cf20b45d38f1ba5d52e2ea177a07
2223 F20101108_AABSFD ritorto_m_Page_041.txt
a3a82490170433d56e41a2f1df9c28ff
089b378ace373d89e849ba7aaac15193492f8b48
523 F20101108_AABSEP ritorto_m_Page_018.txt
cdb526a8b2719348a6a6a77d881a656a
c0870a36cde687faa5ab295c8a587f46a875694a
F20101108_AABRZK ritorto_m_Page_102.tif
ed61f4d761d2ae9dee296912fca3374d
9c90976d25c72a695df2bb51e17aef93486e823d
F20101108_AABRYV ritorto_m_Page_076.tif
3d7f0eb4caafb8c2002c4bc7fbb0cc86
72591568dea7a3984eb568b32fefbdea2a83ce2b
2141 F20101108_AABSFE ritorto_m_Page_042.txt
ca7fb7267f0c5db664fbf52498217a6d
9ed2e0b84deb053212ea97173117c678b0fbcf4d
1980 F20101108_AABSEQ ritorto_m_Page_019.txt
2960a2019b13be30077d93fdf0f16357
8b11890fa7d2ce0cbee739d338a90f4cb0fd302c
F20101108_AABRZL ritorto_m_Page_103.tif
db1fd986e6647b5117aa67c3a1cd619a
0473aafbabab492e4295284dbf358aecf1ac2f51
F20101108_AABRYW ritorto_m_Page_077.tif
6c02e31a64452241c02b984258e6ec4f
88e82f1a77980d0a089385c5f39441e990d0e3b0
2047 F20101108_AABSFF ritorto_m_Page_045.txt
c1e1401d0dcdb7eb9b00e76c006da696
2456e61d611c7bf3723820ed5e41f3e5aa4b84cf
2208 F20101108_AABSER ritorto_m_Page_020.txt
a973102e3d4b9a4164b696eed8e20b00
fc8e5c64fb438888e8593ac0d858c1deda6becb4
F20101108_AABRZM ritorto_m_Page_104.tif
d4d26f14bf25274cc35ba04e9abf7208
9951ddb08c8a1612fba3bd806bafba74def40c45
F20101108_AABRYX ritorto_m_Page_078.tif
a03228b93d9efa6bd17b7ce78096862b
240d0b24b39516aaeff3c84e3802fcbe710458cc
2270 F20101108_AABSFG ritorto_m_Page_046.txt
31761e2f1bc827ac3dcfb0c743735d86
7c6ad440d5e9040d820f5264b051e486c4604d49
2121 F20101108_AABSES ritorto_m_Page_021.txt
1f7656af406f17e2755dddd6ab7352b1
ceff144db1f617b272131ddd5d2ec8952393b1b8
F20101108_AABRZN ritorto_m_Page_105.tif
c718c9764e09cfd04c56507933fe5b3a
24eec5afa9481e4ba0ea01516fabf90a16f98175
F20101108_AABRYY ritorto_m_Page_079.tif
7d982f5f9e5f4d6db07a74c718c1c24e
2685cfefab26a5b68d4de6d63ef06b99e864b35d
2080 F20101108_AABSFH ritorto_m_Page_047.txt
06dab86a9c9430c08109a4b888cf117c
f9bb4b6b71acac28175a4641caa8f410d1561e77
2061 F20101108_AABSET ritorto_m_Page_022.txt
d1831a312bdd55902ab1a7da7a1e313d
20ae36f01d3142f58c6b0893f892ec63fd2f8e33
F20101108_AABRYZ ritorto_m_Page_081.tif
5bdb7194b18b4af2169c3f3d6af7f5af
7fad3de21966a7a48b7b89456bfd1362c2dc5946
2108 F20101108_AABSFI ritorto_m_Page_050.txt
77329fd57ec7a53202c6b148f1396c79
4cce5c91a39df0f5957cccdc7cfafb83c8aa8197
2197 F20101108_AABSEU ritorto_m_Page_023.txt
28fefbc1c61035a179d11451e3843166
fd9ad701fe67adabe17a3b49765b196ea60fd259
F20101108_AABRZO ritorto_m_Page_106.tif
f1f8d32ba7263d11498f7219650793af
01ea8083077c89fca050308e9804c8681ca1a0c9
1854 F20101108_AABSFJ ritorto_m_Page_052.txt
3e195f771c3f118007d3e9c8b87966db
5272d31d5295a37bbd05fd9e7821981b0b9160dc
2203 F20101108_AABSEV ritorto_m_Page_024.txt
a32f0e9937d93e76cf8bd46864784b32
ee82801c52e30546e201114ca8636eab0ab820a7
F20101108_AABRZP ritorto_m_Page_107.tif
ba6c487431b98dfecfc24d86d07271a3
117d575855c11cebce13cb10260686083f466d3b
92259 F20101108_AABRCI ritorto_m_Page_131.jpg
ce6442238830afebc120198ac558ec5a
ba8c69f1081ab573be64fad56591db05a21c61a9
487 F20101108_AABSFK ritorto_m_Page_054.txt
fc56ce92b6b4afacd396d9df7e9cf603
914fa1eb067400cc9258344839b8b48131b42db6
2148 F20101108_AABSEW ritorto_m_Page_025.txt
f223e6de8088fc21d143532e7dbcfa7e
28e03d428014e20f4a511b1ac1f3cf5c4aeff200
F20101108_AABRZQ ritorto_m_Page_108.tif
bef423628a435d3db2c37291b6adbfa5
90565f7502295c025407da533811b6663714ccaf
507 F20101108_AABSGA ritorto_m_Page_081.txt
98f032f159ae6993333449d1dcddda34
0f236c578eeece2628ad0b0891d9d06e2ce75d39
90668 F20101108_AABRCJ ritorto_m_Page_069.jpg
bfc69f6efab9afa594017c8c446565c0
f36a4d913e52c15f237cb5349bfb9f47a83c4f2e
502 F20101108_AABSFL ritorto_m_Page_055.txt
ae7603e6de55df46da2aba2fc2cc0cf1
874c080f4bf71cdc17a9a0c20c5546b912eb7da6
2311 F20101108_AABSEX ritorto_m_Page_026.txt
79e1a8cb2cec98dd1bb40f1e24402a59
49fa3f371ab718860ba18d5a76e29be16a50ef62
F20101108_AABRZR ritorto_m_Page_110.tif
60229e6a9a87e8d92cbce125a4968e6c
305e8b72f50fd508a075e53960f8892165efb232
92444 F20101108_AABRCK ritorto_m_Page_036.jpg
b7d84f2e2d0a2d90570a033afb70f0ba
e8fb0a17b52cf04df85a96dde86173d615450640
2137 F20101108_AABSFM ritorto_m_Page_059.txt
2128dc0368c32d75e89074dfb85de435
f26d7203fec91a3518b6c4f64a6355d9e9b6b2cb
2184 F20101108_AABSEY ritorto_m_Page_029.txt
007aa4707a66440b45f33a13783b7772
29b76cf4b3316ada5ea3618b7e7767f5daea20dc
F20101108_AABRZS ritorto_m_Page_111.tif
e2c2a98cc248f6c461172c03af9d8d2e
f23dc9e80c073e5e77b79592de137d482cc0cf7d
373 F20101108_AABSGB ritorto_m_Page_082.txt
52912ba85cfd083dc563e4beb1629e78
7999fee6ae98877996ea45da6287fcf11f45af9d
51659 F20101108_AABRCL ritorto_m_Page_059.pro
c4dfeb25d0ed46ca7c731191ea3bcb18
6a0f8ed4ceaead0bd526b0cafd8004f0513e63ef
2119 F20101108_AABSFN ritorto_m_Page_061.txt
51f3ec9557e5b71b49d11f0f215138de
bb09823df6aaff2a82677b3029893ffd8a995cec
2262 F20101108_AABSEZ ritorto_m_Page_030.txt
d1022a68261f468340b1a616a798080b
71e74137521281344217137560e2000ceb7d6f7b
F20101108_AABRZT ritorto_m_Page_113.tif
aa3f83e480049f28df5626d43b60a5cd
1ae5ee4519ba8ceb483eb8dea4073e1ffe80ee02
500 F20101108_AABSGC ritorto_m_Page_083.txt
cd26f71b37b47b86018aad666e27ca9a
aaa10e00153195c6d27fbd332701354eed57f967
38653 F20101108_AABRCM ritorto_m_Page_102.jpg
d0add9506a5abb914446761a410ff175
55b793ba301f092ca5a19034044c425b4d864ca4
2135 F20101108_AABSFO ritorto_m_Page_062.txt
779ea9305f8122f2ab557fd3783a7d28
4a5e10fd85ce522550cc15e1ff5fa2f4e2043890
F20101108_AABRZU ritorto_m_Page_114.tif
809f33bb76ccfa503751b059c0fd8ccf
d94d4252688b75ee7aec1c36f61a165c7b22e500
111851 F20101108_AABRDA ritorto_m_Page_047.jp2
f00347d653a60e8b1a03c85df9b42f02
f59c6030231c2eb503b9ced857224ae185e01e4c
402 F20101108_AABSGD ritorto_m_Page_087.txt
ed047e835601d92c3bb2c52bfa403f71
48516f06dff01e1c7f175697b3d859d0316c4413
7575 F20101108_AABRCN ritorto_m_Page_116.QC.jpg
de777c9438b28b5a0206bb474a586f14
b1151f5ebb6c40157cf3043bb35e5911c9a97e4c
2278 F20101108_AABSFP ritorto_m_Page_064.txt
7cad1ae6f460ac0b2dfd9e89e74db007
34c03ab2c9c44212699ccfc0b058b390069c7fec
3549 F20101108_AABRDB ritorto_m_Page_006thm.jpg
4bfb11e233a6c52b0a742178fdb5e8e0
f3e4d5e1b93c20dff8ba7612f1a12db50d437c65
369 F20101108_AABSGE ritorto_m_Page_090.txt
5844b1ffc3d32dcd4810b19ccab8035d
74d3890d271bea6267db66e9b6703f7b114092d5
28595 F20101108_AABRCO ritorto_m_Page_023.QC.jpg
53b56cc6f6e80d3213b04b77530abaf9
a31ef00967e0f89d5e4b3d1116a0adf68e947f29
F20101108_AABSFQ ritorto_m_Page_067.txt
d57752eb13c8b9d1ab5eba9fe013ac77
93dbc7542280fddf7417903187c625d2dab816f5
F20101108_AABRZV ritorto_m_Page_115.tif
fe1f7200447189e7ad9768c62abb7590
2ea9dcc0b5762f538e24c6edfe76f7a310b6023e
F20101108_AABRDC ritorto_m_Page_112.tif
b7496d85644af1c610127e66fe0bd12f
9e5decaaf31e061b0c604c7a55f8775b12974bc1
316 F20101108_AABSGF ritorto_m_Page_092.txt
16d64f6cf592f7508e30fe56e97eebb0
0261f3369a502a0d1c28bcf8c6876667a8a380d5
5466 F20101108_AABRCP ritorto_m_Page_048thm.jpg
a22afb4a5664f845852d834c7a18ef7d
acfd120543a6fba9d4e3e0f281ad2b2d28a28794
2136 F20101108_AABSFR ritorto_m_Page_069.txt
45a0d76868719e699e4f6dfd77dfbd2d
2bb63740814f33cb3f93283236c888fd170541d3
F20101108_AABRZW ritorto_m_Page_116.tif
80c006cca9c340fa2f8cb1c9156703d8
9ec930aaf0b8a643bda51ecc009a7f136b29d663
140542 F20101108_AABRDD ritorto_m_Page_027.jp2
d921467e90504cab4ee8dd447b01c2ce
25fc0c5094d97af84a3f00e5d62525fdb4425f4e
437 F20101108_AABSGG ritorto_m_Page_093.txt
f7a72592532b33574180514602853589
15c6928d73e5d7f830aca7e97aa3b73b4cc42a05
34282 F20101108_AABRCQ ritorto_m_Page_093.jpg
617e79a5f6b9fb2c9a77fe2176daa72e
85f1ada773acaec8e6150cb175294c2affc7e62f
2236 F20101108_AABSFS ritorto_m_Page_072.txt
1fba328f9d7acb3f330e8eb544c0f353
d8a5384aeef3fff670070c4afc8e6d6799415bb7
F20101108_AABRZX ritorto_m_Page_118.tif
404d08f0978fa8c26708747154177a3b
0a3fe1bbd1a85787088e5a01bf8b271d8e417eea
10050 F20101108_AABRDE ritorto_m_Page_092.QC.jpg
7161a7b64763d9ae59547915dfcf5a56
edc0715a24b9724f1255fc2865d6c7c136f67961
764 F20101108_AABSGH ritorto_m_Page_094.txt
e698fea3925e3141d559690ae9117b86
6a13b63d82ef9342084b21b7cc91aa517c94eef3
52267 F20101108_AABRCR ritorto_m_Page_022.pro
3e3d9f45f510e57cc94fbea38f463dc4
3af1de8fcd3f998f9e0d4cb18aaec5eaa6517680
2216 F20101108_AABSFT ritorto_m_Page_073.txt
da8c496b99aac243c064aa592e9d5c5f
1406c9621ed2ed99adb15998d675bbbb643a0ff3
F20101108_AABRZY ritorto_m_Page_120.tif
237db5312333b60d7a98a4028cc4f030
f08059c2222b42cddd7d7f31b64a0e6c77c3f0cb
6978 F20101108_AABRDF ritorto_m_Page_066thm.jpg
c9da20d23cb8f9dd879799de824737c4
1447d24c417ddaf6d333c5acea4117e532cda794
755 F20101108_AABSGI ritorto_m_Page_097.txt
e0b5b126334e75f6a12a3e4390878e99
2102800383a08af4aec303a98a805796902dc676
3159 F20101108_AABRCS ritorto_m_Page_092thm.jpg
72415c16be421fef1089b699df2f2249
2fe5de7d95b39dcfbc9f122f88f6d7a388f6d7fd
2164 F20101108_AABSFU ritorto_m_Page_074.txt
bdaeceed75593955b3fbd4da3955a884
383077622d041cba70ba4193174222e1877e6f50
F20101108_AABRZZ ritorto_m_Page_122.tif
087c319b22728c223e869d7e36c3bad5
2d382026308a9da5c063dfcdcecf86c6e85e2113
3444 F20101108_AABRDG ritorto_m_Page_115thm.jpg
662f7639504e3352959e9fdaa6544257
8e63b9fe65aeec72dd48fe335af2914e5a06f0e0
165 F20101108_AABSGJ ritorto_m_Page_099.txt
5e66681b677ebd756756410eca7a15de
2abe5e48771254f86844d4aa083d7c6d4db5d71f
10260 F20101108_AABRCT ritorto_m_Page_114.QC.jpg
dccbb79ac051e7087f25ae9d4295d1a1
6c1f4e23362e819c316e156719f6d57ea546c863
2259 F20101108_AABSFV ritorto_m_Page_075.txt
434aa4d454f903a2ff0aea79a629be3c
2b6f089efcc2137292244afbea2e64bf35a6cc52
3022 F20101108_AABRDH ritorto_m_Page_009.txt
57bd35146474c2dc478f22a34828b611
f0d534629159a32533b93b35ecf96681d067985d
390 F20101108_AABSGK ritorto_m_Page_100.txt
0bc87f12dd8ad3865c9f6986876ebc42
e984497db3714d1ac0c54e95a5ad02e403cc9bc9
126 F20101108_AABRCU ritorto_m_Page_035.txt
f60e78059665bc3cfb3aa9ff0c71a983
bdd31fb4e18e0197914e0577b4dc590f6ccc660d
1987 F20101108_AABSFW ritorto_m_Page_076.txt
1878c7bc910d599342e72363db976bbe
038a3bd9408f35c9948f77ccb32fe29d959f1a31
F20101108_AABRDI ritorto_m_Page_091.tif
463f61867087263f1cc89938f0628b2b
349b478221fcac6916df70632c6967ce04ac6610
2242 F20101108_AABSHA ritorto_m_Page_125.txt
682c772aa27918d9ef237820dab89e29
f6171cbddb614a13c2323b78d7639121a3b41d9d
2272 F20101108_AABRCV ritorto_m_Page_065.txt
05a9c0392c49c831d6b6788b85fbad8f
cb8b0061be1544a0c8231db434c79ec522b9e6c7
723 F20101108_AABSFX ritorto_m_Page_078.txt
adfcee27360c3a91863f4941afb6ee44
4c4d06b7f4075205bf2b44cac5d0938cc78f39eb
F20101108_AABRDJ ritorto_m_Page_063.tif
79646bef22fd36fc3e8dda17d3ae618d
ae53679c61f864ae110257e8dd1efe2cc83cdccb
433 F20101108_AABSGL ritorto_m_Page_102.txt
ee8ea6cdbed703a0cc5f19783009894b
bd587321a9be46101d720f1f2f42c40cfab84218
2117 F20101108_AABSHB ritorto_m_Page_127.txt
4d823002859aba6d7275e7818b07a231
4906653e9ae18ed2b26a76dd4567ca67a1ef81b5
13460 F20101108_AABRCW ritorto_m_Page_107.QC.jpg
c50b8d67b2f22012d45dbc3d94d4665f
b227ff74b4284710e21d44456ece892899505263
278 F20101108_AABSFY ritorto_m_Page_079.txt
30df869879a71164a6d4928fe376f8ca
6ceaa12ad69e630c5422db871c5243256aedc5ff
24242 F20101108_AABRDK ritorto_m_Page_081.jpg
3560e5488dd60fb89007746774f275a2
b7f26b3d5396d08486f845cdbc90e7b7094104e1
451 F20101108_AABSGM ritorto_m_Page_103.txt
b5924366206a88856f739ebcfec7b5e6
0ef51aaf9570d2880f469a9debab41698410674a
55478 F20101108_AABRCX ritorto_m_Page_060.pro
159b5f44ed6cac184823eb1d85a0723b
cdecf0e222bc3338ea6e44920ccfd4219ec12cc5
388 F20101108_AABSFZ ritorto_m_Page_080.txt
8815bffe76902eb21f44d89303307bbb
7c2d56f0e09fcde5d7d7e716a4465a8ed316a560
52238 F20101108_AABRDL ritorto_m_Page_062.pro
e85d3933f620f181b5c9baabde264cfb
9a76e631f76eae00f4daebcb07e871508f31b287
464 F20101108_AABSGN ritorto_m_Page_105.txt
f03d4dc6b4c0ab9373b289b1c01d06f7
926e273c845cfcac1e6183c249c83242d5c674c0
2319 F20101108_AABSHC ritorto_m_Page_129.txt
84cb662d8b916935b0c2288de147de07
ec31fc754b8e1d0e5b1d6c5c37be0a0a4cc3c616
524 F20101108_AABSGO ritorto_m_Page_107.txt
40a247d42e6cc27ca78235f0557e2770
aef69de6524a9f055e46f2b615e521be11bfee0f
F20101108_AABRCY ritorto_m_Page_032.tif
5947857ae0ceb1af417ae02548631ae3
06f3c02b6391043d4c398d45f38ab82536cee00e
F20101108_AABREA ritorto_m_Page_002.tif
4e80321325782540b8bf7306b04f773e
fa2d665df7d5ef7fdf72ba7540c6c5eb6a9b6e09
14723 F20101108_AABRDM ritorto_m_Page_099.jpg
4a05c385c324e0bde7d4922e554ed6d3
83f507967d9cf950136a7da1e0c7aeb4659b4981
2128 F20101108_AABSHD ritorto_m_Page_132.txt
95dd7b9ff6dd76a069e1e4ccec0ece31
6d477db3864c9ab2ba9cd92d19bc54660d49325e
459 F20101108_AABSGP ritorto_m_Page_108.txt
f042985dbc1af81a103083ed86fd3072
1ea84da212f169655a0b07d8a46e574e46e43747
30516 F20101108_AABRCZ ritorto_m_Page_064.QC.jpg
6775f3b9c12f5b055ad7aa6fb9293a30
f47f3cfaa9c4a7b7baf8b09e460fe1336dc8bf7a
85831 F20101108_AABREB ritorto_m_Page_043.jpg
29a8a1a0a83e680f9293f98a7617582f
fe489268049680a917636ac6a3f0699c059d7376
6594 F20101108_AABRDN ritorto_m_Page_132thm.jpg
1a77ccb3c77a33e4039489394a1c6ff6
bdec84ac2eb3e14dc5c1d1ebf6db352903274c58
2277 F20101108_AABSHE ritorto_m_Page_134.txt
d2865b1b9dbb6e65627f7506b3d54dbc
c0d4ea3f8f9bfe97c5e8be1f0a9ca28064ee4110
448 F20101108_AABSGQ ritorto_m_Page_110.txt
f08e7137a6f18f56b093a2d4ef623b8f
09ddd3a5598f1c2ff63d766b45566b534a578029
47267 F20101108_AABREC ritorto_m_Page_019.pro
43a07a17d3c5e1c834a8132403b90669
fdf64039d0a7de55fb5d4e540939f121c1216fd6
571897 F20101108_AABRDO ritorto_m_Page_011.jp2
953969f6fe3db2e214f085345d595f39
954da4923671f23215a96f35223d832f2ab84b17
2116 F20101108_AABSHF ritorto_m_Page_135.txt
11fc99cc1ae1e39e2a595a7e75a11a49
c99839c16e1aa80024a08a8e5422e4852d42cba6
430 F20101108_AABSGR ritorto_m_Page_111.txt
dd267ad979a5fd1ff1e93690626b82ec
eea82b591c77eb41a8d41b3d11f4787ed4cfc666
252 F20101108_AABRED ritorto_m_Page_084.txt
d2957a95d6f663f00d64d32abdfc61b3
41b0407d36a2eed49e72185b02262b21eec7b9e3
28741 F20101108_AABRDP ritorto_m_Page_024.QC.jpg
fdc5731f18b6dda7cd31b037500e8c9a
a9334b19f10663deade1ca488f455053bbf14a57
2066 F20101108_AABSHG ritorto_m_Page_139.txt
7a7981cbc7bc11351168b5b1da35a3f8
fb5715fea1a1b80f977f1d44e6304ae9b324ef4b
438 F20101108_AABSGS ritorto_m_Page_112.txt
81015116991fc2b306e3526b3b93d9c7
6e954144a7173d37de6d3db7baa16a07f308cb81
28791 F20101108_AABREE ritorto_m_Page_004.jp2
3d98289fe087f287a46d227610d32c87
d30108dcade97ec305d548c7439ecfb96a06df5a
55118 F20101108_AABRDQ ritorto_m_Page_074.pro
da5a1e2e98ae1779ed39e53c7edd797b
6654c6af648bf445b1fcff2274b6484e59f8d44f
F20101108_AABSHH ritorto_m_Page_141.txt
9b194c2786fdc905d041510251682c0a
78452e7741398cdfd39623494fc24b10a7d87fca
539 F20101108_AABSGT ritorto_m_Page_113.txt
3f56778daec6b05afc99d300ef4c3035
da6130c3f8ce4e98298d229c2ebc8016856208ec
114314 F20101108_AABREF ritorto_m_Page_133.jp2
57b46aba1c87a92c3e21428a09107e08
37ba726ac20ba144a39c52c4e8830e83739ef95e
214577 F20101108_AABRDR ritorto_m_Page_056.jp2
465dce67b8f63e44dc86213b37157d61
6c002a618d041dede2baf70e4e5aa44c3c74ef71
2284 F20101108_AABSHI ritorto_m_Page_142.txt
bb0cf3aa9a46efa0fe8d8ac367c88844
94eba12ac9f6e41c51450de083d983b780dea354
634 F20101108_AABSGU ritorto_m_Page_115.txt
bff9b2fc22390d9a305d4e64bcdfaba3
c11d6ec6957dac7cfd0f792cf476964b1fe75518
684142 F20101108_AABREG ritorto_m_Page_108.jp2
edb126ed889be3c5a960947a1912b6c7
077216d775df14ecb0957a02f2e4150a9942b528
306928 F20101108_AABRDS ritorto_m_Page_116.jp2
1995b57358a1f98059a1439ab0568176
b1b6fd91539d837e5c421ce59b8d5c4e91d1899e
2529 F20101108_AABSHJ ritorto_m_Page_143.txt
867dacb6acdab690124f3c4097311058
59c0056c8ca6d1b2ca63b940cc13a3c80c3d5233
395 F20101108_AABSGV ritorto_m_Page_116.txt
8a6b715923c2392cff7ce5cc8acd55d8
3c29998e80d418b563a046e04402d2a8aee32663
26129 F20101108_AABREH ritorto_m_Page_141.QC.jpg
5f942ef70a463931148b877bbcc5182c
58bf8e7806dccec1a6ca67d405d2bc2f2ab7c111
2179 F20101108_AABRDT ritorto_m_Page_014.txt
035b1752f12fc796c082392b59c1bb4b
a963d23163ebb414ae335e22ff8303d1341607e3
1354 F20101108_AABSHK ritorto_m_Page_144.txt
b715c4ce67dc78a6ac6a956e083990ff
22b1610d635d5853b1c6c6bea8cc152a8145f077
832 F20101108_AABSGW ritorto_m_Page_118.txt
3b1fb7e71bb2cd5661180140a5d16969
8ca91bf3675db99515e3c9414373153948403b26
473 F20101108_AABREI ritorto_m_Page_089.txt
5a89ac38621f6fbc5988bf241cf2b67c
8e6e8158cd05bac311e32f0cf1f54001a9068594
1688 F20101108_AABRDU ritorto_m_Page_117.txt
fe35f0b3589708083b98f3bc8e21cb80
daa1c69674378b476550404958602d8968902b8f
3979 F20101108_AABSIA ritorto_m_Page_101thm.jpg
675d00b62b7b7ea3941a6e4d92420599
c908964ee5d51f3e76adbc3322f77badc8f8b553
698 F20101108_AABSHL ritorto_m_Page_145.txt
2a3a62153863634d35f4db1482a0051d
3e83733ac8cbb1f80e5895789298f67d5137c0e1
1112 F20101108_AABSGX ritorto_m_Page_120.txt
5542caddbcce0ea8202f3f80fa61dac5
4f4c0449de50d7e5e1730440c85b08e9f2de982d
93070 F20101108_AABREJ ritorto_m_Page_130.jpg
e462aa85fd8a4c790f610f9ff8528802
8c0b2ab10d51408c81a2f0ddcb3ccd0d48ae89a0
F20101108_AABRDV ritorto_m_Page_017.tif
a00e6c9391b2bfa8bb09668e6d0818b5
4e054fea66ddce8808af45a9368625fc5a3f62ec
1782 F20101108_AABSIB ritorto_m_Page_001thm.jpg
76876a5ef89e1888815a305ef8ee56c1
7ac02006d5584e288c90eba93b413c9d5482fdeb
29992 F20101108_AABSHM ritorto_m_Page_009.QC.jpg
8824e87649ff5989ba0b62a24d283e57
ba5d78a772ac1b22f943a51399c80208db2dda5d
2202 F20101108_AABSGY ritorto_m_Page_123.txt
bbb026294a6f0d21c5826273e087a8cb
6f8f894cfafcd2c5b88f2cbdedc7c70c18132d1b
2758 F20101108_AABREK ritorto_m_Page_100thm.jpg
8cc53d4f490c1aa098a044de1f2f860f
5da0798e731997dbcd5e6fa18a90f0f97879bb58
91196 F20101108_AABRDW ritorto_m_Page_077.jpg
bfafc09b4177d4227da375ca30c1c910
0549e46ce16d0d9bc234bc2998e77337d2d3140f
6354 F20101108_AABSIC ritorto_m_Page_045thm.jpg
71d815b5e7fca8e399de7e8a3b6fd72c
35e48791a788425cef96a8133ad4d34095245c21
1549 F20101108_AABSHN ritorto_m_Page_118thm.jpg
1ea86808c0ffdf61bf13d9b73d8ea581
5abf00a7216d12d4d442b4243bce454b1fe8acb3
2280 F20101108_AABSGZ ritorto_m_Page_124.txt
f36b58eca777b76a41c6dffe8cdcb2a4
3042da611b689a9af21fbc45ee3d903df72c96ae
1801 F20101108_AABRFA ritorto_m_Page_051.txt
fb39a36b24a49408ac4f751811003c19
cb6de0f97012eeef7552372a37d64b3e12d6a01b
3611 F20101108_AABREL ritorto_m_Page_096thm.jpg
6dab1180391862076258acf07918c2cb
dff6ab904f3145de8d4706e968c3ef076c2beb72
6564 F20101108_AABRDX ritorto_m_Page_116.pro
61818aa31fa6296c7814eac8a6905023
795d32dc1c7ada33e82744be241c68bc5317c11d
6819 F20101108_AABSHO ritorto_m_Page_143thm.jpg
6c5ffec50586d02dcc73bf8ba56c727f
b28e6eec3b7378f389304c1cc2d44d533b2b0528
2252 F20101108_AABREM ritorto_m_Page_130.txt
4b4189e517d0f65c4bec86c3aab0a975
09018b05a2f4c8b933ef2503b773c239c8a2b1db
568732 F20101108_AABRDY ritorto_m_Page_115.jp2
cb0c5b4933fd0057bad63665bc67b192
7d3003481144a059dbc24260aa307cba920cf7b1
11577 F20101108_AABSID ritorto_m_Page_103.QC.jpg
3c8698df2b1346eea9052a5b0087d806
adbeab2ff80a1a53806d2096b279892621248cf2
3172 F20101108_AABSHP ritorto_m_Page_055thm.jpg
894c7ff0d6025c946f8653bc60d5ff3e
2845a7dc4e3d389a4c091c49410c6cae7c2625e4
2797 F20101108_AABRFB ritorto_m_Page_027.txt
2c29d9c1244e590b4da1bd502cda2ff6
582c47ce26441a8f256642603983d2395f45e960
92034 F20101108_AABREN ritorto_m_Page_052.jp2
a0d652fdecd0fa2ac59323bb1221a4ef
2a4cb1db5a227ca21fed09390bddcf9ea0a3e57e
F20101108_AABRDZ ritorto_m_Page_061.tif
11f99d8d86b097cf9d501a2994728b91
514513d8c358de5fa75e40500e872621c7ef8edd
2332 F20101108_AABSIE ritorto_m_Page_121thm.jpg
be00410142af2a840557114d18e353dc
0b0a6385fb789fa02b01a67b7fc9351b5fcd5aa8
4579 F20101108_AABSHQ ritorto_m_Page_007thm.jpg
4efdbaadfa48949846c8a599279e3358
19a67a0702b2a26d5e576c0b1e870a38d0439417
72901 F20101108_AABRFC ritorto_m_Page_034.jpg
ed35aa72196dfd21b55e038f0d636404
aa362454f02e4d7303ed17a22e928cb77a3b6758
112263 F20101108_AABREO ritorto_m_Page_067.jp2
3ac2f260762cc8596f02e391229ff377
dfab1f71489fc8a33e4c53aca8481de6f0328477
25717 F20101108_AABSIF ritorto_m_Page_139.QC.jpg
17958028b1d7a6e19c8e7a6ad6498af6
11967e9c57ea85b7a2b461cc15670d31040e788d
6912 F20101108_AABSHR ritorto_m_Page_130thm.jpg
63db9e0d781ae0f0eca20fad56a6ef75
74036e5bc3847fe10ddb97ac862cea56aeb42e84
1778 F20101108_AABRFD ritorto_m_Page_007.txt
9bc5b3ff9d1a347463c2bda7a08d5886
22739e4470298c1122d3b05c537fc6f81f9b4ee9
793748 F20101108_AABREP ritorto_m_Page_106.jp2
14c91b1333ce5fc74435958168dc996d
94e8f08d2af5b84bde75ca31bd488b6bccfd155a
22886 F20101108_AABSIG ritorto_m_Page_051.QC.jpg
a12966e189681c596968cc4e92de15e5
edb9673befa4cbab79e5f131549f4e324ec705d8
13101 F20101108_AABSHS ritorto_m_Page_112.QC.jpg
2209163672b56d7f2a70515c81ea2eef
b052e2b2b510a5d4d8403f2ba1e129b41928d72d
1125 F20101108_AABRFE ritorto_m_Page_138.txt
cb80f2ad836108083fc863e6a0773525
b9c23880ab3f2560ea7ffadc0962852a70c18e05
1739 F20101108_AABREQ ritorto_m_Page_121.txt
88baa11ed50b863c4fce1525d1132ace
e1e498adeadf526dab43bcb55f34084400019393
2555 F20101108_AABSIH ritorto_m_Page_145thm.jpg
ec1b4ecf713a0eb007860115f3f5f074
8a299a956bd94fb1c839b690fc1e11d1ceff6e24
3963 F20101108_AABSHT ritorto_m_Page_109thm.jpg
6687def2cdfccb42e19ecc8b726d6c24
030e01d790f1249e7091c5147622cfea890f167c
14843 F20101108_AABRFF ritorto_m_Page_006.QC.jpg
bc996af00352219513df25024bfebf2f
5295c25cf45274927abec6adea2c345d72ae49c7
1774 F20101108_AABRER ritorto_m_Page_004thm.jpg
15f1e59cce9727f8b623eeb98943b91a
86bb2e6298a06c9b4c56e06a5aea423845b51c23
18408 F20101108_AABSII ritorto_m_Page_007.QC.jpg
f911cd8ac67fb9f5b950bd639681fe6d
5706c92d632a7fd7cc8a60fa05f162a0a65b706d
9894 F20101108_AABSHU ritorto_m_Page_145.QC.jpg
e013b7e620d24eca7a9bdc87f7bbad55
179eaec14eeeb26e4f266bf2e1a3d72dccceedb1
2265 F20101108_AABRFG ritorto_m_Page_078thm.jpg
ce3d0a97ca5c867323a9a4703c222892
5378f760fe814a732a27a4ff945b241653a198c5
F20101108_AABRES ritorto_m_Page_083.tif
bb3ba4f489277c5af149c64448cc407c
e26e960baccc0c8bf09c14a889b95a112c61e8c0
6669 F20101108_AABSIJ ritorto_m_Page_078.QC.jpg
630ee37800f8a769d1c8a9397d196b1b
0fe201021d7cc1298cbc2824f60efcd27c0d1765
4541 F20101108_AABSHV ritorto_m_Page_099.QC.jpg
df07e86545dd293881049f063e3b7118
f104370eabb37429eb1777fb810001330992793c
1521 F20101108_AABRFH ritorto_m_Page_140.txt
d99b19b5a9c446804485941c8dbdb773
815ce744eb99faa3552c4f649eadfa953d74839b
F20101108_AABRET ritorto_m_Page_095.tif
0dde0d730438f7312e2f1a37b8cbd5f4
a47128be3e0ee8cc1b7aef129b181ce3289b7dc1
6967 F20101108_AABSIK ritorto_m_Page_065thm.jpg
fb0359543ae4401fc0a70cd680db86d8
da931f34364d10c1bf0c47334ae97cf8a353b4ab
6932 F20101108_AABSHW ritorto_m_Page_046thm.jpg
d3eeec249d1e6d398972ea65603a0499
b4be3279426dbbf3b48cec05a2d5d5574f6a0f1c
53632 F20101108_AABRFI ritorto_m_Page_071.pro
d3f6a59191b81c73d7a9b47a3b1e9ed9
38e2e77df5efe22d74c2825e2980599190235dde
108295 F20101108_AABREU ritorto_m_Page_139.jp2
05d9e8d1f39ea354a1009ab951f5f3d4
7232a559297e9f9e44dfec8b8585586f3da93b7d
1084 F20101108_AABSJA ritorto_m_Page_003.QC.jpg
e7a70cef1818427932fabab3d4932cc1
a4e3df9a4084c2da91af0ec12a8f4b4a29db4b8f
10174 F20101108_AABSIL ritorto_m_Page_093.QC.jpg
abcc2b524f7911ac6c9fae856a8534d0
2e851dac5056a71635ea8b5fd73b73aafbcb992a
13596 F20101108_AABSHX ritorto_m_Page_105.QC.jpg
800e00c6db0ae41fdbabf1761a6448da
e0e1bb400712c0c81da3838c5ab12184501ee12f
7735 F20101108_AABRFJ ritorto_m_Page_098.QC.jpg
54b3813e24b049f9e87855465615bde7
2c194c0259c25d5f9631fd49c163310d773e2946
28409 F20101108_AABREV ritorto_m_Page_050.QC.jpg
0390cf67a72019315235949f16aab9a9
8d45cec1554a7cf8f2dfc6758d6210d3a67ff589
6741 F20101108_AABSJB ritorto_m_Page_138.QC.jpg
58ff8a389cd38c0995f7c1eca574c7d5
4d8748525a470d7af74198fd64f7521f370b3485
6817 F20101108_AABSIM ritorto_m_Page_042thm.jpg
4c819862a3348463d581edd88461f1bc
b53f0db4360a56ea686b0c589f10b67c8861a8b9
12031 F20101108_AABSHY ritorto_m_Page_096.QC.jpg
e42599062c328d55cf28733d5200b7fd
806d54429366a5298523079c79edcb87cc35afd4
92561 F20101108_AABRFK ritorto_m_Page_033.jpg
c31ae4c8de60897eeb81d3c6cdb717a4
714b141048d73273c8e49c204234106da725c2ef
85616 F20101108_AABREW ritorto_m_Page_031.jpg
cd1a4c618be7b908e418765f8c5baa05
55665f0bde66942bb276df4953963c6d418f67c1
29930 F20101108_AABSJC ritorto_m_Page_030.QC.jpg
bf8855f8faf8bb560da8ac2bfa1fc462
ac7fc494ef26d5776502391d10088de24cb7ccf9
F20101108_AABSIN ritorto_m_Page_070thm.jpg
6c4e32d068e48343deba4566c7b00de9
309110e2bcd8f7cc7a2858c58c41e47e73def4be
6416 F20101108_AABSHZ ritorto_m_Page_031thm.jpg
c55febe03a7511921a5fb74d7a7d42b1
28d9df479994b3b22e9deedfcd0431eaab2b4634
114732 F20101108_AABRFL ritorto_m_Page_014.jp2
caeee2e922b1b070b23b46fbf0dd6ad8
3e12c32da984e66b797f70a8b6c9390871c42b11
F20101108_AABREX ritorto_m_Page_031.tif
3d805ace143ff1436026a9f0a4744f39
2c6493a13b98f1f44caf5caa6e0064c7871bf0dd
2154 F20101108_AABRGA ritorto_m_Page_032.txt
9b1a061713f1fac22f698211e280189a
c3d551612ed3897974fc80933d5b9d4294e37c2d
4269 F20101108_AABSJD ritorto_m_Page_106thm.jpg
d7646eaeed841b3aa46284c8542d3a62
8ca25d04c7ad66a211b25990883cc40b912709d9
6651 F20101108_AABSIO ritorto_m_Page_141thm.jpg
7f1b5490370ec93287a7ce73b42ce0e4
d3ebb5bab4eca66d202d1631695abf3bc50f6931
8089 F20101108_AABRFM ritorto_m_Page_054.pro
71e08f6638f9a28a986c7fa9f66fca51
3215ade726ca69fa19b7bf3144c9e9dd8976c6c5
566 F20101108_AABREY ritorto_m_Page_095.txt
fb923535d09e1fbb16454804c14aeafb
b2634bf2a36d716f02a7a609787b866007d94196
211 F20101108_AABRGB ritorto_m_Page_086.txt
36faa0c4ce6315a01b99429626aafcad
d1dbc53a72654663e7230e5688baa469108e31ce
3942 F20101108_AABSIP ritorto_m_Page_095thm.jpg
ea8ea01b73850360c26690416949f1a4
b075d25776a9679d239ba46ff6439b895ebe5be0
2201 F20101108_AABRFN ritorto_m_Page_128.txt
e6b9eb5e6dca3c70d904d2f090e25a69
bbd38dbc274f2641d76e04307bd49c9c798e5e8f
12222 F20101108_AABREZ ritorto_m_Page_037.QC.jpg
22d3f37c108225be32ba4ea2000be438
1f87ac426ff3692858218f04eb9d13ab69f48470
12393 F20101108_AABSJE ritorto_m_Page_110.QC.jpg
f87b3c619907aae2ea6054f9dc33bff7
ca42b14ce5365a63fda6ac9fa6660392a884fb54
27223 F20101108_AABSIQ ritorto_m_Page_043.QC.jpg
7926477e80907a98034e1c9e0b13be5f
09568a6b37809ee4b1c28978f9bd04e5ba068f6c
17598 F20101108_AABRFO ritorto_m_Page_118.jpg
36d7bd4610cd991e1a530a32cce63926
898f0ba54485ff0d614503453c2d79cee65ba0f6
F20101108_AABRGC ritorto_m_Page_060.tif
6e1ba885c32273209cdb01bbc93054e8
67eb8b3e661d9b6a6d886046f56ae6f8a832a61f
27442 F20101108_AABSJF ritorto_m_Page_070.QC.jpg
7e0e464158935c8d4f6d5893c3537bfa
0cf5cfacf35e6236b820c1e545f44af0c220a9e8
13958 F20101108_AABSIR ritorto_m_Page_095.QC.jpg
b7065b589de19dfb2a3347e11a3ab155
a64b9b74e4016be2fc0d10737f98294a6180623d
43867 F20101108_AABRFP ritorto_m_Page_048.pro
38e3e1182e0670971f36723c5952cde9
4c34f3590b717dfa78d3d7bf4dc2cbc505158263
F20101108_AABRGD ritorto_m_Page_046.tif
d2c997f183193cb3f24b249b8bba119a
efcf70694fc9d8197a41841af7746ba8186f7601
5046 F20101108_AABSJG ritorto_m_Page_013thm.jpg
27d4cf0eb2c02cfeb119365e76feefc0
83667d1b0d3cb20d63afa3c07782dd22d2818856
3058 F20101108_AABSIS ritorto_m_Page_089thm.jpg
d7ea4cc45aaff4562bf9176cd1b464e5
404d7f6ac6ff521f7435c41f8fc59c78ab6f3acd
52730 F20101108_AABRFQ ritorto_m_Page_043.pro
4f10a397edc435e85a7afd0d75611d92
b44ef858110bd6d1fd9ee7ebaf296c485f5bddf2
2112 F20101108_AABRGE ritorto_m_Page_122.txt
1c6b0473e339a299f2f66846e6a98b74
832a60007eacf200ae6ba00e2efe9965b8b1c118
6861 F20101108_AABSJH ritorto_m_Page_023thm.jpg
66a76be22d27eb00e683eb5dba87729d
59ebb30062f6e37375adee8fe41888e5f57e2075
28644 F20101108_AABSIT ritorto_m_Page_016.QC.jpg
5aafa64a7fe84365d06d534ad8be49f4
01f1961ff6f79ce906c4b57b92c5b1cda1005f9f
6386 F20101108_AABRFR ritorto_m_Page_126thm.jpg
67359352b57d20bc4ccf540034c06ede
4b771eb0dc3087b887e0c069af076dcc2f189358
54435 F20101108_AABRGF ritorto_m_Page_069.pro
307a8a665089df28d2df8037ca6041ed
4c9f6a0e8ed805bb928f89de9d8f55141212ca8a
6928 F20101108_AABSJI ritorto_m_Page_134thm.jpg
cf3a59beb13c3cdd66cecf684b8a61af
22bf07cb8971b842d2a58b0a2dfdce9943915ca3
28102 F20101108_AABSIU ritorto_m_Page_133.QC.jpg
883ec3c5d2e9cfe659adb27d36d8ccd2
62257987c4c5b40592cdbb862cc31ea7bbbd68a8
2167 F20101108_AABRFS ritorto_m_Page_038.txt
de1c0ea1ad5ffb61d31f848f961d69b1
5532f991b01258e032ae4264f2af021e56aaac25
3068 F20101108_AABRGG ritorto_m_Page_057thm.jpg
0bf69d917538357ed706acbfc2300b06
a56407ad829568d82d6fa53fe78f2fd818e87da8
27675 F20101108_AABSJJ ritorto_m_Page_071.QC.jpg
50c071e6d7b8628721735cfcc55048b9
671a675cdbca88180a39ff46ec4a9b76befe07a2
4345 F20101108_AABSIV ritorto_m_Page_105thm.jpg
975f31d319e43443d22da5997fa27753
ba63d4ad2f2eecc4aedfe9a15de584f63979797b
5000 F20101108_AABRFT ritorto_m_Page_056.QC.jpg
d4bef9e14eac7ebffcc45a3be62097ba
79c0ac89e3c908215661b52eecbc79c31011aac4
311 F20101108_AABRGH ritorto_m_Page_056.txt
3ee6016e57bb8009196a7e6d776a3f5d
26bafe4a23a663ec294311728969f72c617b1d7b
27678 F20101108_AABSJK ritorto_m_Page_059.QC.jpg
6d2d5afd5b785aa902b4941df3066087
b56e0e194c74b449ada2c353223e5eb1c0077ca0
6726 F20101108_AABSIW ritorto_m_Page_073thm.jpg
8ea57856cedf818f1f8a134f0f8bf201
9e09531c05c51500cce457caec00dec9d8f6a0f7
29126 F20101108_AABRFU ritorto_m_Page_033.QC.jpg
7620b489d208774bd7bb83cbe1329b1d
efaa0e291d1506db1c0e6c9fba5d5de27ed899d7
26534 F20101108_AABRGI ritorto_m_Page_047.QC.jpg
80db46c2534368dfc1dd0c2ada1a808a
df0594916448cc3487adc337b0ad7e028c656c35
28912 F20101108_AABSKA ritorto_m_Page_125.QC.jpg
1d083bb3e870e807a087545cf0f82f8a
d146da279c96f3eb0a5df968a9709c2762f0a237
6531 F20101108_AABSJL ritorto_m_Page_062thm.jpg
d88ff70c76dd82d947296f5a7b5aec94
da77012ca48093aae92ef05d467279a181593877
5638 F20101108_AABSIX ritorto_m_Page_012thm.jpg
0fbf67de8a5225899fa23a30d11dda88
d3d4179f312607283b81104b75094f63512a3156
F20101108_AABRFV ritorto_m_Page_065.tif
1da39446d74d8fdda2c2b401e5dd23d5
63b389e07eb2b355b566b766cac45beda220e0aa
42349 F20101108_AABRGJ ritorto_m_Page_110.jpg
c215bd965bb6aeff438ab16b2f09a567
fe07753054e6f3f08dedfbb8919f85cae57a15d5
3912 F20101108_AABSKB ritorto_m_Page_097thm.jpg
9e0c7f545f4e6f24963082b5a60f79ae
3d8e9355cb4b419eaf645787aeace51f595f17fe
7346 F20101108_AABSJM ritorto_m_Page_009thm.jpg
d6c8d9cb3456eb42e84b5e9043065804
2517d9cfbdb553b90bc66f0dce8fb9ad5fae474a
28071 F20101108_AABSIY ritorto_m_Page_074.QC.jpg
386a9368e4abfe4e575c26ee317d3c06
4e1c40ef6705c04a5140e42faebb1440b034a31a
5135 F20101108_AABRFW ritorto_m_Page_052thm.jpg
1104fd169196b2e2dc52ba32ad7b0a0b
caa45c7c892393184503751fd528cb2372d398b2
90472 F20101108_AABRGK ritorto_m_Page_029.jpg
6e5b00aa52299a0012651819df4548e9
df35636146df8c1da5811b540bf1bf8f7d57ee46
6696 F20101108_AABSKC ritorto_m_Page_075thm.jpg
fc1adb079cfb41f0a714249219e5eaec
de32445b9c0dc56ee19de10a565f9a76f86b03ad
7697 F20101108_AABSJN ritorto_m_Page_083.QC.jpg
a1732a2cce19ee996d5c95e2796c3614
ca87d23ef4f4583afdaf4b869cb8a95605d8746a
4174 F20101108_AABSIZ ritorto_m_Page_112thm.jpg
6dd82beaa8cd649bbc1e25fea4d5d572
b958787514ffe78b0a455f224c91748c589e9b6a
34436 F20101108_AABRFX ritorto_m_Page_114.jpg
f96ca934809ef99cfbffe21aca563c43
0784da88eb78bbef9978f3600a08c9faaacac3a7
F20101108_AABRHA ritorto_m_Page_086.tif
bafcd14d06c35a1135d579f6d3e15e23
1de2656302d661118a90a1da0d65ecd1fce18bc1
92159 F20101108_AABRGL ritorto_m_Page_041.jpg
c5a955225c633bd5ffb5905eaadb73f1
c1d915d294495506fb293eda25827cff00223784
6702 F20101108_AABSKD ritorto_m_Page_131thm.jpg
6b187ef60f0b0d7d428a2a7422bdf22f
3d1ce2c6e951489bb54c139e1d2357aff5a68f34
27380 F20101108_AABSJO ritorto_m_Page_126.QC.jpg
853127a7c7f6f6da1b7a56e5da6a9d48
24d92bdd4b578682349671c59d8e4d3b3cd1695f
23716 F20101108_AABRFY ritorto_m_Page_004.jpg
3396c2328e8c325d0d7915777ea42b26
ea8ebe4aa8a77b0e0ca152fe06dab250033007ad
44099 F20101108_AABRHB ritorto_m_Page_096.jpg
1387f4dfb721ad913cb9f0a006d3cb78
19a12d5fcffb95f6175c1e624f90a4bd2ba52558
F20101108_AABRGM ritorto_m_Page_117.tif
baa1675601e7f6f42ffee90df2911240
3282a5da235a6d2999f15a07ca6eb60cb59656bc
6644 F20101108_AABSKE ritorto_m_Page_040thm.jpg
01998f7c42f5180d63bf8826830d37b2
b0d23eefb4b99d322dd4c99df5eefe9548518866
4199 F20101108_AABSJP ritorto_m_Page_144thm.jpg
6786fe89426576913c78467bb0d1439f
086d43de0130e3013f66e9e6e7fc7b8e785871e9
2209 F20101108_AABRHC ritorto_m_Page_077.txt
d989199c358caff91e0c50b485e8d809
bde20f9086f5cfb6e7acd98a278f8aacc9331f00
689851 F20101108_AABRGN ritorto_m_Page_097.jp2
63015a7c996ec4187dbed8d7052882a7
bbd2690219c021475ab51c5d71f0993823e698f9
6858 F20101108_AABRFZ ritorto_m_Page_032thm.jpg
1b0a919f6f05447b3c1ecd81a03d55ed
032105875fdc3ce19aaecba179027ccf4657271c
6866 F20101108_AABSJQ ritorto_m_Page_029thm.jpg
3ec8ed9b4b73d00d78625b00aa85f0ae
3b03241b19a4311d3ca4f14697aee12adc73f88c
110696 F20101108_AABRGO ritorto_m_Page_008.jpg
83d9e1c3560d655b8d6bd2ef1b8213db
f9b64e8b158f0d3d9c04a3ad699f0bc336483407
5978 F20101108_AABSKF ritorto_m_Page_058thm.jpg
362dea99b89cd963436dce13452fdc6f
af206f17d58ff316c6cab5b2ae0bb22d8f48f230
10310 F20101108_AABSJR ritorto_m_Page_121.QC.jpg
38472d19c99e09458e45e88e1572c48d
f64ca856022bae1af5d0a45b5b1a562420b0ecc9
1051957 F20101108_AABRHD ritorto_m_Page_023.jp2
8f87f199f4c17602a6f843aa1a6696b0
919789c42eacaf2dabff26a8cd55065d086c0097
57197 F20101108_AABRGP ritorto_m_Page_016.pro
52b64ad2337b80b3a4cffa6551c35d5d
67df8468ae6d85511aacb935709dd859d840cd3e
12908 F20101108_AABSKG ritorto_m_Page_113.QC.jpg
19af3be196d0cb1031141465f0b56806
53d385d0b38b3013e050d48a07a2fa9840ca363b
29096 F20101108_AABSJS ritorto_m_Page_046.QC.jpg
74595cb97cd309284e01567d92ba9386
4f036c75b55920c890f2c5584ceb9c5d1c036e46
2102 F20101108_AABRHE ritorto_m_Page_070.txt
0da8edb98d202902032b369036872948
8f2d1f181ee52e2ba167323cb611a255f9394623
2115 F20101108_AABRGQ ritorto_m_Page_071.txt
455987825f2163b8d92624af1cd30e97
60c17a08cde74d4ec30a862129c5a7029af69c9e
5021 F20101108_AABSKH ritorto_m_Page_005thm.jpg
c89be7cdd0b720dc78a71a683c21d03f
0b2c8bfb959b590aac8c2950f4457560f8eccf53
29706 F20101108_AABSJT ritorto_m_Page_039.QC.jpg
c4ea7f25c6d16c621ba1d45d25459e76
ef252f03c9a62d8284cf35d2f4c6d292eed15a91
2150 F20101108_AABRHF ritorto_m_Page_040.txt
d05f292cd22c55d7d3b623790c2f50ec
8fa3584afd6f9ee2cede92a9e53c98f374c3623c
636 F20101108_AABRGR ritorto_m_Page_096.txt
e779fd52db11b7d379cfdf5566448dbd
f76d33196f8c2b1c19966fecca0c8b26c11cb866
7171 F20101108_AABSKI ritorto_m_Page_011.QC.jpg
5421a97b9849688e14ec88ce941eba06
5aa8cbfad7a536879a918e84b927ff615d8bd426
2718 F20101108_AABSJU ritorto_m_Page_053thm.jpg
bef5dfae7955dc4b5ef81cbe1cdd9b18
c075ea60121d7d489263ee6912e0e3af168ea75d
F20101108_AABRHG ritorto_m_Page_101.tif
d7c14b2920882efba141819a35eebdd4
aebe74c5be0fc07af6d1da90fe3f6af7b758cd93
23326 F20101108_AABRGS ritorto_m_Page_120.pro
9ded1d4a711be0848ec75a983f2ea387
d6f34b73f44231a567dfc1d6b1f46d6f2d16a8fe
7450 F20101108_AABSKJ ritorto_m_Page_087.QC.jpg
c96eb4c70a4c133bea404bb7699ed5ed
4d0f7da45fffb4927074c2e946bed0414d10bbad
25059 F20101108_AABSJV ritorto_m_Page_019.QC.jpg
fbf41399a7a1ed76e2fbf10aafec19fe
f93d5226424d80b03290d2fbd05109299f4ba42e
12491 F20101108_AABRHH ritorto_m_Page_018.pro
593d34a80af7f366ee487b9f8b280b57
2923ef3ed23d29e22aac86480f61a2884ad55b6a
10873 F20101108_AABRGT ritorto_m_Page_035.QC.jpg
93a490402f084ffea9b5e3785f7edaab
1545a78e66ba0a652b394b87331e394b8af8ff26
26973 F20101108_AABSKK ritorto_m_Page_022.QC.jpg
592a0805a75f5d19f047d21d873ed624
dc02fa70ff85428751fa2d7a50a6bcbfa14bf97a
2835 F20101108_AABSJW ritorto_m_Page_054thm.jpg
a66316e666a83157b3e07b32d8507285
6d122f3168569792dc331042208f3352c5ef7a86
95044 F20101108_AABRHI ritorto_m_Page_141.jpg
8c396442d62c0b68c8f2fbfbb0184fcb
9e8148475d8051ff23df2b679d3ca23743d4d2cc
3792 F20101108_AABRGU ritorto_m_Page_099.pro
04802fa1f9947cb612343064a4aa8be1
41459ef14664800a31daa83bcabd4f65b2e3c916
3573 F20101108_AABSLA ritorto_m_Page_035thm.jpg
6e71a01b673ecd96956f43ef648aa00d
de2c8c3ae83c0859872d12f1384701112304477a
6807 F20101108_AABSKL ritorto_m_Page_039thm.jpg
da54ff220dcaf0d3629ddb9ec2f43bba
a7e2cb1ed0db975fda43c1ec6d097a1c657f448d
28379 F20101108_AABSJX ritorto_m_Page_077.QC.jpg
85b3dc3c80598a2bdb2ed6e6437a27c5
493b670a9e51c53da94538ee9e09f992a1bba32a
2106 F20101108_AABRHJ ritorto_m_Page_126.txt
a2ab9f5fd126e75ed524300022dee4a4
0f2edd0791f591a926bdcc10399ce0cf63877386
518424 F20101108_AABRGV ritorto_m_Page_114.jp2
48fcd7e0be9aef2c7c9204463d154172
bb0c6cb46374c33103f15de3854d02456f9866cf
28797 F20101108_AABSLB ritorto_m_Page_060.QC.jpg
3bb4f9c45cf47ca47019fe1ebfc59f4a
32d1c8bd07d7ee1d00ccecf6f6db4f0d34b2d2e8
2545 F20101108_AABSKM ritorto_m_Page_119thm.jpg
cc7c3c036eb3f3aa0f3b9a52a34eaf48
ea26961ab046f7807cc429a6ecec35f00e2dd356
6582 F20101108_AABSJY ritorto_m_Page_127thm.jpg
187b7a402d2e6baa5a158bcefc90947d
2dd370254de9db602720e044d7983c1d12bee7e0
F20101108_AABRHK ritorto_m_Page_130.tif
3de1a89df172f90765d639043bdc0f02
50c33773cdff852f408aa65118986aed993f3cd5
2049 F20101108_AABRGW ritorto_m_Page_119.txt
bc978297512e9e5dae2cfe5b8f8fb7b7
2aeb33e843199dbe83af364dc0b5e0274a1dc981
25979 F20101108_AABSLC ritorto_m_Page_122.QC.jpg
f676c4ad8463c16606d3eaa02c3c0ab0
f1fe7363978d6ae6e391bffd83be7e04b9c38b4c
34368 F20101108_AABSKN ritorto_m_Page_010.QC.jpg
8955c4cb9696f227fb652f6089e03d33
cfc385bb55f561161f2de9ab6831e3bd2a082aac
4043 F20101108_AABSJZ ritorto_m_Page_108thm.jpg
2d730e5b276be33f5b719b35165d2a1c
5f86a3eac03cf7b552002591ef6231361561a045
6374 F20101108_AABRHL ritorto_m_Page_036.pro
71fe9ed89ebf032c7a4f1ec8267d4807
fc08bd6c0ac8715bc1b8a56aa5cfef4fdb49e399
5082 F20101108_AABRGX ritorto_m_Page_140thm.jpg
132d38adba86558bfeea7ecf6bac0ee2
cf0b1dc7d43495722ea07a6e2b5cfb08e0dfabe9
69714 F20101108_AABRIA ritorto_m_Page_007.jpg
404bd12ae358c9b9d4773f9e4f7e86f4
ec774a8561251a2e2117475ad3abcae271d017b2
4597 F20101108_AABSLD ritorto_m_Page_117thm.jpg
b06fa73fd63012be45a5352353b1390a
11b78ab09d57a0dcc4edc2d7becdd1673e225acf
2975 F20101108_AABSKO ritorto_m_Page_137thm.jpg
7778f421c82a989a3981a19a31210f34
fda35a9989fde44e3065755ca648d25205b2e5a6
6895 F20101108_AABRHM ritorto_m_Page_008thm.jpg
e101ad5f150420300853a4c1cdb9de40
2646e77296c97ac97ceb3d8d3afd1f28eadcd513
F20101108_AABRGY ritorto_m_Page_052.tif
ad236323d7778f2fe4ae9ac5baa72001
a68cf44fa7d609e8bf0c7c1140f4dfcff03ae153
23677 F20101108_AABRIB ritorto_m_Page_018.jpg
8363c8e1ed146f186b8ec67f39b6193f
b726f1ed9dd22470d396f5bf7fdc32795ab8127b
3855 F20101108_AABSLE ritorto_m_Page_110thm.jpg
75727eb394db2e132c7d067137284fdc
bc1ae6809828527d4751acd6aeeee5c4dcc40cb2
6870 F20101108_AABSKP ritorto_m_Page_016thm.jpg
1313be0740cadde9b11b67e3f924cf6e
0767b00d4af83896f10f761a3eb7fd47c2b94a62
116216 F20101108_AABRHN ritorto_m_Page_021.jp2
33cf7a5bd3b0977d8eb563e840e5a344
d763ca4623c63a7d85dee27d53d66f64b735b165
F20101108_AABRGZ ritorto_m_Page_140.tif
2db6dd7737a84135617bcd7c28f83bb6
ab76f5569e85660cf3d85f452f5c53bc4348d876
935 F20101108_AABRIC ritorto_m_Page_003.pro
69b14f5d92854ec829a6496817e195d4
2bffd8d01be98fb595eb2d356824f435563a3275
11381 F20101108_AABSLF ritorto_m_Page_115.QC.jpg
93adc5a8ab6ec5f997fa047a667b2de5
8c5721aadfb479736641b4b3e7a2db8598d66187
F20101108_AABSKQ ritorto_m_Page_018thm.jpg
2fcc74906199ca797301964fa89c86c2
fffc6feeb3e41ef7ea1e6adf97e66c75d174c7a1
46079 F20101108_AABRHO ritorto_m_Page_137.jp2
0df38765b02b9ddd389576222dad67ab
de2a396a38339ce480d6ff44583657f67bca6f5f
8619 F20101108_AABRID ritorto_m_Page_055.pro
3a9d1a22bfff230c9d75c9f500afca96
6f85f46653fe8a26330635eb31b58abf5c3d5303
28955 F20101108_AABSKR ritorto_m_Page_066.QC.jpg
52e8c42db09fb652648820cf2c662a71
0056310041977f6dae18f86b656a3c0ac3c5f0c2
2239 F20101108_AABRHP ritorto_m_Page_016.txt
f7b770600f4394a5c432ce5783825037
0d322e1c92f3319f7f3c7cc6b2987be3a4afccb5
29928 F20101108_AABSLG ritorto_m_Page_134.QC.jpg
9a0393206ee44a49a6b209edeffdcabf
bdb98923dfd2f1a0dce7db093d81c54d1e1dd339
F20101108_AABSKS ritorto_m_Page_050thm.jpg
aa40f58d49cd33a1f23f2899ca8713f1
c1928cf55abea44536c9a15b612aa8b0f7207e43
86209 F20101108_AABRHQ ritorto_m_Page_022.jpg
2ca4016d93df8dbdcfc160fc15d00d79
43aafc0445ce4085f7df87dcd7ee990b6cc786c0
2158 F20101108_AABRIE ritorto_m_Page_068.txt
677f00f71e320fa593d56dd471128795
ca7717bad7de4992bb2286b58b7115a0d142fa41
29502 F20101108_AABSLH ritorto_m_Page_124.QC.jpg
9f18114d24c82fad73a6c99a1fd37bfd
794237101f06ae6c1564d07cd57f0cb61e80988a
3132 F20101108_AABSKT ritorto_m_Page_090thm.jpg
686ab5c08726e173682c44e8bfd9cd77
29a47d6967360a5f2553d4987aa8527f5ebacba8
11856 F20101108_AABRHR ritorto_m_Page_078.pro
a05e3c9808ec281e064bfd8b28a547b1
a9073770d2f11cfb4d36cacfd3694201dc25623a
43135 F20101108_AABRIF ritorto_m_Page_108.jpg
3ad49f86ff0a6567bd891255bc8ed01e
c85ea8d886c66397708a1055b85fa3a3e85e45ea
28628 F20101108_AABSLI ritorto_m_Page_069.QC.jpg
d2c4024e021b66f74e715bd21254948f
579a0563c76e3881146f525839992c11346d7301
6383 F20101108_AABSKU ritorto_m_Page_067thm.jpg
d010641befb97c59b8626bb42e745aae
e2190c6f7f0aaa1c02b7d669039b1b8b3dc2cce9
13446 F20101108_AABRHS ritorto_m_Page_011.pro
33e0ba3c9ac4c4807e2ecd60cb52bbed
2dc9897031f412c24b8619248944134110e123cd
1975 F20101108_AABRIG ritorto_m_Page_058.txt
23d4c13305a6995f4daee40a42037ffe
587e4f6a4b4154cdeb04d721a94ace0db890c7c7
27491 F20101108_AABSLJ ritorto_m_Page_132.QC.jpg
6e804ef033e844e877e2f584a33ed488
eaadb88257c2ee9b96cdf3fb4c49682503bba312
22736 F20101108_AABSKV ritorto_m_Page_048.QC.jpg
a8cbf0aeb1db7621998708fdfc7396ce
e163153143ddd93cebd37972af977db661ff7f20
40639 F20101108_AABRHT ritorto_m_Page_145.jp2
7ec9a199324e387b36c58ee872bf94bd
15eec02b35a25f970b244d4278dcc034401a3cd5
8615 F20101108_AABRIH ritorto_m_Page_111.pro
f0b62dc717d40ccda44e7d3592573944
6d08e6bf96f1e3b68155d78b2918bb10f865ebe6
28384 F20101108_AABSLK ritorto_m_Page_029.QC.jpg
6ef7e58e555bea2908ea248c20f14fe4
8b9a7abe30f26fae2acb678e08ec914d74d1b5ff
6884 F20101108_AABSKW ritorto_m_Page_064thm.jpg
f47e6b7e87c295ec0ddbf68e40871b2d
036a2f6ebfa4925ec03fc51ddd0a1a81ab1e9145
5932 F20101108_AABRHU ritorto_m_Page_019thm.jpg
0bb58557cddfcb00a55bce9209971357
2ace65ba919b4ce003bcd0fc36115197647a8fbb
3935 F20101108_AABRII ritorto_m_Page_079.pro
75ff2a40ad3440e2ce01d2b54baed51f
04706548d2a4a567726e3b2520bf774962f73769
28643 F20101108_AABSMA ritorto_m_Page_008.QC.jpg
e9958560da28f2e9d36600f2265081b6
1cbadde7ca6aa9b441a5172935bac3d25fb14767
6859 F20101108_AABSLL ritorto_m_Page_123thm.jpg
5c8260af9a52fb5cc01eb2d62f1a7cb4
fa92ff93bc6b27100167e8a15977d866a928361b
1873 F20101108_AABSKX ritorto_m_Page_086thm.jpg
53b911ed22f3632c011477c981a43251
4880d4a9abb3aeb016aa8b73deef3d2f172ced4b
27618 F20101108_AABRHV ritorto_m_Page_061.QC.jpg
7067b73eacf55caaf4bc40b1efed02b1
d5657ca384e8f4318ec91e3da71f6c78d4ba6330
3567 F20101108_AABRIJ ritorto_m_Page_103thm.jpg
6179ed11d5647c06d7946c30536d3b96
f8aa3779c7632e5e00cb1785a53bd5ae6e115de3
7788 F20101108_AABSMB ritorto_m_Page_010thm.jpg
9d1f9803edc5981c3e65e05ebf6c850b
ce15fa3afefbef3413406c9d649cf0f86e11f5bc
9497 F20101108_AABSLM ritorto_m_Page_088.QC.jpg
6f3a96200ff37ae59ed3a825c3292c90
bc7bfbd45fb084685bac9addc2001092b5d6b486
6778 F20101108_AABSKY ritorto_m_Page_041thm.jpg
100f67f75b1f1ee3c6f14bdaa83537ea
5ecbf68b79f5e8daa9ef63ca7eabb24132277ac7
89138 F20101108_AABRHW ritorto_m_Page_028.jpg
da83a7d4841bd9deae6320fb351f80db
62bf0d44cec3b4ecc4c8179890be2e0df8286663
34809 F20101108_AABRIK ritorto_m_Page_117.pro
183e80981ee1a89514d9dc56e7b1a7c8
b8a4844938f932db47db859c0679a79221d1a7c5
1945 F20101108_AABSMC ritorto_m_Page_011thm.jpg
3bbeb48cad23e7408096c9c547778063
c309a49bd781a192c305fc485cd8646c307aef6c
530 F20101108_AABSLN ritorto_m_Page_003thm.jpg
bcbe5a1c2f94a310f52bd7d9732d48e2
9ec16bdc08482cb943e1d66c583ef2faf8ee85c2
27047 F20101108_AABSKZ ritorto_m_Page_067.QC.jpg
10d8f3c4a895a9b829e0f9aa670f218e
c79524117c7d8a91b947b8509834535275b2ee7a
10484 F20101108_AABRHX ritorto_m_Page_095.pro
b72f46b6e336e8dda34c80ff81d18716
c7460241b8fc4b33efa8e18a8eafd1386673d96a
86535 F20101108_AABRJA ritorto_m_Page_015.jpg
488956a9d6123005f22e809f5acd69cb
2bc8b4d9bd76d83dc7d7c563215b35605e3891ee
514 F20101108_AABRIL ritorto_m_Page_104.txt
7b234686821562f378775deefb52134d
1cb0362507d326dbd73d9f5563ef0ca46462dbc3
27611 F20101108_AABSMD ritorto_m_Page_014.QC.jpg
9ac95806434395f56b1de78478ad142a
3595ea9ef3bd3db4e904f6ec482326f48d9e1fc9
25982 F20101108_AABSLO ritorto_m_Page_076.QC.jpg
89eca8edcbac659ecae5bb2da444f7ab
565444d128b8a47b9c5e72b6e97b7da9e176722f
60658 F20101108_AABRHY ritorto_m_Page_117.jpg
45f180e30ec763fe609377094ef11389
e1452049611893aa3ac27c3027f5ecfb8eaef30b
27409 F20101108_AABRJB ritorto_m_Page_062.QC.jpg
0cff167b73a7e11e0e91b341b24ce9a4
be0f27a69ffff8e7128820e69873703c237eb549
24747 F20101108_AABRIM ritorto_m_Page_011.jpg
10f0dde5bd001dcb36e55e5eaae57fef
d9f9605bc2796d8dfd3ef3bf384e2d5e67257297
6535 F20101108_AABSME ritorto_m_Page_014thm.jpg
662e0355ef92912ab0084570b708eaad
e98a75f13873f5c487bb2d79de8875cf3f9ab3bc
3235 F20101108_AABSLP ritorto_m_Page_093thm.jpg
ebb50a26bfc84fc123f180c6970352bc
48d24ad31dd7af6f5f8d338bede399d938f93b32
118139 F20101108_AABRHZ ritorto_m_Page_136.jp2
8c0ad7552cb5f696669ca5ed37e490b2
2f2de6e3695bec5e1cd31ff6eed244c008462e7b
F20101108_AABRJC ritorto_m_Page_013.tif
9f73f92714fab507ba5ad9d6c07bc7cd
0b0dbff2766ec766ea9d0568e94911b4911e7a23
2006 F20101108_AABRIN ritorto_m_Page_049.txt
13362a911ca5f84168910c8db444655e
eee1711f42f337e24944ab3c4c7ba09fb537b2fc
6721 F20101108_AABSMF ritorto_m_Page_015thm.jpg
1fd508951f2844a4e8dc81e91b659df1
15813c234f6164f220e3f6647fea2b86808ebe53
26481 F20101108_AABSLQ ritorto_m_Page_058.QC.jpg
726ca17496c3360c1f2af910583af8e9
433dd9ca745fff2b09e965914334799dfd79babe
107883 F20101108_AABRJD ritorto_m_Page_076.jp2
52baf313cc9f0e440b2cc4750f3b8613
cd5a20b1cd3ba372f6e1d0d84a525b3cfb67ce97
6796 F20101108_AABRIO ritorto_m_Page_033thm.jpg
8b457a97ed46046fd926d8d9aead3d3b
566442d75af46ff03e3eddac4d08e02f0ad2fac3
6472 F20101108_AABSMG ritorto_m_Page_017thm.jpg
b413d41728c93a515684101ba6abf03c
6b6af438c5c3b86acf97fffea22d3092ce1406d4
2584 F20101108_AABSLR ritorto_m_Page_081thm.jpg
d2dd3f12f65fad714288582160f28cb9
3f41a6abdcbee7a168eed50e0a0c7d882f9a890e
6762 F20101108_AABRJE ritorto_m_Page_022thm.jpg
1eddf7f802baf612895f1937c341a587
2035bd1cd18391877d2bfe064ca58dc282a90e6b
F20101108_AABRIP ritorto_m_Page_131.tif
87cf5cf2d42f9e72ee1b0995d1556fe9
eed6816c9cbedc3b3c59d0cc766d8bdd4fe32f0b
7996 F20101108_AABSLS ritorto_m_Page_082.QC.jpg
8ae64f21bc2cdcafabd9efe5bee8efd8
74eb1ce0319e8d909ebb308e457f5df41a093186
92692 F20101108_AABRIQ ritorto_m_Page_066.jpg
f354ff629a3055d720676b62b36d5668
2291671b739a7a5e410dbb8d2b37922d2454e631
7726 F20101108_AABSMH ritorto_m_Page_018.QC.jpg
55980ace5c156681d6cfc2484e671a2b
f5825487a472dde98eaf2bc62fb5293e679383fc
20210 F20101108_AABSLT ritorto_m_Page_140.QC.jpg
d630f864fc44e07950ac94cdbea1d64a
735915b4fbf7a96d67cbfba0462585a1e51a48ae
1997 F20101108_AABRJF ritorto_m_Page_012.txt
884a9a6298fb9cf23ee6ffa984066972
53401094a019a77c9b517b7b298e7aeaab207c3b
15503 F20101108_AABRIR ritorto_m_Page_057.pro
cd4b1bfaf4fb555243e627eda28decad
924a8dffc368c6088e62026e27b7a31945d61a27
29245 F20101108_AABSMI ritorto_m_Page_020.QC.jpg
e7f1310d66c94924771c53ef8a3fbb05
9e8e21a6249d54bf5c825bd6c2c52c826ffd7083
25123 F20101108_AABSLU ritorto_m_Page_142.QC.jpg
938876f9f688860282b93ee219abed36
36c939d722abeb331447d296e618cda132665093
6445 F20101108_AABRJG ritorto_m_Page_038thm.jpg
4fb2e10bb666352db3d248b077c07041
842c6828cd5e733308a43a9f834c7cac9a34f4b3
F20101108_AABRIS ritorto_m_Page_080.tif
6a6f4f76a83f0f77ca3d457ac8e2e4ad
4d2ba9568a0685047daeaf8b9ed92f371cc8986d
6787 F20101108_AABSMJ ritorto_m_Page_020thm.jpg
8830037a6f195e60f724e274030953e2
1786b45940ad61967065d5e8a9cfc7ceaa6a8cba
30136 F20101108_AABSLV ritorto_m_Page_027.QC.jpg
bcd60c1cbc8e6bbfb064dd5306592de7
2e4f27936ef69289a6d0f48d5ab7767967cbf8b0
27828 F20101108_AABRJH ritorto_m_Page_026.QC.jpg
e2a535ce6998f07baeaae23fd51a3885
a64145a12e3c421c7fdd77252ab10f9342b63f0e
43272 F20101108_AABRIT ritorto_m_Page_034.pro
f3db0e911ef6fe1a8c7565155918cecb
a5199768d9325eb46605a854fa85930ec10c1aa1
28381 F20101108_AABSMK ritorto_m_Page_021.QC.jpg
a34e82a7aa423e1010e16f5b9681965f
b30ffa18936da7ba5b59305559a8a42e20ccd8b5
F20101108_AABSLW ritorto_m_Page_056thm.jpg
cd51e13e2a637ad728c1d6fa7f525065
add9b6761fdd55f8f3f3de124d0b906749ee8947
2211 F20101108_AABRJI ritorto_m_Page_136.txt
78fce83cb7b40e2f45e59809073d62d7
b8bb3400bd6f7ef59adf3f1052e2e00250bef1ca
121707 F20101108_AABRIU ritorto_m_Page_026.jp2
94d2be2cf4819a19c17e904406570938
118f000971323d13009836a6cd0ee8c8cb4e59c4
25511 F20101108_AABSNA ritorto_m_Page_049.QC.jpg
571ec6d493d1eeab2151f0db99f188e0
50fd943e4bef83059145e986c1fdfd64a92496cd
6882 F20101108_AABSML ritorto_m_Page_024thm.jpg
b84aaf752a24446238cdb5524e772ed9
72cea02a94139b0b6af758ff37b40a5dec4a2cbc
217953 F20101108_AABSLX UFE0021715_00001.xml FULL
0623bed3e0626223b7d119368833e72b
3c964b1013d727d3b054fc9e3792044f563a1a3b
11272 F20101108_AABRJJ ritorto_m_Page_106.pro
46a20e13b8cf8fd85e3bbf638fddbde8
8c4106c8166ad29ce0150499ccc0fef657e91ccc
27334 F20101108_AABRIV ritorto_m_Page_127.QC.jpg
e97d7595b9756d98d7106fe9f3f0c971
574584527eb20abf7fe954cc9320215b28b04923
5782 F20101108_AABSNB ritorto_m_Page_051thm.jpg
4d38a3ab941cc3fb793e58a71b58bd1a
d8daa9b7920e37242d791083c919349a0df5d8da
6624 F20101108_AABSMM ritorto_m_Page_025thm.jpg
c4ab85fb2eede9e620a478c511c13eee
cc9b657dbf6c0a334e1679a874d143f9f3a44aa4
7411 F20101108_AABSLY ritorto_m_Page_004.QC.jpg
598ce2166758a53fd98b85abc0ba3104
175f7854cc73c4c1ed74a868b1f197bfb8feb115
56425 F20101108_AABRJK ritorto_m_Page_077.pro
2ea74457e8ca7560e595a9b80051f2b1
fffa08b93f5c4df460c8660fb80c81dbd4589969
113661 F20101108_AABRIW ritorto_m_Page_127.jp2
0c395c87da4f0e2a43064e4ca3fa17d0
ae9e04096a8eca159bbeae7723321ae27d8cb162
9500 F20101108_AABSNC ritorto_m_Page_053.QC.jpg
9be2fe94c3c2f5fc9101054fe65374a0
d48776ae050be26d8bb444f651202b97d11da524
6635 F20101108_AABSMN ritorto_m_Page_026thm.jpg
0512b47a03c910296a8e2cd9e076e5ab
b0faaa5a50f17ef38ab044974054ddc0dabcd9c5
20160 F20101108_AABSLZ ritorto_m_Page_005.QC.jpg
f3d806e1d0e77a38e0c7817b18242720
f8dfcf31601fd24d3bd6aedf35da3da835f7df43
7155 F20101108_AABRKA ritorto_m_Page_036thm.jpg
8d3c36c0166d22666d6051fe7b14c28e
534139ded4abb7f5d1ec312a6bf011be45d44588
6466 F20101108_AABRJL ritorto_m_Page_098.pro
819eae013d74663c5213c0f88aa64672
5a5094f6935619a19f2a421808841b9e2854ccc7
51738 F20101108_AABRIX ritorto_m_Page_017.pro
82fb543c50bd7533abf2347c59ed2197
d7de3e7f1481db2b938a3b1f1dd9c0722852909d
9425 F20101108_AABSND ritorto_m_Page_054.QC.jpg
824ae84e895a642eb3de79427320ba3e
4092564ce60e52963258a37781c167005ba936fd
27989 F20101108_AABSMO ritorto_m_Page_028.QC.jpg
c58a1471362e65adb70e2a6029e5c840
5b1e949893124a734ff999c5809d1ff007b210c4
56060 F20101108_AABRKB ritorto_m_Page_024.pro
2375c60b4775fa2b21261c7b83a10584
7d438d03115c1cf615f00badc8e0a9deacd5798e
66928 F20101108_AABRJM ritorto_m_Page_013.jpg
5092936de63dc74ab7910185ea22fd23
7ae72c04bea6ec2f77a8d9ff4a4078cb40284d8f
F20101108_AABRIY ritorto_m_Page_014.tif
0c7e098b64e351de62f9348610d0353a
09451a6f7ac2d0e74bf83d8506e7f798444fac87
9764 F20101108_AABSNE ritorto_m_Page_055.QC.jpg
283df121a36f7daa267a40a352e4d165
8a99359746d31160a004ee3c49b78b48c6b30c7b
6713 F20101108_AABSMP ritorto_m_Page_028thm.jpg
5ca09a930b0e1defc9aa267c31eb36d3
a0769a4340b2578cee6edf9c0332161c6e5e2d13
28016 F20101108_AABRKC ritorto_m_Page_042.QC.jpg
9c3e8984493da5f9c2cbfd627cfe6d7d
2e1346eba29ca48aedb916b478c73b7bd2d44294
F20101108_AABRJN ritorto_m_Page_066.tif
5ae459a5d95be070f7bbfa0998fed65c
5fb80f477ce42e5c5b2f56675de5953ea6b75031
780207 F20101108_AABRIZ ritorto_m_Page_107.jp2
8ec4eccbb0cfdba97b0a6f3e49b36f8d
477ec1f2189ca53498d7ed6694cf1111531caece
13441 F20101108_AABSNF ritorto_m_Page_057.QC.jpg
16c31d1d3dd1f64203d14933a4d6a5f2
03d41846e0c509e5788381c755cf0b6814c7d730
6749 F20101108_AABSMQ ritorto_m_Page_030thm.jpg
3de0265f4fbe5623dda1495aa7c79540
f3354548e106398250d989b19413f42297437764
54207 F20101108_AABRKD ritorto_m_Page_132.pro
781f15e6daef2d236a16a3e3c3af2861
d4cc7d8cb8ca3568631a6dd300c0c7dc20090de6
347 F20101108_AABRJO ritorto_m_Page_091.txt
358534b610e6126acb2466a715288d57
cd65edfd2356ffa79b3ccca8864823071cae6ff6
6475 F20101108_AABSNG ritorto_m_Page_059thm.jpg
f14bbed738ab5faed9886610197004d4
ca60ad30b94fd17065fed106a4070ffa1fc5393c
27914 F20101108_AABSMR ritorto_m_Page_032.QC.jpg
a7782bca418331e52902e69b6015c3e1
7bd52a61c053471cc3fcccb72a6f5e13cfcb12c9
23265 F20101108_AABRKE ritorto_m_Page_012.QC.jpg
24c6cae8d8de9809b0bfe8979de2c8e8
90671204a0b374d01686813c14d7b06f27dcb313
1834 F20101108_AABRJP ritorto_m_Page_044.txt
007f5a3315c41204eaec67693bc1c8d7
6de0d0c72c9ba898364fcd02e2c7f9e898760476
6897 F20101108_AABSNH ritorto_m_Page_060thm.jpg
96b623b019b70039b6073c06ba095d01
27ef00eae6dfddc22c38b31f4cef90eb4c2f8b39
5452 F20101108_AABSMS ritorto_m_Page_034thm.jpg
f287fbca6a9b19da85b681d41beb2340
89cfe0fda413b736cc6b1eb90e54041047fec192
F20101108_AABRKF ritorto_m_Page_037.tif
567b9dbfda708d25693a7f8c6f0a11d4
e507d7dbf958654995b9b779c4efbf73dcc70e9b
1849 F20101108_AABRJQ ritorto_m_Page_048.txt
089f32d8d724a13971834bc5d917af56
df19ef3c295383ce2e47aa0ff6225b628995c9e6
27188 F20101108_AABSMT ritorto_m_Page_036.QC.jpg
828241a7578fd16521a3109a6095e88d
e38ea9b14c405fa69079e6642904c4be5e2145c3
3680 F20101108_AABRJR ritorto_m_Page_003.jpg
aa298e15346cf310f1e1557f22349574
428efca406e7ac419cdc824b66b1fbd14764791f
6744 F20101108_AABSNI ritorto_m_Page_061thm.jpg
238d73894b23682eccc363d598803c13
b4b30cdb0315ab1e3c7882483a1fc3fe537e9e18
3112 F20101108_AABSMU ritorto_m_Page_037thm.jpg
e087f1a6dbf979b7f3e7dea56e4da982
1516555b8bd820cc499ada0a8808de4911f5c11e
11693 F20101108_AABRKG ritorto_m_Page_137.QC.jpg
3630ccf7a994f9651994028c0975c1c5
71c7c71c60dbc512d5229de31b7819138f4bfed9
F20101108_AABRJS ritorto_m_Page_012.tif
4eb9b6ec0c9804f40183c390e595d6e8
ef81ac3ebe5b7c16a27f5d0719862705c4fd36da
28905 F20101108_AABSNJ ritorto_m_Page_063.QC.jpg
75b6c9704a702911a5b3e98e472b6b90
cf28d5f8759641532ab67735fa978daae0e1c70a
27613 F20101108_AABSMV ritorto_m_Page_038.QC.jpg
a3d3c640207e939ecf92119ea7451926
650dc6485e1b9ae38627b118ca99ddbe33285747
7004 F20101108_AABRKH ritorto_m_Page_129thm.jpg
504541e2490dbac4aa80e632fd729364
0250b05d69a2982040ce67ace74f5bfced8f121c
1594 F20101108_AABRJT ritorto_m_Page_013.txt
d4c2f856008e41af1100f4274d63374b
74f8e4beb3adefda4d6b19ed9bb33ada7a13d92b
6913 F20101108_AABSNK ritorto_m_Page_063thm.jpg
2ec8e191e3213f42d68bb8c1860cea10
605b7ece961bf6ae2a99bccee04f61b517e28a3f
6508 F20101108_AABSMW ritorto_m_Page_043thm.jpg
11e77c1e63d52e197fb5065231a60d2a
372795df872088f921af608936d8fdb5ee0269a1
59218 F20101108_AABRKI ritorto_m_Page_129.pro
5a0de2a67111ff93383d0728cdf65d2e
634ea7782a12ba17c11c103d69bb12f565cf0a7e
121701 F20101108_AABRJU ritorto_m_Page_016.jp2
5b726b55a9accec727d886bfec93db4f
74d77abfa2febd02f1d9c361878a7964d5911763
6411 F20101108_AABSOA ritorto_m_Page_084.QC.jpg
6c8c0d164349410dd0881065fc3bc39e
d45cd08433fd6aeac08547f1597abb0d380dca3b
30359 F20101108_AABSNL ritorto_m_Page_065.QC.jpg
d235279c323ce6f754c8a042d9fa5493
5f0c0173dac80d8f704d500559a3c0ee6fec1825
20055 F20101108_AABSMX ritorto_m_Page_044.QC.jpg
dbd3227b8e5f9adbe6c829eed4243b84
3518eb10ebde8f520acadfac6ef5679b4a5cbe1a
2118 F20101108_AABRKJ ritorto_m_Page_133.txt
46fbcce31e8c97e2ae11e657a27a9249
fb24965eabedf660d0164b44095469c3f3a60bbe
112681 F20101108_AABRJV ritorto_m_Page_062.jp2
c781e67d329a5af5cc58e8542d28b9dd
361fc28a8494693a97b6119fcfad4a8aaaa9f885
1915 F20101108_AABSOB ritorto_m_Page_084thm.jpg
b5d9821ea18aaeb64af34dc5a06ff6f8
6dbf527ce1983122c8b41fa6293590d363070273
28765 F20101108_AABSNM ritorto_m_Page_068.QC.jpg
6b558cdec899f06b5ff664d378809b7c
11c007359c0cabc7489aa4c0282509db49683c5a
25797 F20101108_AABSMY ritorto_m_Page_045.QC.jpg
1a884def8453deff63dd41603d8dd320
dbce4e0fa2b297bfe2df5953d1eb3881188fdeea
F20101108_AABRKK ritorto_m_Page_145.tif
adb298b0508f2f87e9a0919c7a55d745
79c8cf058240e9c4d656061450d032ea8b588fb5
7481 F20101108_AABRJW ritorto_m_Page_001.QC.jpg
122f6dc1d2a835b1503899d223f38df4
434fcca2aed01e4136a8f39164c13afc7b65c5a3
1919 F20101108_AABSOC ritorto_m_Page_085thm.jpg
7b31fd9ac279d54b8a77833e3da019b8
448681d1c22458075c23677ed82ada22084aaf3f
6927 F20101108_AABSNN ritorto_m_Page_068thm.jpg
4868f6365a7beefc3c2fbe1e1f7883ec
fb7ef04c9d1f91c08883daf2808de3c02ccd2c24
6464 F20101108_AABSMZ ritorto_m_Page_047thm.jpg
fcaeac8742e045a0dcb899208cb65ea8
d147f57d2dff5b5cebc051f843bc000e4062b84b
4962 F20101108_AABRKL ritorto_m_Page_044thm.jpg
f081ff2864d6bbfcfd00bd32affa153e
a3c083fdb9e169b9dd41113b1d16f7a945e704ee
53058 F20101108_AABRJX ritorto_m_Page_061.pro
c8a239a7093cd1f6d399ab1583d7cf9a
9a289aa5017a69f375282d000790e381fb254b4d
1624 F20101108_AABRLA ritorto_m_Page_099thm.jpg
7a120f1b29b6e3daa9d0bcf749b1f277
d6ce47554887a2927c3a80abb69877e18e261865
5676 F20101108_AABSOD ritorto_m_Page_086.QC.jpg
68cf0db2dcd5dc5d484b6aa41b919901
ad958ea4ed7e1a5aee6625ba413dde86d8fa68f4
6849 F20101108_AABSNO ritorto_m_Page_069thm.jpg
cb3b2cb63406b414224bf40f5ee7ed1f
1ba294eb0d83045174ae69241aaad6b86d6bf457
28010 F20101108_AABRKM ritorto_m_Page_143.QC.jpg
5801cfb416f6babf5a074ace9d7f37f5
532ad82c675db21e20359fcfc5a18542a27278f0
180 F20101108_AABRJY ritorto_m_Page_053.txt
628c46d679c659642c310c7014feec3b
90a2c15136318de05b2a3094a3e700d70c133fe2
2292 F20101108_AABRLB ritorto_m_Page_039.txt
3d7e4b00c974470aa67935528aed97fd
717f35696b2445a39a6055b4d085fb6dd451cb7d
2620 F20101108_AABSOE ritorto_m_Page_087thm.jpg
24842edd5f587fc920dc38334a5ff639
6bc423ef4ed40b82dfa69b32aef3db96440a0cff
6598 F20101108_AABSNP ritorto_m_Page_071thm.jpg
9969212c56716e7543806e87e3bb006d
ab9b174056b43e3be2c873fa4abcc6e8d4517c36
2625 F20101108_AABRKN ritorto_m_Page_116thm.jpg
0665d20074f346eb29b46120637cb145
ad1f6eafdb1b1db4276c05343317a9b6e7423a64
2042 F20101108_AABRJZ ritorto_m_Page_017.txt
916f8154e5f1c6b3ad6910673e0c1025
43407dce00303c9a369024a25eafccfdfb875024
100459 F20101108_AABRLC ritorto_m_Page_143.jpg
482c1b3b5131de82f954c0481b99542a
823398cac849ab0d6f4d8727b3f89e0586d3e9c8
2981 F20101108_AABSOF ritorto_m_Page_088thm.jpg
08c8bd958f5486b025014a2b723ed765
b1a8eb7960e47ed278e92e3827f6b1e7d7d8ea0c
28893 F20101108_AABSNQ ritorto_m_Page_072.QC.jpg
05cc1d0ddde685a9bd5d91a9fd10613b
d8d8f7bff39ea19abb94190ff8d6b7d264f8f95a
58166 F20101108_AABRKO ritorto_m_Page_064.pro
1fcfa0e6211f3d4b3b6f63611fbe99f8
c4151e8d1f4d493af11e2cef432da921f2e2180e
26902 F20101108_AABRLD ritorto_m_Page_015.QC.jpg
f0a91658dbb0061d7b00706993eeba5a
4a0630e39ce95112727e2c7cfaa2267001219a76
9890 F20101108_AABSOG ritorto_m_Page_089.QC.jpg
68851211245c6d79e0db15824c99a3b2
9cc5d7ceeacc32de9173b0ed63adb3c38136c3aa
F20101108_AABSNR ritorto_m_Page_072thm.jpg
5da1e2083203dbf71476fa29d1fdc331
2ba8cb9d6dd830cea1518a73ef5000350b4d96b2
F20101108_AABRKP ritorto_m_Page_028.txt
99c5d8148e385b402889ed93cf09fdd8
b885495cd982749c10f7d036ddc6b3a695f9abf6
21237 F20101108_AABRLE ritorto_m_Page_013.QC.jpg
e058919ba38465c391b949ae4ea9e391
ad0a7c010acad956225244d9822819031bc80e20
9884 F20101108_AABSOH ritorto_m_Page_091.QC.jpg
c3464c052ccce543588631bf0c134ece
438b28a6757159c1e1b1bfc9653e05bb47db149b
29486 F20101108_AABSNS ritorto_m_Page_073.QC.jpg
19f538d7b8cc1ddc642125fcb6b33412
fcd1f54cb32fe2a29ffc51d70b82e62fe894ea85
2178 F20101108_AABRKQ ritorto_m_Page_063.txt
aa1ace48523ae67bb44018e8c9aa4d0f
0e86d79add37de4a64136ce80495d13a8fa65f79
109262 F20101108_AABRLF ritorto_m_Page_122.jp2
9d9399f5b91e33428a6bea106d15303c
9a9e4d722d3595da0d313ef8e6614377c4426f56
3163 F20101108_AABSOI ritorto_m_Page_091thm.jpg
6999a942cda2c9e34a59b8b260534bea
92b52497f667b6a0693c6a5f400a8500dbf9f72a
6415 F20101108_AABSNT ritorto_m_Page_074thm.jpg
def5bf66735d7728d38ce7da52572ebe
aca337f5f657f36abc3db78f07f5db553abe8716
26639 F20101108_AABRKR ritorto_m_Page_031.QC.jpg
5e2737deb76f320e8568b284a9fdb64e
f0fdb688da79304dd1a825aaa36aa758d84eb522
97800 F20101108_AABRLG ritorto_m_Page_051.jp2
5ea0143ff6f258232b514fc050b604d8
a90a063b7f36d83fa3c69bf9e14bb5be8cb88447
28793 F20101108_AABSNU ritorto_m_Page_075.QC.jpg
9f3bbc741aad23b00c8b1f23590d9979
c4fdf1f6e0c6e57a5ce21e25ace459c58f60c5c7
338 F20101108_AABRKS ritorto_m_Page_098.txt
cba924e7ce0a5eaa159ab09c6243271f
4592ecfe53794495f7392b76563a5d52ef31f678
4181 F20101108_AABSOJ ritorto_m_Page_094thm.jpg
7899e14b507eb5b23f71e76e463d7e5d
e043519e5fcaef5e46de9ecd723a74caf7791642
6809 F20101108_AABSNV ritorto_m_Page_077thm.jpg
a3d3e9726723d7d462f7aff9b75673a6
3a5be28e819177a508263e3ce625fab05b05231f
5256 F20101108_AABRKT ritorto_m_Page_118.QC.jpg
a538c2eccfcfe41793042a6a439d704b
c986868207be8f93d6f558873887812fad809b33
2820 F20101108_AABRLH ritorto_m_Page_080thm.jpg
7977c5710fa130e51c4e1baa846dac40
f459c52b33b63506f31b2034d97128442b074f12
2487 F20101108_AABSOK ritorto_m_Page_098thm.jpg
d11ab2d2013d0b111edbedee00273ce5
b32be59484e9da4abb3a6ea183d295c0c38b1ccf
9231 F20101108_AABSNW ritorto_m_Page_079.QC.jpg
342ef2dc49b9592500bdb3a5c78b4065
5eb28bc405eb08674dcfb5538d69734bb45fbd0d
28193 F20101108_AABRKU ritorto_m_Page_025.QC.jpg
22997fd7a69ce3bedf790d4625d58902
cd52d2120ba3ec39899e07ff636df35380c08224
554 F20101108_AABRLI ritorto_m_Page_106.txt
5eb4a4b933b4a9d367ae6dae12fea05c
b2c864885103aaa9042c48e0799c6512485e9fee
7073 F20101108_AABSPA ritorto_m_Page_124thm.jpg
8cb925a556deff62fd110127ba98138e
ddb6308a4caec053c5994a566498840f4d70bde3
8263 F20101108_AABSOL ritorto_m_Page_100.QC.jpg
e3d101b99ba2d23d34a3ceea9c5657b5
e0a3751eb7c47452dda87ef121586704c845b6a6
2929 F20101108_AABSNX ritorto_m_Page_079thm.jpg
77e7ceb74a946db8a11730800b2600bc
be5da9b5ed676beda523010ad58fe220d354cabb
14015 F20101108_AABRKV ritorto_m_Page_097.pro
5f09301f2ab00ed5c1b2f90876eda235
791531a4ab37b4c641e174e639981f945ec7e03f
6378 F20101108_AABRLJ ritorto_m_Page_091.pro
a74b3308671e3ec880a0eac0fa6b55db
08cefdf2b82055b4c5fe70f3f651342146e51ad2
6918 F20101108_AABSPB ritorto_m_Page_125thm.jpg
302c2b86f3d68a9d72b13a73f9aa882f
593fb95406169b094bb0ba0ffd674f43746f075f
12483 F20101108_AABSOM ritorto_m_Page_101.QC.jpg
9ef5ff326a54e627f4fc072cdefd3e87
e53fe1708470f51528033b8e8e76e8ef779e00d8
2618 F20101108_AABSNY ritorto_m_Page_082thm.jpg
a792fc6dedd53b5515111096f31450a5
43af50fb3b8d192251f9a879bfc240f8d1b071e6
41927 F20101108_AABRLK ritorto_m_Page_037.jpg
b66f0e4a6b6916d1159d59a31e749143
077fa40e2ae300f233e0f33a083bd29925217037
21136 F20101108_AABRKW ritorto_m_Page_052.QC.jpg
f5a1b61a1ccc6c19e41fd7ccb92dbcea
831cfe3a488ed2f892e0af16889a4bad5a0e0aee
28562 F20101108_AABSPC ritorto_m_Page_128.QC.jpg
84c08fd156bd57f96b92af501d63800c
22f115c7a1cad532d476236e13a1c3e988c2d83c
3704 F20101108_AABSON ritorto_m_Page_102thm.jpg
af44685656172ade0a1b832e4019c00d
5a08763b87090b8dfa1858eda5ebebbf566eb594
2657 F20101108_AABSNZ ritorto_m_Page_083thm.jpg
38674d5e7ac869f30abb0d595c08d033
2c3e33e67661115ab700c3286833355bffe464fc
71329 F20101108_AABRMA ritorto_m_Page_009.pro
ec24397031fbf502dd73f96ea2a495ce
b0e272d1daf91ede2007b57a14d90c46f8e6ef73
F20101108_AABRLL ritorto_m_Page_143.tif
f62815885582298284f2c6495659d29f
065491deaca983e88bf9c2387b6fb86d272a718d
F20101108_AABRKX ritorto_m_Page_092.tif
5f3f993e82e9862fa142f5c0e9f6a886
10446b0f8c5a269acc93ee58e754a169e355f929
6830 F20101108_AABSPD ritorto_m_Page_128thm.jpg
e1f6a8328ee1390890543fd58b1543cf
6e5eb3b79a83c42670787b57bb395a7fbffb4b1e
11791 F20101108_AABSOO ritorto_m_Page_104.QC.jpg
0580149351528df4eaad9bc79f10f788
23da49d6866c9a4d16b8049f106f5d952396998e
1051983 F20101108_AABRMB ritorto_m_Page_010.jp2
f1f5922109932b5b700445000af89b91
1d3f44b6f72a7d27d871919b86be1c3892bc5006
6014 F20101108_AABRLM ritorto_m_Page_049thm.jpg
42383117b870547868d6fa268687bf52
385a896a21812b98e734d01e59b4ac0a536c6614
51542 F20101108_AABRKY ritorto_m_Page_045.pro
4c1f042a185a7ff14e51c99ea2a1b16e
da5d12ed86d2646a79abfa52334113bf5df960e2
30797 F20101108_AABSPE ritorto_m_Page_129.QC.jpg
7c59ef4d75cead46ebba29810d64d028
bb1ab6d8b48b808651148f349412bb39848b7d0c
3899 F20101108_AABSOP ritorto_m_Page_104thm.jpg
350502b8601806c5ae09b5e657ad7fdb
9e7ba6fd93332475beeeb33c4dbde72ea90edf3c
769 F20101108_AABRMC ritorto_m_Page_114.txt
d36a36e7109290a7287635ba3774b220
9b1855070f67d29e58c2b67e7117f5400eb6c52d
115364 F20101108_AABRLN ritorto_m_Page_050.jp2
5495cd582d9e38d3ac06324303f741b7
dd8b750f79575c48a1cd163de58ff70e18d89b4b
466359 F20101108_AABRKZ ritorto_m_Page_054.jp2
55d3e3e5d3af33718c57a85bd9d52318
4458cc99541fcf9cb3c9ff7613d0da0882bfb9a9
29623 F20101108_AABSPF ritorto_m_Page_131.QC.jpg
9f902b3d31de3e1d4e527035b0c66074
63f65000e586f481f9fbc72267f110ec82382598
13613 F20101108_AABSOQ ritorto_m_Page_106.QC.jpg
8d21b22796483b9455e7aa54ed6ff135
7d06b5d8bc84f4d77fe1ab2775b8823b026fbafe
6254 F20101108_AABRMD ritorto_m_Page_076thm.jpg
20b84b29c3b66190f298fdc1e63dd87e
a206bb7849f3978b4f7657b5f7be3431e7f18950
55510 F20101108_AABRLO ritorto_m_Page_063.pro
e997d47c64accc6caa07d22ebe990ba7
53f771406695a717b59b09b66f0db21e61653c72
6373 F20101108_AABSPG ritorto_m_Page_133thm.jpg
139608d4655eb7a8fadc46c20d19d725
8650b8f807971802599708db3803b8ecebd5d657
4261 F20101108_AABSOR ritorto_m_Page_107thm.jpg
592718b6da1867b49d05ef83806a6a1a
58d8b120058a47a662940c7ca28fc7d9131f5b45
99790 F20101108_AABRME ritorto_m_Page_012.jp2
539f773980681ee0c01d6671ff9ecc3d
7d86a5fe105763b63d26694277e44d16b7fec283
87801 F20101108_AABRLP ritorto_m_Page_070.jpg
9506923398ce0d94a4044b55f5d7ae05
2fb4b39fce8b873537b21ace855aaad7ce7d37b4
6522 F20101108_AABSPH ritorto_m_Page_135thm.jpg
e68e24e02a8ec1cb44e16bac1cdecf94
320cd511d449c969c98cf7afd0f96a4ca94368cc
12888 F20101108_AABSOS ritorto_m_Page_108.QC.jpg
0d78186d69597a56a5e9db411b5dbba1
cfccd5401d8ccfcf643edb2d3f697b9e5d7aa6d8
27975 F20101108_AABRMF ritorto_m_Page_040.QC.jpg
4d32f512eb9c6561c871a1d163ae5868
6ce18007e01d984fb95ece3711999d4a7b713cc7
189 F20101108_AABRLQ ritorto_m_Page_085.txt
886e14c89a09a46b2837709a326f0c38
b69e86698131b04a28d3fd9cf8c956c627ff329c
F20101108_AABSPI ritorto_m_Page_136.QC.jpg
9573b2b5f7feeab7d27f4ba3663f43fb
8d96ef277a32ede84ae4557a357302c24aab5dc9
12862 F20101108_AABSOT ritorto_m_Page_109.QC.jpg
5eabd6daaae2be5cb882b6aefd58ede9
55a6dd302805c23aa5d51dc9d8fe4b844f0a5679
56931 F20101108_AABRMG ritorto_m_Page_072.pro
ec4096c5a71556edb83bced077205ca2
56e885e5787b50e80f63e154bb1ce2639a9516ba
5775 F20101108_AABRLR ritorto_m_Page_085.QC.jpg
4dd9191135c7c549745d08e7df0e6ed5
a47ee3eeff540a21a89369b2c301e708a58f056d
6920 F20101108_AABSPJ ritorto_m_Page_136thm.jpg
80d34bcb811c9c9c040392e85f7a7ea0
8f5c6cad86c9e1a10838855ebccec123167f4d2c
13205 F20101108_AABSOU ritorto_m_Page_111.QC.jpg
b77e1f68b52531604e874d81dd5a4841
194b5436c9b54b760dfa764cd8ee8460d4b963b0
11038 F20101108_AABRMH ritorto_m_Page_109.pro
6321616252b763cfaef5a998cf99fe35
adf7ec73dc0daed6699feed29160b46a145afc86
42412 F20101108_AABRLS ritorto_m_Page_007.pro
e07c12059de9af1e866a7330f0823253
5682c33c967b47563a6ebf1aadfa36725298031d
4358 F20101108_AABSOV ritorto_m_Page_111thm.jpg
1ce3fc4b586b7ff86eb24eb38ae818a7
f239b04303152ad3a50d867c9e82c3db3f78a2e2
6630 F20101108_AABRLT ritorto_m_Page_088.pro
1f71d82178c6cade401d28f1485e1937
d9ef4f59ee3d11e9aa7f00b16d18516a724f8f6c
1824 F20101108_AABSPK ritorto_m_Page_138thm.jpg
7778022cbef815cb582c877e9cf65baf
07f05a6a150c5fa5753ef3197dde212bff0e05ca
4242 F20101108_AABSOW ritorto_m_Page_113thm.jpg
736e338550663fcbb2e89fe5960ec662
b0aaa96fd6861888e431ecafa6acba49de0aeb6a
511 F20101108_AABRMI ritorto_m_Page_004.txt
b4ea4a55eb6cec68183ad10e4aebd84f
4835cc2b5fbb639566d537b27933e49c3f56c7e1
682985 F20101108_AABRLU ritorto_m_Page_080.jp2
95da31fa212a4130b222f0b3202f431a
96f0f512f94c3ed76e83b67c7ca6750d732225fe
6156 F20101108_AABSPL ritorto_m_Page_139thm.jpg
f5ba33ad57591fac06e874e1921cc5a0
7681fa7adb68ba88b012e8e5d813b094ea4f2d61
3046 F20101108_AABSOX ritorto_m_Page_114thm.jpg
64770ca0bbee3aefc3f32e2c901bb1b7
b6403238e185f8a1feaed72aa923714d15b4a846
736177 F20101108_AABRMJ ritorto_m_Page_082.jp2
5c1baa915f78b4421140093a86f41701
f7e5f5dfff4414ee0daa7a025cca20018323f488
55244 F20101108_AABRLV ritorto_m_Page_131.pro
d7abff976f58e13e0fbba50f3c10fc77
5b29e52301317a4818806025f7f5f57e707a25bf
6412 F20101108_AABSPM ritorto_m_Page_142thm.jpg
0f67a9bf11bf37ca1b18e3de11c32c59
3550ddffa80b66dd5b84728608f8c8b29d06b8a9
9667 F20101108_AABSOY ritorto_m_Page_119.QC.jpg
65ed3fb8d8732926280a1f9efab04db4
0449dc5209075675a583035347d1302e5724d4f9
30527 F20101108_AABRMK ritorto_m_Page_055.jpg
c458104489c65917cc2c33fd35f02ad2
7dc09890e48d25d2edabb4a940ab64b248f3c8b1
29176 F20101108_AABRLW ritorto_m_Page_041.QC.jpg
a31bb2ddd349b2e8459b1042765310aa
5a43bc7435c69c23a89c5c5857d782154ad0fbcf
16144 F20101108_AABSPN ritorto_m_Page_144.QC.jpg
20e67e6e27da7abb3e049a890c324993
9c04949c6ed62c88cf3f0683d91160448133ca5c
2344 F20101108_AABSOZ ritorto_m_Page_120thm.jpg
071d358da86a910d056762abe35899f4
a3c9f382ee85aeab0b134a0a1f25e61dd2296775
6799 F20101108_AABRML ritorto_m_Page_027thm.jpg
464828908ce9c226d350bf21cfcb5e9a
fdb8d5ee00c08efcda32e2bf6279ae5403b2ac58
791924 F20101108_AABRLX ritorto_m_Page_094.jp2
0a1e0d492cd0d8c08fc8a94b331b1cb1
02e8b5972c2762e95385b7c31525d84d0811572d
F20101108_AABRNA ritorto_m_Page_109.tif
1ae1f312bed256e68b02fe3005614e92
8db3dd198c2c0ddf3f2ffa4aa79ad030c544bc41
87863 F20101108_AABRMM ritorto_m_Page_071.jpg
4ed6b211abb1b5fe7f61e27bcd0d48c2
4b4d819f96f22a0e0bc68e19c56c0f442799822a
709240 F20101108_AABRLY ritorto_m_Page_112.jp2
9e27012d3ed47083bab2eac5e218f4f8
bad1d1e295162916eb6a1c1b81310edac07f59cc
427 F20101108_AABRNB ritorto_m_Page_101.txt
591072ef4b85a927d213e6bb65dd3b24
daeb466236704d1a497d2b9f5a0d6351c6e1543f
41565 F20101108_AABRMN ritorto_m_Page_101.jpg
7d0fd97983f5a9ca6c7dd00c719bcb36
838f1fc0cc765fa11e08f6b4ce5dc2e13056b54e
376954 F20101108_AABRLZ ritorto_m_Page_086.jp2
2fee210fddd6ab909ef11776c25ca034
430ea420f9203fab63464ccfe5e9b7b4210cf7ec
111305 F20101108_AABRNC ritorto_m_Page_043.jp2
72f50666caf58674c54aec7fe929c2a5
58b649324ce8771ac1f6d32ccd7134abbca802e8
6642 F20101108_AABRMO ritorto_m_Page_021thm.jpg
7ec0ffb134a3baa4c24df8ad039e76f5
15da31a10d98a649cb3e4116720f0d0af67d274c
533 F20101108_AABRND ritorto_m_Page_109.txt
1c924defd165356ab91b2a684e964b0e
0e7a1999b89dfddde3395d5397c474dcf2fa4c5f
5397 F20101108_AABRMP ritorto_m_Page_003.jp2
5b18353bac9b6922af9c417c7ce42518
8bb15c0b3bb1b57f562d10ac1020b063342e98e3
490865 F20101108_AABRNE ritorto_m_Page_084.jp2
a42350b689817f3ef7891c320ddc1a74
44862e5bf2cf4321811c4c89b12700ab47a3a41a
F20101108_AABRMQ ritorto_m_Page_087.tif
99305e4c3955130f14b8fa3bc4f5a25b
01379b60d088467f11cd5aca55807794e2248c78
2229 F20101108_AABRNF ritorto_m_Page_033.txt
fc2ce90b06be3797b6343ed0495d553a
083b306591f5700bf28d85baff7527bbfa1f018a
11627 F20101108_AABRMR ritorto_m_Page_102.QC.jpg
c12c8702d28868e99f4f9fe36a8845b4
95e84798c83280aed15f0d65e5e1409735de496c
2073 F20101108_AABRNG ritorto_m_Page_043.txt
720861193b0fe428cdce3a17a744e4dc
fcb210d3c09ba27f94bf8c5a750760f251ad3e9e
117501 F20101108_AABRMS ritorto_m_Page_040.jp2
4131cf6e86723ae8c79b299cd92bed7f
d55c72fd27fe4441d85fdaac9ffb01fda161b969
13193 F20101108_AABRNH ritorto_m_Page_097.QC.jpg
ebf867e7d77267a5d841318eefeb1518
19c326330126dd3b357085a2759709ba0e54dae7
124081 F20101108_AABRMT ritorto_m_Page_065.jp2
e1c509e542cda3e839d7b74315f359ae
6c09539522ac4279bb3a972a3d79f3ad362a1513
826 F20101108_AABRNI ritorto_m_Page_137.txt
b332515d6d5e4f3aad004c4262446c41
4b762af0cc3ff50a2897fe64e72a3c9f41389f9c
F20101108_AABRMU ritorto_m_Page_082.tif
dc79dd960fa139dbd4e2d15e07e8f284
5be6804abfb2365a1b677fe9a56a76e01c0a8364
26334 F20101108_AABRMV ritorto_m_Page_017.QC.jpg
34685a38e3b2dd0eb51fa6ebb4a01848
15346c3765eeccbdf1cc3dc99d79f5d2f35e126a
27912 F20101108_AABRNJ ritorto_m_Page_135.QC.jpg
00a6a691e714a0434a1eca83576419af
3f084afcd44336d7695930f08693f6adde0df3df
91882 F20101108_AABRMW ritorto_m_Page_024.jpg
ac3811ae9368095a9374f7ee6f790fe0
98276ff5ce1f91dd663e539f76d7758dbf89cd65
F20101108_AABRNK ritorto_m_Page_088.tif
46ebcf4000d0ab3c56bebc42dc357e5d
70702668c851967a55a6cd01d71067ebbeba4300
122024 F20101108_AABRMX ritorto_m_Page_046.jp2
ae5725f7c9592df040a21d98bb0277e6
461d3da7824d74b9a55098f3fc484da80fdf976c
54907 F20101108_AABROA ritorto_m_Page_068.pro
dba2428672462cfbaa3e21e7a7e46956
9835b7c5add247d30760294e76c72aa4ada4d543
50432 F20101108_AABRNL ritorto_m_Page_094.jpg
14462b7d2dd17cc68a929100f4b0df42
c8fa7c6f95d8b10a33a2781061ceef39ac3dd12e
8909 F20101108_AABRMY ritorto_m_Page_080.QC.jpg
3873e3772cda0f31b64f3f757dada63e
3d095dc18fd943352d15446fcd4ec02de18da689
3004 F20101108_AABROB ritorto_m_Page_010.txt
5d0555e2d3494ccc2e1df48a9b75fd68
45e38642befc009a57aee009b5057cbf09373336
84476 F20101108_AABRNM ritorto_m_Page_140.jp2
7c661d6ce5b698d30710d91da76886bb
5e95ab0eb9b1e40c5cde0890754e7298b18dc3e7
58008 F20101108_AABRMZ ritorto_m_Page_046.pro
f439059b80f575ea1e1a843cf2194ac1
7d3a2fde96ce239fb397b6d60c96d32935d8d887
1044 F20101108_AABROC ritorto_m_Page_002.QC.jpg
7b6cd8431ff997650c44d6da86849131
285041503805ba7d1199730264a3d400d9253f23
30215 F20101108_AABRNN ritorto_m_Page_079.jpg
8acbe2b3f9d094517d6e0d9586855ce8
4864fbbafa029afc041143f296177febb692d298
28563 F20101108_AABROD ritorto_m_Page_123.QC.jpg
07bc2f6af52fe51502f56168a76089ae
a2f418ccadf27bd827b633bca5553389878dd14e
F20101108_AABRNO ritorto_m_Page_018.tif
390d6d795754ce3836936ac16019560e
3d0ba0f137682e3e4028bd1e6694bf1044c0cfc3
15324 F20101108_AABROE ritorto_m_Page_094.QC.jpg
55d991bff4c94cbb7d31552ae7906ccf
552aa55a1f587659ce339a82471877707dfe0e98
F20101108_AABRNP ritorto_m_Page_128.tif
b2ba20b94920c6da799db6a465532841
3d11fd9ab234617d7de0cb8217c19d3eaa9daab1
454 F20101108_AABROF ritorto_m_Page_002thm.jpg
211ab83fcac157648cbb064e5134d536
9f83d0696ec610e340f19fdc695d3cb8cb73ae49
55021 F20101108_AABRNQ ritorto_m_Page_020.pro
774b61c69a56041c80a5a42946325bee
f910454ad493d2ecb80893ac0193fd535c72b97f
130020 F20101108_AABROG ritorto_m_Page_143.jp2
1d5506313f50a6daf8d4f9b6f8206010
8ce400ddf930764e40837552a835796f27efcbef
F20101108_AABRNR ritorto_m_Page_027.tif
281f3e8da6a616ac508e49f8fe0ab1a6
b46bc7c06afe961cd30958fcc679e37c1f516e37
386 F20101108_AABROH ritorto_m_Page_088.txt
15b6c469876b331b4395393495089b63
a06001a7ae114086379d5849d6ee061e684009bb
28916 F20101108_AABRNS ritorto_m_Page_130.QC.jpg
731f3f8e47bad0dc0a17002525c0126e
b3fd49b5a10e8ba1061d1798a708e21fc587c852
6457 F20101108_AABROI ritorto_m_Page_122thm.jpg
0aab35976cb1c638bb22799e983e4cd0
147fab4f94a376775eaadb065f918c40dd4e5181
2104 F20101108_AABRNT ritorto_m_Page_031.txt
1c01dc42310a022cc5ea34efef574566
f9665356d3b58b2d61c9a7592d1846419b8a2aa4
116611 F20101108_AABROJ ritorto_m_Page_142.jp2
a3b124e74c50b805157db63253c4c93a
49ee786295c4d7fe24372e288b630087b372cd17
2212 F20101108_AABRNU ritorto_m_Page_066.txt
c2b3e2f3234998ca6bfde877169e1996
a306ccacb87ec54039d68c3bb9691c16a588e003
18010 F20101108_AABRNV ritorto_m_Page_117.QC.jpg
e4b92576c46d8fe46f99388cdad8f593
13cf46ee77b1fc186791a65385a062102e5b70f7
F20101108_AABROK ritorto_m_Page_119.tif
fc298865660338702a193ed2ea52e22c
36de795ce46d3e4cc41ccf663dece0af8014ff09
2171 F20101108_AABRNW ritorto_m_Page_131.txt
8c33207d953189a83e18acc5defe1f86
3190c88fc5fe498814768b0b486bd0d31a1f1487
63776 F20101108_AABRPA ritorto_m_Page_006.jpg
a968976e12d591dee50b98d2d89b1908
233915b9c6cecad65457dcc18bded5f017a57033
F20101108_AABROL ritorto_m_Page_127.tif
99ad4b9f2cbcc88c2adb620132209c2f
9a1d6f0f29f57ec3e183b0b159376a136b026719
743 F20101108_AABRNX ritorto_m_Page_057.txt
9dc80ae0cc20e7aa8bbf63a85c4e0943
61174b01847ba086465d0392c99b5c3ff56a3dcb
112370 F20101108_AABRPB ritorto_m_Page_009.jpg
a9907c0a50ba05df273f4cbcd3886c78
11eb8aa82ac537b74d152b5f9d20b95438d6aa86
9510 F20101108_AABROM ritorto_m_Page_001.pro
8c312850a083c9cd8df05570304c5c10
dd5163b9a61305afa870acf358c93d251ef65da7
7485 F20101108_AABRNY ritorto_m_Page_120.QC.jpg
d90e4e3fe53df8f7529353a146559b20
d07dca1a7750d9e1f5ab6fa563cc48ee5d18b09e
129490 F20101108_AABRPC ritorto_m_Page_010.jpg
5343df37229b548922ec02eb208342c1
58556b7d19e5c53857191ca7ca98b492692042f4
22644 F20101108_AABRON ritorto_m_Page_034.QC.jpg
0ddada28f1fa152baa22eea72c93f76f
bfbf5add72edd25b39e1de2dc6a87de758932763
7607 F20101108_AABRNZ ritorto_m_Page_081.QC.jpg
a3d2db183868ea038a3fe31fec7a2f98
60e9b7a7cb2015cfa1107fbc7ba5addf41add829
78579 F20101108_AABRPD ritorto_m_Page_012.jpg
18521fda826ee3c5a2eddf659545898c
fe531b23304dfb445a9e163c15458683e4e204e0
118814 F20101108_AABROO ritorto_m_Page_141.jp2
28b079ded8c07418fe6e6f0e4bcc6bad
8d58495712d864ac76f41395bfa1ea5880f66525
87631 F20101108_AABRPE ritorto_m_Page_014.jpg
40a24b3258277804f741cc7dea4d072e
b05a39e4a8df9316f9b7626d968338e6b97e82ad
F20101108_AABROP ritorto_m_Page_094.tif
ae099d34d9b94b6d97a82a5fd7298f2c
8cf9fcaac8e80642df751a3209dcf3eba23fcfec
92442 F20101108_AABRPF ritorto_m_Page_016.jpg
090d898a1fd86caaefe2660351366298
52553fdc49523bc16d9101ff82229314480339fc
2173 F20101108_AABROQ ritorto_m_Page_060.txt
9a2af266a87e0d1c2430feed6ca45102
c4bbbe59b812ad7fb2cb746364d1ae413f9af5be
85806 F20101108_AABRPG ritorto_m_Page_017.jpg
8f0cbe633db5e769768c764631ebe53c
e2564499390c3b1f52a13b3dfc5cde2deb61b8b8
F20101108_AABROR ritorto_m_Page_121.tif
40c7ba62071df8aefa64a13ef1ef7f7d
d1bbbd456f760d863c7e3c9ede4f4acc2e592704
81381 F20101108_AABRPH ritorto_m_Page_019.jpg
a6f8765b371fc44b397c327696324175
edb4eab00f755041c5152c55eff40b700a191bee
6136437 F20101108_AABROS ritorto_m.pdf
7a023cc0c28663753343a183bf38bb69
61d8c143cc242e838737b36c9a092672ed048405
93403 F20101108_AABRPI ritorto_m_Page_020.jpg
e2e9515afc6d463a227a9e88f5b0976a
bb28828528384e69875409efab5b74f3abdd1f67
9978 F20101108_AABROT ritorto_m_Page_090.QC.jpg
183efc94e92258751e8892b83bda6535
f490e02c0e64d5d36c2c4acf18cb7cccb39d0de6
89246 F20101108_AABRPJ ritorto_m_Page_021.jpg
402f9048c7fb5e3f833e63a5bf873272
3ab5fe90a13baad15caace7a386e084bce5840d8
168132 F20101108_AABROU UFE0021715_00001.mets
c9078d143279bf90f76b9beac6f0dd93
8c731006624ae394b177ac573f8054f50b25d4d9
91317 F20101108_AABRPK ritorto_m_Page_023.jpg
7b50468d5f375a3f4b86eca18f882023
358912dee18177de80db21384fcc7832bf818cda
90520 F20101108_AABRPL ritorto_m_Page_025.jpg
aeda2975ecaad537a5190d390c1becca
846a31714a7bd83393924780fd8e0dd626fc423a
25733 F20101108_AABROX ritorto_m_Page_001.jpg
7dc7a0c445365ec13cbbade490dce68d
e5ee0e5d444d7c129423cf56c364ac5807059216
81482 F20101108_AABRQA ritorto_m_Page_049.jpg
6ab47908c783c66776b5fa537be75c53
279de782c911796852e82db8a0b45b193bcfaef5
93537 F20101108_AABRPM ritorto_m_Page_026.jpg
d1196fe1fe5bf418657b39f23ef49df7
7481dbb6212d15026a83ccce91cce565a623b3a6
3568 F20101108_AABROY ritorto_m_Page_002.jpg
5f7991398f34eec973eb352a4528af86
6823525b404d47c5028e332833f5c3b114eee922
88370 F20101108_AABRQB ritorto_m_Page_050.jpg
92f8c0bf4beb9b3c304076d86695034e
f1feed4b99e9f916afaa8f76fd02a8791ec1acc2
107886 F20101108_AABRPN ritorto_m_Page_027.jpg
f5d12144ba9f33cb0cd13a3bb935091b
34d303c3a717a96d7c43e53f37aae9926a0ca6d3
93453 F20101108_AABROZ ritorto_m_Page_005.jpg
effc497f668e375288df6dec9cb1b265
7e6e9fd64247fc3b0178053285fa54421c8ee16f
75348 F20101108_AABRQC ritorto_m_Page_051.jpg
400a159b75ef9ab28103f59f94c0d5b3
3cd5c2fc3b2bfc2cd89ddfca684dedc4e7f402a3
94013 F20101108_AABRPO ritorto_m_Page_030.jpg
80eb4b133877646e8b13c237611bfa18
efe4d7a11e243e3f44335f34670ee33621f2d759
68859 F20101108_AABRQD ritorto_m_Page_052.jpg
d996181b4026f35137197cd392986a45
7757cf41e2cdba46258aaf2e8ca3d76374d194f8
89889 F20101108_AABRPP ritorto_m_Page_032.jpg
c0c74712760a598a3a0651ad84ef8743
9b25f73e3094c8f2a1821bb37875ec77f4ab3f87
27825 F20101108_AABRQE ritorto_m_Page_053.jpg
6bbce0c03c1f17a57f5a38f3b28ae471
3225aab2d1e5d5de5097282721662b151e412192
30254 F20101108_AABRPQ ritorto_m_Page_035.jpg
d1f03e74e9df9c99cd3d911d271e9eec
017861232ca0ee34fc5e98f6b96c578d4ce2e90e
29193 F20101108_AABRQF ritorto_m_Page_054.jpg
10115d95e5089411b765d3ed22c87a72
b291aa828a7f219c6ed6a978de2d2b3cb8f5b10e
87051 F20101108_AABRPR ritorto_m_Page_038.jpg
4bf10c4dc07d2d6b8a43bb46b36d4872
2fd7b13c3c93db16dd9f89cbb3bc1b82d29c841b
17937 F20101108_AABRQG ritorto_m_Page_056.jpg
7db5b9fe6bf8f62c3b37cc0ab5e528f0
6bab873654339a0790d2f9f836efdf08c60072c4







IMPACTS OF DIFFUSE RECHARGE ON TRANSMISSIVITY AND WATER BUDGET
CALCULATIONS INT THE UNCONFINED KARST AQUIFER OF THE SANTA FE RIVER
BASINT




















By

MICHAEL RITORTO


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF SCIENCE

UNIVERSITY OF FLORIDA

2007











































O 2007 Michael Ritorto



































To my Grandpa Joe Ritorto









ACKNOWLEDGMENTS

I would like to thank my advisor, Elizabeth Screaton, for giving me the opportunity to

learn from her expertise and for supporting and encouraging me along the way. I would also like

to thank PJ Moore and Abby Langston for their time and efforts helping me complete the field

work for this project. Lastly, I would like to thank my parents for their love and support and all

of my friends who kept me motivated along the way.












TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS .............. ...............4.....


LIST OF TABLES ............ ...... ._ ._ ...............7....


LIST OF FIGURES .............. ...............8.....


AB S TRAC T ............._. .......... ..............._ 12...


1 INTRODUCTION .............. ...............14....


2 BACKGROUND .............. ...............19....


Study Area .................. ...............19.......... .....
Geologic Background ................... ............ ..... ...............20.......
Previous Investigations in the Santa Fe River Basin ................. ...............21..............
Diffuse Recharge in Other Study Areas .............. ...............26....
Aquifer Parameters in Other Study Areas .............. ...............30....

3 M ETHODS .............. ...............38....


Data Collection .............. ...............38....
W ater Level s ........._..... ...._... ...............3 9....
Sink and Rise Discharge ........._..... ...._... ...............41....
Evap otran spi rati on ................. ...............44......... ....
R echar ge .............. ............ ..............4
W ater Bud get M ethod ................ ...............47...
Water Table Fluctuation Method............... ...............47.
Chloride Concentration Method ....._.._................. ........_.._ ......... 4
M atrix Transmissivity............... ........ .........4
Analytical Method without Recharge ........._...... ......_.._.............__._ ............4
1-Dimensional Model with and without Recharge in MODFLOW ........._...... ..............50
Recession Curve Analysis ............. ...... __ ...............51...


4 RE SULT S .............. ...............59....


Precipitation and Evapotranspiration............... ...........5
W ater Levels ............ ..... .._ ...............61...
Sink and Rise Discharge ........._.. ........_. ...............62.
R echar ge .............. ... ................. .. .... ...... .......6
Diffuse Recharge Calculated From Water Budget ....._._._ ........___ ........_.......62
Diffuse Recharge during Large Storm Events.................. ...............6
Diffuse Recharge Calculated From Chloride Concentrations .........._........ ....._.........64
Diffuse Recharge Estimated Using the Water Table Fluctuation Method.............._._....65
Transmi ssivity ................. ...............67........... ....












Analytical Method without Recharge ................. ............. ............ ....... ........ 6
1-Dimensional Model with and without Recharge in MODFLOW ............... .... .........._.69
Sensitivity to Specific Yield ............ ..... .._ ...............73..
Recession Curve Analysis .............. ...............74....
Basin Area Calculations .............. ...............75....


5 DI SCU SSION ............_ ..... ..__ ...............122...


Applicability of Different Recharge Calculation Methods and Effect of Conduit
Boundaries ...................... .... .. .... ... ... .. ........2

Daily Versus Monthly Recharge and Evapotranspiration Calculations ............... ...............126
Transm i ssivity .............. .......... ... .. .. ..... ....... .. ........2
Influence of Diffuse Recharge on Storm Event Hydrographs and Transmissivity
Calculations. ........._...... .. ......_ ..... .._.._.......... .........2
Influence of Scale Effects on Transmissivity Calculations ........_.._.. ... ......_.._.. .....129
Influence of Storm Event Magnitude on Transmissivity Calculations .........................131
Choosing the Appropriate Method to Calculate Transmissivity ........._.._... ................133
Diffuse Recharge Component and Possible Implications for Dissolution ...........................134

6 SUM M ARY ................. ...............13. 9...............


LIST OF REFERENCES ................. ...............141................


BIOGRAPHICAL SKETCH ................ ............. ............ 145...










LIST OF TABLES


Table page

2-1 Geologic and Hydrogeologic Units of the Santa Fe River Basin .........._.................37

3-1 Summary of Monitoring Wells ........._._. ......._. .....__. ....__................57

3-2 Location and dates of data collection............... ...............5

4-1 Annual precipitation, annual evapotranspiration, and annual calculated recharge from
2002 through 2006 ........... ..... ._ ...............117....

4-2 Precipitation, potential evapotranspiration, and calculated recharge for three large
storm events ........._.__ ..... ._ ...............117....

4-3 Chloride concentrations in shallow wells from various sampling trips ........._................117

4-4 Recharge results for 2006 using chloride concentration factors .........._..._.. ........._......117

4-5 Calculated diffuse recharge for two non-conduit influenced events using the water
budget and water table fluctuation methods ..........._._ ........ ........._.........18

4-6 Summary of calculated transmissivities (m2/day) with different methods in the Santa
Fe Sink-Rise System. ........._._ ..... .__ ...............119..

4.7 Sensitivity of 1D model results to specific yield ........... ..... ._ .......__.......2

4-8 Area calculations using recession curves............... ...............121

5-1 Elevations of the ground surface, top of limestone bedrock, and water table before
and at the peak of various large storm events at different well locations (in meters
above sea level) ........... ..... .._ ...............138..










LIST OF FIGURES


Figure page

2-1 Location of the Santa Fe River. ................ ...............35.......... ...

3- River Sink rating curve produced by the Suwannee River Water Management
District for the stage gauge located in O'Leno State Park ................. .......................53

3-2 River Sink rating curve produced in this study using sink elevations measured by the
data logger correlated with stage at the River Sink gauging station ................. ...............54

3-3 River Rise rating curve produced by Screaton et al. (2004) using data from the
Suwannee River Water Management District ................. ...............55...............

3-4 Function assuming a maximum soil-moisture deficit of 10 cm used to calculate leaf
conductance in the Penman-Monteith evapotranspiration method. After Stewart
(1988). .............. ...............56....

4-1 Daily precipitation records from O'Leno State Park during this study. Three storm
events are highlighted. ............. ...............78.....

4-2 Daily calculated evapotranspiration during this study using the Penman-Monteith
Model. Spaces in the data indicate a value of zero. ........._.._.. ...._... ......_.._... ....79

4-3 Monthly calculated potential evapotranspiration using the Thornthwaite method and
monthly actual evapotranspiration using the Penman-Monteith method ..........................80

4-4 Long term record of River Rise water levels. Storm events that have been cited in
previous work and in this study are highlighted. ............. ...............81.....

4-5 Water levels of various monitoring wells during this study and River Rise water level
for comparison. Three storm events are highlighted. ............. ...............82.....

4-6 River Sink and River Rise discharge during this study. Three storm events resulting
in increased discharge are highlighted. The time period where the River Rise gauge
was warped is represented with the blue highlight. .............. ...............83....

4-7 Daily calculated diffuse recharge during this study using the water budget method.
Recharges associated with the three maj or storm events are highlighted. ................... ......84

4-8 Precipitation and calculated diffuse recharge with the water budget method for the
small event in June 2006............... ...............85..

4-9 Precipitation and calculated diffuse recharge with the water budget method for the
small event in July 2006. ............. ...............86.....










4-10 River Rise water level during summer 2006. Small events in June and July are
highlighted. ............. ...............87.....

4-11 Water level perturbation at Well Sa after a small rain event in June 2006 with
recession curve projected. .............. ...............88....

4-12 Water level perturbation at Well 6a after a small rain event in June 2006 with
recession curve predicted. .............. ...............89....

4-13 Water level perturbation at Well 7a after a small rain event in June 2006 with
recession curve projected. .............. ...............90....

4-14 Water level perturbation at Well Sa after a small rain event in July 2006 with
recession curve projected. .............. ...............91....

4-15 Water level perturbation at Well 6a after a small rain event in July 2006 with
recession curve projected. .............. ...............92....

4-16 Water level perturbation at Well 7a after a small rain event in July 2006 with
recession curve projected. .............. ...............93....

4-17 Curve matching results at Well 4 for the September 2004 storm event using the
analytical method without recharge. .............. ...............94....

4-18 Curve matching results at Well 6 for the September 2004 storm event using the
analytical method without recharge. .............. ...............94....

4-19 Curve matching results at Well 7 for the September 2004 storm event using the
analytical method without recharge. .............. ...............95....

4-20 Curve matching results at Well 4 for the March 2005 storm event using the analytical
method without recharge............... ...............95

4-21 Curve matching results at Well 6 for the March 2005 storm event using the analytical
method without recharge............... ...............96

4-22 Curve matching results at Well 7 for the March 2005 storm event using the analytical
method without recharge............... ...............96

4-23 Curve matching results at Well 4 for the December 2005 storm event using the
analytical method without recharge. .............. ...............97....

4-24 Curve matching results at Well 6 for the December 2005 storm event using the
analytical method without recharge. .............. ...............97....

4-25 Curve matching results at Well 7 for the December 2005 storm event using the
analytical method without recharge. .............. ...............98....

4-26 1-Dimensional model with recharge grid ................. ......... ......... ..........9










4-27 Results from of the analytical method without recharge and the 1D model without
recharge for a hypothetical storm event ................. ...............100..............

4-28 Curve matching results at Well 4 for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference. ........101

4-29 Curve matching results at Well 6 for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........102

4-30 Curve matching results at Tower Well for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........103

4-31 Curve matching results at Well 1 for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........104

4-32 Curve matching results at Well 4 for the September 2004 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........105

4-33 Curve matching results at Well 6 for the September 2004 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........106

4-34 Curve matching results at Well 7 for the September 2004 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........107

4-35 Curve matching results at Well 4 for the March 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........108

4-36 Curve matching results at Well 6 for the March 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........109

4-37 Curve matching results at Well 7 for the March 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........110

4-38 Curve matching results at Well 4 for the December 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........111

4-39 Curve matching results at Well 6 for the December 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........112

4-40 Curve matching results at Well 7 for the December 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference .........113

4-41 Recession curve breakdown at Well 4 for the hydrograph produced by the March
2003 storm event .......... __... ......... ...............114....

4-42 Recession curve breakdown at Well 4 for the hydrograph produced by the September
2004 storm event ....._ ................. ........._._.........11











4-43 Recession curve breakdown at Well 4 for the hydrograph produced by the March
2005 storm event ................. ...............115................

4-44 Recession curve breakdown at Well 4 for the hydrograph produced by the December
2005 storm event ................. ...............115................

4-45 Example of using two portions of the recession curve that results from the December
2005 event at Well 6 to calculate basin area ................. ...............116........... .









Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Science

IMPACTS OF DIFFUSE RECHARGE ON TRANSMISSIVITY AND WATER BUDGET
CALCULATIONS INT THE UNCONFINED KARST AQUIFER OF THE SANTA FE RIVER
BASINT

By

Michael Ritorto

December 2007

Chair: Elizabeth Screaton
Major: Geology

In karst aquifers, high transmissivities make it difficult to conduct traditional aquifer tests.

An alternative method to evaluate formation transmissivity on a large scale is passive monitoring

of observation well response to changes in stream stage or conduit hydraulic head during high

discharge events. During these events, increases in water levels at observation wells are

dependent on the stream or conduit boundary condition, but are also influenced by the amount of

diffuse recharge arriving at the water table. This study used data from the karstic Floridan aquifer

in northern Florida to evaluate the effect of neglecting diffuse recharge on the transmissivity

estimate. At the Santa Fe River Sink/Rise system, the Santa Fe River enters the Floridan aquifer

at the River Sink, and re-emerges downstream at the River Rise. Between these locations, cave

divers have mapped over 7 km of conduits. Diffuse recharge was estimated as rainfall minus

evapotranspiration calculated using the Penman-Monteith method. The transit time of recharge

through the unsaturated zone is not known; given the relatively thin unsaturated zone and its

expected high hydraulic conductivity, the simplifying assumption was made that recharge

reaches the aquifer on the same day as rainfall occurs. Hydraulic heads were monitored through









time at locations within the conduit system and in wells monitoring the surrounding aquifer. Four

storm events were recorded by the monitoring network.

Aquifer responses to storm events were first predicted using an analytical solution created

by Pinder et al. (1969) that does not include diffuse recharge. For comparison, aquifer response

was then simulated using a one-dimensional model implemented in MODFLOW incorporating

the calculated diffuse recharge during the storm event. In the model, the time-varying hydraulic

head at the conduit was simulated as a specified head boundary condition. Calibration of the

model for individual monitoring well hydrographs yields estimated transmissivity values for the

aquifer between the monitoring well and conduit. Results suggest that neglecting to include

diffuse recharge in the solution results in overpredicted transmissivity values. Results also

suggest that the magnitude of diffuse recharge in comparison to head change in the conduit

determines the relative effect of the diffuse component on the observed well hydrograph after a

storm event. In terms of total water volumes, results show that diffuse recharge can be more than

twice the input received from the conduit boundary during small storm events, and less than half

the input received from the conduit boundary during large storm events. These results suggest

that in future calculations it will be important to consider diffuse recharge as an influence on

dissolution across the entire Santa Fe River basin.









CHAPTER 1
INTTRODUCTION

Considering that more than 25% of the world's population either lives on or obtains its

water from karst aquifers (Karst Water Institute, 2006), it is no surprise that a growing effort to

model karst aquifers has prevailed in recent years. In the United States alone, 20% of the land

surface is karst and 40% of the groundwater used for drinking comes from karst aquifers (Karst

Water Institute, 2006). In order to preserve the useable water from these aquifers, we must start

by building a solid understanding of the hydrologic processes that affect karst terrains.

Karst aquifers are characterized by dissolution-generated conduits that permit the rapid

transport of ground water (White, 2002). These conduits may range in aperture from as little as 1

cm to about 0.5 m; at which point the pathway becomes a cave large enough for human

exploration (White, 2002). Under normal gradients, conduit flow behaves in a turbulent manner

(White, 2002). Conduits are often connected to the surface via openings such as sinkholes and

sinking streams. These connections allow for recharge to the conduit, and mixing of surface and

ground water (Martin and Dean, 2001). Connections to surface waters are important because

they allow conduits to quickly transmit pollutants without filtration over long distances (Mylroie,

1984).

In the early work on karst hydrology, the focus remained mostly on determining the

properties of the conduit systems. However, karst researchers have migrated towards discussing

karst aquifers based on the triple permeability model. This model evaluates a karst system based

on conduit flow, fracture flow, and intergranular flow, and the interactions between each.

Fractures are much smaller than conduits (aperture ranging from 50 500 Cpm) and usually occur

from mechanical joints and bedding plane partings (White, 2002). Intergranular flow occurs on

even smaller scales, and is often referred to as matrix flow. For the purpose of this study, we will









use the term matrix flow to encompass both fracture and intergranular flow. This is because it is

often difficult to distinguish between fracture and intergranular flow and because flow in these

regions is generally laminar.

Mixing between conduit and matrix water is controlled by various factors including

matrix permeability and matrix transmissivity. These factors are often related to the geologic

background of the study area. To date, much of the existing understanding of fluid flow in karst

aquifers is based on work in extensively cemented and recrystallized Paleozoic and Mesozoic

carbonates (White, 1999). In these regions, termed telegenetic karst, the karst is developed on

and within ancient rocks that are exposed after the porosity reduction of burial diagenesis

(Vacher and Mylroie, 2002).

Understanding of fluid flow is less advanced for aquifers with high intergranular

porosity, termed eogenetic karst (Vacher and Mylroie, 2002). In eogenetic karst regions, the land

surface is evolving on, and the pore system developing in, rocks undergoing meteoric diagenesis

(Vacher and Mylroie, 2002). Eogenetic regions, like the karst limestones of the unconfined

Floridan Aquifer, have not been deeply buried and thus have significantly higher matrix

permeability than telegenetic regions (Florea and Vacher, 2005).

Understanding transmissivity in karst aquifers is important for better assessment of the

potential for flow of contaminants through the aquifer. Determining transmissivity in karst

aquifers is difficult due to heterogeneity of hydrologic properties and the high flow rates

necessary to perturb the water table. Multiple studies have concluded that transmissivity

measurements in highly heterogeneous aquifers appear scale dependent. This is because water is

likely to travel through preferential flow paths over long distances. Bradbury and Muldoon

(1990) worked in glacial outwash sediment and observed three to Hyve times greater









transmissivity in regional pumping tests as compared to smaller scale slug tests. Rovey and

Cherkauer (1995) used slug, pressure, and pumping tests, along with two numerical models, to

determine the dependency on scale of hydraulic conductivity measurements in units of a

carbonate aquifer. Martin et al. (2006) used results from passive monitoring of water level

fluctuations to calculate transmissivities at least two orders of magnitude different between wells

at various locations away from a known conduit in the Floridan aquifer.

In an unconfined, eogenetic karst aquifer, the aquifer will be recharged by both a diffuse

component and possibly an allogenic, focused component. Allogenic recharge is defined by

White (2002) as surface water inj ected into the aquifer at the swallets of sinking streams. Diffuse

recharge was defined by White (2002) as rainfall directly onto the karst surface and from there

entering the aquifer as infiltration through the soil and matrix permeability of the underlying

carbonate bedrock. A crucial factor that influences the amount of diffuse recharge that enters the

aquifer is evapotranspiration, a collective term for all the processes by which water in the liquid

or solid phase at or near the earth's land surfaces becomes atmospheric water vapor (Dingman,

2002). In an unconfined aquifer with no runoff, all of the precipitation that is not lost to

evapotranspiration will recharge the aquifer system. If the unsaturated zone of the unconfined

aquifer is composed of sands with high hydraulic conductivity, then diffuse recharge will occur

rapidly. This increases potential to quickly move contaminants into the aquifer. In carbonate

aquifers, diffuse recharge should be considered an important source of water that can potentially

have a maj or influence on the karstifieation (dissolution) of carbonate bedrock. Methods that

have been used to calculate the diffuse recharge component in unconfined aquifers include

analyses of basin water budgets and stream flow (Grubbs, 1998), using fluctuations in the water

table level with a known specific yield (Healy and Cook, 2002), direct quantifieation from










lysimeters (Healy and Cook, 2002), and numerical modeling of flow systems that calculate

various inputs and outputs (Bush and Johnston, 1988).

The Santa Fe River Sink/Rise system lies within the unconfined Floridan aquifer and

offers a unique area to investigate transmissivity and diffuse recharge. In this system, the Santa

Fe River is captured by a sinkhole known as the River Sink, and then continues to flow

underground for about seven kilometers. The river re-emerges at the River Rise and continues to

flow across the landscape. In previous research, passive monitoring of storm pulse events has

yielded results about the hydrologic properties of the system (Martin et al, 2006). These studies

have been complemented with studies of the hydrogeochemistry of the system (Martin and Dean,

2001; Screaton et al, 2004).

This study further investigated the transmissivity of this system by using new storm event

data and by applying a numerical modeling method. This study will investigate how the diffuse

recharge component affects transmissivity estimates made using passive monitoring methods.

Passive monitoring can be extremely valuable in karst aquifer regions in order to assess the

hydrologic behavior of an aquifer on a larger scale than is capable with typical laboratory and

well tests. However, using passive monitoring to determine transmissivity is difficult in regions

where there are multiple sources of water influencing the shape of a flood hydrograph.

Understanding the effect that recharge has on the calculation of transmissivity will shed some

insight on whether applying a passive monitoring method is worth the effort in comparison to

smaller scale tests. Also, quantifying the diffuse recharge, along with the area of the basin

contributing to discharge, will allow for comparison of total volumes input to the system from

the allogenic, focused recharge and total volumes input to the system from the diffuse recharge.










These volumes have potential implications for karstification in the region. The two main goals of

this research are as follows:

Goal 1: Determine the influence of the diffuse recharge component on storm event

hydrographs and evaluate the effect of diffuse recharge on the calculation of transmissivity

by passive monitoring.

Goal 2: Evaluate the total volume of conduit allogenic input and the total volume of diffuse

recharge input to the unconfined aquifer system.









CHAPTER 2
BACKGROUND

Study Area

The Santa Fe River, a tributary to the Suwannee River, is located in north-central Florida

(Figure 2-1). The Santa Fe River basin covers an area of roughly 3500 km2 (Hunn and Slack,

1983). From its origin in Lake Santa Fe, the Santa Fe River flows west for approximately 50 km

(Hisert, 1994). At this point the river flows into a 36-m deep sinkhole known as the River Sink

(Martin and Dean, 2001). The river continues to flow underground for approximately 7 km and

reemerges at a first magnitude spring called the River Rise. Between these two locations, the

river reappears at various intermediate karst windows, most notably at Sweetwater Lake (Figure

2-2).

The Santa Fe River Basin flows across two maj or physiographic provinces known as the

Northern Highlands and the Gulf Coastal Lowlands. The Northern Highlands generally consist

of gently sloping plateaus in the interior regions and marginal slopes that are well drained by

dendritic streams (Grubbs, 1998). The Gulf Coastal Lowlands have a noticeable lack of surface

streams and generally consist of terraces and ancient shorelines that slope gently toward the coast

(Grubbs, 1998). The Northern Highlands have elevations that are in excess of 30 meters above

sea level (masl), where the Gulf Coastal Lowlands are at elevations of less than 15 masl.

Dividing the two provinces is the marginal zone. Within the marginal zone is an escarpment

known as the Cody Scarp, which in this region represents the erosional edge of the Miocene

Hawthorn Group. Most streams that cross the Cody Scarp either sink completely into the

subsurface or sink into the subsurface and re-emerge (Hunn and Slack, 1983). The Cody Scarp

marks the boundary between the confined and unconfined Floridan aquifer.









The Santa Fe Sink/Rise system is located within the boundaries of O'Leno State Park and

River Rise State Preserve, which encompasses nearly 6,000 acres of state land. The closest

significant city to O'Leno State Park is Lake City (~ 24 miles away), which has an average

annual temperature of 20.5oC (Hunn and Slack, 1983). Seasonal temperatures range from 4 to

10.C during the winter months, and from 25 to 35oC in the summer months (Grubbs, 1998).

Average annual precipitation typically ranges from 130 to 150 cm/yr, with nearly half of the

rainfall occurring between June and September (Grubbs, 1998). Average annual

evapotranspiration ranges from 90 to 105 cm/yr (Bush and Johnston, 1988).

Geologic Background

Three principle hydrogeologic units occur in north central Florida: the surfieial aquifer

system, the intermediate confining unit, and the Floridan aquifer system (Table 2-1). The

surfieial aquifer occurs throughout the Northern Highlands province, and more locally in the

Gulf Coastal Lowlands (Grubbs, 1998). Where present, the surficial aquifer system is contiguous

with the land surface and consists primarily of unconsolidated sediments (Grubbs, 1998). In the

Santa Fe River Basin, surficial sediments are composed of white to gray Eine to coarse sand, and

are Pleistocene and Holocene age (Hunn and Slack, 1983). The thickness of the surfieial aquifer

system in the Santa Fe Basin is approximately 5 m, and the water table is reached approximately

3 m below the surface (Hunn and Slack, 1983).

Beneath the surfieial aquifer lies the intermediate confining unit known as the Hawthorn

Group. This layer is composed primarily of siliciclastic rocks and is Miocene in age (Hunn and

Slack, 1983). Regionally, this unit acts as a confining unit that restricts the exchange of water

between the overlying surficial and underlying Floridan aquifer systems (Grubbs, 1998).

The Floridan aquifer system underlies an area of about 260,000 km2 including all of

Florida, southeast Georgia, and small parts of Alabama and South Carolina (Bush and Johnston,









1988). The Floridan aquifer is the principal source of water for municipal and industrial use in

the Santa Fe Basin (Hunn and Slack, 1983). The aquifer system is divided into the upper and

lower aquifer by a layer of less permeable carbonate rock in the lower Avon Park formation

(Bush and Johnston, 1988). The Upper Floridan aquifer is extremely permeable and capable of

transmitting large volumes of water (Grubbs, 1998). From oldest to youngest, the Upper Floridan

aquifer is divided into the Avon Park formation, Ocala Limestone, and Suwannee Limestone.

These layers range from Eocene to Oligocene in age, and range from 90 to 240 m in thickness

(Hunn and Slack, 1983). In the unconfined portion of the Santa Fe River Basin, the Ocala

limestone is the uppermost unit (Hunn and Slack, 1983). The aquifer thickness is smallest in the

region near the Santa Fe River at about 300 ft (~100 m), but extends to almost 800 ft (~240 m)

just north and south of this reach (Hunn and Slack 1983). Specific yield of the unconfined

Floridan aquifer has been estimated from porosity measurements and ranges between .10 and .45

(Palmer, 2002).

Previous Investigations in the Santa Fe River Basin

One of the earliest comprehensive studies in the Santa Fe River Basin was performed by

Hunn and Slack (1983), which involved mapping the hydrogeologic units, identifying how water

moves from one unit to another and to streamflow, and determining aquifer parameters such as

transmissivity and recharge rates. Mapping of the potentiometric surface allowed Hunn and

Slack (1983) to infer that groundwater flow is towards the Santa Fe River. Hunn and Slack

(1983) also concluded that annual fluctuations of water level are due largely to variations in

rainfall amounts. Aquifer tests in the western part of the Santa Fe River Basin, where the

Floridan aquifer is unconfined, produced transmissivity values ranging from 3,000 m2/day to

50,000 m2/day; Signifieantly lower transmissivity was observed in the eastern confined region of









the Floridan aquifer, where transmissivity ranged from 1,900 m2/day to 3,300 m2/day (Hunn and

Slack, 1983).

Using the gaseous tracer SF6, Hisert (1994) attempted to make connections between the

surface water locations and the karst windows of the Santa Fe River Sink-Rise system. A link

was verified between the River Sink, seven intermediate karst windows, and Sweetwater Lake,

and subsequently between Sweetwater Lake and the River Rise. No direct connection was made

from the River Sink to the River Rise. Tracer travel times suggested rapid flow through the

system; however precise velocity determinations were not possible because of the method of

inj section. Hisert' s (1994) work suggested the existence of large conduits that carry the maj ority

of flow through the system.

In 1995, cave divers began exploring the conduits in the Santa Fe River region. Since

then, a single conduit has been mapped that connects Sweetwater Lake to the River Rise, along

with multiple conduits that connect the River Sink to various downstream karst windows (Figure

2-2). A large conduit that feeds from the east of O'Leno State Park has also been mapped. As of

summer 2006, no direct physical connection has been made between the Rise and Sink; however,

over 7 km of conduits have been mapped between the two locations (M. Poucher, unpublished

data). Divers estimate that conduits can reach widths of 24 m.

Martin and Dean (1999) used temperature as a natural tracer of water through the

system.Vrelocities increased with increasing river stage and rapid velocities of greater than one

kilometer per day indicated conduit flow. Martin and Dean (2001) also investigated the exchange

of water between conduits and matrix in the Santa Re River basin. Water samples were collected

from three locations along the river (River Sink, Sweetwater Lake, River Rise) at varying flow

conditions. During low-flow conditions, different compositions of water at the River Sink from









those at Sweetwater Lake and River Rise suggest that little water discharging from the River

Rise is derived from the River Sink (Martin and Dean, 2001). During flood conditions, when the

river stage is elevated, water at Sweetwater Lake and River Rise were found to retain

compositions that are similar to the River Sink (Martin and Dean, 2001). Martin and Dean also

took groundwater samples from two observation wells in the area. It was observed that

concentrations of conservative solutes sodium and chlorine were decreasing over a period of five

months following a flood event. This suggests that the dilution was occurring from water that

entered the matrix after the flood event (Martin and Dean, 2001).

Screaton et al. (2004) worked on the Santa Fe River Sink-Rise system looking at the

exchange of water between conduits and matrix. By comparing discharge data into the River

Sink and out of the River Rise, the direction of flow between the conduit system and the aquifer

could be determined. For example, during times of baseflow, the discharge is greater at the River

Rise and groundwater flows from the aquifer matrix to the conduit. At the peak of storm events,

the discharge is greater at the River Sink and the conduit system loses water, probably to the

surrounding aquifer matrix. The idea that the Santa Fe River loses water to the surrounding

matrix is important because it shows the vulnerability of the aquifer system water quality to a

flood event (Screaton et. al, 2004). Screaton et al. (2004) also used arrival times of temperature

signals to estimate average velocities in the Santa Fe River Sink-Rise system. Using temperature

as a tracking signal allows for tracing an individual water packet entering the system after a

storm event. Using the discharge data and the velocity determined from the temperature signals,

the average cross sectional area of the conduit system was calculated to be ~3 80 m2. This also

allows for the calculation of conduit diameter, which would be 22 m if a single conduit exists.

This calculation is consistent with observations made by cave divers.









Screaton et al. (2004) were also able to make some first approximation calculations of the

dissolution resulting from a storm pulse passing through the Santa Fe system. A small storm

event from late 2001 was used from which the amount of water lost from the conduit to the

matrix was calculated. Chemistry data from the River Sink in 1998 was used to estimate the

degree of undersaturation for a river stage similar to that which occurred during the storm event.

The amount of rock needed to dissolve in order to return the system to equilibrium was

calculated to be 0. 13 moles of calcite per liter of river water, or a total of 2.86 x 10s moles of

calcite. This amounts to a total rock volume of 1.10 x 104 m3 Or a denudation of 4.60 x 10-4 m in

the region surrounding the conduit. This value was an order of magnitude greater than

denudation rates from previous studies in this area (Opdyke et al, 1984), which indicates that

dissolution in this area may have previously been under predicted.

Martin et al. (2006) used an analytical method for a semi infinite aquifer developed by

Pinder et al. (1969) to provide a first approximation of the matrix transmissivity between known

conduits and observation wells in the unconfined Floridan aquifer. Although this analysis did not

account for heterogeneity of the aquifer, or head changes at the wells due to diffuse recharge,

initial estimates of transmissivity were possible. Hydraulic conductivity was determined by

dividing the best fit transmissivity by an assumed aquifer thickness of 275 m. Calculated

transmissivity values ranged between 900 m2/day and 500,000 m2/day and thus hydraulic

conductivity ranged from 3 m/day to 1,818 m/day. These high values are consistent with other

studies of the Floridan aquifer. For example, Palmer (2002) estimates hydraulic conductivity of

the Floridan to range between 86.4 m/day and 864 m/day, which is four orders of magnitude

greater than the hydraulic conductivity of 0.086 m/day reported for recrystallized Paleozoic

aquifers of PA and NY. Bush and Johnston (1988) looked at aquifer tests performed at 1 14









locations throughout the state of Florida, mostly from the Upper Floridan aquifer. The closest

location to O'Leno State Park was in Lake City, where the transmissivity was estimated to be

3,344 m/day. Results from these tests showed that transmissivity values of the Upper Floridan

aquifer are directly related to the thickness and lithology of its upper confining unit. The

confined areas generally had lower transmissivity values than the unconfined areas, with the

highest transmissivity occurring in karstic north-central Florida (Bush and Johnston, 1988).

Another important result from the Martin et al. (2006) study is that transmissivity differed

by two orders of magnitude for wells located at various distances from the conduit. Estimated

transmissivity of wells within 150 m of the conduit were ~ 900 m2/day, while wells at distances

greater than 500 m from the conduit produced transmissivity values of greater than 100,000

m2/day. These results were attributed to a scaling effect. Scale effects result from the likelihood

that large-scale tests will include preferential paths that dominate a larger percentage of

groundwater flow, and have been previously documented in the literature (Rovey and Cherkauer,

1995).

Recharge in the Santa Fe Basin has been investigated by Grubbs (1998). His study

provided an overview of recharge rates to the Upper Floridan aquifer using a simple water

budget analysis. In the lower Santa Fe River basin, runoff is generally negligible because very

little channelized surface drainage is present, the soils are permeable, and the slope of the land

surface is gradual to flat. Water-budget analysis in the unconfined region placed average-annual

recharge rates between 45 to 60 cm/yr (Grubbs, 1998). Other methods including chloride tracing,

hydrometric base flow, and ground-water level changes showed a range of average-annual

recharge between 20 and 80 cm/yr (Grubbs, 1998). Recharge rates of the confined Upper

Floridan aquifer were found to range based on confinement of the aquifer, and water budgets









indicated that recharge is less than 30 cm/yr in areas where there is confinement by an overlying

layer of sediments with low permeability (Grubbs, 1998).

Grubbs (1998) water budget calculations used evapotranspiration values for Florida

calculated by Bush and Johnston (1988). Evapotranspiration was calculated with a method that

uses biotemperature and precipitation. Biotemperature is defined as the annual sum of hourly

temperatures between 32oF and 86oF divided by the numbers of hours in the year.

Biotemperature was linearly related to potential evapotranspiration in order to estimate a quantity

of water that would be given up to atmosphere within a zonal climate. Long-term average

evapotranspiration from the region was estimated to be between 90 and 105 cm/yr (Bush and

Johnston, 1988).

Diffuse Recharge in Other Study Areas

In addition to the work previously completed in the Santa Fe River basin, much of the

methodology behind studying diffuse recharge has been commonly applied to karst and non-

karst locations around the world. A debate often ensues about which method is best suited for

individual study areas. For example, in 2005 the USGS published a scientific investigations

report from east-central Pennsylvania which compares multiple methods for estimating ground-

water recharge in the same watershed area. Recharge was estimated on a monthly and annual

basis using three methods; (1) unsaturated-zone drainage collected in gravity lysimeters, (2)

daily water balance, and (3) water-table fluctuations in wells (Risser et. al, 2005). The following

briefly summarizes the different methods used in their study and their respective results:

1) Gravity lysimeters directly measure the vertical flow of water through a large section
of unsaturated zone at a depth below the root system. Drainage from lysimeters
represents water that has not yet reached the water table, however, because water has
passed beneath the root zone, it is assumed to represent the water that will reach the
water table (Risser et al, 2005). The mean-annual recharge from seven lysimeters in
the Pennsylvania study area was 30.9 cm.










2) The water balance method uses the general water budget equation: R = P (ET + RO
+ AS) where R is recharge, P is precipitation, ET is evapotranspiration, RO is runoff,
and AS is change in storage (Risser et al, 2005). This method is easily applied to
locations where precipitation data is available; however, the accuracy of the estimated
recharge relies on the accuracy of other budget terms. The mean annual recharge
estimated from the budget method over the same seven year period was 31.2 cm.

3) Water table fluctuation is based on the rise of water levels in a well multiplied by the
specific yield of the aquifer (Rasmussen and Andreasen, 1959) to estimate recharge.
This method assumes that water-level rise is solely caused by recharge arriving at the
water table and that specific yield remains constant. It is therefore necessary to know
the correct value of specific yield for the study area. Specific yield in the
Pennsylvania watershed was determined to be 0.013 using the equation Sy = S / Ah
where Sy is specific yield, S is streamflow volume during a recession period
consisting of only ground-water discharge over the watershed area, and Ah is the
average decline in water-table level during a recession period. The mean annual
recharge estimated from the fluctuation method over the seven year period was 25.1
cm.

A study by Rasmussen and Andreasen (1959) in the Beaverdam Creek Basin, Maryland

provides another example of the application of the water table fluctuation method for estimating

recharge. This is a relevant study because the unconfined aquifer study site has similar

characteristics to the unconfined aquifer region in the Santa Fe River Basin. In the Beaverdam

Creek Basin, surficial sands and silt overlie Tertiary sand aquifers, and the water table is

generally within a few meters of the surface (Rasmussen and Andreasen, 1959). Water levels

were measured in observation wells on a weekly basis and precipitation was measured at various

sites in the basin. Specific yield was estimated by the water budget method to be 0. 11

(Rasmussen and Andreasen, 1959). The rise in water level was estimated as the difference

between the peak water level during the storm event and the extrapolated antecedent water level

prior to the rise. Results produced a two year record of monthly groundwater recharge from April

1950 to March 1952.

Jones et al. (2000) provide an example of how chemistry can also be used to quantify

recharge. Jones et al. (2000) measured 68"O isotopes of water in the Pleistocene limestone









aquifer of Barbados. Oxygen isotopes are useful in this region because the islands' small size,

low relief, and tropical climate allow for constraint of the factors that influence spatial and

seasonal variations of rainwater and groundwater isotopic compositions (Jones et al, 2000). In

this study, groundwater was sampled from 29 wells from 1961-1992. A small range of

groundwater 618O (-2.5 %o to -4.5 %o) values were consistently observed, which suggested that

evaporation likely had little influence on 68"O values prior to recharge. This is most likely an

affect of rapid diffuse infiltration through highly permeable limestone. The oxygen isotope

method predicted recharge to be between 15% and 45% of the mean monthly rainfall. These

values were greater than recharge predicted by comparing rainfall and potential

evapotranspiration, which was about 6% of the mean monthly rainfall. This result occurred

because the use of potential evapotranspiration in the recharge estimate overestimates actual

loses to evapotranspiration by assuming that the actual evapotranspiration is limited only by the

amount of rainfall (Jones et al, 2000). Recharge amounts predicted from standard methods such

as the water-balance method were highly variable and proved unsuccessful in this region. This

was attributed to the idea that coastal discharge rates can be affected by groundwater withdrawal

and return flow in adj acent catchments, resulting in artificial redistribution of groundwater

recharge (Jones et al, 2000).

Vacher and Ayers (1980) recognized that increased chloride concentrations on the oceanic

island of Bermuda should be a direct result of evapotranspiration. This happens because there are

no surface streams and thus all rainfall is transmitted back to the atmosphere as

evapotranspiration, or recharged and passed to the shoreline via the subsurface.

Evapotranspiration that transports water back to the atmosphere concentrates the rain-derived

chloride in the soil. Comparing the ratio of fresh rainwater chloride concentration versus









concentrated groundwater chloride concentration allows for a concentration factor to be

calculated. If the precipitation has been measured, than the concentration factor can be used to

calculate the amount of evapotranspiration, and thus recharge can be calculated by subtracting

the evapotranspiration from the precipitation. In Vacher and Ayers (1980) study, recharge was

calculated on an annual basis to be 37 cm/yr.

In addition to quantifying the diffuse recharge component in order to better understand

the volumetric input of water into a karst system, many have used the diffuse component to

characterize the behavior of the vadose zone of karst aquifers. For example, Batiot et al. (2003)

collected drip water from Nerj a Cave in Spain, and investigated the use of Total Organic Carbon

(TOC) as a tracer of diffuse infiltration. In this region, the highest concentrations of TOC in drip

water occurred during the summer months. This was likely caused by rainwater entering after the

dry season and transporting carbon from the soil when it has the highest content of organic

matter (Batiot et. al, 2003). Mean residence time of water in the karstified unsaturated zone over

the cave was determined by comparing TOC of rainfall and drip water. Residence time varied

significantly (2 to 8 months) which was likely dependent on the amount of water stored in the

unsaturated zone over the cave prior to rainfall events and also the highly variable thickness of

the unsaturated zone (4 to 90 m) (Batiot et. al, 2003).

Maher and DePaolo (2003) used Sr isotopes to determine vadose zone infiltration rates at

a site in Hanford, Washington. Sr can be used as a tracer of diffuse infiltration because

groundwater entering a specific rock or soil environment typically contains dissolved Sr in low

concentration (~ 100 Cpg/L total Sr) with an isotopic ratio (Srs7/Sr86) that is generally different

from mineral phases in soil matrices that are undergoing weathering (Maher and DePaolo, 2003).

Weathering results in dissolution of primary minerals and introduces Sr to the pore fluid, thus









shifting the Sr ratio towards that of the dissolving minerals. Infiltration rates are of concern in the

Hanford region because substantial quantities of radioactive materials have been disposed into

the surface soils and vadose zone sediments, with the expectation that ion exchange processes

would retain most of the radionuclides and prevent them from ever reaching groundwater (Maher

and DePaolo, 2003). Infiltration rates were determined by matching isotopic profiles from two

models (steady state and non steady state finite difference models) to isotopic profiles from pore

waters and bulk sediment analysis from a 72 m borehole. Model results suggest that the

infiltration flux ranges from 0.4 to 1 cm/yr, which would result in a transit time from water on

the surface to the water table between 600 and 1600 years (Maher and DePaolo, 2003). These

transit times are significantly longer than can be expected in the vadose zone of the Santa Fe

River basin, where surficial sediment is composed of fine to coarse sands and the water table is

close to the surface.

Aquifer Parameters in Other Study Areas

The most commonly used technique for determining hydraulic parameters is an aquifer

test. In an aquifer test, a well is pumped and the rate of decline of the water level in nearby

observation wells is noted (Fetter, 1994). Pumping tests typically last until the water level

reaches a state of equilibrium; that is there is no further drawdown with time (Fetter, 1994). The

result of an aquifer test is a set of data with the drawdown given at various times after the start of

pumping (Fetter, 1994). This data can be graphically matched to a set of type curves originally

developed by Theis. Curve matching allows for a graphical means to solve a set of equations for

the various aquifer parameters (Fetter, 1994). Jones (1999) demonstrates an example of an

aquifer test in a carbonate aquifer. In his test, a water supply well in West Virginia was pumped

for five hours and significant drawdown (~2.5 m) was observed at a monitoring well 315 m

away. Transmissivity and storage coefficients were calculated using the Jacob time-drawdown









method (a simplification of the Theis solution), and varied by an order of magnitude in different

locations of the carbonate aquifer. These results were attributed to the varying locations of

fractures within the aquifer system relative to the pumping location.

An aquifer test would be difficult in our study area of the Santa Fe River Sink-Rise

system, primarily because of the difficulty to produce a large enough drawdown from pumping.

This results from both the high transmissivity of the Floridan aquifer, and also the lack of high

pumping capacity. Pumping rates during aquifer tests in karst aquifers are commonly greater

than 100 gal/min. We can estimate the pumping capacity needed to observe drawdown in an

observation well in our study area by using the modified Theis solution for nonequilibrium radial

flow in an unconfined aquifer:

T = (Q / 4II (ho h))*"W(ua, ub, r

T = Transmissivity (m2/day)
Q = pumping rate (m3/day)
ho h = drawdown (m)
W(ua, ub, F) = well function (unitless)

In order to determine the well function for the aquifer, the radial distance to the pumping

well must be known. The closest observation wells in this study area are approximately 50 m

apart. The hydraulic conductivity and initial saturated aquifer thickness are also needed to

determine the well function. A value of 5 m/day was applied as an low end estimate of hydraulic

conductivity in this area, and a value of 100 m was used for the aquifer thickness. The calculated

well function with these values is 2. With a transmissivity of 500 m2/day, a pumping rate of 314

m3/day or 57.61 gal/min is needed in order to observe a small drawdown of .1 m (~.33 ft). In

order to observe a larger drawdown of 0.3 m (~ 1 ft), a pumping rate of 942 m3/day or 172.8

gal/min would be needed. Our current pumping ability maxes out at a rate of 9 gal/min and thus

is much lower than the rates needed to produce significant drawdown. Because the wells are









constructed from PVC that has only a 2-inch diameter, we are limited in the size of the pump that

can be used in the well. If the hydraulic conductivity is actually higher, then a greater pumping

rate would be needed to observe drawdown.

An alternative field method to pumping tests is slug tests. In a slug test, a known amount

of water is quickly drawn from or added to a monitoring well, and the rate at which the water

levels fall or rise in the same well is measured (Fetter, 1994). Commonly, the water column in

the well casing is induced to rise by rapidly lowering a solid slug into the well below the static

water level. This is equivalent to adding a volume of water that is equal to the volume of the slug

(Fetter, 1994). The recovery data are then analyzed using an appropriate method, such as the

Hyorsley Slug test method. These tests allow for determining the hydraulic conductivity in the

vicinity of a monitoring well. The difficulty with performing a slug test in the highly

transmissive unconfined Floridan Aquifer is that results will most likely represent localized

parameters, rather than representing parameters on a regional scale. Results of recent slug tests

near the Santa Fe River Sink/Rise system yielded hydraulic conductivities that ranged from 8 x

10-1 m/day to 8 x 10-3 m/day (Langston, personal communication).

Powers and Shevenell (2000) describe a passive monitoring technique to calculate

transmissivity in an unconfined karst aquifer based on analysis of well hydrographs. They

observed that the recession limb of a well hydrograph responded similarly to those of karst

spring hydrographs. Recession limbs in karst aquifers commonly produce two or more

"segments" with different slope values that represent different portions of the ground water

system (Powers and Shevenell, 2000). The first and steepest slope segment was inferred to

represent the drainage of conduit features, while the following segments were inferred to

represent flow through fractures and matrix (Powers and Shevenell, 2000). The solution method









used by Powers and Shevenell (2000) adapts the Bernoulli equation for flow through a conduit to

a spring to determine the slope of these recession segments, which can then be translated to

theoretical flow values that correspond to the various types of flow (conduit, fracture, matrix).

Using these theoretical flows along with a valid estimation of the specific yield and distance

from recharge area, transmissivity can be calculated. Powers and Shevenell (2000) applied their

solution to calculate transmissivity from 49 recession curves from wells within three different

Midwestern field sites. They went on to compare their results to slug and pumping tests results,

and found that results matched within an order of magnitude between the different methods.

Differences in transmissivity values may have occurred based on the difference in the scale of

measurement between the aquifer tests and the passive monitoring method (Powers and

Shevenell, 2000). Results from this study show that determining aquifer parameters passively

provide an alternative to the commonly used aquifer and slug methods, and also provide

information about aquifer parameters during natural flow conditions.

Pinder et al. (1969) describe another passive monitoring technique which allows for

calculation of aquifer diffusivity (ratio of transmissivity to the coefficient of storage) based on

responses to fluctuations in river stage (e.g. storm pulse). In the Pinder et al. (1969) method, a set

of type curves are generated for various observation wells around the river location. Each set of

curves represents the computed change in head in the aquifer due a change in river stage when a

selected diffusivity is assumed (Pinder et. al, 1969). The best fit diffusivity of the aquifer is

determined by choosing the curve that best matches the response in the observation well. The

analytical solution can be applied to a flood stage hydrograph of any shape. In their study, the

analytical solution was applied to an aquifer at Musquodoboit Harbour, Nova Scotia. Values

calculated with the solution were on the same order of magnitude with those obtained from










pumping tests. Values were different for different observation well locations, which was

attributed to the heterogeneity of the aquifer between the river and the observation well (Pinder

et al, 1969).

Another alternative method using borehole lithologies to estimate aquifer transmissivity

was demonstrated in the unconfined region of the Columbia aquifer (Andres and Klingbeil,

2006). Boreholes were drilled to penetrate the entire thickness of the unconfined aquifer, and

then lithologies were classified into categories (sand, sand and gravel, silt, etc). Lithologies were

assigned a typical hydraulic conductivity value based on previous studies of sediment in the

region. In areas where the borehole did not reach the base of the aquifer, a hydraulic conductivity

value was assigned that was representative of the local average of all materials. Transmissivity

was then calculated as the sum of the products of material thicknesses (b) and hydraulic

conductivity (k). This method would be difficult and expensive to apply in the O'Leno study area

due to the costs of drilling a hole through the entire unconfined Floridan aquifer. Data from one

borehole exists from the upper portion of the unconfined aquifer in the study area; however, not

enough of the core was recovered during the drilling to make an accurate estimate of hydraulic

conductivity. Another concern with this method would be that using typical values for hydraulic

conductivity will underestimate and overestimate values in localized areas (Andres and

Klingbeil, 2006), and would likely not represent the scale effect that is common in karst aquifers

at the basin scale.










































Figure 2-1. Location of the Santa Fe River.















































































Iniil


Figure 2-2. Map of O'Leno State Park with locations of surface waters, mapped conduits, and
observation wells


t-' '5; 'j~~
r
Two ~ole


rammmmn naewm
il+lL que Ikae s


iu~
rru~
r~ r_


P)AWAS














Table 2-1. Geologic and Hydrogeologic Units of the Santa Fe River Basin based on Hunn and
Slack (1983), and Hisert (1994).
Seris .r:i- ... Hydlrogeologic Lir...l: 1:;,:Descrtiprin Thi~citess
UJnit Untit(m
Sinkhole fill, flui
Holcrcene Uadriffe~rentiate~d
Plcsocerne siedurnents
anfical sand
Rrddish-utite snds
Plelstocene Alachuar
WeIaontedat withdy csanch~clays,
to.,::.,< Fonnation
--lo'''-Pl--and pholsphisrte pebbles a3
Middle to s. .lfnir. Unn .; law.l.: clavr3j siand-
Hawr~horn sandly L. 1. Ea rlrvnl
Lower
C-tou~tp amount ot Fullea Earth~
Muiocene
andt caroonaste

Suwatnnee moderarely rrndurated.
Limnestlone I- 1 -.porlous. :ossil-tich
Floriancs carrlc~atenntg.
.-. asIf .. "Ieri :permieable v:1n e to
Oc ala Ls. I L blaclaistic 5-0
Eow Av~on Park; Ls. limestone
OlJdernar Ls. Doltmnutice hmtestone &P
Io ave Finonr~ dnolonne
Lea Cedatr]1 .-'.inte Limetstone rome 30-
Pal~o~cene Fonnation etvapofrite antd clsrq









CHAPTER 3
METHOD S

Data Collection

Data collected during this study includes daily precipitation, river stage, and ground

water levels. Daily precipitation was recorded at the O'leno State Park automated gauging station

and was accessed through the Suwannee River Water Management district database. The

automated station records precipitation on an hourly basis and total precipitation for a 24 hour

period can be viewed in real time for 30 days; archived daily precipitation data is available in the

database. The current record of precipitation downloaded for this and previous studies dates back

to July 2000. River stage measurements were recorded daily by O'leno State Park personnel

from one of two staff gauges located approximately 0.5 km upstream of the River Sink at the

O'Leno State Park suspension bridge. The stage gauges at this location are divided into a lower

and upper gauge, with a gap of 2. 10 m where no readings are available. The top of the lower

gauge stops at 12.41 meters above sea level (masl), while the upper gauge is available between

14.50 mas1 and 16.55 masl. For days when the gauge was not read, or during time periods when

the water level was between the two gauges, data records were filled in using a long term

correlation between stage at the gauge location and water level elevation recorded by a data

logger at the River Sink (described below).

Data were collected from twelve monitoring wells placed at various distances away from

the conduit system. Eight of these wells were installed to be at approximate conduit depth, 30 m

below ground surface (bgs) (Table 3-1). Six of these eight wells were installed in January 2003,

while Well 5 was installed in March 2003 and Well 8 was installed in May 2004. The other four

wells were installed in January 2006 and are finished to depths just below the water table at their

respective locations. Each of these wells is constructed of 2-inch PVC piping with a screened









interval of 20 ft (6. 10 m) for the deep wells and 10 ft (3.05 m) for the shallow wells. Wells were

completed above the ground surface and installed with a protective casing to prevent outside

disturbances. The elevations of the top of the well casing were determined from surveying from a

known benchmark. Deep wells installed in 2003 were professionally surveyed, while shallow

wells installed in 2006 were surveyed by members of this proj ect from the adj acent deep well.

Ground surface elevations were determined from measuring from the top of the well casing to

the ground using the survey rod. In this area, bedrock was assumed to be the first consolidated,

carbonate rich material encountered during drilling (Table 3-1). All well locations with the

exception of Well 1 have less than 6 m of unconsolidated sands that lay above the carbonate

rock. Estimated depth to bedrock may vary between shallow and deep wells (Table 3-1) at the

same location due to wells being drilled by different drill teams with different descriptions of

rock encountered during drilling. Deeper depths most likely represent the consolidated limestone

rock, as opposed to the first appearance of unconsolidated pieces of limestone; therefore the

deeper depths will be used at locations where there are discrepancies.

Water Levels

Water levels were recorded at surface water locations and observation wells using three

different types of automated water level recorders: In-Situ Minitroll Loggers (accuracy of +/- .02

m), Van Essen Divers (accuracy of +/- .005 m), or Van Essen shallow/deep CTD Divers (10 m

divers accuracy of +/- .01 m, 30 m divers accuracy of +/- .03 m). CTD divers with 10 m range

were installed in shallow wells and surface water locations, while CTD divers with 30 m range

were installed in deep wells. Data loggers placed at the River Sink, River Rise, and Sweetwater

Lake were installed in screened 2-inch PVC pipes that extend below the surface water and are

accessible from the bank. Data loggers in the wells were attached to the well cap via coated

stainless steel wire and allowed to hang in the well. All data loggers were set to record data at









10-minute intervals. Since October 2005, data were downloaded on a four to five week basis to

minimize instrument drift. Before then, time between data downloads varied with up to eight

weeks between downloads. The long term records of water levels vary for the locations within

the study area (Table 3-2). This is mainly a result of the addition of new equipment over the

course of the data collection, moving equipment to new wells after they were constructed, and

the malfunction of existing equipment over the course of data collection. Water levels at the

River Rise have been measured nearly continuously since August 2001 about 250 m downstream

from the conduit outflow location and have been referenced to gauge values at the River Rise.

For each recording interval, water pressures measured by the data loggers were corrected

using the ambient barometric pressure recorded by a Baro Diver (accuracy of +/- .0045m). Water

pressures were then referenced to the water elevation determined manually at the time the data

were downloaded. Manual water levels were measured using an electronic probe in the wells, or

by reading a staff gauge and/or surveying at the surface water locations. Instrument error as a

result of drift and/or shifting of the loggers within the PVC casing was calculated for each

sampling period by subtracting the water level calculated from the logger data at the start of the

interval from the manually measured water level at the start of the interval. Errors at the well

locations were consistently less than 0.03 m, while errors at surface water locations were

generally less than 0.03 5 m. Errors due to shifting of the loggers are generally expected to be

lower at the well locations because data loggers have less ability to move around within the well

casing. Surveying error is estimated to be 0.015 m at the surface water locations. This is

determined by surveying from the known benchmark to the water level, then completing the

survey loop back from the water level to the benchmark. Water levels were only surveyed during









times when the staff gauge was unreadable or the surface water was below the known

benchmark.

Sink and Rise Discharge

Discharge rates into the River Sink were calculated by converting water levels to

discharge using one of two rating curves. The primary rating curve (Figure 3-1) is based on a

rating table created by the Suwannee River Water Management District for the station located on

the Santa Fe River at O'Leno State Park (Rating No. 3 for Station Number 02321898). This

curve was used during all periods of record where the water level from the gauge at the O'Leno

State Park bridge was measured. The second rating curve (Figure 3-2) was used as a backup for

those periods where the O'Leno gauge data was not available, such as days when the gauge was

not read. The second curve was created by correlating the water levels measured at the O'Leno

State Park gauge with the water levels measured by the logger ~1000 m downstream at the River

Sink for records since 2002. The correlation was done by plotting 142 data points of O'leno

Stage versus sink logger water level and then fitting a 2nd order polynomial trend line through the

data (Figure 3-2). The trend line produces a R2 Value Of .9994. It is important to note that the

maj ority of the data points used for the correlation are between logger water levels of 10 to 12

meters, and therefore this is the most accurate portion of the curve. Above 12 meters there are

only a handful of data points, and thus the curve is limited in its predictive capability. This is not

a problem for this study, as discharges were mainly determined from the primary rating curve,

with the second curve serving only to fill in a few scattered missing points over a few years.

Discharge rates out of the River Rise were calculated based on the rating curve produced

by Screaton et al. (2004) (Figure 3-3). This rating curve was calculated from measured water

level elevations and unpublished discharge data provided by the Suwannee River Water

Management district. All of the data used to create this rating curve were at or below water levels









of 12.5 masl, with the maj ority of the points between 9.5 m and 1 1.5 m. This happens primarily

because it is difficult to make discharge measurements during high water levels, which usually

coincide with large storm events. The result is that discharge will be less accurate at high water

levels where there are less data. It is also important to note that the discharge curves may change

over time if the morphology of the region around the River Sink and Rise changes. For example,

discharge may decrease if the river channel were to widen because there would be a larger area

for flow.

Over the course of time, various reference points have been used to determine the water

level at the River Rise. This occurred as a result of reference points becoming altered from natural

events, or unusable from equipment malfunction. For example, starting in mid 2003 the stage

gauge at the River Rise became increasingly warped, making it extremely difficult to obtain an

accurate reading. After the collapse of the gauge in mid 2005, a new gauge was installed by the

Suwannee River Water Management District in February 2006. Errors like the slow warping of

the gauge at the River Rise are difficult to assess because they are not always noticeable right

away and can compound over time. For this study period, the warped gauge may have

contributed to error in the observed water level at the River Rise between July 2003 and June

2005. Because there is no clear correction for the slow warpage of the gauge, discharge values

during this time will not be used for calculations.

Water levels observed at the River Rise after the installation of the new gauge have shown

some inconsistencies when compared to the wells in the surrounding areas. During late 2006,

water levels were reaching the end of a long recession, and by early 2007 the water table has

essentially flatlined at Wells 3, 4 and 5. Comparing the water level of the wells and the Rise

during this time indicates that the Rise is slightly higher than the wells. Hydraulically this would









indicate that water is flowing from the rise towards the wells during very low flow. However,

this is likely not the case as water was still flowing out of the River Rise during this time. The

possible cause of this error was thought to be the survey of the newly installed gauge. To

investigate this possible error, water level was recently surveyed from the most accessible well

location (well 6) to the river rise, and compared to the reading from the new gauge. Results of

the survey indicated that the surveyed water level was 0. 11 ft less than what was read from the

stage gauge at the same time. With a difference of 0. 11 ft, the water level of the rise would be

even or less than the surrounding wells. This would be consistent with flow towards the rise

during very low conditions.

During early 2007, the River Sink was experiencing a long recession, and reached water

levels similar to those observed after a long recession during the first half of 2002. Comparing

the manually measured water levels of the Sink and the Rise during both these times indicates

that the River Rise was up to 0.3 ft higher in 2007 than it was in 2002 relative to the River Sink.

If the relationship between the River Sink and River Rise has not changed over time, than this

suggests that the error in the Rise water level could potentially be as great as 0.3 ft. It is

important to note that this error could also be caused in part if there has been a lowering of the

benchmark at the River Sink. This benchmark has not yet been resurveyed. Because an error of

0.3 ft in the Rise water level would be enough to cause a noticeable decrease in Rise discharge,

all calculations made using the discharge values were completed for observed values of

discharge and for hypothetical discharges based on a water level decreased by 0.3 ft from the

measured value. This potential survey error will only affect the discharge and area calculations,

while the rest of the work only relies on changes in head.









Evapotranspiration

Evapotranspiration (ET) was calculated using the Penman-Monteith Model for

determining water lost to the atmosphere from a vegetated surface. This method adapts Penman's

original model for evaporation from a free-water surface by incorporating a canopy conductance

factor (Monteith, 1965):


a R,z + pa c, C,, ea ( (- W,)
ET =
Pw 1, -.[a +7 y (1+ Ca,/ Com, 1


A = slope of relation between saturation vapor pressure and temperature
Rn = net radiation input
pa = density of air
ca = heat capacity of air
Cat = atmospheric conductance
ea* = saturation vapor pressure in air
Wa = relative humidity
pw = density of water
hv = latent heat of vaporization
Ccan = canopy conductance
y = psychrometric constant

Required measured daily inputs for this model include temperature, relative humidity,

average solar radiation, and wind speed. These data were recorded at a weather station located in

the city of Alachua (approximately 15 miles from O'Leno State park), operated by the Florida

Automated Weather Network (FAWN). These data were archived from 1997 and can be

accessed through the FAWN database generator (http://fawn.ifas.ufl .edu). Along with the daily

measured inputs, the most important part of the Penman-Monteith model is the calculation of

canopy conductance. Canopy conductance assumes that a reasonably uniform vegetated surface

can be represented as a single "big leaf' whose total conductance to water vapor is proportional









to the sum of the conductances of millions of little leaves (Dingman, 2002). To calculate canopy

conductance, the following equation is used:

Cean = fs LAI Clear

Leaf Area Index (LAI) is the relative size of the hypothetical "big leaf and is determined

from the total area of leaf surface above a ground area. LAI can range from 1.0 to 6.0 based on

the type of climate and vegetation. Federer et al (1996) provides typical values of LAI from

various environments; a value of 3.0 was assumed for the O'Leno State Park study area based on

the similarity to a savannah/shrub type land cover with an average height of vegetation near eight

meters (Federer et al., 1996). Shelter factor (fs) accounts for the fact that some leaves are

sheltered from the sun and wind and thus transpire at lower rates. Shelter factor ranges from 0.5

to 1.0 and a value of 0.5 was chosen as an estimate for a completely vegetated area (Allen et al,

1989). Leaf conductance (Clear) is determined by the number of stomata per unit area and the

size of the stomatal openings. The most important controlling factor of leaf conductance is the

soil-moisture deficit (Stewart, 1988). The soil-moisture deficit is represented by a non-linear

function in the solution for leaf conductance (Figure 3-4).

Soil moisture deficit used for the Penman-Monteith solution was determined by tracking

the soil moisture storage over time. This first required estimating the maximum amount of soil

moisture possible in this study area. Different types of soil have the ability to retain different

amounts of water. This amount is commonly represented by the term called field capacity, which

is defined as an index of water content that can be held against the force of gravity (Dingman,

2002). Field capacity (volume of water in a soil/volume of soil) is usually determined in the

laboratory, and is represented as a unitless number. Field capacity can range from 0.1 for sands









to 0.3 for clays (Dingman, 2002). The surfieial aquifer in the Santa Fe Sink-Rise study area is

mostly composed of sands, and therefore was represented by a field capacity of 0. 1.

The soil moisture storage is also dependent on the vegetation that pulls water from the

vadose zone. This requires understanding the average depth of the root zone of the fauna in the

study area. Gilman (1991) discusses common tree care, and points out that the maj ority of tree

roots occur in the upper foot of the soil. This is supported by Rindell (1992) who discusses

common misconceptions about tree roots, and notes that most tree roots rarely grow beneath four

feet. Stewart (1988) applied the Penman- Monteith model to Thetford forest in Norfolk, England

and determined that tree roots do not exceed one meter. Vegetation in the O'Leno study area

consists of a mix of oak, pine, and palmettos trees, along with many cypress trees along the river

bank. Vegetation ranges in height, girth, and leaf size, and therefore in this area most likely has a

tree root depth zone that does not exceed one meter on average. Therefore, the maximum soil

moisture storage used in the model (10 cm), was calculated by multiplying the estimated depth of

the root zone (1 m) by the field capacity (0.1). In order for the calculated evapotranspiration to

occur on a given day, there must be enough water available in the soil moisture storage. When

soil-moisture deficit reaches a value of 10 cm there is no longer any water available for

evapotranspiration. The effect of the soil moisture deficit on the evapotranspiration calculation is

governed by the function for a soil moisture deficit of 10 cm (Figure 3-5). Sensitivity analysis of

the 10 cm estimate indicates that overestimating the maximum possible soil moisture deficit

would cause little error to the ET calculation during the maj ority of times. This is most likely

because there is usually plenty of water available for evapotranspiration in this region.

Overestimating the maximum possible soil moisture deficit during very dry times would result in

under predicting the amount of evapotranspiration. This can be as significant as a few cm per










year if the soil moisture remains close to maximum deficit throughout the year, to less than 1 cm

per year if the soil moisture was only near maximum deficit sporadically throughout the year.

This was determined by increasing the maximum possible soil moisture during both a very dry

year and a very wet year and comparing the differences.

Recharge

Water Budget Method

In the unconfined Santa Fe River basin, nearly all of the basin drainage occurs through

the subsurface, and runoff does not generally occur. This happens because little channelized

surface drainage is present, the soils are permeable, and the slope of the land surface is gradual to

flat (Grubbs, 1998). Therefore, in this setting recharge is dependent only on the amount of

precipitation and evapotranspiration that occurs, and the maximum soil-moisture storage

possible. Daily precipitation is added to a daily running total of soil moisture. Recharge is only

possible when the soil moisture storage exceeds its estimated maximum capacity of 10 cm. Once

the soil moisture is full, recharge is calculated by subtracting the amount of calculated

evapotranspiration from the measured precipitation. Recharge was calculated on a daily basis

allowing for quantifieation of recharge over individual storm events (days) and also over the long

term (annually) by summing the daily values.

Water Table Fluctuation Method

Minor water level perturbations observed in the hydrograph record of the shallow wells

were common after small rain events. These small rain events occur sporadically during periods

of extended dryness, and do not significantly alter the shape of the River Rise receding

hydrograph. During these low-flow periods, there is no movement of water from the conduit to

the matrix, and thus the small change in water level should be attributed solely to water arriving

at the water table via diffuse recharge. The water table fluctuation method assumes that the










change in water level is related to the specific yield of the aquifer (Healy and Cook, 2002). It

also assumes that all of the recharge arrives at the water table, with no loss of water due to lateral

flow through the unsaturated zone:

R = AhS,

Ah = Change in head
S, = Specific yield
R = Recharge

When applying the water table fluctuation method, the maximum change in head during an

observed perturbation was calculated by projecting the recession curve path and determining the

maximum distance between the peak increase and proj ected path. If lateral flow through the

aquifer occurs at the same time scale as water movement downward through the vadose zone,

this method will underestimate recharge.

Chloride Concentration Method

In this region, the only source of chloride entering the unconfined Floridan aquifer is

from rainwater. Chloride concentrations measured at shallow well locations are consistently

much higher than typical chloride concentrations found in rainwater. These concentrated values

are a direct result of evapotranspiration occurring from the top of the water table. The amount of

evapotranspiration is quantifiable by first calculating a concentration factor, which is the ratio of

chloride concentration in rainwater to the chloride concentrations measured at each of the

shallow wells:

Concentration factor = [Cl- rainfall]
[Cl- well]

The concentration factor can be then used to determine how much of the precipitation has

actually been evaporated. For example, a concentration factor of 0.25 indicates that the

evapotranspiration process has caused a 4 times increase in chloride concentration, and thus 0.75









of the rainwater must have evapotranspired. Recharge can then be calculated by subtracting the

calculated evapotranspiration from the measured precipitation amount.

Matrix Transmissivity

Analytical Method without Recharge

The analytical method developed by Pinder et al (1969) was used to make a first

approximation of matrix transmissivity in the Santa Fe River Basin. This method does not

include recharge but was applied so that a comparison to results produced by Martin et al. (2006)

was possible. This method calculates theoretical changes in head at an observation well a

distance away from a known conduit. The input signal, which is the water level at the conduit

boundary, is broken into increments and then the incremental change in head is calculated. The

total change in head is calculated by summing the increments:






Ahm= change in head of well per time step (m)
AHin= change in head of conduit per time step (m)
x = distance from well to conduit (m)
v = diffusivity (T/S,) (m2/d) (S, assumed to be 0.2)
t = time step (days)

Distances from the conduit to the well locations were determined using the measure tool

on a georeferenced map created in Arcview GIS. Distances represent the shortest distance from

the observation well to mapped/inferred conduit locations. The greatest uncertainty in distances

occurs in areas where conduits are not mapped. This is most applicable to the wells north of Well

8 (Figure 2-2) where the exact conduit location has not yet been mapped. A value of 0.2 was

chosen by Martin et al. (2006) to represent the specific yield of the unconfined Floridan aquifer

and therefore was also used in these calculations. This value was based on porosity estimates in

the Floridan aquifer by Palmer (2002) that range from 0. 1 to 0.45. If the specific yield is higher









than the predicted value of 0.2, then the calculated transmissivity would decrease. The time step

chosen for the calculations was one day.

This analytical curve matching method allows for the use of any shaped flood stage

hydrograph. Transmissivity is determined by adjusting values until calculated changes in head

most closely match measured changes in head at observation wells. In unconfined aquifers,

transmissivity will vary with time as the saturated thickness of the aquifer changes with changes

in head. However, the error introduced by using a constant transmissivity in this model is most

likely minimal when comparing the relatively small head changes we see over the course of a

storm event (maximum of 3.5 m in matrix wells) and the assumed total thickness of the Upper

Floridan aquifer (100 m). Transmissivity results may be significantly different when using a

time-lag to match curves as compared to trying to match the amplitude of the curve (Martin,

2003). For this study results are reported for both time-lag and amplitude curve matching

techniques.

1-Dimensional Model with and without Recharge in MODFLOW

The analytical method used by Pinder et al. (1969) uses head change in a boundary

condition (conduit) to predict head change in observation wells at a known distance away. This

solution allows for the calculation of diffusivity (transmissivity/storage), but does not include a

recharge component. In order to observe how much of a difference including the recharge

component makes on the calculation of transmissivity, a one-dimensional, transient finite-

difference model with recharge was created in MODFLOW. A numerical method was selected

for ease of use during calibration and sensitivity tests.

Similar to the analytical method, the model uses head change at a boundary condition to

predict head change at observation wells. However, the difference in the numerical model is the

addition of recharge in the groundwater flow equations solved in MODFLOW. The ground water









flow equation in one dimension for an unconfined aquifer with a recharge source is:(modified

from Schwartz and Zhang, 2003):

8 (Kx Sh) + Q(x) = Sy Sh
dx dx at


Including recharge in the solution allows for separation of the storage component (Sy)

from the diffusivity. The result is that any value of recharge (represented by the Q in the

groundwater flow equation) can be entered per time step. Change in head per time step is thus a

result of the hydraulic conductivity (K), recharge (Q), and specific yield (Sy).

A benefit of the model interface (Groundwater Vistas) is that it allows for adjustment of

parameters to determine the best curve match. The error statistic of root mean square error was

used as a calibration tool. Root mean square error is the average of the squared differences in

measured and simulated heads, and is usually thought to be the best measure of error if errors are

normally distributed (Anderson and Woessner, 2002). The values used for the initial conditions,

model parameters, boundary conditions, and other details are further explained in the next

chapter.

Recession Curve Analysis

Transmissivity was also calculated in this study area using a recession curve analysis

method based on the work of Shevenell (1996) and Powers and Shevenell (2000). Recession

curves from observation well hydrographs were broken down into three segments. The first and

steepest slope represents the dominant effects of drainage of the larger karst features (conduits),

while the second slope represents the emptying of well-connected fractures (Shevenell, 1996).

The third and broadest slope represents the slowest drainage of the matrix rock. Each of these









segments have a characteristic slope (h) for any given storm event, and the slope is defined by

the equation:

h= In f(Y2)_ I(1 Q2)
t2-tl t2-tl

h = slope of recession segment (m/day)
Y1 and Y2 = water levels at time 1 and time 2 (m)
tl and t2 = time after onset of recession corresponding to water levels (Day)
Q1 and Q2 = aSsociated theoretical flows (discharges) corresponding to the water
levels (m3/day)

Solving for Q1/Q2 in the above equation allows for the ratios of the theoretical flows to be

calculated, where the ratio of Q1 to Q2 TepreSents conduit-dominated drainage, the ratio of Q2 to

Q3 TepreSents fracture-dominated flow, and the ratio of Q3 to Q4 TepreSents matrix-dominated

flow. The theoretical flows (Q values) are dependent on the water levels, and therefore the ratio

between two respective theoretical flows can be determined from the ratio of two respective

water levels. The ratio of Q3 to Q4 can be applied in the solution adapted by Atkinson (1977) for

non-conduit transmissivity of an unconfined aquifer from a baseflow recession curve:

Log (Q3 Q4) = T/S (t2 -tl) 1.071/L2

T = Transmissivity (m2/day)
S = Storage coefficient (unitless)
L = distance between monitoring well and the groundwater divide (m)

In our study area, the aquifer is unconfined and thus the storage coefficient is

approximately equal to the specific yield. A value of 0.2 was used for specific yield in the

calculations. A value of 5000 m was used as an estimated distance between the monitoring wells

and the groundwater divide. Transmissivity was calculated by solving for T in the equation.





















S in O Stag masil~ i

Figur 3-.ie ikrtigcrepoued by th uane ie aerMngmntDsrc
fo h saegag octdinOLnoSae ak














140-
v = 4.1363x2 63.642x + 224.59
120 -'=099

S 100-






40-

20-


9.5 10 10.5 11 11.5 12 12.5 13 13.5 14 14.5

Sink Elevation (rnasl)



Figure 3-2. River Sink rating curve produced in this study using sink elevations measured by the
data logger correlated with stage at the River Sink gauging station.












100


80-

cZy = 545.18 -127.23x +7.36 x2

60 R =0.98


S40--


20 -

*Data fr'om SR WM~D


9 9.5 10 "10.5 11 `1 1.5 "12 12.5
Rise elevation (masl)

Figure 3-3. River Rise rating curve produced by Screaton et al. (2004) using data from the
Suwannee River Water Management District.



















fo (a8) = 1- .00119*exp (.61*A8)


Soil moisture deficit A0 (cm)


Figure 3-4. Function assuming a maximum soil-moisture deficit of 10 cm used to calculate leaf
conductance in the Penman-Monteith evapotranspiration method. After Stewart
(1988).










Table 3-1. Summary of Monitoring. Wells
Ground surface
Completed Depth Screened Depth to bedrock elevation
(mbgs Interval (m bgs (mbgs (masl)
Well 1 22.85 22.85-16.75 17.07 14.45

Well 2 30.48 30.48-24.38 6.10 15.96
Well 3 28.35 28.35-22.25 3.05 17.87

Well 4 29.57 29.5 7-23 .47 4.57 17.89
Well 4a 9.75 9.75-6.70 5.18 17.96

Well 5 29.87 29.87-23.77 5.48 16.22
Well Sa 8.23 8.23-5.18 3.05 16.20

Well 6 31.09 31.09-24.99 4.88 13.51
Well 6a 5.49 5.49-2.44 3.96 13.55

Well 7 29.87 29.87-23.77 5.48 15.22
Well 7a 7.62 7.62-4.57 2.43 15.19

Well 8 30.48 30.48-24.38 3.05 13.32














Table 3-2. Le cation and dates of data collection. D = Diver, C = CTD Diver, T = Minitroll

logger, B = Baro Diver. Periods of no data indicate time when loggers were either
malfunctioning or not i stalled
Well Well Well Well Well Well Well Well Well Well Well Well Tower River River Sweet-
1 2 3 4 4a 5 Sa 6 6a 7 7a 8 well Sink Rise water
1/22/03-
2/26/03 D D D C C C

2/26/03-
3/27/03 D D D C C C
3/27/03-
5/14/03 D D B C C
5/14/03-
7/24/03 D D B C C C
7/24/03-
9/12/03 D D B C C C
9/12/03-
11/21/03 D D B C C C
11/21/03
2/9/04 D D B C C C
2/9/04 -
4/7/04 T D T D B T T T C C C

4/7/04 -
6/9/04 T D T D B T T T C C T
6/9/04 -
7/7/04 T D T D B T T T C C T
7/7/04 -
9/2/04 T D D B T T T C C T
9/2/04 -
2/4/05 T D B T T T C C
2/4/05 -
5/1 8/05 T T D B T T C
5/18/05-
7/1 8/05 T T B D T T C
7/18/05-
10/27/05 T T B D T T C
10/27/05-
12/13/05 T T B D T T C
12/13/05
-1/17/06 T T B D T T C
1/17/06 -
2/21/06 T T B D T T C
2/21/06 -
4/11/06 T T B T C C C C C C C C
4/11/06 -
5/1706 T T B T C C C C C C C C C
5/17/06 -
6/1 5/06 T T B T C C C C C C TB C C C
6/15/06 -
7/1 2/06 T T B T C C C C C C TB C C C
7/12/06 -
8/1 7/06 T T B T C C C C C C TB C C
8/17/06 -
9/1 5/06 T T B C T C C C C C C TB C C
9/15/06 -
10/20/06 T T B C T C C C C C C TB C C
10/20/06-
11/29/06 T T B C T C C C C C C TB C C
11/29/06
-1/10/07 T T B C T C C C C C C TB C C
1/10/07 -
2/1 9/07 T T B C T C C C C C C TB C C
2/19/07 -
3/1 9/07 T T B C T C C C C C C TB C C









CHAPTER 4
RESULTS

Precipitation and Evapotranspiration

Daily precipitation records from O'Leno State Park are plotted from April 2004 through

April 2007 (Figure 4-1). Average annual precipitation during this period was 130 cm/yr (Table

4-1). This value is slightly less than the average precipitation of 137 cm/year recorded in the

Santa Fe River Basin by Hunn and Slack (1983) between the years of 1900 and 1976. Higher

than normal precipitation occurred in 2004, when 187 cm were recorded. This was primarily a

result of two September hurricane events that were responsible for 50 cm of precipitation. In

contrast, only 84 cm of precipitation occurred at O'Leno State Park during 2006, which is

significantly lower than average values. During this time the north-central Florida region was

experiencing a significant drought.

Evapotranspiration (ET) was calculated on a daily basis from January 1, 2002 through

April 2007 using the Penman-Monteith method and is plotted for this study from April 2004

through April 2007 (Figure 4-2). A value of zero on the plot indicates that there was not enough

water available for evapotranspiration to occur. Monthly and annual ET values were calculated

by summing daily values. The average annual ET calculated since 2002 was 76 cm/yr (Table 4-

1). The lowest annual calculated ET was 66 cm, which occurred in 2003. The highest annual

calculated ET was 85 cm, which occurred in 2005. These values are lower than values of 90 to

105 cm/yr calculated at 96 rainfall stations in south-central Georgia and across the state of

Florida by Bush and Johnston (1988). The lower calculated values in this study are most likely

caused by a combination of the differences in the methods used and the scale of the calculation.

The method used by Bush and Johnston (1988) used the mean annual temperature to estimate

potential evapotranspiration, and then estimated the actual evapotranspiration based on an









estimated ratio of potential evapotranspiration to precipitation. The Penman-Monteith model

used in this study is not only dependent on temperature, but there are many other components

included in the calculation. One of the most important is the soil-moisture deficit to track the

amount of water available per day for evapotranspiration. The Penman-Monteith model will

produce a value of zero when there is not enough water available for evapotranspiration from the

soil moisture storage. This will likely cause a more accurate prediction of the actual water that

evaporates. The scale of the calculation also is important because Bush and Johnston (1988)

calculated evapotranspiration for a much larger region than this study. In their calculation,

evapotranspiration was averaged from land and open water regions across south Georgia and

north Florida. The Penman-Monteith model incorporates evapotranspiration only from the land

surface and was applied only to the area around O'Leno State Park.

Monthly potential ET (PET) was calculated using the Thornthwaite method for

comparison (Figure 4-3). For the three year period during this study, the average annual PET

calculated with the Thornthwaite method is 102 cm/yr. This value agrees with the values

calculated by Bush and Johnston (1988). This is most likely a result of both methods being based

on mean temperatures. Monthly calculations of PET and ET using the respective methods show

distinct seasonal trends, with increased values in the summer months and less in the winter

months (Figure 4-3). Comparing both methods shows that PET calculated with the Thornthwaite

method is greater than calculated ET with the Penman-Monteith method during the summer

months, by up to 5 cm in July and August. Results also show that during the winter months the

ET calculated by the Penman-Monteith method is sometimes estimated to be higher than the

PET calculated with the Thornthwaite method. This is likely a result of the differences in the

method. Xu and Chen (2005) worked with three methods that calculate ET and compared them









to four methods that calculate PET, one of which was the Thornthwaite method. Although the

Penman-Monteith method was not included in their research, results indicated that ET calculated

with other methods can sometimes be greater than PET calculations made with the Thorthwaite

method. Lemur and Zhang (1990) evaluate three methods of calculating ET, and their results

demonstrate that the evapotranspiration calculated with Penman-Monteith reflects values

calculated from a water balance better than the other two methods. The Penman-Montieth model

has become the most widely used approach at estimating evapotranspiration from land surfaces

(Dingman, 2002), and therefore would likely represent the best estimate of evapotranspiration in

this region.

Water Levels

River Rise data have been used in various prior studies since 2001 (Screaton et al, 2004;

Martin et al, 2006) (Figure 4-4). Water levels were also recorded at the River Sink and

Sweetwater Lake locations during this time; however, records were interrupted at various points

due to logger malfunction and/or lack of logger availability, and therefore the record is much less

complete (Table 3-2). Water levels recorded specifically during this study were between April

2004 and April 2007 (Figure 4-5). During this recording period, there were three storm events

distinctly displayed by three large peaks in the River Rise hydrograph (Figure 4-5). These storm

events were also observed in fluctuations of water levels at various monitoring wells (Figure 4-

5). Increase in head at the monitoring wells ranged from as high as 3.5 m during the largest

September 2004 event, to 1 m during the smaller December 2005 event. Similar increases in

head were observed at the River Rise. Water levels at Well 1 are greater than the River Rise.

This is because Well 1 is influenced by the change in head at the River Sink, which is upgradient

of the River Rise and thus has higher water levels. The maximum water level at the River Rise









occurs slightly before the maximum water level at most of the monitoring wells, indicating that

there is a small time-lag for the head change to reach the wells.

Sink and Rise Discharge

Stage at the River Sink and the River Rise were converted to daily discharge rates based

on the appropriate rating curve. Daily discharge rates are plotted for this study between April

2004 and April 2007 (Figure 4-6). The blue highlighted area of the figure indicates the time

period during this study where the gauge at the River Rise was becoming increasingly warped.

Discharge rates were also corrected for a possible 0.3 ft error and are also plotted in figure 4-6.

The corrected water level values shift the Rise discharge curve slightly lower.

During this time period, there were three storm events which increased discharge (Figure

4-6). During these storm events, the discharge into the River Sink is equal or greater than the

discharge out of the River Rise. This indicates that some of the storm water which entered the

Sink is leaving the conduit, presumably to the matrix. The idea that water moves into the matrix

during these large events is also supported by the fact that change in head at the monitoring wells

during storm events is much larger than the amount of calculated recharge divided by the

specific yield. In contrast, during periods of low river stage, the discharge out of the River Rise is

noticeably higher than the discharge into the River Sink. This indicates that the conduit is

gaining water along its flow path.

Recharge

Diffuse Recharge Calculated From Water Budget

Excess precipitation was calculated daily from 2002 through April 2007 by subtracting

evapotranspiration calculated with the Penman-Monteith method from precipitation. Because

runoff is minimal in the unconfined Floridan aquifer, recharge was calculated using this excess

precipitation and accounting for soil moisture storage (Table 4-1). Daily recharges values are










plotted for the study period of April 2004 to April 2007 (Figure 4-7). Most of the recharge during

this time occurred within three time periods, which are associated with the three large storm

events (Figure 4-7). The largest single recharge peak (~20 cm) occurred during the September

2004 storm event. During 2006, there are very few instances where recharge occurs (Figure 4-7).

This is caused by less precipitation occurring during this time (Figure 4-1). Comparing the daily

recharge values with the precipitation record also indicates that much of the precipitation never

actually makes it to recharge. Instead, this precipitation is evapotranspired or occupies soil

moisture storage and is later evapotranspired. Annual recharge relies heavily on the amount of

precipitation that the area receives over the course of a year.

Diffuse Recharge during Large Storm Events

During large storm events, precipitation is significantly greater than evapotranspiration,

and soil moisture storage is likely to reach capacity quickly. The combination of a large diffuse

recharge component combined with an input from the conduit, results in the largest increases in

observation well hydrographs in response to these events (Figure 4-5). Diffuse recharge is shown

for each of three events occurring after April 2004 (September 2004, March 2005, and December

2005), as well as the March 2003 event analyzed by Martin et al. (2006) (Table 4-2). Each storm

event varied in the amount of precipitation and ET, based on the time of the year and the

magnitude of the storm. The calculated recharge during the September 2004 hurricane events

was 41 cm, which occurred over a relatively short period of time (24 days). In contrast,

calculated recharge for the March 2003 storm event was 27 cm and for the December 2005 storm

event was 24 cm; both of these results are nearly half the amount of recharge spread out over

double the amount of time (50 days). The smallest event was in March and April 2005, when

only 10 cm was recharged over 28 days. The duration of the storm events were chosen based on









the water level records, with a at least three to four days included before the increase in water

levels and at least five to ten days included during the recession.

Diffuse Recharge Calculated From Chloride Concentrations

Over the course of 2006, hydrographs. from conduit locations and well locations both show

a consistent recessional to flat trend (Figure 4-5). During these recessions, water may flow from

the shallow well locations towards the conduit, and thus it is likely that the shallow wells are not

receiving any conduit water. Therefore, concentrations of chloride in the various shallow well

locations should primarily reflect the evapotranspiration process, and can be used to make an

estimate of the evapotranspiration occurring at the well locations. Water samples were available

from six sampling trips beginning in April 2006 and ending January 2007 (Moore, unpublished

data). Shallow well bulk water chemistry was analyzed after each sample period, and chloride

concentrations are available for the four shallow wells (Table 4-3).

The average chloride concentration for the six sampling periods was used to calculate a

concentration factor for each shallow well location (Table 4-3). Concentration factors were

calculated assuming a chloride concentration value of 1 mg/L for the precipitation that occurs at

O'leno State Park. This value was chosen from a contoured map created by Junge and Werby

(1958) with data collected at various locations across the United States between July 1955 and

July 1956. Chloride concentrations in Florida from Junge and Werby's (1958) study range from

0.66 mg/L over the Florida panhandle, to 2.44 mg/L near Miami. The closest location to O'Leno

State Park was near St. Augustine, with a chloride concentration of 1.02 mg/L. The 1.0 mg/L

value was chosen based on the 1.0 mg/L contour that falls over the O'Leno State Park study area.

The highest concentration factor determined from the chloride concentrations was .917, which

occurred at well Sa. Concentration factors of well 4a and 5a are almost identical, and all of the

concentration factors are within 6% of each other. If the chloride concentration is actually lower









than is estimated from the contour map, then the resulting concentration factors would be higher.

Differences in concentration factors may be a result of differences in the local vegetation near

the respective well location, which directly affects the amounts of evapotranspiration that would

occur. Differences may also be a result of the thickness of the vadose zone and thus the amount

of water that can evapotranspire from the top of water table.

Evapotranspiration was calculated for 2006 by comparing the chloride concentration

factors determined at each well location to the total amount of observed precipitation (Table 4-

4). Average evapotranspiration for the four wells was 75 cm, which is consistent with the

average calculated value of 76 cm/yr using the Penman-Montieth method for the five-year

period. This value is slightly higher than the 66 cm calculated for 2006 with the Penman-

Monteith model. Annual recharge was then calculated for 2006 by taking the difference between

total precipitation during 2006 and the calculated evapotranspiration using the concentration

factor. Calculated recharge values range between 7 cm and 12 cm, which is lower than the 17 cm

calculated with the water budget method. Both results support that 2006 diffuse recharge was

well below the average of 45 to 60 cm/yr calculated by Grubbs (1998).

Diffuse Recharge Estimated Using the Water Table Fluctuation Method

The water table fluctuation method is often applied as a comparative method to other

recharge calculations (Healy and Cook, 2002). This method is commonly used in regions where

change in water level is attributed solely to water arriving at the water table via diffuse recharge.

The water table fluctuation method assumes that the change in water level observed at the water

table is related to the amount of precipitation that reaches the water table divided by the aquifer

storage (specific yield). During low-flow periods in the O'leno study area, discharge out of the

River Rise is larger than discharge into the River Sink. We expect that there is little to no

movement of water from the conduit to the matrix. During these times, the overall hydrographs









are experiencing a recession. It is common that a small rain event will trigger a minor "blip" in

the hydrograph, but will not change the long term recessional trend. Water levels at shallow well

locations Sa, 6a, and 7a, which are screened at the water table, were monitored from the time of

their installation on February 21, 2006 through April 2007. Because most of 2006 was dry, the

overall trend of the hydrographs at the wells was a recession. However, over the summer of 2006

there were two rainfall events on record that caused small perturbations to water levels. The first

event occurred in June 2006 when 12.3 cm of precipitation was recorded over three days. The

diffuse recharge calculated with the water budget method for this event was 3.2 cm. The second

event occurred in July of 2006 when 7.8 cm of precipitation was recorded over three days. This

event produced 0 cm of recharge calculated with the water budget method (Figure 4-9). During

the time when these events occur, the River Rise water level continued to experience a recession

(Figure 4-10). During the recession, the River Rise water level increases very slightly (<.05 cm)

after the June event, which may suggest that the change in water levels observed in the matrix

wells might be influenced by the conduit.

For each of these small events, the raw 10 minute data were smoothed using an interval

of 50 points (Figures 4-11 through 4-16). The recession curve was then proj ected from the

smoothed line, and the difference was calculated between the proj ected curve and the maximum

spike in water level. Recharge was estimated by taking the maximum change in head and

dividing by the estimated specific yield of 0.2. Recharge calculated using this method is

compared to recharge calculations using the water budget method (Table 4-5). For the June 2006

event, the water table fluctuation method predicts significantly less recharge at all three shallow

well locations than the 3.18 cm predicted by the water budget method. In contrast, for the July

2006 event, the water table fluctuation calculations show a small amount of recharge (between









0.38 and 0.43 cm), while the water budget calculation indicates that no recharge should have

occurred (0 cm) (Table 4-5).

Transmissivity

Analytical Method without Recharge

Martin et al. (2006) investigated transmissivity in this study area using an analytical

curve-matching method described by Pinder et al. (1969) that estimated monitoring well

response to head changes in the conduit system; diffuse recharge was not included. This method

was applied to monitoring well hydrographs from a storm event in March 2003 (Figure 4-4).

Results showed that transmissivity between the conduit and various monitoring wells ranged by

orders of magnitude, and was highly dependent on the distance of the monitoring well to the

conduit. For example, calculated transmissivity to Tower well (approximately 3700 m from the

conduit; Figure 2-2) and Well 1 (approximately 500m from the conduit; Figure 2-2), were

160,000 m2/day and 97,000 m2/day respectively, while calculated transmissivity to Well 4

(approximately 115 m from the conduit; Figure 2-2) and Well 6 (approximately 140 m from the

conduit; Figure 2-2) were 950 m2/day and 900 m2/day respectively (Martin et al, 2006). The

difference in transmissivity between the various well locations was attributed to the fact that

water traveling a further distance through a highly karstic system will be more likely to

encounter preferential flow paths along the way.

Martin (2003) also observed that calculated transmissivity could vary at a single well

location depending on whether the calculated data was matched to the observed data by

correlating the timing of the hydrograph peak or by correlating the amplitude of the hydrograph

peak. For example, when using the time lag as the matching point, the estimated transmissivity at

Tower Well was 120,000 m2/day, which is significantly less than the 550,000 m2/day calculated

when matching the curve based on amplitude.









In this study, the Pinder et al. (1969) method was applied to the three storm events that

occurred after March 2003. Transmissivity was calculated between the River Rise and Wells 4,

6, and 7 (Table 4-6). Transmissivity could not be calculated between the River Sink and Tower

Well or Well 1 because the data logger was not operating at the River Sink during these times.

However, Well 7 is located at a distance of 1025 meters from the conduit, which is intermediate

in distance between Tower Well and Well 1. Results from the curve matching between actual

head change and calculated head change for the three storm events are shown in the Eigures:

September 2004 (Figures 4-17, 4-18, 4-19), March 2005 (Figures 4-20, 4-21, 4-22), and

December 2005 (Figures 4-23, 4-24, 4-25). In these Eigures, head change refers to the difference

between water level on a given day and the water level on the first day of the time period being

analyzed. For all three events, at least 1-3 days of small head change (< .02 m) exist prior to the

start of simulations. For events that have two peaks in the hydrograph, the amplitude match was

made with the first peak because the first peak on the hydrograph should represent the initial

pressure head change in the conduit that reaches the well after the onset of the event from a base

flow condition. The following peaks occur as a result of precipitation that happens after the

hydrograph has already increased significantly, and thus is not starting from the base flow initial

conditions.

Transmissivity calculated for these events varied by orders of magnitude depending on

the distance of the observation well from the conduit, as previously observed by Martin et al.

(2006). For example, using the time lag match, the September 2004 event produced

transmissivity values from Wells 4 and 6 (115 m and 140 m from the conduit) of 800 m2/day and

1200 m2/day respectively. These values are significantly less than the transmissivity of 150,000

m2/day determined from Well 7 (1025 m from the conduit). Transmissivity also varied









significantly when using a time-lag match as compared to an amplitude match. For the same

September event, the amplitude-match produced a transmissivity of 350,000 m2/day from well 7

as compared to the time-lag calculation of 150,000 m2/day.

1-Dimensional Model with and without Recharge in MODFLOW

The Pinder et al. (1969) analytical method allows for prediction of ground water

fluctuations at monitoring wells based on water level changes in the nearby conduit, but does not

include diffuse recharge. To judge how much diffuse recharge contributes to storm event

hydrographs and in the calculation oftransmissivity, a one-dimensional, transient finite-

difference model with a recharge component was created in MODFLOW. The model grid has 50

columns, with an initial spacing of 10 m close to the conduit and an increasing cell size by a

multiplier of 1.1 away from the conduit (Figure 4-26). The spacing is refined near the conduit

because this is the region where the model is most affected by changes in the conduit water level

(Figure 4-26). The total length of the model is 1 163 8 m. The total thickness of the aquifer was

assigned to be 100 m. The stress period was set at 1 day with 10 uniform time steps applied to

each stress period. The total time of the model was altered to work with each of the storm events,

with the longest model duration being 70 days for the September 2004 event, and the shortest

model being 29 days for the March 2005 event. The conduit was simulated as a transient

constant-head boundary with water level values from the River Rise imported for each day of the

event. In order to account for the differences between the conduit head and the head at the

observation wells at the beginning of the simulation, change in head from the onset of the storm

event was used at both locations. Thus, the initial head throughout the model at the beginning of

all the simulations was set at zero. The boundary condition at the other end of the model (furthest

point from conduit) was assigned to be a no flow boundary.









The specific yield of the unconfined aquifer was set at 0.2 for the model runs. This value

was chosen as a best estimate of specific yield and to also to be consistent with all other

calculations; sensitivity to the selected specific yield was tested as part of the modeling. In the

model, three observation points were located at 110 m, 140 m, and 1020 m from the constant

head boundary to represent Wells 4, 6, and 7 respectively. Observed data from each well for each

storm event were compared with simulation results. The curve match that produced the lowest

root mean square error was used to determine the best fit hydraulic conductivity. Transmissivity

was then calculated by multiplying the best fit hydraulic conductivity determined in the model by

the assigned aquifer thickness of 100 m (Table 4-6).

The 1D model was first simulated for all four storm events without including calculated

recharge. Except for effects from the discretization of space and time, and the no flow boundary

on one side, simulating the model without recharge should produce results identical to solving

the Pinder et al (1969) analytical solution without recharge. This is verified by results of a

hypothetical simulation of the Pinder et al (1969) analytical solution and the numerical model

without recharge (Figure 4-27). In the hypothetical simulation, both methods were simulated for

a hypothetical observation well 1500 m from the conduit location, with a three meter change in

head in the conduit and a transmissivity of 100,000 m2/day. Figure 4-27 shows that predicted

curves with both methods produce results that are very similar.

Estimated transmissivity values calculated from the storm events with the analytical

solution without recharge and the model without recharge differ (Table 4-6). This is a result of

the curves being matched differently between the two methods. The model was calibrated to find

a best fit hydraulic conductivity by using the lowest root mean square error between the observed









and calculated heads at the observation well. The analytical solution was calibrated using a

visual match of amplitude or time-lag.

After simulating the system without recharge and determining the best fit transmissivity,

the model was then rerun and recalibrated with the calculated recharges included. To simplify

the modeling, recharge was assumed to reach the aquifer on the day of occurrence and thus was

applied to the time step when recharge was recorded. The surficial sediments that make up the

unsaturated zone are mostly sands, and are less than 6 m thick at all well locations except Well 1

(~17 m thick). The maximum depth to the water table at the peak of the smallest event

(December 2005) was 6.38 m, while during the larger events the water table came extremely

close to the surface (<. 1 m during the September 2004 event). The small thickness of the surficial

sediments and the proximity of the water table to the surface support the idea that water should

move through the unsaturated region and quickly become recharge. This is also supported by the

small June 2006 event where water level changes occur on the same day as estimated recharge

occurs (Figure 4-8). However, model results (Figure 4-28 through 4-40) indicate that for three of

the four storm events, including the diffuse recharge component on the same day results in

predicted curves that are pointed and jagged. These results are not consistent with the smooth

curves that are seen in the observed data, and suggest that the timing of the recharge input may

be more spread out than is applied in the model simulation. The assumption that recharge arrives

at the water table on the same day as precipitation occurs maximizes the impact of diffuse

recharge in the calculation because the recharge is arriving in a lump sum rather than over a few

days.

The addition of recharge to the model simulations results in predicted curves at Well 7

during the September 2004 and December 2005 events that plot higher than the observed data for









these respective events (Figure 4-34, 4-40). Although these predicted curves produce the lowest

root mean square error in the curve match, it seems unlikely that the predicted curve should vary

so significantly from the observed data. Two possible explanations are available. If the recharge

is overpredicted, then the model would produce an unreasonable shift of the curve above the

observed data. This is unlikely, considering that the same recharge has been applied to other well

locations and this result is not observed. The second possible explanation may be that the

influence of the conduit is overpredicted at Well 7. Based on the location of Well 7 with respect

to the River Rise (Figure 2-2), this is possible if Well 7 is actually being influenced by the Santa

Fe River at a location downstream of the River Rise; at this location head changes in response to

storms may be less and thus we would expect the calculated transmissivity to also be less.

Comparing 1D model simulations with and without recharge indicates that the best fit

transmissivity values are higher when the recharge is neglected. This result is seen in 10 out of

12 model simulations, with the only exception being the matches possibly in error at Well 7

(Table 4-6). These results suggest that in order to match observed and simulated curves, the

calculated transmissivity must adjust for the lack of recharge by increasing. Therefore, the curve

matching techniques which neglect the recharge component will overpredict transmissivity from

storm events. For the storm events in this study, the transmissivity required to make up for the

lack of recharge can be up to four times greater than when recharge is included in the calculation

(e.g. Well 6 during the September 2004 event).

Consistent with results of the analytical method, results from both the 1D model with and

without recharge indicate that calculated transmissivity can be one to two orders of magnitude

higher for Well 1, Well 7, and Tower Well, than for wells closer to the conduit (Table 4-6). For

example, model results at Well 7 for the September 2004 event are 300,000 m2/day, while the









highest transmissivity calculated from Wells 4 and 6 during all simulations was 3,000 m2/day.

Both the model with and without recharge also show that the storm event with the highest change

in head and most diffuse recharge (September 2004) produces higher calculated transmissivity

values than the smaller events. For example, Wells 4 and 6 have a calculated transmissivity of

3000 m2/day from the September 2004 event, while the transmissivity values calculated at these

locations from the other events does not exceed 1000 m2/day.

Sensitivity to Specific Yield

For all the calculations performed to date, a specific yield of 0.2 was used as a best

estimate for the highly karstic, unconfined Floridan Aquifer based on porosity estimates from

Palmer (2002). If this is not an accurate estimate, then the calculated transmissivity values would

also be in error. To test the effect of changing specific yield on the transmissivity calibration and

to observe whether the specific yield value affects the quality of the curve match, the 1D model

with recharge was simulated with a smaller specific yield of 0. 1, and a larger specific yield of

0.3. Including recharge in the solution allows us to separate transmissivity and specific yield.

The best fit transmissivity was once again chosen from the root mean square error of the curve

matching (Table 4-7).

Decreasing the specific yield to a value of 0. 1 results in significant changes to the

calculated transmissivity from all four storm events (Table 4-7). For example, calculated

transmissivity between the conduit and Well 4 during the March 2003 storm event is three times

as much with the lower specific yield, while during the March 2005 event the transmissivity to

Well 4 is about four times less. However, more important than the change in transmissivity is

that decreasing the specific yield to 0.1 significantly increases the root mean square errors for all

model simulations. This indicates that the specific yield of 0.2 yields a better fit to the data than a

lower specific yield of 0. 1. Results are perhaps more interesting when increasing the specific










yield to a value of 0.3. For almost half of the simulations, the root mean square errors decrease as

specific yield is increased from 0.2 to 0.3. These results suggest that specific yield might vary

between the conduit and the various well locations. For the purpose of calculations at this scale,

the model results generally confirm that the original assumption of an average specific yield of

0.2 is reasonable. Ideally, a sample set of more well locations spaced at various points

throughout the matrix would provide a better estimate of how specific yield varies across the

study area. Finer sensitivity analysis would help pin down a more exact estimate of the specific

yield at these locations.

Recession Curve Analysis

Unlike the Pinder et al. (1969) analytical curve matching technique and the 1-

dimensional model with recharge, the recession curve analysis method does not require a conduit

signal as an input. Therefore, the benefit of this method is that it can be applied to calculate

transmissivity at all well locations where an uninterrupted well hydrograph recession curve is

available. However, it is also important to note that this method is used to calculate "matrix

transmissivity" or what is described by Shevenell (1996) as the slower hydrologic response of

the intergranular porosity. This transmissivity is mainly derived from the third, and broadest

sloped, segment of the recession curve. Therefore, these values do not necessarily indicate flow

through all of the preferential flow paths, and therefore do not represent transmissivity on the

formation scale.

For this study, 14 segments from various recession curves were available for analysis.

These recession curves spanned all four storm events since 2003 (Table 4-6). Length of recession

varied for the different storm events, with the shortest recessions coming after the March 2003

event (~40 days), and the longest recessions coming after the December 2005 event (~55 days).

A specific yield of 0.2 was used for all of the calculations. Inflection points were chosen visually









on the recession curve at the first and second noticeable change of slope (Figures 4-41 4-44).

Finding the inflection points was difficult for the March 2005 event, where a few small rainfall

events caused slight perturbations in the recession curve (Figure 4-43).

Results show that calculated transmissivity varies per storm event, with the highest

transmissivity values being calculated for the larger 2003 and 2004 events (Table 4-6). The

transmissivity calculated for the larger events is on average twice the transmissivity calculated

for the two events in 2005. This is similar to what is observed with the 1D model with recharge,

where transmissivity values calculated from the September 2004 storm event were greater than

the other three events. Perhaps more interesting is that transmissivity is generally consistent at all

the well locations on a per storm event basis. For example, results do not show a transmissivity

at Well 7 that is orders of magnitude larger than at Wells 4 and 6, as observed in the previous

analyses. This is likely a result of the method used, as the recession curve values are only

indicative of the matrix transmissivity, as opposed to the effective transmissivity of the entire

formation between the conduit and well.

Basin Area Calculations

Recession curves were also used in this study to estimate the aquifer surface area

upgradient of the River Rise which contributes to the discharge of the Santa Fe River Sink-Rise

system. This has previously been a difficult value to assess because groundwater divides are not

clearly defined by water table mapping in this area. Knowing the surface area of the groundwater

basin that discharges to the River Rise is valuable because it allows quantification of a

volumetric input to the system via the diffuse recharge component. These values can then be

compared to the volume that the aquifer receives from conduit losses during storm events. For

these calculations, recharge to the confined portion of the Floridan aquifer within the basin is

neglected because all calculations are made using the assumed specific yield value of 0.2 for the









unconfined aquifer. Making this assumption underestimates overall area because the confined

region will have a specific yield that is much smaller than the unconfined region. However, this

assumption is acceptable in this study because we are only concerned with recharge to the

unconfined portion of the aquifer.

Recession curves following the storm events in March 2005 and December 2005 were

used for the calculations. The time periods used for the analyses are November through

December 2005 and March through December 2006. The recession curves following the

September 2004 event were not used because of the possible error caused by the warpage of the

River Rise water level gauge. The slopes of the recession curves were determined by finding the

change in head per number of days of recession. For the recession curves occurring as a result of

the December 2005 storm event, the curves was broken into segments which clearly varied in

slope (Figure 4-45). The average slope of all the recession segments was 0.0078 m/day, and

ranged from 0.0053 m/day to 0.01 16 m/day depending on the storm event and the segment of the

curve. In total there were 14 recession segments available to calculate area. Change in volume

per day of the recession was calculated by taking the daily difference between the Rise and Sink

discharge. Basin area was calculated using the equation:

Basin area = AV / Sy~h

AV = Change in volume (m3)
Sy = Specific Yield
Ah = Change in head (m)

The surface area draining to the River Rise calculated from the average of 14 recession

segments is 4.44 x 10s m2 Or 444 km2 (Table 4-8). For the recession curves resulting from the

December 2005 event, the calculated area varied for the upper and lower portions of the

recession curves. The first portion (upper half before first inflection point) of the recession









curves yielded calculated areas that were less than the second portion (lower half after first

inflection point). The average area for the first portion of the curves was 356 km2, while the

average area for the second portion of the curves was 547 km2. A possible explanation for this

result is that the specific yield is changing as the saturated thickness of the aquifer decreases

during recession. During storm events, the water level reaches well into the surficial sediments.

If these sediments have a different specific yield than the carbonate aquifer, then using a value of

0.2 for specific yield for both the upper and lower portions of the curve would result in the

apparent difference in area. If the inferred change in area is due to changes in specific yield, we

should see the curve inflection point occur at the same water level for each respective well

location. However, this is not the case, as the inflection points occur at different water levels

after each storm event. Another possible explanation for the changing area is that there is a non-

stationary groundwater divide. If the groundwater divide moves away from the discharge point

as the curve recedes, then we would see the increased area from the lower portion of the

recession curves.

The area calculations are sensitive to the accuracy of the discharge measurement. If the

stage at the River Rise is incorrect, the resulting discharge and calculated area would change. To

estimate a possible range of error in the area calculation due to suspected problems with the rise

gauge, a calculation was made to see how the area changes if the stage at the River Rise was

overestimated by 0.3 ft. Calculations show that a decrease of 0.3 ft decreases the average

discharge at the Rise and thus increases the difference between the Sink and Rise discharge. This

only has a small effect on the calculated area, which would be 3.35 x 10s m2 Or 335 km2. If Some

of the error is due to changes at the River Sink, we would expect a similar affect. Increasing the

River Sink stage by .3 ft would result in a calculated area of 3.25 x 10s m2, Or 325 km2
















































QQ
=r


0
0 0
Ob
~I
NcO 0
~I
0 d
~I


Date


Figure 4-1. Daily precipitation records from O'Leno State Park during this study. Three storm events are highlighted.


111 1111


,i, ,,,.


111


II I i YLIIIIIIYIII. 1 111 11 II IIYill 1~111 IIII( IYIYIIIIII I II IIYIIIl IlYII I Ii IY IIIIIIII1IIII I1 11 1111 111 II


d
d d
88~a
C~rC?~
r cu
oo


~n ~n~n
~n ~n 00
0 00 0
$ $ ~no,
0
cu cu
0
















0.6



0.5



0.4
C


S0.3



S0.2



0.1



0n~


~~n ~nDate

Fiue -. alyclclte vpornsiato drngtissuy sngte ema-onethM dl.Sacsinte aa nict
value~N of ero















SActual ET
SPotential ET


C
O






co LI.J
0o


0000000000000000000


Month



Figure 4-3. Monthly calculated potential evapotranspiration using the Thornthwaite method and monthly actual evapotranspiration
using the Penman-Monteith method
























Cr 13 -- -- -- -- -- -- -- -- -------------- ---------------







10








4/1/01 4/1102 4(1103 4/1/04 4(1105 4(1106 4/1/07


Date



Figure 4-4. Long term record of River Rise water levels. Storm events that have been cited in previous work and in this study are
highlighted.


























10 r

4/1 /04


1 0/1 /04 4/1 /05 1 0/1 /05 4/1 /06 1 0/1 /06 4/1 /07


Date


Figure 4-5. Water levels of various monitoring wells during this study and River Rise water level for comparison. Three storm events
are highlighted.



















- Rise (-0.09 m)


2 107


1.5 107


1 107


5 106




0 100 =-
4/1/04


10/1/04 4/1/05 10/1/05 4/1/06 10/1/06 4/1/07


Date



Figure 4-6. River Sink and River Rise discharge during this study. Three storm events resulting in increased discharge are highlighted.
The time period where the River Rise gauge was warped is represented with the blue highlight.






















20


E
0 15
a,


a, 10





0 ,1 I IdI



Dat
Figre -7.Daly alclatd iffse ecarg duingths sudyusng h ae ugtmto.Rcare soitdwt h he
majorb stor event arihlgtd









I Prccipi~t~ation(cm)


Date


Figure 4-8. Precipitation and calculated diffuse recharge with the water budget method for the small event in June 2006.


I1


5










I ~Rech~arge (c~m)


2

. .


Figure 4-9. Precipitation and calculated diffuse recharge with the water budget method for the small event in July 2006.


III I

date


















10.05


10


E 9.95


9.9


9.85


9.8




5/1 5/06 6/1 5/06 7/1 5/06 8/1 5/06 9/1 5/06 10/15/06

Date





Figure 4-10. River Rise water level during summer 2006. Small events in June and July are
highlighted.














Well Sa June 2006


9.98


9.96


9.94


9.92


9.9


9.88


9.86
6/6/06 6/9/06 6/1 2/06 6/1 5/06 6/1 8/06 6/21/06 6/24/06

Date


Figure 4-11. Water level perturbation at Well Sa after a small rain event in June 2006 with
recession curve proj ected.














Well 6a June 2006
10


9.98


J 9.96


-9.94


9.92





9.88


9.86
6/6/06 6/9/06 6/1 2/06 6/1 5/06 6/1 8/06 6/21/06 6/24/06

Date


Figure 4-12. Water level perturbation at Well 6a after a small rain event in June 2006 with
recession curve predicted.














Well 7a June 2006


10

9.98


9.96

9.94

9.92

9.9

9.88

9.86

9.84

9.82
6/6/06


II


6/9/06 6/1 2/06 6/1 5/06 6/1 8/06 6/21/06

Date


6/24/06


Figure 4-13. Water level perturbation at Well 7a after a small rain event in June 2006 with
recession curve proj ected.














Well Sa July 2006


9.92



9.9


9.88



9.86


9.84


9.82
7/1 6/06


7/20/06 7/24/06 7/28/06


8/1/06


Date


Figure 4-14. Water level perturbation at Well Sa after a small rain event in July 2006 with
recession curve proj ected.














Well 6a July 2006


9.86


9.84



9.82



9.8


9.78


9.76
7/1 6/06


7/20/06


7/24/06


7/28/06


8/1/06


Date


Figure 4-15. Water level perturbation at Well 6a after a small rain event in July 2006 with
recession curve proj ected.














Well 7a July 2006
9.84



9.82



E 9.8



9.78



9.76



9.74
7/1 6/06 7/20/06 7/24/06 7/28/06 8/1/06

Date


Figure 4-16. Water level perturbation at Well 7a after a small rain event in July 2006 with
recession curve proj ected.





























Well 4 Observed
0.5 Well 4 Amplitude match
Well 4 time-lag match

9/2/04 9/7/04 9/1 2/04 9/1 7/04 9/22/04 9/27/04

Date


Figure 4-17. Curve matching results at Well 4 for the September 2004 storm event using the
Pinder et al. (1969) method.


0
9/2/04


9/7/04 9/1 2/04 9/1 7/04 9/22/04 9/27/04


Date


Figure 4-18. Curve matching results at Well 6 for the September 2004 storm event using the
Pinder et al. (1969) method.































0
9/2/04


9/7/04 9/1 2/04 9/1 7/04 9/22/04 9/27/04


Date


Figure 4-19. Curve matching results at Well 7 for the September 2004 storm event using the
Pinder et al. (1969) method.


O Y
3/21/05 3/26/05 3/31/05 4/5/05 4/10/05 4/15/05 4/20/05 4/25/05


Date


Figure 4-20. Curve matching results at Well 4 for the March 2005 storm event using the Pinder
et al. (1969) method.















































Swell 7 observed
- Well 7 amplitude match
- Well 7 time-lag match


0
3/21/05 3/26/05 3/31/05 4/5/05 4/10/05 4/15/05 4/20/05 4/25/05


Date


Figure 4-21. Curve matching results at Well 6 for the March 2005 storm event using the Pinder
et al. (1969) method.


0
3/21/05


3/26/05 3/31/05 4/5/05 4/10/05 4/15/05 4/20/05 4/25/05


Date


Figure 4-22. Curve matching results at Well 7 for the March 2005 storm event using the Pinder
et al. (1969) method.















- Well 4 observed
-Well 4 amplitude match
-Well 4 time-lag match


Figure 4-23. Curve matching results at Well 4 for the December 2005 storm event using the
Pinder et al. (1969) method.






Well 6 observed
-Well 6 amplitude match
-Well 6 time-lag match
1.5 P--


0 ~
12/12/05


12/22/05 1/1/06 1/11/06 1/21/06 1/31/06


Date


O
12/12/05


12/22/05 1/1/06 1/11/06 1/21/06 1/31/06


Date


Figure 4-24. Curve matching results at Well 6 for the December 2005 storm event using the
Pinder et al. (1969) method.
















- Well 7 observed
-Well 7 amplitude
-Well 7 time-lag


0 -
12/12/05 12/22/05 1/1/06 1/11/06 1/21/06 1/31/06


Date



Figure 4-25. Curve matching results at Well 7 for the December 2005 storm event using the
Pinder et al. (1969) method.










Observation W~ells: tower 7


16 4


Conduit water
level


No flow boundary


Figure 4-26. 1-Dimensional model with recharge grid
















-1D Model without recharge
-Analytical solution without recharge


12.5


12


11.5


11


10.5


10


15 20


Day

Figure 4-27. Results from of the Pinder et al (1969) analytical solution without recharge and the
1D model without recharge for a hypothetical storm event














I _


I Precipitation (m)
I Calculated recharge (m)


2 I


- Well 4 observed
-Well 4 with recharge (T=150)
- Well 4 without recharge (T=450)


10 20 30 40


0.25


0.15


0.1


0.05


Figure 4-28. Curve matching results at Well 4 for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference.















2 1


Well 6 observed
-Well 6 with recharge (T=250)
-Well 6 without recharge (T=1300)


I Precipitation (m)
I Calculated recharge (m)


10 20 30 40


Day


0.25


0.15


0.1


0.05


Figure 4-29. Curve matching results at Well 6 for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference


c
















STower well observed
- Tower well with recharge (T=60000)
STower well without recharge (T=330000)


I Precipitation (m)
I Calculated recharge (m)


0.5




0 10 20


0.25


0.15



0.1



0.05


Figure 4-30. Curve matching results at Tower Well for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference












































I Precipitation (m)
- Calculated recharge (m)


xWell
- Well
- Well


1 observed
1 with recharge (T=50000)
1 without recharge (T=90000)


10 20 30 40


0.25



0.2



0.15



0.1



0.05



0


Figure 4-31. Curve matching results at Well 1 for the March 2003 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference









































I Precipitation (m)
I Calculated recharge (m)


0 10 20 30 40 50 60 70 80

Day


0.25


0.2 ~


0.15


0.11


0.05


I


I I


....I .I


Figure 4-32. Curve matching results at Well 4 for the September 2004 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference














































I Precipitation (m)
I Calculated recharge (m)


- Well 6 observed
- Well 6 with recharge (T=3000)
- Well 6 without recharge (T=12000)


0 10 20 30 40 50 60 70 80


0.25



0.2



0.15



0.1



0.05


1.


I I


...r .L...


Figure 4-33. Curve matching results at Well 6 for the September 2004 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference


.M a. h














































I Precipitation (m)
I Calculated recharge (m)


1

0.5









0.25 -i



0.2



0.15



0.1



0.05


10 20 30 40 50 60 70 80

Day


I


I I


.... r a


Figure 4-34. Curve matching results at Well 7 for the September 2004 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference


. M ..llh








































Precipitation (m)
Calculated recharge (m)


1.25


1


0.75


0.5


0.25


0


5 10 15 20 25


0.25


0.15


0.1


0.05


0 -


-


-


Figure 4-35. Curve matching results at Well 4 for the March 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference





























- Well 6 observed
- Well 6 with recharge (T=65010)
-Well 6 without recharge (T=30


I Precipitation (m)
I Calculated recharge (m)


1.25


1


0.75


0.5


0.25


0


5 10


15 20

day


0.25


0.15


0.1


0.05


.L


I


Figure 4-36. Curve matching results at Well 6 for the March 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference










































I Precipitation (m)
I calculated recharge (m)


1.25


1


0.75


0.5


0.25


0


5 10 15 20 25


0.25


0.15


0.1


0.05


.L


I


Figure 4-37. Curve matching results at Well 7 for the March 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference



































Precipitation (m)
Calculated Recharge (m)


1.25

1

0.75

0.5

0.25

0


10 20 30 40


0.25


0.15


0.1


0.05


1


II.


Figure 4-3 8. Curve matching results at Well 4 for the December 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference



































Precipitation (m)
Calculated Recharge (m)


1.25

1

0.75

0.5

0.25

0


10 20 30 40


Day


0.25


0.15


0.1


0.05


1


II.


Figure 4-39. Curve matching results at Well 6 for the December 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference























Well 7 observed
Well 7 with recharge (T=125000)
-Well 7 without recharge (T=62500)

10 20 30 40


Precipitation (m)
Calculated Recharge (m)


1.25

1

0.75

0.5

0.25

0


Day


0.25


0.15


0.1


-.1


0.05,


1LI1Lj


Figure 4-40. Curve matching results at Well 7 for the December 2005 storm event using the 1D
model with measured precipitation and calculated diffuse recharge for reference






















S11.5






10.5-



10
3/2/03 3/1 6/0 3 3/30/03 4/1 3/03 4/2 7/0 3 5/1 1/03 5/25/03

date


Figure 4-41. Recession curve breakdown at Well 4 for the hydrograph produced by the March
2003 storm event





13.5

13 JZ


12.5

12

11.5

11


10
9/27/04 10/7/04 10/17/04 10/27/04 11/6/04 11/16/04 11/26/04

date


Figure 4-42. Recession curve breakdown at Well 4 for the hydrograph produced by the
September 2004 storm event



114






















10.8

10.6

10.4


10.2 -~

10

9.8
6/30/05 7/30/05 8/30/05 9/30/05 10/30/05

date



Figure 4-43. Recession curve breakdown at Well 4 for the hydrograph produced by the March
2005 storm event





11.4


10.2


10
2/5/06


2/19/06 3/5/06 3/19/06 4/2/06 4/16/06 4/30/06 5/14/06

date


Figure 4-44. Recession curve breakdown at Well 4 for the hydrograph produced by the
December 2005 storm event
















10.8



10.6



10.4



10.2


10



9.8
3/12/06 3/26/06 4/9/06 4/23/06 5/7/06 5/21/06 6/4/06 6/18/06

Date



Figure 4-45. Example of using two portions of the recession curve that results from the
December 2005 event at Well 6 to calculate basin area












Table 4-1. Annual precipitation, annual evapotranspiration, and annual calculated recharge from
2002 through 2006
Year Precipitation Evapotranspiration Diffuse Recharge


(cm)
124.7
142.1
187.4
132.7
84.1


(cm) j
81.6
65.9
81.0
84.8
66.7


38.8
81.2
102.3
47.8
17.8


2002
2003
2004
2005
2006


Table 4-2. Precipitation, potential evapotranspiration, and calculated recharge for three large
storm events
Storm Event Dates Precipitation Evapotranspiain DfueRcag


(c m)
31.5
44.5
15.4
28.6


(c m)
3.3
6.2
8.4


(c m)
26.9
41.0
9.6
24.2


2/5/2003
9/2/2004
3/24/2005 -
12/14/2005


3/26/2003
9/25/2004
-4/21/2005
- 1/24/2006


Table 4-3. Chloride concentrations in
Sample Date Cl- concentration
Well 4a (mg/L)
4/11/2006 11.7
6/15/2006 12.1
7/12/2006 10.0
8/28/2006 12.2
10/12/2006 12.0
1/17/2007 12.0
Average 11.7


shallow wells from various sampling trips
Cl~ concentration Cl~ concentration Cl~ concentration
Well 5a (mg/L) Well 6a (mg/L) Well 7a (mg/L)
12.8 6.5 9.7
12.7 6.2 9.2
11.0 6.0 7.0
12.0 7.4 8.0
12.0 8.0 8.0
12.0 7.0 7.0
12.1 6.9 8.2


Table 4-4. Recharge results for 2006 using chloride concentration factors
Well 4a Well 5a Well 6a Well 7a
Concentration Factor .914 .917 .854 .877
Evapotranspiration (cm) 76.9 77.1 71.8 73.8
Recharge (cm) 7.2 7.0 12.3 10.3












Table 4-5. Calculated diffuse recharge for two non-conduit influenced events using the water budget and water table fluctuation
methods


Precipitation
(cm)
12.3
12.3
12.3


Calculated
ET (cm)
0.4
0.4
0.4


Amount that enters
soil moisture (cm)
8.7
8.7
8.7


Recharge water budget
method (cm)
3.2
3.2
3.2


Change in
head (cm)
2.2
2.8
7.4


Specific
yield
0.2
0.2
0.2


Recharge water table
fluctuation (cm)
0.4
0.6
1.5


Well #
5a
6a
7a


Date
6/12/2006 6/14/2006
6/12/2006 6/14/2006
6/12/2006 6/14/2006


7/22/2006 7/25/2006
7/22/2006 7/25/2006
7/22/2006 7/25/2006


2.2 0.2
1.9 0.2
2.4 0.2











Table 4-6. Summary of calculated transmissivities (m2/day) with different methods in the Santa Fe Sink-Rise System. Root mean
square errors associated with model simulations are included. A value of (-) indicates no data was available for
calculations. Results for March 2003 analytical method are from Martin (2003).
Storm event Well # Distance Pinder et al. Pinder et al. 1D model Root mean 1D model Root mean Recession
to well (1969) (1969) without square error with square error curve
(m) analytical analytical recharge in for model recharge in for model w/ analysis
time lag amplitude MODFLOW w/o recharge MODFLOW recharge
match match
March 2003 1 470 97,000 97,000 90,000 -50,000
March 2003 2 1050 ------1,898
March 2003 4 110 950 950 450 0.060 150 0.173 2,646
March 2003 6 140 900 -1,300 -250
March 2003 Tower 3750 120,000 550,000 330,000 0.069 60,000 0.187

September 2004 4 110 800 10,000 9,000 0.330 3,000 0.205 2,102
September 2004 6 140 1,200 16,000 12,000 0.432 3,000 0.256 2,233
September 2004 7 1020 150,000 350,000 300,000 0.176 300,000 0.356 2,013

March 2005 2 1050 ------991
March 2005 4 110 500 800 900 0.124 450 0.090 1,234
March 2005 6 140 600 800 1,300 0.123 650 0.089 966
March 2005 7 1020 25,000 26,000 30,500 0.056 13,500 0.076 616

December 2005 1 470 ------1,635
December 2005 2 1050 ------1,035
December 2005 4 110 700 2,200 2,500 0.274 200 0.114 1,211
December 2005 6 140 1,100 2,200 4,000 0.183 1,000 0.062 1,011
December 2005 7 1020 40,000 40,000 62,500 0.074 125,000 0.266 1,034












Table 4.7. Sensitivity of 1D model results to specific yield


Sy = .1


Sy = .2


Sy = .3


Best match
Transmissivity (m2/day)


root mean Best match
square error Transmissivity (m2/day)


root mean Best match
square error Transmissivity (m2/day)


root mean
square error


March 2003
Well 4
Tower Well


450
540,000



9,000
7,500
1,500,000


90
130
675



1,600
10,000
1,500,000


0.472
0.603



0.324
0.338
0.440



0.140
0.143
0.137



0.262
0.244
0.323


150
60,000



3,000
3,000
300,000


450
650
13,500


200
1,000
125,000


0.173
0.187



0.205
0.256
0.356



0.090
0.089
0.076



0.114
0.062
0.266


75
195,000



4,500
6,000
300,000


900
1,300
27,000


600
1,100
37,500


0.105
0.115



0.216
0.305
0.270



0.097
0.096
0.061



0.158
0.062
0.157


September 2004
Well 4
Well 6
Well 7

March 2005
Well 4
Well 6
Well 7

December 2005
Well 4
Well 6
Well 7





Table 4-8. Area calculations using recession curves
Well # of Head Head
Date range # Days (max) (min)
10/17/05 -11/19/05 1 33 10.40 10.18
10/12/05 -11/20/05 2 40 10.52 10.28
10/12/05 -11/25/05 4 45 10.42 9.99
10/12/05 -11/25/05 6 45 10.36 10.07
10/12/05 -11/25/05 7 45 10.20 9.96


Slope
(Ah/day)
0.0067
0.006
0.0096
0.0064
0.0053

0.0086
0.0105
0.0095
0.0116
0.0092

0.0065
0.0063
0.0067
0.0057


Average Q
(m3/day)
5.99 x 10"
5.84 x 10"
5.86 x 10"
5.86 x 10"
5.86 x 10"

6.92 x 10"
6.98 x 10"
6.98 x 10"
6. 98 x 10"
6. 98 x 10"

6. 80 x 10"
6.77 x 10"
7.05 x 10"
6.80 x 10"


Average =


Average Q (-0.09m)
COrrection (m3/day)
4.42 x 10"
4.24 x 10"
4.24 x 10"
4.24 x 10"
4.24 x 10"


Area (m2)
4.50 x 10
4.87 x 10
4.55 x 10
3.07 x 10
5.49 x 10

4.00 x 10
3.31 x10
3.02 x 10
3.69 x 10
3.79 x 10

5.22 x 10
5.29 x 10
5.36 x 10
6.02 x 10

4.44 x 10


Area (m2)
3.31 x108
3.53 x 108
3.29 x 108
2.21 x 108
3.97 x 108

3.04 x 10"
2.52 x 10"
2.29 x 10"
2.80 x 10"
2.88 x 10"

4.03 x 10"
4.15 x 10"
4.14 x 10"
4.65 x 10"

3.35 x 10s


ah (m)
0.22
0.24
0.43
0.29
0.24


5.24 x 10"
5.31 x10"
5.31 x10"
5.31 x10"
5.31 x10"

5.26 x 10"
5.53 x 10"
5.23 x 10"
5.26 x 10"


3/22/06 4/12/06
3/17/06 4/23/06
3/17/06 4/23/06
3/17/06 4/23/06
3/17/06 4/23/06

4/23/06 5/15/06
4/23/06 5/11/06
4/23/06 6/9/06
4/23/06 5/15/06


10.50
10.82
10.59
10.63
10.43

10.42
10.23
10.19
10.08


10.31
10.42
10.23
10.19
10.08

10.27
10.11
9.87
9.95


0.19
0.40
0.36
0.44
0.35

0.15
0.12
0.32
0.13


m2 Average = m2









CHAPTER 5
DISCUSSION

Applicability of Different Recharge Calculation Methods and Effect of Conduit Boundaries

Calculated recharge values for the small June 2006 event using the water table fluctuation

method at wells Sa, 6a, and 7a are 0.4 cm, 0.6 cm, and 1.5 cm respectively. Water table

fluctuation calculations at all three locations predict less recharge than the 3.2 cm calculated with

the water budget method. These results may have occurred if the water budget method is

overpredicting the amount of water that actually makes it to the water table. In a highly karstic

aquifer, it is possible that water making it past the root zone and soil moisture storage may flow

laterally in the unsaturated zone before recharging the water table. If water were to flow laterally

at a similar time scale to which recharge enters the vadose zone, then perhaps we are not seeing

the full recharge pulse in the water table fluctuation at the wells. Instead the water will flow

towards the conduit location and recharge at some point along the way.

It is also possible that since the volume of recharge is small, the loss of water to the

conduit may occur in the saturated zone on a similar time scale as the recharge is raising the

water table. The effect that fluctuations of water level in the conduit boundary has on the water

level change at matrix monitoring wells has been difficult to assess. The most apparent effect

from the conduit boundary comes after significant changes in conduit water level related to large

storm events. It is during these times that the water table fluctuation method is least applicable;

this is because it does not differentiate between sources of water and therefore during large storm

events the water table method would overestimate the amount of diffuse recharge. However

during small events, where the changes in head at both the wells and the boundary are small, the

effect of the conduit is less obvious.









In the case of the June 2006 small event, the River Rise water level does increase very

slightly (~0.005 m), suggesting that there may be some conduit influence contributing to the

fluctuation of the water table. To test the conduit influence, the analytical solution was used by

substituting the 0.005 m change in head at the conduit over a three day period and observing if

this can cause changes in head in the observation wells. Transmissivity values determined from

model results at the different wells were used for their respective well locations. The calculated

effect of the conduit is less than .00001 m at all three observation well locations, which is

significantly less than the change in head observed at these locations during the small event.

Therefore, we are confident that the change at the observation well during the small events is

primarily related to diffuse recharge arriving at the water table.

The change in head at the observation well during the small events may actually be

underpredicted if water is flowing away from the well towards the conduit at the same rate as

water is arriving at the water table. We can assess this situation in the same way as we

approached the effect of the conduit on the wells. An average head change of 0.022 m was

observed at shallow Wells 4 and 6 during these two storm events. Applying this head change

over the same three day period, and using a transmissivity value of 500 m2/day, the distance the

conduit would need to be away from the well before it would stop having an effect is ~200 m.

This result suggests that Wells 4 and 6 can potentially be losing some amount of recharge to

lateral flow towards the conduit. This would explain why the water table fluctuation method

predicts less recharge during the small events when compared with the water budget method.

Risser et al (2005) discussed the effect of proximity to a stream boundary on the water

table fluctuation method applied in a small watershed of eastern Pennsylvania. Risser et al

(2005) used a 1D model in MODFLOW to simulate changes in head at observation wells as a









result of diffuse recharge. Three observation wells were simulated which are located at three

different distances away from a stream boundary. The result was that the observation well

furthest from the stream boundary produced the greatest change in head after the diffuse recharge

was applied. The observation well that was closest to the stream boundary produced a head

change that was only 20% of what was observed at the furthest well. Results from the study by

Risser et al (2005) and from this study indicate that the accuracy of the water table fluctuation

method will be compromised in regions where hydrologic boundaries may influence the results.

The small event that occurred in July of 2006 was similar to the June event in that it

lasted only four days and the soil moisture storage was minimal at the beginning of the storm.

However, in contrast to the June 2006 event, we see the opposite results for the two recharge

methods applied to the July 2006 event. For this storm, the water budget method predicts zero

recharge, while the water table fluctuation predicts a small amount of recharge (<0.5 cm) at all of

the well locations. We can be confident that water is arriving to the water table, based on the

increase in the well hydrograph. We can also be confident that water is not arriving from the

conduit because there is no increase in the River Rise water level. A possible explanation for

why the water budget method does not predict any recharge is that the soil moisture storage

capacity may have been over predicted. When the soil moisture has reached its wilting point (i.e.

the point when soil moisture content is so low that the surface tension of the soil-water interface

exceeds the osmotic pressure of the roots and water will no longer enter the roots (Fetter, 1994)),

then the estimated 10 cm of precipitation is needed before any recharge can occur. The July 2006

event resulted in a little less than 8 cm of recharge, thus was not able to account for the soil

moisture storage. If the storage capacity is actually closer to 8 cm, then we would see the small

recharge observed in the water level. Another possible explanation to consider is that since the









amount of recharge calculated with the water table fluctuation is so small, it is possible that a

small amount of water might bypass the soil moisture storage. This could happen if some

recharge were to hit a fast flow path (fracture) which allows it to move quickly through the

unsaturated zone, and thus avoiding the soil moisture storage. This amount of water could be

affected by the intensity of the rainfall. For example, if all the rain falls within a ten minute

period, it is more likely that some might bypass the storage as opposed to a slow and steady

rainfall over the course of a three day period. Because precipitation was recorded on a daily basis

during this study, the differences in intensity can not be evaluated.

In the case of these small summer events, it is difficult to assess the accuracy of the water

table fluctuation method. If the conduit is truly not contributing any water, as demonstrated by

the receding trend in water level at the River Rise (Figure 4-10) and the calculations of conduit

effect, then we know that the slight change in water level must be a result of water arriving from

the diffuse component. If water is arriving from any other sources, then the diffuse recharge

portion of the head change will be over predicted using the water table fluctuation. However, if

water is flowing away from the well location to the conduit at the same rate that it is recharging

the water table, then the water table fluctuation results may be an under estimate. Fluctuations in

the water table can be potentially valuable in this system for seeing the timing of recharge;

however, a calculation of the effect of a nearby boundary (such as a stream or a conduit) should

be considered before applying the water table fluctuation method to estimate volumes of diffuse

recharge. In this study area, there are no wells that are far enough away from a boundary

condition that we can be totally confident in the water table fluctuation method.

The chloride method provides an alternative long term method for quantifying recharge.

This method is valuable in systems where there is only once source of chloride and where










evapotranspiration is abundant. This method works well in dry years (e.g. 2006) in the

unconfined Santa Fe River basin because diffuse recharge and the conduit component are not

constantly diluting the chlorine concentrations. Since the concentrations are dependent on the

precipitation and evapotranspiration, the more measurements of chloride concentrations that can

be made will result in the most accurate calculation of recharge. The chloride concentration is

also good as a method to determine the long term differences in recharge at the various well

locations.

Daily Versus Monthly Recharge and Evapotranspiration Calculations

Because precipitation records are readily available and usually easily accessible, they are

often used as a substitute to approximate recharge, rather than calculating recharge on a daily

basis. Daily recharge calculations require determining soil moisture storage and adjusting for

daily evapotranspiration, and thus require significantly more time and effort. However, results

from this study indicate that in this study area daily precipitation (Figure 4-1) and daily recharge

records (Figure 4-7) differ tremendously. It is not uncommon for an afternoon thunderstorm in

the summer to supply more than 5 cm of precipitation, yet none of that water is calculated to

become recharge. For example, this happens after the small storm in July of 2006 when there

was 8 cm of precipitation, but no calculated recharge. The recharge does not exactly mimic

precipitation because so much water is either stored in soil moisture, or is evapotranspired back

to the atmosphere; this result would be expected in any region with a high soil moisture capacity

and high evapotranspiration rates.

The recharge values calculated on a daily basis were especially important to this study

because they were calculated on the same time scale as the 1-dimensional model. Daily recharge

calculations allowed for calculation of recharge for any time-step of the model. If the recharge









could be calculated on even a finer time scale, we could reduce the time step of the model, and

this may improve the calculated transmissivity results.

Potential evapotranspiration is commonly calculated on a monthly basis to emphasize the

seasonal differences throughout a year. This is a relatively easy calculation with very few inputs

using the standard Thornthwaite method (Schwartz and Zhang, 2003). In this study,

evapotranspiration was calculated on a daily basis with the much more detailed Penman-

Monteith equation. The monthly calculated values of evapotranspiration tend to be consistent

with the summed daily calculations (Figures 4-2 and 4-3). Although this Penman- Monteith

method requires more data input and some assumptions, the daily evapotranspiration calculations

are very valuable. For example, more than half of the precipitation that falls in this region every

year is lost to evapotranspiration, and therefore daily values of evapotranspiration are necessary

to produce an accurate daily recharge calculation. Daily values can also be used to indicate dry

times in the record, as can be seen by the periods of zero calculated evapotranspiration in figure

4-2. In this study area, this commonly happens during the month of May, when there is not

enough precipitation to produce recharge.

Transmissivity

Influence of Diffuse Recharge on Storm Event Hydrographs and Transmissivity
Calculations

Comparing results of the 1D model with and without recharge showed that including

diffuse recharge in the solution will significantly change the predicted transmissivity. These

results suggest that transmissivity may be overpredicted with the methods that do not include the

diffuse recharge. This happens because an artificially high transmissivity is needed to create a

good match in methods that disregard diffuse input. This was the case in Martin et al. (2006),

where transmissivities were calculated using only the Pinder et al. (1969) analytical solution









which disregards the diffuse component. Calculated transmissivity values from the same storm

were three times more when calculated with the 1D model without recharge. This suggests that

choosing a method which does not include recharge is a crucial omission when using a large

scale natural event to estimate transmissivity through passive monitoring. In local karst regions,

dissolution of the landscape leads to high susceptibility to pollution via groundwater recharge.

Overpredicted transmissivity values suggest faster movement of pollutants through the system

than what may actually be observed.

Although the inclusion of diffuse recharge changes the calculated transmissivity values, it

is important to point out the difficulties that go along with including the recharge. Three of the

four model simulations including the diffuse recharge component result in predicted curves that

are unrealistically pointed and jagged when compared to the smooth curves that are produced in

the observed data (Figures 27-30, 35-40) This effect is most noticeable during the small

December 2005 event, and is not seen in the large September 2004 event. Assuming the total

amount of calculated recharge per storm event is correct, then the jagged predicted curves may

be displaying a lag time effect. This happens if the full amount of the calculated recharge per

day did not fully reach the well during that same calculated day. Rather, the recharge actually

reaches the aquifer slowly over the course of a few days after the rain event. To correct for this in

the model, recharge would have to be spread across the time steps in varying amounts.

Determining how to spread the recharge across the model is difficult because the excess

precipitation calculation can not be refined to better than a daily time scale. It is also difficult

because the influence that the diffuse recharge has on the hydrograph varies per storm event. For

example, the j aggedness of the predicted curves during the March 2005 storm event is much

more apparent than during the September 2004 storm event. This may be a result of most of the









recharge (>30 cm) occurring over a short two day period. In contrast, the other storms have

recharge occur sporadically throughout the course of the event. The heavily focused recharge

coincides with the large storm pulse arriving via the conduit during the September 2004 event,

and results in a quick, sharp increase in the storm hydrograph (3.5 m in 10 days). This helps

mask the effect of the diffuse recharge component, as a smooth hydrograph is predicted. The

time lag of diffuse recharge to the water table is influenced by the location of the water table as

compared to the surface during the various storm events. During larger events like in September

2004, the water table came extremely close to the surface (Table 5-1), and thus it is expected that

recharge occurs quickly. In smaller events like March 2005 and December 2005 the water table

starts low and does not increase too far into the surficial sediment (Table 5-1); the greater depths

to the water table would result in greater amount of time for recharge to reach the water table as

compared to the September event. The time lag of the diffuse recharge is also affected by the

storage that occurs in the unsaturated zone. The ability to temporarily store more water in this

area will increase the time it takes for recharge to reach the water table.

Choosing one pattern of water table response to rainfall would also be difficult in this

region because of the varying flow that may occur in the unsaturated zone produced by different

magnitude storms. When soil is not saturated, soil moisture flows downward by gravity flow

through interconnected pores (Fetter, 1994). With increasing water content, more pores will fill

and the rate of downward water movement increases (Fetter, 1994). Because the vertical

unsaturated hydraulic conductivity is not constant, the rate of flow will be affected by the amount

of diffuse recharge entering the unsaturated zone. This amount varies with intensity of the storm.

Influence of Scale Effects on Transmissivity Calculations

Both the analytical method without recharge and the 1D model simulations show the effect

that scale has on the transmissivity calculation. It has been noted in previous work on karst










aquifers (e.g., Bradbury and Muldoon, 1990; Rovey, 1994; Martin et al, 2006) that when

increasing scale, preferential pathways tend to dominate a larger percentage of groundwater

flow, thus increasing transmissivity. This seems to be the case around the Santa Fe Sink-Rise

system, where transmissivity to the wells furthest from the mapped or inferred conduit (Well 1,

Tower Well, and possibly Well 7) are two to three orders of magnitude greater than

transmissivity of wells near the mapped conduit (Well 4 and Well 6). The scale effect is seen

with all four storm events since 2003. The fact that transmissivity increases so much with

increasing distance away from the river supports the idea that the preferential flow paths exist,

and thus the karstic limestone of the Floridan aquifer is highly heterogeneous in nature.

In contrast to both of the curve matching techniques, the recession curve analysis does not

show the scaling effect. However, the recession curve analysis results should be interpreted in a

different manner than results produced by the two curve matching methods. The recession

analysis method yields insight into the transmissivity of the intergranular, matrix rock. This does

not include the conduit and fracture zones that make up the preferential flow paths in the highly

karstic parts of the Floridan aquifer. This is most obvious for locations at a far distance away

from the conduit (Well 1, Well 2, and Well 7) where the recession curve transmissivities are

significantly lower than those calculated by the analytical solution and model (Table 4-6).

Transmissivity calculated with the recession curve analysis at well locations that are close to the

conduit (Well 4 and Well 6) yield very similar results to those calculated with the other methods.

This is most likely because the amount of preferential flow paths encountered in the short

distance between the wells and conduit is much less than over the whole formation, so therefore

the transmissivity of the formation between these points mimic the matrix transmissivity around

the wells. These results suggest that the analytical solution without recharge and 1-dimensional









model with recharge are more useful for predicting how the system will respond to a storm pulse

arriving from the conduit location, while the recession curve analysis is better for understanding

how a particular well responds after the storm pulse. For example, if the goal was to understand

how a contaminant will move from the conduit to the matrix during a storm event, then one

would be more concerned with the understanding the maximum transmissivity of the whole rock

formation. In contrast, if a contaminant was dumped at a point source (e.g. near an observation

well), one might be concerned with how that contaminant will move near the region in which it

was dumped.

Prior to this study, air permeability tests were performed on cores recovered from Well 5 in

the O'Leno State Park study area. In regions where at least 50% of the core was recovered,

permeabilities averaged about .4 darcys, which is equivalent to a hydraulic conductivity of about

.297 m/day (Moore and Florea, unpublished data). Slug tests were also preformed in this study

area at well locations 3, 4, 5, 6 and 7 by Hamilton (unpublished data) and Langston (unpublished

data). Slug test results analyzed using the Bouwer and Rice Slug Test Method yielded average

hydraulic conductivity values between 0.864 m/day and 3.89 m/day at the six well locations

(Hamilton, unpublished data). Average hydraulic conductivity values from this study ranged

from 1.5 m/day at Well 4 during the March 2003 event to 3000 m/day at Well 7 during the

September 2004 event. These values are greater than those calculated from permeability and slug

tests, which is a common result in hydraulic conductivity determined from laboratory and well

tests as compared to those determined from passive monitoring. These values further support the

scale effect in this region.

Influence of Storm Event Magnitude on Transmissivity Calculations

The September 2004 storm event was the largest event recorded during this study, with

head changes greater than 3.5 m at the monitoring wells. Calculated transmissivity values from









the September 2004 storm pulse are significantly greater than calculated transmissivity values

from the other three storm events. For example, transmissivity values from Wells 4 and 6 using

the September event were 3000 m2/day, while calculated transmissivity did not exceed 1000

m2/day from the other storm events. Transmissivity for the September 2004 event at Well 7 was

300,000 m2/day, which is about 20 times greater than the transmissivity of 13,500 m2/day

calculated from Well 7 during the March 2005 event.

In an unconfined aquifer, the level of saturation rises or falls with the amount of water that

is in storage (Fetter, 1994). One possible explanation for the higher calculated transmissivity

values is that the amount of preferential flow paths encountered by the storm pulse is increasing

as the saturated thickness of the unconfined aquifer increases into the surficial sediments during

a storm event. This can also happen if the permeability of surficial sediments in this region are

higher than the limestone. Based on descriptive analysis of a core taken during the installation of

well location 5, surficial sediment to depths of about 2 m are lightly consolidated fine to coarse

sands with interspersed chunks of gravel. During the September 2004 storm event, water levels

were less than 2 m below the surface on average at wells 4, 6, and 7, and were as little as 0.1 m

from the surface at Well 6 specifically (Table 5-1). This indicates that water was almost

completely saturating the surficial sediments in some areas. In contrast, water levels at Wells 4,

6, and 7 were greater than 3 m below the surface on average during the March 2003 event and

greater than 4 m below the surface during the other two storm events (Table 5-1). Since the

boundary between the limestone bedrock and surficial sediments lies at about 3 4 m below the

surface on average, during the small events the groundwater levels rarely reach above the

limestone. If the saturated thickness of the aquifer increases above the limestone/surficial









sediment boundary, then the calculated transmissivity may increase as a result of flow through

this zone of higher porosity.

If it is true that transmissivity can change relative to the increasing storm size, then

determining the maximum transmissivity of this system requires a storm event that fully

saturates the system. The storm event in September 2004 is represents this scenario, as water

levels were at or near the surface. One interesting point to note is that overland flow was

observed in some areas at the O'Leno field site during the September 2004 event (Moore,

personal communication). Overland flow might induce an artificially high calculation of

transmissivity if water were to flow faster overland and infiltrate downward into the wells. This

happens because overland flow decreases the distance from the conduit to the monitoring wells.

Overland flow may have occurred near Well 6 during this storm, as can be seen by the water

level being extremely close to the surface. However, overland flow is not likely at Wells 4 and 7,

which still had water levels about 2 m below the surface at the peak of the event. Overland flow

could potentially be a possible source of error in any passive monitoring calculation of

transmissivity, and should be noted during large events where it may occur.

Choosing the Appropriate Method to Calculate Transmissivity

Three methods were used to calculate transmissivity throughout this study. The most

important differences between the three methods are whether the recharge is included in the

method and the type of transmissivity that is being calculated (i.e. matrix versus whole rock

formation). Because of these differences, comparison between methods should be approached

with caution.

As previously discussed, the 1-dimensional model without recharge and the Pinder et al.

(1969) analytical method without recharge essentially produce the same results (Figure 4-27).

The only real difference between the two methods in this study is the way in which curve









matching was performed. With the analytical method, curves are matched visually by time-lag or

by amplitude. In the model, the curves are calibrated numerically from deviation between

predicted and observed curves via root mean square error. The difference that this has on the

calculation of transmissivity in some cases can be significant (Table 4-6). The model is most

likely more accurate because the best fit curve is determined numerically as opposed to making a

visual estimate. Methods like these that neglect the diffuse recharge component are adequate to

use in regions where little diffuse recharge occurs, such as a confined karst aquifer.

The 1-dimensional model with recharge is useful in regions similar to the unconfined Santa

Fe River basin, where diffuse recharge is significant. The model used in this study maximizes the

effect of incorporating recharge by assuming that all of the calculated excess precipitation arrives

as recharge to the water table on the same day it was calculated for. Results show that

disregarding the diffuse component in such regions can lead to overpredicted transmissivity

values. Creation of the very basic 1-dimensional numerical model used in this study is a similar

amount of work compared to the computing iterations involved with solving the analytical

solution. Therefore, it would be hard to argue that the analytical solution is any easier to apply.

Also, the modeling allows for simple calibration of various parameters with numerical based

sensitivity analysis of those parameters.

Diffuse Recharge Component and Possible Implications for Dissolution

An important part of understanding the potential karstification of the unconfined Floridan

aquifer region in response to storm events is being able to accurately make water budget

calculations. This requires quantification of the volumetric inputs and outputs of the system;

most importantly the volume contributed by diffuse recharge in comparison to the volume of

water received from the conduit. Previously, quantifying the diffuse recharge input had been a

challenge in this study area because the basin area contributing to the system had been difficult









to assess. However, using the area estimates made from recession curves in this study (Table 4-

8) we now have the ability to make some comparisons between these two components. Possible

sources of error to the overall budget include if there are other inputs and outputs to the system.

Besides the Sink and the diffuse recharge, a third possible input to the system may be from a

feeder conduit that has been mapped to the east of the O'Leno State Park study area (Figure 2-2).

Other possible inputs include at Vinzant' s landing just north of the River Sink, and any recharge

that might occur to the confined Floridan aquifer. Other possible outputs include springs that are

located downstream of the River Rise.

Screaton et al. (2004) discuss possible local karstification in the Santa Fe River system

resulting from inflow of conduit water to the aquifer after a storm event. A small storm event

from late September 2001 was used (Figure 4-4), which produced a little more than seven cm of

precipitation over 16 days. The amount of water lost to the matrix during this time was 2.2 x 106

m3, calculated from the difference between River Sink and River Rise daily discharges. The

degree of undersaturation was estimated from chemistry data collected in 1998 at the River Sink,

for a stage similar to the stage recorded during the 2001 event. Because the residence time of the

conduit water in the matrix was not known, the calculation was simplified by assuming all the

water lost from the conduit remains in the matrix rock sufficient time to reach equilibrium. It was

calculated that the volume of rock needed to dissolve in order to bring the calcium concentration

back to equilibrium was 1.06 x 104 m3. An estimated area of dissolution surrounding the conduit

of 3.0 x 10' m2 was used, which implied regional dissolution of 4.6 x 10-4 m. This regional

dissolution rate was an order of magnitude greater than rates of 2.63 x 10-5 m reported by

Opdyke et al (1984) from measurements of dissolved calcium in Florida's springs.









The calculations of possible dissolution in Screaton et al. (2004) do not account for

dissolution occurring from diffuse recharge. Using the calculated basin area of 4.44 x 10s m2

from above, we can now calculate the total volume of water that enters the matrix via diffuse

recharge. For the same 2001 storm event, assuming that precipitation recharges evenly across the

entire basin area, this volume would be equal to 3.20 x 10' m3. If We USe the low end of the

calculated basin area of 3.3 5 x 108 m2, than this volume of water would be equal to 2.3 8 x 107

m3. These calculated volumes of diffuse recharge are an order of magnitude higher than the

volume contributed to the system by the conduit. Determining the dissolution caused by this

volume will be possible once chemistry data from precipitation in the study area is determined.

However, the fact that the total volume of water input from the diffuse recharge during this small

event is much larger than the conduit input, suggests that dissolution will actually be greater than

calculated by Screaton et al. (2004).

With the addition of more storm event data since the original calculations in Screaton et

al. (2004) were performed for the 2001 event, it is now possible to make a similar calculation of

dissolution for a larger event. The March 2003 event discussed in Martin et al (2006) had a head

increase of near three meters, and a full record of daily discharge values. During the 19 days of

this storm event, the conduit lost 2.83 x 10' m3 Of walter to the matrix based on the Sink and Rise

daily discharges. This volume is an order of magnitude higher than the amount of water lost to

the matrix during the smaller 2001 event. Using the same chemical data from the River Sink

collected in 1998 (Screaton et al, 2004), .3 moles of calcite per liter of water is needed to

dissolve in order to reach equilibrium at a River Sink stage of 13.43 meters. This would equal a

dissolved rock volume of 3.13 x 10' m3 fTOm the conduit component. These results can be

converted to regional denudation rates using the same estimated area of 3.0 x 107 m2. The









regional denudation resulting from this storm would be 1.04 x 10-2 m, Which is two to three

orders of magnitude higher than the original reports of 4 x 10-5 m/yr by Opdyke et. al (1984) and

4.60 x 10-4 m by Screaton et al (2004) for the smaller 2001 event.

During this same large 2003 storm event there was 12 cm of diffuse recharge. This would

equal a total diffuse recharge input of 5.33 x 107 m3 Of walter into the matrix using an area of 4.44

x 10s m2, and an diffuse recharge input of 4.02 x 107 m3 USing an area of 3.3 5 x 10s m2. Even if

we assume the lower calculated value for diffuse input is correct, this is still much greater than

the total amount contributed by the conduit during the storm. This would increase the estimated

dissolution over the entire basin area.











Table 5-1. Elevations of the ground surface, top of limestone bedrock, and water table before and at the peak of various large storm
events at different well locations (in meters above sea level)
Water Table Water Table Water Table Water Table at Water Table Water Table Water Table Water Table at
before at peak of before peak of before at peak of before peak of
Well limestone March 2003 March 2003 September 2004 September 2004 March 2005 March 2005 December 2005 December 2005
# Surface bedrock event event event event event event event event
1 14.45 -2.62 10.17 13.81 10.50 13.25 10.25 12.10
4 17.89 13.32 10.17 12.23 9.85 13.13 10.56 11.89 9.94 11.28
6 13.51 8.64 10.11 13.38 10.57 11.89 10.07 11.28
7 15.22 9.73 9.97 13.05 10.41 11.46 9.98 10.95









CHAPTER 6
SUMMARY

Understanding aquifer parameters in karst regions is an important concern

because of the reliance on these regions for potable water and because of the possibility

for contamination across these landscapes. There are many methods available to evaluate

aquifer parameters, including laboratory tests and aquifer tests. However, these methods

are difficult to apply in karst regions because the small scale often neglects to incorporate

the heterogeneous nature of the whole rock formation. In this study, passive monitoring

of water levels was used as a method to evaluate transmissivity between a conduit

location and observation wells at varying distances away from the conduit.

The Pinder et al. (1969) analytical solution was first applied in this study area by

Martin et al. (2006) to storm event data from 2003. This method predicts water levels at

observation wells based on response in a nearby boundary (conduit) location, but does

not include a recharge component in the solution. Results indicated that transmissivity

values range based on vicinity to the conduit location, with wells at a further distance

away having the highest calculated transmissivity values. These scale effects were

consistent with results in this study using the Pinder et al. (1969) analytical method

applied to three more storm events occurring after 2003.

Three methods including a water budget, chloride concentrations, and water table

fluctuation were used to calculate diffuse recharge in the study area. Results indicate that

diffuse recharge is quantifiable and especially significant during large storm events, and

thus should not be neglected in the transmissivity calculation. A 1-dimensional model in

MODFLOW was created to test the influence of diffuse recharge on the transmissivity

calculation. Like the analytical solution, the model predicts water levels in observation









wells based on change in nearby boundary conditions; however, the model also allows for

recharge to be entered for any time step. Calculated recharges using the water budget

were entered into the model on the day of occurrence. Models were calibrated to

observed data and a best fit transmissivity was determined using root mean square error

as the criteria for fit. Results indicated that neglecting to include diffuse recharge will

result in overpredicted calculated transmissivity. Results also showed some difficulties

with adding recharge to the model. These include the possibility that a lag time may exist

from when the initial portion of diffuse recharge reaches the water table to the final total

amount of diffuse recharge, and also that the effect of recharge on the observed

hydrograph may be influenced by the storm event magnitude.

Calculation of the basin area contributing to the discharge of the Santa Fe River

Basin upstream of the River Rise, in combination with quantifying the diffuse recharge,

allows for comparing the components of the water budget in the system. Results show

that the total input of diffuse recharge during a storm event, as compared to input from

the conduit boundary, will vary based on the magnitude of the storm. Diffuse recharge

should be an important part of the potential dissolution of limestone in this study area,

and should be included in future calculations when chemistry data become available.










LIST OF REFERENCES


Allen, R.G., Jensen, M.E., Wright, J.L. and Burman, R.D., 1989. Operational Estimates
of Reference Evapotranspiration. Agronomy Journal, 81: 650-662.

Anderson, M.P. and Woessner, W.W., 2002. Applied Groundwater Modeling. Academic
Press, San Diego, 381 pp.

Andres, A.S. and Klingbeil, A.D., 2006. Thickness and Transmissivity of the Unconfined
Aquifer of Eastern Sussex County, Delaware. In: D.G. Survey (Editor), Newark.

Atkinson, T.C., 1977. Diffuse Flow and Conduit Flow in Limestone Terrain in the
Mendip Hills, Somerset. Journal of Hydrology, 35: 93-110.

Batiot, C., Linan, C., Andreo, B. and Emblanch, C., 2003. Use of Total Organic Carbon
(TOC) as a Tracer of Diffuse Infiltration in a Dolomitic Karstic System: The
Nerja Cave (Andalusia, Spain). Geophysical Research Letters, 30(22).

Bradbury, K.R. and Muldoon, M.A., 1990. Hydraulic Conductivity Determinations in
Unlithified Glacial and Fluvial Materials. In: D.M. Neilson and A.I. Johnson
(Editors), Hydraulic Conductivity and Waste Contaminant Transport in Soils.
American Society for Testing and Materials, Philadelphia, pp. 138-151.

Budd, D.A. and Vacher, H.L., 2004. Matrix Permeability of the Confined Floridan
Aquifer, Florida, USA. Hydrogeology Journal, 12: 531-549.

Bush, P.W. and Johnston, R.H., 1988. Ground-Water Hydraulics, Regional Flow, and
Ground-Water Development of the Floridan Aquifer System in Florida and in
Parts of Georgia, South Carolina, and Alabama. U.S. Geological Survey
Professional Paper, 1403(C).

Dingman, S.L., 2002. Physical Hydrology. Prentice Hall, Upper Saddle River, New
Jersey, 646 pp.

Federer, C.A., Vorosmarty, C. and Fekete, B., 1996. Intercomparison of Methods for
Calculating Evaporation in Regional and Global Water Balance Models. Water
Resources Research, 32: 2315-2321.

Fetter, C.W., 1994. Applied Hydrogeology. Prentice Hall, Upper Saddle River, New
Jersey.

Florea, L.J. and Vacher, H.L., 2005. Springflow Hydrographs: Eogenetic vs. Telogenetic
Karst. Groundwater: 1-10.

Gilman, E.F., 1991. Community Tree Care. University of Floridan IFAS extension,
Gainesville, Florida.










Grubbs, J.W., 1998. Recharge Rates to the Upper Floridan Aquifer in the Suwannee
River Water Management District, Florida. In: U.S.G. Survey (Editor). Water-
Resources Investigations Report 97-4283, Tallahassee.

Healy, R.W. and Cook, P.G., 2002. Using Groundwater Levels to Estimate Recharge.
Hydrogeology Joumnal, 10: 91-109.

Hisert, R.A., 1994. A Multiple Tracer Approach to Determine the Ground and Surface
Water Relationships in the Western Santa Fe River, Columbia county, Florida.,
University of Florida, Gainesville, 212 pp.

Hunn, J.D. and Slack, L.J., 1983. Water Resources of the Santa Fe River Basin, Florida.
In: U.S.G. Survey (Editor). Water Resources Investigations Report 83-4075,
Tallahassee.

Jones, I.C., Banner, J.L. and Humphrey, J.D., 2000. Estimating Recharge in a Tropical
Karst Aquifer. Water Resources Research, 36(5): 1289-1299.

Jones, W.K., 1999. Pump Tests of Wells at the National Training Center Near
Sheperdstown, West Virginia. Karst Waters Institute Special Publication, 5: 259-
262.

Junge, C.E. and Werby, R.T., 1958. The Concentration of Chloride, Sodium, Potassium,
Calcium, and Sulfate in Rain Water over the United States. Journal of
Meteorology, 15: 417-425.

Katz, B.G., Coplen, T.B., Bullen, T.D. and Davis, J.H., 1997. Use of Chemical and
Isotopic Tracers to Characterize the Interactions between Ground Water and
Surface Water in Mantled Karst. Groundwater, 35(6): 1014-1028.

Lemeur, R., and Zhang, L., 1990. Evaluation of Three Evapotranspiration Models in
Terms of their Applicability for an Arid Region. Journal of Hydrology, 114:395-
411

Maher, K. and DePaolo, D.J., 2003. Vadose Zone Infiltration Rate at Hanford,
Washington, Inferred from Sr Isotope Measurements. Water Resources Research,
39(8).

Martin, J.B. and Dean, R.W., 1999. Temperature and a Natural Tracer of Short Residence
Times for Groundwater in Karst Aquifers. Karst Waters Institute Special
Publication, 5: 236-242.

Martin, J.B. and Dean, R.W., 2001. Exchange of Water Between Conduits and Matrix in
the Floridan Aquifer. Chemical Geology, 179: 145-165.










Martin, J.M., 2003. Quantification of the Matrix Hydraulic Conductivity in the Santa Fe
River Sink/Rise system with Implications for the Exchange of Water between the
Matrix and Conduits, University of Florida, Gainesville, 80 pp.

Martin, J.M., Screaton, E.J. and Martin, J.B., 2006. Monitoring Well Responses to Karst
Conduit Head Flucuations: Implications for Fluid Exchange and Matrix
Transmissivity in the Floridan Aquifer. Geological Society of America Special
Paper, 404.
Monteith, J.L., 1965. Evaporation and Environment. Proceedings of the 19th Symposium
of the Society for Experimental Biology. Cambridge University Press, New York,
205-233 pp.

Mylroie, J.E., 1984. Hydrologic Classification of Caves and Karst. In: R.G. LaFleur
(Editor), Groundwater as a Geomorphic Agent. Rensselaer Polytechnic Institute,
Troy, New York.

Palmer, A.N., 2002. Karst in Paleozoic Rocks: How Does it Differ from Florida? Karst
Waters Institute Special Publication, 7: 185-190.

Pinder, G.F., Bredehoeft, J.D. and JR, H.H.C., 1969. Determination of Aquifer
Diffusivity From Aquifer Response to Flucuations in River Stage. Water
Resources Research, 5(4): 850-855.

Powers, J.G. and Shevenell, L., 2000. Transmissivity Estimates from Well Hydrographs
in Karst and Fractured Aquifers. Groundwater, 38(3): 361-369.

Rasmussen, W.C. and Andreasen, G.E., 1959. Hydrologic Budget of the Beaverdam
Creek Basin, Maryland. In: U.S.G. Survey (Editor). Water Supply Paper.

Rindels, S., 1992. Tree Root Systems, www.ipm.iastate. edu/ipm/hortnews/1 992/4-1-
1992/treeroot.html. Iowa State University Department of Horticulture, Ames,
iowa, pp. 43-44.

Risser, D.W., Gburek, W.J. and Folmar, J., 2005. Comparison of Methods for Estimating
Ground-Water Recharge and Base Flow at a Small Watershed Underlain by
Fractured Bedrock in the Eastern United States. In: U.S.G. Survey (Editor). U.S.
Department of the Interior, Reston, Virginia.

Rovey, C.W. and Cherkauer, D.S., 1995. Scale Dependency of Hydraulic Conductivity
Measurements. Groundwater, 33(5): 769-780.

Rushton, K.R., Eilers, V.H.M. and Carter, R.C., 2005. Improved Soil Moisture Balance
Methodology for Recharge Estimation. Journal of Hydrology, 318: 379-399.

Schwartz, F.W. and Zhang, H., 2003. Fundamentals of Ground Water. John Wiley and
Sons, Inc. ,New York, 583 pp.











Screaton, E., Martin, J.B., Ginn, B. and Smith, L., 2004. Conduit Properties and
Karstification in the Unconfined Floridan Aquifer. Groundwater, 42(3): 338-346.

Shevenell, L., 1996. Analysis of Well Hydrographs in a Karst Aquifer: Estimates of
Specific Yields and Continuum Transmissivities. Journal ofHydrology, 174: 331-
355.

Stewart, J.B., 1988. Modelling Surface Conductance of Pine Forest. Agricultural and
Forest Meteorology, 43: 19-35.

Vacher, H.L. and Ayers, J.F., 1980. Hydrology of Small Oceanic Islands Utility of an
Estimate of Recharge Inferred from the Chloride Concentration of the Freshwater
Lenses. Journal of Hydrology, 45: 21-37.

Vacher, H.L. and Mylroie, J.E., 2002. Eogenetic Karst from the Perspective of an
Equivalent Porous Medium. Carbonates and Evaporites, 17(2): 182-196.

White, W.B., 1999. Conceptual Models for Karstic Aquifers. Karst Waters Institute
Special Publication, 5: 11-15.

White, W.B., 2002. Karst Hydrology: Recent Developments and Open Questions.
Engineering Geology, 65: 85-10.

XU, C.Y. and Chen, D., 2005. Comparison of Seven Models for Estimation of
Evapotranspiration and Groundwater Recharge Using Lysimeter Measurement
Data in Germany. Hydrological Processess, 19: 3717-3734.









BIOGRAPHICAL SKETCH

Michael Ritorto was born in Long Island, NY and lived there for 18 years. His

family includes his parents Frank Ritorto and Denise Ritorto, and his sister Brittany

Ritorto. After graduating from Commack High School in 2001, Michael moved to Ann

Arbor, Michigan to attend the University of Michigan (UM). Michael received his

bachelor' s degree in environmental geosciences from UM in December 2004. Michael

went on to pursue his master' s degree in geology at the University of Florida, starting in

September 2005. Michael is currently seeking employment in the environmental and

geological science field.





PAGE 1

IMPACTS OF DIFFUSE RECHARGE ON TRANSMISSIVITY AND WATER BUDGET CALCULATIONS IN THE UNCONFINED KA RST AQUIFER OF THE SANTA FE RIVER BASIN By MICHAEL RITORTO A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007 1

PAGE 2

2007 Michael Ritorto 2

PAGE 3

To my Grandpa Joe Ritorto 3

PAGE 4

ACKNOWLEDGMENTS I would like to thank my a dvisor, Elizabeth Screaton, for giving me the opportunity to learn from her expertise and for supporting and encouraging me along the way. I would also like to thank PJ Moore and Abby Langston for their tim e and efforts helping me complete the field work for this project. Lastly, I would like to th ank my parents for their love and support and all of my friends who kept me motivated along the way. 4

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES.........................................................................................................................8 ABSTRACT...................................................................................................................................12 1 INTRODUCTION................................................................................................................. .....14 2 BACKGROUND........................................................................................................................19 Study Area..............................................................................................................................19 Geologic Background.............................................................................................................20 Previous Investigations in the Santa Fe River Basin..............................................................21 Diffuse Recharge in Other Study Areas.................................................................................26 Aquifer Parameters in Other Study Areas..............................................................................30 3 METHODS.................................................................................................................................38 Data Collection.......................................................................................................................38 Water Levels...........................................................................................................................39 Sink and Rise Discharge.........................................................................................................41 Evapotranspiration............................................................................................................. .....44 Recharge.................................................................................................................................47 Water Budget Method.....................................................................................................47 Water Table Fluctuation Method.....................................................................................47 Chloride Concentration Method......................................................................................48 Matrix Transmissivity.......................................................................................................... ...49 Analytical Method without Recharge..............................................................................49 1-Dimensional Model with and without Recharge in MODFLOW................................50 Recession Curve Analysis...............................................................................................51 4 RESULTS...................................................................................................................................59 Precipitation and Evapotranspiration......................................................................................59 Water Levels...........................................................................................................................61 Sink and Rise Discharge.........................................................................................................62 Recharge.................................................................................................................................62 Diffuse Recharge Calculated From Water Budget..........................................................62 Diffuse Recharge during Large Storm Events.................................................................63 Diffuse Recharge Calculated From Chloride Concentrations.........................................64 Diffuse Recharge Estima ted Using the Water Table Fluctuation Method......................65 Transmissivity................................................................................................................. ........67 5

PAGE 6

Analytical Method without Recharge..............................................................................67 1-Dimensional Model with and without Recharge in MODFLOW................................69 Sensitivity to Specific Yield............................................................................................73 Recession Curve Analysis...............................................................................................74 Basin Area Calculations........................................................................................................ .75 5 DISCUSSION...........................................................................................................................122 Applicability of Different Recharge Ca lculation Methods an d Effect of Conduit Boundaries........................................................................................................................122 Daily Versus Monthly Recharge a nd Evapotranspiration Calculations...............................126 Transmissivity................................................................................................................. ......127 Influence of Diffuse Recharge on Storm Event Hydrographs and Transmissivity Calculations................................................................................................................127 Influence of Scale Effects on Transmissivity Calculations...........................................129 Influence of Storm Event Magnitude on Transmissivity Calculations.........................131 Choosing the Appropriate Method to Calculate Transmissivity...................................133 Diffuse Recharge Component and Possible Implications for Dissolution...........................134 6 SUMMARY..............................................................................................................................139 LIST OF REFERENCES.............................................................................................................141 BIOGRAPHICAL SKETCH.......................................................................................................145 6

PAGE 7

LIST OF TABLES Table page 2-1 Geologic and Hydrogeologic Units of the Santa Fe River Basin .....................................37 3-1 Summary of Monitoring Wells..........................................................................................57 3-2 Location and dates of data collection.................................................................................58 4-1 Annual precipitation, annual evapotranspira tion, and annual calculated recharge from 2002 through 2006...........................................................................................................117 4-2 Precipitation, potential evapotranspirati on, and calculated recharge for three large storm events................................................................................................................... ..117 4-3 Chloride concentrations in shallow wells from various sampling trips...........................117 4-4 Recharge results for 2006 using chloride concentration factors......................................117 4-5 Calculated diffuse recharge for two nonconduit influenced events using the water budget and water table fluctuation methods....................................................................118 4-6 Summary of calcula ted transmissivities (m2/day) with different methods in the Santa Fe Sink-Rise System........................................................................................................119 4.7 Sensitivity of 1D model results to specific yield.............................................................120 4-8 Area calculations using recession curves.........................................................................121 5-1 Elevations of the ground surface, top of limestone bedrock, and water table before and at the peak of various large storm events at differe nt well locations (in meters above sea level)............................................................................................................... .138 7

PAGE 8

LIST OF FIGURES Figure page 2-1 Location of the Santa Fe River..........................................................................................35 3River Sink rating curve produced by the Suwannee River Water Management District for the stage gauge located in OLeno State Park.................................................53 3-2 River Sink rating curve produced in this study using sink elevations measured by the data logger correlated with stage at the River Sink gauging station..................................54 3-3 River Rise rating curve produced by Sc reaton et al. (2004) us ing data from the Suwannee River Water Ma nagement District....................................................................55 3-4 Function assuming a maximum soil-moisture deficit of 10 cm used to calculate leaf conductance in the Penman-Monteith evapotranspiration method. After Stewart (1988).................................................................................................................................56 4-1 Daily precipitation records from OLeno State Park during this study. Three storm events are highlighted........................................................................................................78 4-2 Daily calculated evapotranspiration dur ing this study using the Penman-Monteith Model. Spaces in the data indicate a value of zero............................................................79 4-3 Monthly calculated potential evapotrans piration using the Thornthwaite method and monthly actual evapotranspiration us ing the Penman-Monteith method..........................80 4-4 Long term record of River Rise water leve ls. Storm events that have been cited in previous work and in this study are highlighted................................................................81 4-5 Water levels of various monitoring wells during this st udy and River Rise water level for comparison. Three storm events are highlighted.........................................................82 4-6 River Sink and River Rise discharge duri ng this study. Three storm events resulting in increased discharge are highlighted. Th e time period where the River Rise gauge was warped is represented with the blue highlight............................................................83 4-7 Daily calculated diffuse recharge during this study using the water budget method. Recharges associated with the three major storm events are highlighted..........................84 4-8 Precipitation and calculated diffuse recharge with the water budget method for the small event in June 2006....................................................................................................85 4-9 Precipitation and calculated diffuse recharge with the water budget method for the small event in July 2006....................................................................................................86 8

PAGE 9

4-10 River Rise water level during summer 2006. Small events in June and July are highlighted.........................................................................................................................87 4-11 Water level perturbation at Well 5a af ter a small rain event in June 2006 with recession curve projected...................................................................................................88 4-12 Water level perturbation at Well 6a af ter a small rain event in June 2006 with recession curve predicted...................................................................................................89 4-13 Water level perturbation at Well 7a af ter a small rain event in June 2006 with recession curve projected...................................................................................................90 4-14 Water level perturbation at Well 5a af ter a small rain event in July 2006 with recession curve projected...................................................................................................91 4-15 Water level perturbation at Well 6a af ter a small rain event in July 2006 with recession curve projected...................................................................................................92 4-16 Water level perturbation at Well 7a af ter a small rain event in July 2006 with recession curve projected...................................................................................................93 4-17 Curve matching results at Well 4 for the September 2004 storm event using the analytical method without recharge...................................................................................94 4-18 Curve matching results at Well 6 for the September 2004 storm event using the analytical method without recharge...................................................................................94 4-19 Curve matching results at Well 7 for the September 2004 storm event using the analytical method without recharge...................................................................................95 4-20 Curve matching results at Well 4 for the March 2005 storm event using the analytical method without recharge....................................................................................................95 4-21 Curve matching results at Well 6 for the March 2005 storm event using the analytical method without recharge....................................................................................................96 4-22 Curve matching results at Well 7 for the March 2005 storm event using the analytical method without recharge....................................................................................................96 4-23 Curve matching results at Well 4 for the December 2005 storm event using the analytical method without recharge...................................................................................97 4-24 Curve matching results at Well 6 for the December 2005 storm event using the analytical method without recharge...................................................................................97 4-25 Curve matching results at Well 7 for the December 2005 storm event using the analytical method without recharge...................................................................................98 4-26 1-Dimensional model with recharge grid..........................................................................99 9

PAGE 10

4-27 Results from of the analytical method without recharge and the 1D model without recharge for a hypothetical storm event...........................................................................100 4-28 Curve matching results at Well 4 for the March 2003 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........101 4-29 Curve matching results at Well 6 for the March 2003 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........102 4-30 Curve matching results at Tower Well fo r the March 2003 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........103 4-31 Curve matching results at Well 1 for the March 2003 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........104 4-32 Curve matching results at Well 4 for the September 2004 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........105 4-33 Curve matching results at Well 6 for the September 2004 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........106 4-34 Curve matching results at Well 7 for the September 2004 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........107 4-35 Curve matching results at Well 4 for the March 2005 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........108 4-36 Curve matching results at Well 6 for the March 2005 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........109 4-37 Curve matching results at Well 7 for the March 2005 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........110 4-38 Curve matching results at Well 4 for the December 2005 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........111 4-39 Curve matching results at Well 6 for the December 2005 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........112 4-40 Curve matching results at Well 7 for the December 2005 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference.........113 4-41 Recession curve breakdown at Well 4 for the hydrograph produced by the March 2003 storm event..............................................................................................................1 14 4-42 Recession curve breakdown at Well 4 for the hydrograph produced by the September 2004 storm event..............................................................................................................1 14 10

PAGE 11

4-43 Recession curve breakdown at Well 4 for the hydrograph produced by the March 2005 storm event..............................................................................................................1 15 4-44 Recession curve breakdown at Well 4 for the hydrograph produced by the December 2005 storm event..............................................................................................................1 15 4-45 Example of using two portions of the r ecession curve that results from the December 2005 event at Well 6 to calculate basin area....................................................................116 11

PAGE 12

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science IMPACTS OF DIFFUSE RECHARGE ON TRANSMISSIVITY AND WATER BUDGET CALCULATIONS IN THE UNCONFINED KA RST AQUIFER OF THE SANTA FE RIVER BASIN By Michael Ritorto December 2007 Chair: Elizabeth Screaton Major: Geology In karst aquifers, high transmissivities make it difficult to conduct traditional aquifer tests. An alternative method to evaluate formation tran smissivity on a large scal e is passive monitoring of observation well response to changes in stream stage or conduit hydraulic head during high discharge events. During these events, increa ses in water levels at observation wells are dependent on the stream or conduit boundary cond ition, but are also influenced by the amount of diffuse recharge arriving at the water table. This study used data from the karstic Floridan aquifer in northern Florida to evaluate the effect of neglecting diffuse recharge on the transmissivity estimate. At the Santa Fe River Sink/Rise system, the Santa Fe River enters the Floridan aquifer at the River Sink, and re-emerges downstream at the River Rise. Between these locations, cave divers have mapped over 7 km of conduits. Diffuse recharge was estimated as rainfall minus evapotranspiration calculated using the Penman-M onteith method. The transit time of recharge through the unsaturated zone is no t known; given the relatively thin unsaturated zone and its expected high hydraulic conductivity, the simp lifying assumption was made that recharge reaches the aquifer on the same day as rainfall occurs. Hydraulic heads were monitored through 12

PAGE 13

time at locations within the conduit system and in wells monitoring the surrounding aquifer. Four storm events were recorded by the monitoring network. Aquifer responses to storm events were first predicted using an anal ytical solution created by Pinder et al. (1969) that does not include diffuse recharge. For comparison, aquifer response was then simulated using a one-dimensional model implemented in MODFLOW incorporating the calculated diffuse recharge during the storm event. In the model, the time-varying hydraulic head at the conduit was simulated as a speci fied head boundary condition. Calibration of the model for individual monitoring well hydrographs yi elds estimated transmissivity values for the aquifer between the monitoring well and conduit. Results suggest that neglecting to include diffuse recharge in the solution results in overpredicted transm issivity values. Results also suggest that the magnitude of diffuse recharge in comparison to head change in the conduit determines the relative effect of the diffuse component on the observe d well hydrograph after a storm event. In terms of total water volumes, resu lts show that diffuse recharge can be more than twice the input received from the conduit boundary during small storm events, and less than half the input received from the conduit boundary dur ing large storm events. These results suggest that in future calculations it will be important to consider diffuse recharge as an influence on dissolution across the entire Santa Fe River basin. 13

PAGE 14

CHAPTER 1 INTRODUCTION Considering that more than 25% of the world's population either li ves on or obtains its water from karst aquifers (Karst Water Institute, 2006), it is no surprise that a growing effort to model karst aquifers has prevaile d in recent years. In the United States alone, 20% of the land surface is karst and 40% of the groundwater used for drinking comes from karst aquifers (Karst Water Institute, 2006). In order to preserve the us eable water from these aquifers, we must start by building a solid understanding of the hydrologic processes that affect karst terrains. Karst aquifers are characterized by dissoluti on-generated conduits that permit the rapid transport of ground water (White, 2002). These conduits may range in aperture from as little as 1 cm to about 0.5 m; at which point the pa thway becomes a cave large enough for human exploration (White, 2002). Under normal gradient s, conduit flow behaves in a turbulent manner (White, 2002). Conduits are often connected to th e surface via openings such as sinkholes and sinking streams. These connections allow for rech arge to the conduit, and mixing of surface and ground water (Martin and Dean, 2001) Connections to surface wa ters are important because they allow conduits to quickly transmit pollutant s without filtration over long distances (Mylroie, 1984). In the early work on karst hydrology, the focus remained mostly on determining the properties of the conduit systems. However, kars t researchers have migr ated towards discussing karst aquifers based on the triple permeability model. This model evaluates a karst system based on conduit flow, fracture flow, and intergranula r flow, and the inter actions between each. Fractures are much smaller than co nduits (aperture ranging from 50 500 m) and usually occur from mechanical joints and be dding plane partings (White, 2002). Intergranular flow occurs on even smaller scales, and is often referred to as matrix flow. For the purpose of this study, we will 14

PAGE 15

use the term matrix flow to encompass both fractur e and intergranular flow. This is because it is often difficult to distinguish betw een fracture and intergranular fl ow and because flow in these regions is generally laminar. Mixing between conduit and matrix water is controlled by various factors including matrix permeability and matrix transmissivity. These factors are often related to the geologic background of the study area. To date much of the existing understa nding of fluid flow in karst aquifers is based on work in extensively cemen ted and recrystallized Paleozoic and Mesozoic carbonates (White, 1999). In these regions, termed telegenetic karst, the karst is developed on and within ancient rocks that are exposed afte r the porosity reduction of burial diagenesis (Vacher and Mylroie, 2002). Understanding of fluid flow is less advan ced for aquifers with high intergranular porosity, termed eogenetic karst (Vacher and Mylr oie, 2002). In eogenetic karst regions, the land surface is evolving on, and the pore system devel oping in, rocks undergoing meteoric diagenesis (Vacher and Mylroie, 2002). Eogenetic regions like the karst limestones of the unconfined Floridan Aquifer, have not been deeply bur ied and thus have signi ficantly higher matrix permeability than telegenetic regions (Florea and Vacher, 2005). Understanding transmissivity in karst aquifers is important for better assessment of the potential for flow of contaminants through the aquifer. Determining transmissivity in karst aquifers is difficult due to heterogeneity of hydrologic properties and the high flow rates necessary to perturb the water table. Multiple studies have concluded that transmissivity measurements in highly heteroge neous aquifers appear scale depe ndent. This is because water is likely to travel through preferential flow paths over long distances. Bradbury and Muldoon (1990) worked in glacial outwash sediment and observed three to five times greater 15

PAGE 16

transmissivity in regional pumping tests as compared to smaller scale slug tests. Rovey and Cherkauer (1995) used slug, pressure, and pumping tests, along with two numerical models, to determine the dependency on scale of hydrauli c conductivity measurements in units of a carbonate aquifer. Martin et al. (2006) used results from passi ve monitoring of water level fluctuations to calculate transmi ssivities at least two orders of magnitude different between wells at various locations away from a know n conduit in the Floridan aquifer. In an unconfined, eogenetic karst aquifer, the aquifer will be recharged by both a diffuse component and possibly an allogenic, focused component. Allogenic r echarge is defined by White (2002) as surface water inject ed into the aquifer at the swa llets of sinking streams. Diffuse recharge was defined by White (2002) as rainfall directly onto the karst surface and from there entering the aquifer as infiltr ation through the soil and matr ix permeability of the underlying carbonate bedrock. A crucial factor that influences the amount of di ffuse recharge that enters the aquifer is evapotranspiration, a co llective term for all the proce sses by which water in the liquid or solid phase at or near the earths land su rfaces becomes atmospheric water vapor (Dingman, 2002). In an unconfined aquifer with no runoff, a ll of the precipitation th at is not lost to evapotranspiration will recharge the aquifer syst em. If the unsaturated zone of the unconfined aquifer is composed of sands with high hydraulic conductivity, then diffuse recharge will occur rapidly. This increases potential to quickly move contaminants into the aquifer. In carbonate aquifers, diffuse recharge should be considered an important sour ce of water that can potentially have a major influence on the karstification (d issolution) of carbonate bedrock. Methods that have been used to calculate the diffuse rech arge component in unconfined aquifers include analyses of basin water budgets a nd stream flow (Grubbs, 1998), us ing fluctuations in the water table level with a known speci fic yield (Healy and Cook, 2002), direct quantif ication from 16

PAGE 17

lysimeters (Healy and Cook, 2002), and numerical modeling of flow systems that calculate various inputs and outputs (Bush and Johnston, 1988). The Santa Fe River Sink/Rise system lies within the unconfined Floridan aquifer and offers a unique area to investigat e transmissivity and diffuse rech arge. In this system, the Santa Fe River is captured by a si nkhole known as the River Sink, a nd then continues to flow underground for about seven kilometers. The river re -emerges at the River Rise and continues to flow across the landscape. In previous research, passive monitoring of storm pulse events has yielded results about the hydrologi c properties of the system (Mar tin et al, 2006). These studies have been complemented with studies of the hydrogeochemistry of the system (Martin and Dean, 2001; Screaton et al, 2004). This study further investigated the transmissi vity of this system by using new storm event data and by applying a numerical modeling method. This study will investigate how the diffuse recharge component affects tran smissivity estimates made us ing passive monitoring methods. Passive monitoring can be extremely valuable in karst aquifer regions in order to assess the hydrologic behavior of an aquifer on a larger scale than is capable with ty pical laboratory and well tests. However, using passive monitoring to determine transmissivity is difficult in regions where there are multiple sources of water influencing the shape of a flood hydrograph. Understanding the effect that recharge has on th e calculation of transmissivity will shed some insight on whether applying a pa ssive monitoring method is worth the effort in comparison to smaller scale tests. Also, quantifying the diffuse recharge, along with the area of the basin contributing to discharge, will allow for comparison of total volumes input to the system from the allogenic, focused recharge and total volumes input to the sy stem from the diffuse recharge. 17

PAGE 18

These volumes have potential implications for kars tification in the region. The two main goals of this research are as follows: Goal 1: Determine the influence of the diffuse recharge component on storm event hydrographs and evaluate the eff ect of diffuse recharge on the calculation of transmissivity by passive monitoring. Goal 2: Evaluate the total volume of conduit allogenic input and the total volume of diffuse recharge input to the unconfined aquifer system. 18

PAGE 19

CHAPTER 2 BACKGROUND Study Area The Santa Fe River, a tributary to the Suwannee River, is located in north-central Florida (Figure 2-1). The Santa Fe River ba sin covers an area of roughly 3500 km2 (Hunn and Slack, 1983). From its origin in Lake Sa nta Fe, the Santa Fe River flows west for approximately 50 km (Hisert, 1994). At this point the river flows into a 36-m deep sinkhole known as the River Sink (Martin and Dean, 2001). The rive r continues to flow underground for approximately 7 km and reemerges at a first magnitude spring called th e River Rise. Between th ese two locations, the river reappears at various intermediate karst wi ndows, most notably at Sweetwater Lake (Figure 2-2). The Santa Fe River Basin flows across tw o major physiographic provinces known as the Northern Highlands and the Gulf Coastal Lowla nds. The Northern Highlands generally consist of gently sloping plateaus in the interior regi ons and marginal slopes that are well drained by dendritic streams (Grubbs, 1998). The Gulf Coasta l Lowlands have a noticeable lack of surface streams and generally consist of te rraces and ancient shorelines that slope gently toward the coast (Grubbs, 1998). The Northern Highlands have elev ations that are in excess of 30 meters above sea level (masl), where the Gulf Coastal Lowlan ds are at elevations of less than 15 masl. Dividing the two provinces is the marginal zone. Within the marginal zone is an escarpment known as the Cody Scarp, which in this region re presents the erosional edge of the Miocene Hawthorn Group. Most streams that cross the Co dy Scarp either sink completely into the subsurface or sink into the subsurface and re-emerge (Hunn and Slack, 1983). The Cody Scarp marks the boundary between the confined and unconfined Floridan aquifer. 19

PAGE 20

The Santa Fe Sink/Rise system is located within the boundaries of OLeno State Park and River Rise State Preserve, which encompasse s nearly 6,000 acres of state land. The closest significant city to OLeno State Park is Lake City (~ 24 miles away), which has an average annual temperature of 20.5 C (Hunn and Slack, 1983). Seasonal temperatures range from 4 to 10 C during the winter months, and from 25 to 35 C in the summer months (Grubbs, 1998). Average annual precipitation typically ranges fr om 130 to 150 cm/yr, with nearly half of the rainfall occurring between June and September (Grubbs, 1998). Average annual evapotranspiration ranges from 90 to 105 cm/yr (Bush and Johnston, 1988). Geologic Background Three principle hydrogeologic un its occur in north central Fl orida: the surficial aquifer system, the intermediate confining unit, and th e Floridan aquifer system (Table 2-1). The surficial aquifer occurs throughout the Northern Highlands province, and more locally in the Gulf Coastal Lowlands (Grubbs, 1998). Where presen t, the surficial aquife r system is contiguous with the land surface and consists primarily of unconsolidated sediments (Grubbs, 1998). In the Santa Fe River Basin, surficial sediments are compos ed of white to gray fine to coarse sand, and are Pleistocene and Holocene age (Hunn and Slack, 1983). The thickness of the surficial aquifer system in the Santa Fe Basin is approximately 5 m, and the water table is reached approximately 3 m below the surface (Hunn and Slack, 1983). Beneath the surficial aquifer lies the intermediate conf ining unit known as the Hawthorn Group. This layer is composed primarily of silic iclastic rocks and is Miocene in age (Hunn and Slack, 1983). Regionally, this unit acts as a confin ing unit that restricts the exchange of water between the overlying surficia l and underlying Floridan aquifer systems (Grubbs, 1998). The Floridan aquifer system underlies an area of about 260,000 km2 including all of Florida, southeast Georgia, and small parts of Alabama and South Caro lina (Bush and Johnston, 20

PAGE 21

1988). The Floridan aquifer is the principal sour ce of water for municipal and industrial use in the Santa Fe Basin (Hunn and Slack, 1983). The aqui fer system is divided into the upper and lower aquifer by a layer of less permeable car bonate rock in the lower Avon Park formation (Bush and Johnston, 1988). The Upper Floridan aquifer is extremely permeable and capable of transmitting large volumes of water (Grubbs, 1998). From oldest to youngest, the Upper Floridan aquifer is divided into the Avon Park form ation, Ocala Limestone, and Suwannee Limestone. These layers range from Eocene to Oligocene in age, and range from 90 to 240 m in thickness (Hunn and Slack, 1983). In the unconfined portion of the Santa Fe River Basin, the Ocala limestone is the uppermost unit (Hunn and Slack, 1983) The aquifer thickness is smallest in the region near the Santa Fe River at about 300 ft (~100 m), but ex tends to almost 800 ft (~240 m) just north and south of this reach (Hunn and Slack 1983). Specific yield of the unconfined Floridan aquifer has been estimated from poros ity measurements and ranges between .10 and .45 (Palmer, 2002). Previous Investigations in the Santa Fe River Basin One of the earliest comprehensive studies in the Santa Fe River Basin was performed by Hunn and Slack (1983), which involved mapping th e hydrogeologic units, identifying how water moves from one unit to another and to streamflow and determining aquifer parameters such as transmissivity and recharge rates. Mapping of the potentiometric surface allowed Hunn and Slack (1983) to infer that groundwater flow is towards the Santa Fe River. Hunn and Slack (1983) also concluded that annual fluctuations of water level are due largely to variations in rainfall amounts. Aquifer test s in the western part of the Santa Fe River Basin, where the Floridan aquifer is unconfined, produced transmissivity values ranging from 3,000 m2/day to 50,000 m2/day; Significantly lower transmissivity was observed in the eastern confined region of 21

PAGE 22

the Floridan aquifer, where transmissivity ranged from 1,900 m2/day to 3,300 m2/day (Hunn and Slack, 1983). Using the gaseous tracer SF6, Hisert (1994) attempted to make connections between the surface water locations a nd the karst windows of the Santa Fe River Sink-Rise system. A link was verified between the River Sink, seven intermediate karst windows, and Sweetwater Lake, and subsequently between Sweetwater Lake and the River Rise. No direct connection was made from the River Sink to the River Rise. Tracer travel times suggested rapid flow through the system; however precise velocity determinations were not possible because of the method of injection. Hiserts (1994) work s uggested the existence of large conduits that carry the majority of flow through the system. In 1995, cave divers began exploring the conduits in the Santa Fe River region. Since then, a single conduit has been mapped that connects Sweetwater Lake to the River Rise, along with multiple conduits that connect the River Sink to various downstream karst windows (Figure 2-2). A large conduit that feeds fr om the east of OLeno State Park has also been mapped. As of summer 2006, no direct physical conn ection has been made between the Rise and Sink; however, over 7 km of conduits have been mapped betwee n the two locations (M. Poucher, unpublished data). Divers estimate that conduits can reach widths of 24 m. Martin and Dean (1999) used temperatur e as a natural tracer of water through the system.Velocities increased with increasing river stage and rapid velocities of greater than one kilometer per day indicated conduit flow. Martin a nd Dean (2001) also inve stigated the exchange of water between conduits and matrix in the Santa Re River basin. Water samples were collected from three locations along the river (River Sink, Sweetwater Lake, River Rise) at varying flow conditions. During low-flow conditi ons, different compositions of water at the River Sink from 22

PAGE 23

those at Sweetwater Lake and Ri ver Rise suggest that little wa ter discharging from the River Rise is derived from the River Sink (Martin and Dean, 2001). During flood conditions, when the river stage is elevated, water at Sweetwater Lake and Rive r Rise were found to retain compositions that are similar to the River Sink (Martin and Dean, 2001). Martin and Dean also took groundwater samples from two observation we lls in the area. It was observed that concentrations of conservative so lutes sodium and chlorine were decreasing over a period of five months following a flood event. This suggests that the dilution was occurring from water that entered the matrix after the flood event (Martin and Dean, 2001). Screaton et al. (2004) worked on the Santa Fe River SinkRise system looking at the exchange of water between conduits and matrix. By comparing discharge data into the River Sink and out of the River Rise, the direction of flow between the conduit system and the aquifer could be determined. For example, during times of baseflow, the discharge is greater at the River Rise and groundwater flows from the aquifer matrix to the conduit. At th e peak of storm events, the discharge is greater at the River Sink and th e conduit system loses wa ter, probably to the surrounding aquifer matrix. The idea that the Sa nta Fe River loses water to the surrounding matrix is important because it shows the vulnerab ility of the aquifer system water quality to a flood event (Screaton et. al, 2004). Sc reaton et al. (2004) also used arrival times of temperature signals to estimate average velocities in the Sa nta Fe River Sink-Rise sy stem. Using temperature as a tracking signal allows for tracing an indi vidual water packet entering the system after a storm event. Using the discharge data and the ve locity determined from the temperature signals, the average cross sectional area of the conduit system wa s calculated to be ~380 m2. This also allows for the calculation of conduit diameter, which would be 22 m if a single conduit exists. This calculation is consistent with observations made by cave divers. 23

PAGE 24

Screaton et al. (2004) were also able to make some first approximation calculations of the dissolution resulting from a storm pulse passing through the Santa Fe system. A small storm event from late 2001 was used from which the amount of water lost fr om the conduit to the matrix was calculated. Chemistry data from th e River Sink in 1998 was used to estimate the degree of undersaturation for a ri ver stage similar to that which occurred during the storm event. The amount of rock needed to dissolve in orde r to return the system to equilibrium was calculated to be 0.13 moles of calcite per liter of river water, or a total of 2.86 x 108 moles of calcite. This amounts to a to tal rock volume of 1.10 x 104 m3 or a denudation of 4.60 x 10-4 m in the region surrounding the conduit. This value was an order of magnitude greater than denudation rates from previous studies in this area (Opdyke et al, 1984), which indicates that dissolution in this area may have previously been under predicted. Martin et al. (2006) used an analytical met hod for a semi infinite aquifer developed by Pinder et al. (1969) to provide a first approximation of the matr ix transmissivity between known conduits and observation wells in the unconfined Floridan aquifer. Although this analysis did not account for heterogeneity of the aquifer, or head changes at the wells due to diffuse recharge, initial estimates of transmissivity were po ssible. Hydraulic conductivity was determined by dividing the best fit transmissivity by an assumed aquifer thickness of 275 m. Calculated transmissivity values ranged between 900 m2/day and 500,000 m2/day and thus hydraulic conductivity ranged from 3 m/day to 1,818 m/day. Th ese high values are c onsistent with other studies of the Floridan aquifer. For example, Palmer (2002) estimates hy draulic conductivity of the Floridan to range between 86.4 m/day and 8 64 m/day, which is four orders of magnitude greater than the hydraulic conduc tivity of 0.086 m/day reported for recrystallized Paleozoic aquifers of PA and NY. Bush and Johnston ( 1988) looked at aquifer tests performed at 114 24

PAGE 25

locations throughout the stat e of Florida, mostly from the U pper Floridan aquifer. The closest location to OLeno State Park was in Lake City, where the transmissivity was estimated to be 3,344 m/day. Results from these tests showed that transmissivity values of the Upper Floridan aquifer are directly related to the thickness and lithology of its upper confining unit. The confined areas generally had lower transmissivity values than the unconfined areas, with the highest transmissivity occurring in karstic north-central Flor ida (Bush and Johnston, 1988). Another important result from the Martin et al. (2006) study is that transmissivity differed by two orders of magnitude for wells located at various distances from the conduit. Estimated transmissivity of wells within 150 m of the conduit were ~ 900 m2/day, while wells at distances greater than 500 m from the c onduit produced transmissivity va lues of greater than 100,000 m2/day. These results were attribut ed to a scaling effect. Scale e ffects result from the likelihood that large-scale tests will incl ude preferential paths that dominate a larger percentage of groundwater flow, and have been previously docum ented in the literature (Rovey and Cherkauer, 1995). Recharge in the Santa Fe Basin has been investigated by Grubbs (1998). His study provided an overview of recharge rates to the Upper Floridan aquifer using a simple water budget analysis. In the lower Santa Fe River basin, runoff is generally negligible because very little channelized surface drainage is present, the soils are permeable, and the slope of the land surface is gradual to flat. Water-budget analysis in the unconfined region placed average-annual recharge rates between 45 to 60 cm/yr (Grubbs, 1998). Other methods incl uding chloride tracing, hydrometric base flow, and ground-water level changes showed a range of average-annual recharge between 20 and 80 cm/yr (Grubbs, 1998). Recharge rates of the confined Upper Floridan aquifer were found to range based on confinement of the aquifer, and water budgets 25

PAGE 26

indicated that recharge is less than 30 cm/yr in areas where there is confinement by an overlying layer of sediments with low permeability (Grubbs, 1998). Grubbs (1998) water budget calculations us ed evapotranspiration values for Florida calculated by Bush and Johnston (1988). Evapotra nspiration was calculated with a method that uses biotemperature and precip itation. Biotemperature is define d as the annual sum of hourly temperatures between 32 F and 86 F divided by the numbers of hours in the year. Biotemperature was linearly related to potential evapotranspiration in order to estimate a quantity of water that would be given up to atmosphere within a zonal climate. Long-term average evapotranspiration from the region was estimated to be between 90 and 105 cm/yr (Bush and Johnston, 1988). Diffuse Recharge in Other Study Areas In addition to the work previously complete d in the Santa Fe Rive r basin, much of the methodology behind studying diffuse recharge has been commonly applied to karst and nonkarst locations around the world. A debate often ensues about wh ich method is best suited for individual study areas. For example, in 2005 the USGS published a scientific investigations report from east-central Pennsylvania which co mpares multiple methods for estimating groundwater recharge in the same wa tershed area. Recharge was estimated on a monthly and annual basis using three methods; (1) unsaturated-zone drainage collected in gravity lysimeters, (2) daily water balance, and (3) water-table fluctuat ions in wells (Risser et. al, 2005). The following briefly summarizes the different methods used in their study and their respective results: 1) Gravity lysimeters directly measure the ver tical flow of water through a large section of unsaturated zone at a depth below the root system. Drainage from lysimeters represents water that has not yet reached the water table, however, because water has passed beneath the root zone, it is assumed to represent the water that will reach the water table (Risser et al, 2005). The mean-a nnual recharge from seven lysimeters in the Pennsylvania study area was 30.9 cm. 26

PAGE 27

2) The water balance method uses the general water budget equation: R = P (ET + RO + S) where R is recharge, P is precipitati on, ET is evapotranspiration, RO is runoff, and S is change in storage (Risser et al, 2005). This method is easily applied to locations where precipitation data is availabl e; however, the accuracy of the estimated recharge relies on the accuracy of othe r budget terms. The mean annual recharge estimated from the budget method over the same seven year period was 31.2 cm. 3) Water table fluctuation is based on the rise of water levels in a well multiplied by the specific yield of the aquifer (Rasmussen and Andreasen, 1959) to estimate recharge. This method assumes that water-level rise is solely caused by recharge arriving at the water table and that specific yield remains c onstant. It is therefore necessary to know the correct value of specific yield fo r the study area. Specific yield in the Pennsylvania watershed was determined to be 0.013 using the equation Sy = S / h where Sy is specific yield, S is stre amflow volume during a recession period consisting of only ground-water discha rge over the watershed area, and h is the average decline in water-table level during a recession period. The mean annual recharge estimated from the fluctuation method over the seven year period was 25.1 cm. A study by Rasmussen and Andreasen (1959) in the Beaverdam Creek Basin, Maryland provides another example of the application of the water table fluctuation method for estimating recharge. This is a relevant study because th e unconfined aquifer study site has similar characteristics to the unconfined aquifer region in the Santa Fe River Basin. In the Beaverdam Creek Basin, surficial sands and silt overlie Tertiary sand aquifers, and the water table is generally within a few meters of the surf ace (Rasmussen and Andreasen, 1959). Water levels were measured in observation wells on a weekly basis and precipitation was measured at various sites in the basin. Specific yield was estimated by the water budget method to be 0.11 (Rasmussen and Andreasen, 1959). The rise in wa ter level was estimated as the difference between the peak water level during the storm ev ent and the extrapolated antecedent water level prior to the rise. Results produced a two year re cord of monthly groundwater recharge from April 1950 to March 1952. Jones et al. (2000) provide an example of how chemistry can also be used to quantify recharge. Jones et al. (2000) measured 18O isotopes of water in the Pleistocene limestone 27

PAGE 28

aquifer of Barbados. Oxygen isotopes are useful in this region because the islands small size, low relief, and tropical climate allow for constrai nt of the factors that influence spatial and seasonal variations of rainwater and groundwater isotopic compositions (Jones et al, 2000). In this study, groundwater was sampled from 29 wells from 1961-1992. A small range of groundwater 18O (-2.5 to -4.5 ) values were consis tently observed, which suggested that evaporation likely had little influence on 18O values prior to recharge This is most likely an affect of rapid diffuse inf iltration through highly permeable limestone. The oxygen isotope method predicted recharge to be between 15% and 45% of the mean monthly rainfall. These values were greater than recharge pr edicted by comparing rainfall and potential evapotranspiration, which was about 6% of the mean monthly rainfall. This result occurred because the use of potential eva potranspiration in the recharge estimate overestimates actual loses to evapotranspiration by assuming that the actual evapotranspiration is limited only by the amount of rainfall (Jones et al, 2000). Recharge amounts predicted from standard methods such as the waterbalance method were highly variable and proved unsuccessful in this region. This was attributed to the idea that coastal discharg e rates can be affected by groundwater withdrawal and return flow in adjacent catchments, resulting in artificial redistri bution of groundwater recharge (Jones et al, 2000). Vacher and Ayers (1980) recogni zed that increased chloride concentrations on the oceanic island of Bermuda should be a direct result of evapotranspiration. This happens because there are no surface streams and thus all rainfall is transmitted back to the atmosphere as evapotranspiration, or recharged and passe d to the shoreline via the subsurface. Evapotranspiration that transports water back to the atmosphere concentrates the rain-derived chloride in the soil. Comparing the ratio of fresh rainwater chloride concentration versus 28

PAGE 29

concentrated groundwater chloride concentrati on allows for a concentration factor to be calculated. If the precipitation has been measured, th an the concentration factor can be used to calculate the amount of evapotra nspiration, and thus recharge can be calculated by subtracting the evapotranspiration from the precipitation. In Vacher and Ayers ( 1980) study, recharge was calculated on an annual basis to be 37 cm/yr. In addition to quantifying the diffuse rech arge component in orde r to better understand the volumetric input of water in to a karst system, many have used the diffuse component to characterize the behavior of the vadose zone of karst aquifers. For example, Batiot et al. (2003) collected drip water from Nerja Cave in Spain, and investigated the use of Total Organic Carbon (TOC) as a tracer of diffuse infiltration. In this region, the highest concentrations of TOC in drip water occurred during the summer months. This was likely caused by rainwater entering after the dry season and transporting carbon from the soil wh en it has the highest content of organic matter (Batiot et. al, 2003). Mean residence time of water in the karstified unsaturated zone over the cave was determined by comparing TOC of rainfall and drip water. Residence time varied significantly (2 to 8 months) which was likely de pendent on the amount of water stored in the unsaturated zone over the cave prio r to rainfall events and also the highly variable thickness of the unsaturated zone (4 to 90 m) (Batiot et. al, 2003). Maher and DePaolo (2003) used Sr isotopes to determine vadose zone infiltration rates at a site in Hanford, Washington. Sr can be used as a tracer of diffuse infiltration because groundwater entering a specific rock or soil environment typically contains dissolved Sr in low concentration (~ 100 g/L total Sr) with an isotopic ratio (Sr87/Sr86) that is generally different from mineral phases in soil matrices that ar e undergoing weathering (Maher and DePaolo, 2003). Weathering results in dissolution of primary minerals and introdu ces Sr to the pore fluid, thus 29

PAGE 30

shifting the Sr ratio towards that of the dissolving minerals. Infiltration rates are of concern in the Hanford region because substantial quantities of radioactive materi als have been disposed into the surface soils and vadose zone sediments, with the expectation that ion exchange processes would retain most of the radionuclides and prev ent them from ever reaching groundwater (Maher and DePaolo, 2003). Infiltration rates were determ ined by matching isotopic profiles from two models (steady state and non steady state finite difference models) to isotopic profiles from pore waters and bulk sediment analysis from a 72 m borehole. Model resu lts suggest that the infiltration flux ranges from 0.4 to 1 cm/yr, which would result in a transit time from water on the surface to the water tabl e between 600 and 1600 years (Maher and DePaolo, 2003). These transit times are signific antly longer than can be expected in the vadose zone of the Santa Fe River basin, where surficial sediment is composed of fine to coarse sand s and the water table is close to the surface. Aquifer Parameters in Other Study Areas The most commonly used technique for dete rmining hydraulic parameters is an aquifer test. In an aquifer test, a well is pumped and th e rate of decline of th e water level in nearby observation wells is noted (Fette r, 1994). Pumping tests typically last until the water level reaches a state of equilibrium; that is there is no further drawdown with time (Fetter, 1994). The result of an aquifer test is a set of data with the drawdown given at various times after the start of pumping (Fetter, 1994). This data can be graphically matched to a set of type curv es originally developed by Theis. Curve matching allows for a graphical means to solve a set of equations for the various aquifer parameters (Fetter, 1994). Jones (1999) demonstrates an example of an aquifer test in a carbonate aquife r. In his test, a water supply well in West Virginia was pumped for five hours and significant drawdown (~2.5 m) was observed at a monitoring well 315 m away. Transmissivity and storage coefficients were calculated using the Jacob time-drawdown 30

PAGE 31

method (a simplification of the Theis solution), and varied by an order of magnitude in different locations of the carbonate aquifer. These results were attributed to th e varying locations of fractures within the aquifer system relative to the pumping location. An aquifer test would be difficult in our study area of the Santa Fe River Sink-Rise system, primarily because of the difficulty to produce a large enough drawdown from pumping. This results from both the high transmissivity of the Floridan aquifer, and also the lack of high pumping capacity. Pumping rates during aquifer te sts in karst aquifers are commonly greater than 100 gal/min. We can estimate the pumping ca pacity needed to observe drawdown in an observation well in our study area by using the m odified Theis solution for nonequilibrium radial flow in an unconfined aquifer: T = (Q / 4 (ho h))*W( ua, ub, ) T = Transmissivity (m2/day) Q = pumping rate (m3/day) ho h = drawdown (m) W( ua, ub, ) = well function (unitless) In order to determine the well function for the aquifer, the radial distance to the pumping well must be known. The closest observation wells in this study area are approximately 50 m apart. The hydraulic conductivity and initial sa turated aquifer thickness are also needed to determine the well function. A value of 5 m/day wa s applied as an low end estimate of hydraulic conductivity in this area, and a value of 100 m was used for the aquifer thickness. The calculated well function with these values is 2. With a transmissivity of 500 m2/day, a pumping rate of 314 m3/day or 57.61 gal/min is needed in order to observe a small drawdown of .1 m (~.33 ft). In order to observe a larger drawdown of 0.3 m (~ 1 ft ), a pumping rate of 942 m3/day or 172.8 gal/min would be needed. Our current pumping ability maxes out at a rate of 9 gal/min and thus is much lower than the rates needed to pr oduce significant drawdown. Because the wells are 31

PAGE 32

constructed from PVC that has on ly a 2-inch diameter, we are limited in the size of the pump that can be used in the well. If the hydraulic conductiv ity is actually higher, then a greater pumping rate would be needed to observe drawdown. An alternative field method to pumping tests is slug tests. In a sl ug test, a known amount of water is quickly drawn from or added to a monitoring well, and the rate at which the water levels fall or rise in the same well is measur ed (Fetter, 1994). Commonl y, the water column in the well casing is induced to rise by rapidly lowe ring a solid slug into the well below the static water level. This is equivalent to adding a volume of water that is equal to the volu me of the slug (Fetter, 1994). The recovery data are then anal yzed using an appropriate method, such as the Hvorslev Slug test method. These tests allow fo r determining the hydraulic conductivity in the vicinity of a monitoring well. The difficulty with performi ng a slug test in the highly transmissive unconfined Floridan Aquifer is that results will most likely represent localized parameters, rather than representing parameters on a regional scale. Results of recent slug tests near the Santa Fe River Sink/Ri se system yielded hydraulic conduc tivities that ranged from 8 x 10-1 m/day to 8 x 10-3 m/day (Langston, personal communication). Powers and Shevenell (2000) describe a passive monitoring technique to calculate transmissivity in an unconfined karst aquifer based on analysis of well hydrographs. They observed that the recession limb of a well hydrog raph responded similarly to those of karst spring hydrographs. Recession limbs in karst aquifers commonly produce two or more segments with different slope values that represent different portions of the ground water system (Powers and Shevenell, 2000). The firs t and steepest slope segment was inferred to represent the drainage of condu it features, while the following segments were inferred to represent flow through fractures and matrix (Pow ers and Shevenell, 2000). The solution method 32

PAGE 33

used by Powers and Shevenell (2000) adapts th e Bernoulli equation for flow through a conduit to a spring to determine the slope of these recessi on segments, which can then be translated to theoretical flow values that co rrespond to the various types of flow (conduit, fracture, matrix). Using these theoretical flows al ong with a valid estimation of the specific yield and distance from recharge area, transmissivity can be calcu lated. Powers and Sheven ell (2000) applied their solution to calculate transmissivity from 49 reces sion curves from wells within three different Midwestern field sites. They went on to compare their results to slug a nd pumping tests results, and found that results matched w ithin an order of magnitude between the different methods. Differences in transmissivity values may have occurred based on the difference in the scale of measurement between the aquifer tests and the passive monitoring method (Powers and Shevenell, 2000). Results from this study show that determining aquifer parameters passively provide an alternative to th e commonly used aquifer and slug methods, and also provide information about aquifer parameters during natural flow conditions. Pinder et al. (1969) describe another passive monitoring technique which allows for calculation of aquifer diffusivity (ratio of transmissivity to the coefficient of storage) based on responses to fluctuations in ri ver stage (e.g. storm pulse). In th e Pinder et al. (1969) method, a set of type curves are generated for various obser vation wells around the river location. Each set of curves represents the computed change in head in the aquifer due a change in river stage when a selected diffusivity is assumed (Pinder et. al, 1969). The best fi t diffusivity of the aquifer is determined by choosing the curve that best matc hes the response in the observation well. The analytical solution can be appl ied to a flood stage hydrograph of any shape. In their study, the analytical solution was applie d to an aquifer at Musquodoboit Harbour, Nova Scotia. Values calculated with the solution were on the same order of magnitude with those obtained from 33

PAGE 34

pumping tests. Values were different for different observation well locations, which was attributed to the heterogeneity of the aquifer between the river and the observation well (Pinder et al, 1969). Another alternative method usi ng borehole lithologies to esti mate aquifer transmissivity was demonstrated in the unconfined region of the Columbia aquifer (Andres and Klingbeil, 2006). Boreholes were drilled to penetrate the entire thickness of the unconfined aquifer, and then lithologies were classified into categories (sand, sand and gravel, silt, etc). Lithologies were assigned a typical hydraulic condu ctivity value based on previous studies of sediment in the region. In areas where the borehole did not reach the base of the aquifer, a hydraulic conductivity value was assigned that was repres entative of the local average of all materials. Transmissivity was then calculated as the sum of the produc ts of material thic knesses (b) and hydraulic conductivity (k). This method would be difficult a nd expensive to apply in the OLeno study area due to the costs of drilling a hol e through the entire unconfined Flor idan aquifer. Data from one borehole exists from the upper portion of the unconf ined aquifer in the study area; however, not enough of the core was recovered during the dril ling to make an accurate estimate of hydraulic conductivity. Another concer n with this method would be that using typical values for hydraulic conductivity will underestimate a nd overestimate values in lo calized areas (Andres and Klingbeil, 2006), and would likely no t represent the scale effect that is common in karst aquifers at the basin scale. 34

PAGE 35

Figure 2-1. Location of th e Santa Fe River. 35

PAGE 36

Figure 2-2. Map of OLeno State Park with lo cations of surface waters, mapped conduits, and observation wells 36

PAGE 37

Table 2-1. Geologic and Hydrogeologic Units of the Santa Fe River Basin based on Hunn and Slack (1983), and Hisert (1994). 37

PAGE 38

CHAPTER 3 METHODS Data Collection Data collected during this study includes daily precipitation, river stage, and ground water levels. Daily precipitation was recorded at the Oleno State Park au tomated gauging station and was accessed through the Suwannee River Water Management district database. The automated station records precipitation on an hourly basis and total pr ecipitation for a 24 hour period can be viewed in real time for 30 days; archived daily precip itation data is available in the database. The current record of precipitation downl oaded for this and previous studies dates back to July 2000. River stage measurements were recorded daily by Oleno State Park personnel from one of two staff gauges located approximate ly 0.5 km upstream of the River Sink at the OLeno State Park suspension bridge. The stage ga uges at this location ar e divided into a lower and upper gauge, with a gap of 2.10 m where no read ings are available. The top of the lower gauge stops at 12.41 meters above sea level (mas l), while the upper gauge is available between 14.50 masl and 16.55 masl. For days when the gauge was not read, or during time periods when the water level was between the two gauges, data records were filled in using a long term correlation between stage at the gauge location and water level elevati on recorded by a data logger at the River Sink (described below). Data were collected from twelve monitoring wells placed at various distances away from the conduit system. Eight of these wells were in stalled to be at approximate conduit depth, 30 m below ground surface (bgs) (Table 3-1). Six of th ese eight wells were installed in January 2003, while Well 5 was installed in March 2003 and Well 8 was installed in May 2004. The other four wells were installed in January 2006 and are finished to depths just below the water table at their respective locations. Each of these wells is cons tructed of 2-inch PVC piping with a screened 38

PAGE 39

interval of 20 ft (6.10 m) for the deep wells and 10 ft (3.05 m) for the shallow wells. Wells were completed above the ground surface and installed with a protective casing to prevent outside disturbances. The elevations of the top of the we ll casing were determined from surveying from a known benchmark. Deep wells installed in 2003 we re professionally surveyed, while shallow wells installed in 2006 were surveyed by members of this project from the adjacent deep well. Ground surface elevations were determined from measuring from the top of the well casing to the ground using the survey rod. In this area, bedr ock was assumed to be the first consolidated, carbonate rich material encounter ed during drilling (Table 3-1). All well locations with the exception of Well 1 have less than 6 m of unconsolidated sands that lay above the carbonate rock. Estimated depth to bedrock may vary between shallow and deep wells (Table 3-1) at the same location due to wells being drilled by differe nt drill teams with diffe rent descriptions of rock encountered during drilling. De eper depths most likely repres ent the consolid ated limestone rock, as opposed to the first appearance of unc onsolidated pieces of limestone; therefore the deeper depths will be used at locations where there are discrepancies. Water Levels Water levels were recorded at surface wa ter locations and observa tion wells using three different types of automated water level recorder s: In-Situ Minitroll Logg ers (accuracy of +/.02 m), Van Essen Divers (accuracy of +/.005 m), or Van Essen shallow/deep CTD Divers (10 m divers accuracy of +/.01 m, 30 m divers accuracy of +/.03 m). CTD divers with 10 m range were installed in shallow wells and surface water locations, while CTD divers with 30 m range were installed in deep wells. Data loggers pla ced at the River Sink, Rive r Rise, and Sweetwater Lake were installed in screened 2-inch PVC pi pes that extend below the surface water and are accessible from the bank. Data loggers in the wells were attached to the well cap via coated stainless steel wire and allowed to hang in the well. All data loggers were set to record data at 39

PAGE 40

10-minute intervals. Since October 2005, data were downloaded on a four to five week basis to minimize instrument drift. Before then, time be tween data downloads va ried with up to eight weeks between downloads. The long term records of water levels vary for the locations within the study area (Table 3-2). This is mainly a re sult of the addition of new equipment over the course of the data collection, moving equipment to new wells after they were constructed, and the malfunction of existing equipment over the cour se of data collection. Water levels at the River Rise have been measured nearly continuously since August 2001 about 250 m downstream from the conduit outflow location and have been re ferenced to gauge values at the River Rise. For each recording interval, water pressures measured by the data loggers were corrected using the ambient barometric pre ssure recorded by a Baro Diver (accuracy of +/.0045m). Water pressures were then referenced to the water elevation determined manually at the time the data were downloaded. Manual water levels were measured using an electronic probe in the wells, or by reading a staff gauge and/or surveying at the surface water locations. Instrument error as a result of drift and/or shifting of the loggers within the PVC casing was calculated for each sampling period by subtracting the wa ter level calculated from the l ogger data at the start of the interval from the manually measured water level at the start of the inte rval. Errors at the well locations were consistently less than 0.03 m, while errors at surface water locations were generally less than 0.035 m. Errors due to shifti ng of the loggers are generally expected to be lower at the well locations because data loggers have less ability to move around within the well casing. Surveying error is estimated to be 0.015 m at the surface water locations. This is determined by surveying from the known benchmar k to the water level, then completing the survey loop back from the water level to the be nchmark. Water levels were only surveyed during 40

PAGE 41

times when the staff gauge was unreadable or the surface water was below the known benchmark. Sink and Rise Discharge Discharge rates into the River Sink were calculated by converti ng water levels to discharge using one of two rati ng curves. The primary rating curve (Figure 3-1) is based on a rating table created by the Suwannee River Water Ma nagement District for the station located on the Santa Fe River at OLeno State Park (R ating No. 3 for Station Number 02321898). This curve was used during all periods of record wher e the water level from the gauge at the OLeno State Park bridge was measured. The second rating curve (Figure 3-2) was used as a backup for those periods where the OLeno gauge data was not available, such as days when the gauge was not read. The second curve was created by correla ting the water levels measured at the OLeno State Park gauge with the water levels measur ed by the logger ~1000 m downstream at the River Sink for records since 2002. The correlation was done by plotting 142 da ta points of Oleno Stage versus sink logger water level and then fitting a 2nd order polynomial trend line through the data (Figure 3-2). The trend line produces a R2 value of .9994. It is important to note that the majority of the data points used for the correl ation are between logger wa ter levels of 10 to 12 meters, and therefore this is the most accurate portion of the curve. Above 12 meters there are only a handful of data points, and thus the curve is limited in its predictive capability. This is not a problem for this study, as discharges were mainly determined from the primary rating curve, with the second curve serving only to fill in a few scattered mi ssing points over a few years. Discharge rates out of the River Rise were calculated base d on the rating curve produced by Screaton et al. ( 2004) (Figure 3-3). This rating curve was calculated from measured water level elevations and unpublished discharge data provided by the Suwannee River Water Management district. All of the data used to creat e this rating curve were at or below water levels 41

PAGE 42

of 12.5 masl, with the majority of the points between 9.5 m and 11.5 m. This happens primarily because it is difficult to make discharge measur ements during high water levels, which usually coincide with large storm events. The result is that discharge will be less accurate at high water levels where there are less data. It is also important to note that the discharge curves may change over time if the morphology of the region around th e River Sink and Rise changes. For example, discharge may decrease if the river channel were to widen because there would be a larger area for flow. Over the course of time, various reference po ints have been used to determine the water level at the River Rise. This o ccured as a result of reference poi nts becoming altered from natural events, or unusable from equipment malfunction. For example, starting in mid 2003 the stage gauge at the River Rise became increasingly warp ed, making it extremely difficult to obtain an accurate reading. After the collapse of the gauge in mid 2005, a new gauge was installed by the Suwannee River Water Management District in February 2006. Erro rs like the slow warping of the gauge at the River Rise are difficult to asse ss because they are not always noticeable right away and can compound over time. For this study period, the warped gauge may have contributed to error in the obs erved water level at the River Rise between July 2003 and June 2005. Because there is no clear correction for the slow warpage of the gauge, discharge values during this time will not be used for calculations. Water levels observed at the River Rise after the installation of the new gauge have shown some inconsistencies when compared to the wells in the surrounding areas. During late 2006, water levels were reaching the end of a long recession, and by early 2007 the water table has essentially flatlined at Wells 3, 4 and 5. Comparing the water le vel of the wells and the Rise during this time indicates that the Rise is slightly higher than th e wells. Hydraulically this would 42

PAGE 43

indicate that water is flowing from the rise towards the wells during very low flow. However, this is likely not the case as water was still fl owing out of the River Rise during this time. The possible cause of this error was thought to be the survey of the newly installed gauge. To investigate this possible error, water level was recently surveyed from the most accessible well location (well 6) to the river ri se, and compared to the reading from the new gauge. Results of the survey indicated that the surveyed water le vel was 0.11 ft less than what was read from the stage gauge at the same time. W ith a difference of 0.11 ft, the water level of the rise would be even or less than the surrounding wells. This woul d be consistent with flow towards the rise during very low conditions. During early 2007, the River Sink was experiencing a long recession, and reached water levels similar to those observ ed after a long recessi on during the first half of 2002. Comparing the manually measured water levels of the Sink and the Rise during both these times indicates that the River Rise was up to 0.3 ft higher in 2007 than it was in 2002 relative to the River Sink. If the relationship between the River Sink and Ri ver Rise has not changed over time, than this suggests that the error in the Ri se water level could potentially be as great as 0.3 ft. It is important to note that this error could also be caus ed in part if there has been a lowering of the benchmark at the River Sink. This benchmark has not yet been resurveyed. Because an error of 0.3 ft in the Rise water level w ould be enough to cause a noticeabl e decrease in Rise discharge, all calculations made using the discharge va lues were completed for observed values of discharge and for hypothetical disc harges based on a water level decreased by 0.3 ft from the measured value. This potential survey error wi ll only affect the discha rge and area calculations, while the rest of the work only relies on changes in head. 43

PAGE 44

Evapotranspiration Evapotranspiration (ET) was calculated using the Pe nman-Monteith Model for determining water lost to the atmosphere from a vegetated surface. This method adapts Penmans original model for evaporation from a freewater surface by incorporating a canopy conductance factor (Monteith, 1965): )1( )1(* canat vw a aataanCC WeCcR ET = slope of relation between satura tion vapor pressure and temperature Rn = net radiation input a = density of air ca = heat capacity of air Cat = atmospheric conductance ea* = saturation vapor pressure in air Wa = relative humidity w = density of water v = latent heat of vaporization Ccan = canopy conductance = psychrometric constant Required measured daily inputs for this m odel include temperature, relative humidity, average solar radiation, and wind sp eed. These data were recorded at a weather station located in the city of Alachua (approximately 15 miles from OLeno State park), operated by the Florida Automated Weather Network (FAWN). These da ta were archived from 1997 and can be accessed through the FAWN database generator (h ttp://fawn.ifas.ufl.edu). Along with the daily measured inputs, the most important part of th e Penman-Monteith model is the calculation of canopy conductance. Canopy conductance assumes th at a reasonably uniform vegetated surface can be represented as a single big leaf whose total conductance to water vapor is proportional 44

PAGE 45

to the sum of the conductances of millions of little leaves (Dingman, 2002). To calculate canopy conductance, the following equation is used: Ccan = fs LAI Cleaf Leaf Area Index (LAI) is the relative size of the hypothetical big leaf and is determined from the total area of leaf surface above a gr ound area. LAI can range from 1.0 to 6.0 based on the type of climate and vegetation. Federer et al (1996) provides typical values of LAI from various environments; a value of 3.0 was assumed for the OLeno State Park study area based on the similarity to a savannah/shrub type land cover with an average height of vegetation near eight meters (Federer et al., 1996). Shelter factor (fs) accounts for the fact that some leaves are sheltered from the sun and wind and thus transpir e at lower rates. Shelte r factor ranges from 0.5 to 1.0 and a value of 0.5 was chosen as an estima te for a completely vege tated area (Allen et al, 1989). Leaf conductance (Cleaf) is determined by the number of stomata per unit area and the size of the stomatal openings. The most important controlling factor of leaf conductance is the soil-moisture deficit (Stewart, 1988). The soil-m oisture deficit is represented by a non-linear function in the solution for leaf conductance (Figure 3-4). Soil moisture deficit used for the Penman -Monteith solution was determined by tracking the soil moisture storage over time. This firs t required estimating th e maximum amount of soil moisture possible in this study ar ea. Different types of soil have the ability to retain different amounts of water. This amount is commonly represented by the term called field capacity, which is defined as an index of water content that can be held against the force of gravity (Dingman, 2002). Field capacity (volume of water in a soil/ volume of soil) is usually determined in the laboratory, and is represented as a unitless numbe r. Field capacity can ra nge from 0.1 for sands 45

PAGE 46

to 0.3 for clays (Dingman, 2002). The surficial aqui fer in the Santa Fe Sink-Rise study area is mostly composed of sands, and therefore was represented by a field capacity of 0.1. The soil moisture storage is also dependent on the vegetation that pulls water from the vadose zone. This requires understa nding the average depth of the root zone of the fauna in the study area. Gilman (1991) discusses common tree care, and points out that the majority of tree roots occur in the upper foot of the soil. Th is is supported by Rindell (1992) who discusses common misconceptions about tree roots, and notes th at most tree roots rare ly grow beneath four feet. Stewart (1988) applied the PenmanMonteith model to Thetford forest in Norfolk, England and determined that tree roots do not exceed one meter. Vegetation in the OLeno study area consists of a mix of oak, pine, and palmettos tr ees, along with many cypre ss trees along the river bank. Vegetation ranges in height, gi rth, and leaf size, and therefore in this area most likely has a tree root depth zone that does not exceed one meter on average. Therefore, the maximum soil moisture storage used in the model (10 cm), wa s calculated by multiplying the estimated depth of the root zone (1 m) by the field capacity (0.1). In order for the calculate d evapotranspiration to occur on a given day, there must be enough water av ailable in the soil moisture storage. When soil-moisture deficit reaches a value of 10 cm there is no longer any water available for evapotranspiration. The effect of the soil moisture deficit on the evapotrans piration calculation is governed by the function for a soil moisture defic it of 10 cm (Figure 3-5). Sensitivity analysis of the 10 cm estimate indicates that overestimati ng the maximum possible soil moisture deficit would cause little error to the ET calculation during the majority of times. This is most likely because there is usually plenty of water available for evapotranspiration in this region. Overestimating the maximum possible soil moisture deficit during very dry times would result in under predicting the amount of eva potranspiration. This can be as significant as a few cm per 46

PAGE 47

year if the soil moisture remains close to maxi mum deficit throughout the year, to less than 1 cm per year if the soil moisture was only near maximum deficit sporadica lly throughout the year. This was determined by increasing the maximum possible soil moisture during both a very dry year and a very wet year and comparing the differences. Recharge Water Budget Method In the unconfined Santa Fe River basin, nearly all of the basin drainage occurs through the subsurface, and runoff does not generally occur. This happens because little channelized surface drainage is present, the so ils are permeable, and the slope of the land surface is gradual to flat (Grubbs, 1998). Therefore, in this setting recharge is dependent only on the amount of precipitation and evapotranspiration that occu rs, and the maximum soil-moisture storage possible. Daily precipitation is ad ded to a daily running total of soil moisture. Recharge is only possible when the soil moisture storage exceeds its estimated maximum capacity of 10 cm. Once the soil moisture is full, recharge is calc ulated by subtracting the amount of calculated evapotranspiration from the measured precipita tion. Recharge was calculated on a daily basis allowing for quantification of recharge over indivi dual storm events (days) and also over the long term (annually) by summing the daily values. Water Table Fluctuation Method Minor water level perturbations observed in the hydrograph record of the shallow wells were common after small rain events. These small rain events occur spor adically during periods of extended dryness, and do not significantly alter the shape of the River Rise receding hydrograph. During these low-flow periods, there is no movement of water from the conduit to the matrix, and thus the small change in water le vel should be attributed solely to water arriving at the water table via diffuse recharge. The wa ter table fluctuation method assumes that the 47

PAGE 48

change in water level is relate d to the specific yield of the aquifer (Healy and Cook, 2002). It also assumes that all of the recharge arrives at th e water table, with no loss of water due to lateral flow through the unsaturated zone: R = hSy h = Change in head Sy = Specific yield R = Recharge When applying the water table fluctuation me thod, the maximum change in head during an observed perturbation was calcula ted by projecting the recession cu rve path and determining the maximum distance between the peak increase an d projected path. If la teral flow through the aquifer occurs at the same time scale as wa ter movement downward through the vadose zone, this method will underestimate recharge. Chloride Concentration Method In this region, the only source of chloride entering the unconfined Floridan aquifer is from rainwater. Chloride concentrations measur ed at shallow well locations are consistently much higher than typical chloride concentrations found in rainwater. These concentrated values are a direct result of evapotrans piration occurring from the top of the water table. The amount of evapotranspiration is quantifiable by first calculating a c oncentration factor, which is the ratio of chloride concentration in rainwater to the chlo ride concentrations measured at each of the shallow wells: Concentration factor = [Clrainfall] [Clwell] The concentration factor can be then used to determine how much of the precipitation has actually been evaporated. For example, a con centration factor of 0.25 indicates that the evapotranspiration process has caused a 4 times in crease in chloride concentration, and thus 0.75 48

PAGE 49

of the rainwater must have evapotranspired. Recharge can then be calculated by subtracting the calculated evapotranspiration from the measured precipitation amount. Matrix Transmissivity Analytical Method without Recharge The analytical method developed by Pinder et al (1969) was used to make a first approximation of matrix transmissivity in th e Santa Fe River Basi n. This method does not include recharge but was applied so that a comparison to results produced by Martin et al. (2006) was possible. This method calcula tes theoretical changes in he ad at an observation well a distance away from a known conduit. The input si gnal, which is the wate r level at the conduit boundary, is broken into increments and then the incremental change in head is calculated. The total change in head is calculated by summing the increments: hm= change in head of well per time step (m) Hm= change in head of conduit per time step (m) x = distance from well to conduit (m) v = diffusivity (T/Sy) (m2/d) (Sy assumed to be 0.2) t = time step (days) Distances from the conduit to the well locations were determined using the measure tool on a georeferenced map created in Arcview GIS. Distances represent the shortest distance from the observation well to mapped/inferred conduit locat ions. The greatest unce rtainty in distances occurs in areas where conduits are not mapped. This is most applicable to the wells north of Well 8 (Figure 2-2) where the exact conduit location has not yet be en mapped. A value of 0.2 was chosen by Martin et al. (2006) to represent the specific yield of the unconfined Floridan aquifer and therefore was also used in these calculations This value was based on porosity estimates in the Floridan aquifer by Palmer (2002) that range from 0.1 to 0.45. If the specific yield is higher 49

PAGE 50

than the predicted value of 0.2, then the calculate d transmissivity would decrease. The time step chosen for the calculations was one day. This analytical curve matching method allo ws for the use of any shaped flood stage hydrograph. Transmissivity is determined by adju sting values until calcula ted changes in head most closely match measured changes in head at observation wells. In unconfined aquifers, transmissivity will vary with time as the satura ted thickness of the aquifer changes with changes in head. However, the error intr oduced by using a constant transmi ssivity in this model is most likely minimal when comparing the relatively sma ll head changes we see over the course of a storm event (maximum of 3.5 m in matrix wells) and the assume d total thickness of the Upper Floridan aquifer (100 m). Transmissivity result s may be significantly di fferent when using a time-lag to match curves as compared to tryi ng to match the amplitude of the curve (Martin, 2003). For this study results are reported for both time-lag and amplitude curve matching techniques. 1-Dimensional Model with and without Recharge in MODFLOW The analytical method used by Pinder et al. (1969) uses head change in a boundary condition (conduit) to predict head change in observation wells at a known distance away. This solution allows for the calculation of diffusivity (transmissivity/storage), but does not include a recharge component. In order to observe how much of a difference including the recharge component makes on the calculation of transmi ssivity, a one-dimensional, transient finitedifference model with recharge was created in MODFLOW. A numerical method was selected for ease of use during calibrati on and sensitivity tests. Similar to the analytical method, the model uses head change at a boundary condition to predict head change at observation wells. Howeve r, the difference in the numerical model is the addition of recharge in the gr oundwater flow equations solved in MODFLOW. The ground water 50

PAGE 51

flow equation in one dimension for an unconfined aquifer with a recharge source is:(modified from Schwartz and Zhang, 2003): (Kx h ) + Q(x) = Sy h x x t Including recharge in the so lution allows for separation of the storage component (Sy) from the diffusivity. The result is that any va lue of recharge (represented by the Q in the groundwater flow equation) can be entered per time st ep. Change in head per time step is thus a result of the hydraulic conduc tivity (K), recharge (Q), and specific yield (Sy). A benefit of the model interf ace (Groundwater Vistas) is that it allows for adjustment of parameters to determine the best curve match. The error statistic of root mean square error was used as a calibration tool. Root mean square e rror is the average of the squared differences in measured and simulated heads, and is usually thought to be the best measure of error if errors are normally distributed (Anderson and Woessner, 2002) The values used for the initial conditions, model parameters, boundary conditions, and other details are further ex plained in the next chapter. Recession Curve Analysis Transmissivity was also calculated in th is study area using a recession curve analysis method based on the work of Shevenell (1996) and Powers and Shevenell (2000). Recession curves from observation well hydrographs were br oken down into three segments. The first and steepest slope represents the domin ant effects of drainage of the larger karst feat ures (conduits), while the second slope represents the emptying of well-connected fractures (Shevenell, 1996). The third and broadest slope repr esents the slowest drainage of the matrix rock. Each of these 51

PAGE 52

segments have a characteristic slope ( ) for any given storm event, and the slope is defined by the equation: = ln (Y 1 /Y 2 ) = ln (Q 1 /Q 2 ) t2-t1 t2-t1 = slope of recession segment (m/day) Y1 and Y2 = water levels at time 1 and time 2 (m) t1 and t2 = time after onset of recession co rresponding to water levels (Day) Q1 and Q2 = associated theoretical flows (dis charges) corresponding to the water levels (m3/day) Solving for Q1/Q2 in the above equation allows for the ra tios of the theoretical flows to be calculated, where the ratio of Q1 to Q2 represents conduit-dominate d drainage, the ratio of Q2 to Q3 represents fracture-dominated flow, and the ratio of Q3 to Q4 represents matrix-dominated flow. The theoretical flows (Q values) are depend ent on the water levels, and therefore the ratio between two respective theoretica l flows can be determined from the ratio of two respective water levels. The ratio of Q3 to Q4 can be applied in the solution adapted by Atkinson (1977) for non-conduit transmissivity of an unconfined aquifer from a baseflow recession curve: Log (Q3/Q4) = T/S (t2 t1) 1.071/L2 T = Transmissivity (m2/day) S = Storage coefficient (unitless) L = distance between monitoring well and the groundwater divide (m) In our study area, the aquifer is unconfine d and thus the storage coefficient is approximately equal to the specific yield. A valu e of 0.2 was used for specific yield in the calculations. A value of 5000 m was used as an estimated distance between the monitoring wells and the groundwater divide. Transmissivity wa s calculated by solving for T in the equation. 52

PAGE 53

Figure 3-1.River Sink rating curve produced by th e Suwannee River Water Management District for the stage gauge located in OLeno State Park. 53

PAGE 54

Figure 3-2. River Sink rating curve produced in this study using sink elevations measured by the data logger correlated with stage at the River Sink gauging station. 54

PAGE 55

Figure 3-3. River Rise rating cu rve produced by Screaton et al (2004) using data from the Suwannee River Water Management District. 55

PAGE 56

0 0.2 0.4 0.6 0.8 1 024681 0 Soil moisture deficit (cm) ( ) = 1.00119*exp (.61* ) Figure 3-4. Function assuming a maximum soil-moisture deficit of 10 cm used to calculate leaf conductance in the Penman-Monteith evapotranspiration method. After Stewart (1988). 56

PAGE 57

Table 3-1. Summary of Monitoring Wells Completed Depth (m bgs) Screened Interval (m bgs) Depth to bedrock (m bgs) Ground surface elevation (m asl) Well 1 22.85 22.85-16.75 17.07 14.45 Well 2 30.48 30.48-24.38 6.10 15.96 Well 3 28.35 28.35-22.25 3.05 17.87 Well 4 29.57 29.57-23.47 4.57 17.89 Well 4a 9.75 9.75-6.70 5.18 17.96 Well 5 29.87 29.87-23.77 5.48 16.22 Well 5a 8.23 8.23-5.18 3.05 16.20 Well 6 31.09 31.09-24.99 4.88 13.51 Well 6a 5.49 5.49-2.44 3.96 13.55 Well 7 29.87 29.87-23.77 5.48 15.22 Well 7a 7.62 7.62-4.57 2.43 15.19 Well 8 30.48 30.48-24.38 3.05 13.32 57

PAGE 58

Table 3-2. Location and dates of data collection. D = Diver, C = CTD Diver, T = Minitroll logger, B = Baro Diver. Periods of no data indicate time when loggers were either malfunctioning or not installed Well 1 Well 2 Well 3 Well 4 Well 4a Well 5 Well 5a Well 6 Well 6a Well 7 Well 7a Well 8 Tower well River Sink River Rise Sweetwater 1/22/032/26/03 D D D C C C 2/26/033/27/03 D D D C C C 3/27/035/14/03 D D B C C 5/14/037/24/03 D D B C C C 7/24/039/12/03 D D B C C C 9/12/0311/21/03 D D B C C C 11/21/03 2/9/04 D D B C C C 2/9/04 4/7/04 T D T D B T T T C C C 4/7/04 6/9/04 T D T D B T T T C C T 6/9/04 7/7/04 T D T D B T T T C C T 7/7/04 9/2/04 T D D B T T T C C T 9/2/04 2/4/05 T D B T T T C C 2/4/05 5/18/05 T T D B T T C 5/18/057/18/05 T T B D T T C 7/18/0510/27/05 T T B D T T C 10/27/0512/13/05 T T B D T T C 12/13/05 1/17/06 T T B D T T C 1/17/06 2/21/06 T T B D T T C 2/21/06 4/11/06 T T B T C C C C C C C C 4/11/06 5/1706 T T B T C C C C C C C C C 5/17/06 6/15/06 T T B T C C C C C C TB C C C 6/15/06 7/12/06 T T B T C C C C C C TB C C C 7/12/06 8/17/06 T T B T C C C C C C TB C C 8/17/06 9/15/06 T T B C T C C C C C C TB C C 9/15/06 10/20/06 T T B C T C C C C C C TB C C 10/20/0611/29/06 T T B C T C C C C C C TB C C 11/29/06 1/10/07 T T B C T C C C C C C TB C C 1/10/07 2/19/07 T T B C T C C C C C C TB C C 2/19/07 3/19/07 T T B C T C C C C C C TB C C 58

PAGE 59

CHAPTER 4 RESULTS Precipitation and Evapotranspiration Daily precipitation records from OLeno Stat e Park are plotted from April 2004 through April 2007 (Figure 4-1). Average annual precipit ation during this period was 130 cm/yr (Table 4-1). This value is slightly less than the average precipitation of 137 cm/year recorded in the Santa Fe River Basin by Hunn and Slack ( 1983) between the years of 1900 and 1976. Higher than normal precipitation occurred in 2004, when 187 cm were recorded. This was primarily a result of two September hurricane events that were responsible for 50 cm of precipitation. In contrast, only 84 cm of precipitation occurred at OLeno St ate Park during 2006, which is significantly lower than average values. During this time the north-central Florida region was experiencing a significant drought. Evapotranspiration (ET) wa s calculated on a daily basis from January 1, 2002 through April 2007 using the Penman-Monteith method and is plotted for this study from April 2004 through April 2007 (Figure 4-2). A value of zero on the plot indi cates that there was not enough water available for evapotranspiration to occur. Monthly and a nnual ET values were calculated by summing daily values. The average annual ET calculated since 2002 was 76 cm/yr (Table 41). The lowest annual calculated ET was 66 cm which occurred in 2003. The highest annual calculated ET was 85 cm, which occurred in 2005. Th ese values are lower than values of 90 to 105 cm/yr calculated at 96 rainfall stations in south-central Georgia a nd across the state of Florida by Bush and Johnston (1988). The lower cal culated values in this study are most likely caused by a combination of the differences in the methods used and the scale of the calculation. The method used by Bush and Johnston (1988) us ed the mean annual temperature to estimate potential evapotranspiration, and then estimated the actual evapotranspiration based on an 59

PAGE 60

estimated ratio of potential evapotranspiration to precipitation. The Pe nman-Monteith model used in this study is not only dependent on temperature, but there are many other components included in the calculation. One of the most impo rtant is the soil-moisture deficit to track the amount of water available per day for evapot ranspiration. The Penman-Monteith model will produce a value of zero when ther e is not enough water available for evapotranspiration from the soil moisture storage. This will likely cause a mo re accurate prediction of the actual water that evaporates. The scale of the calculation also is important because Bush and Johnston (1988) calculated evapotranspiration for a much larger region than this st udy. In their calculation, evapotranspiration was averaged from land and open water regions across south Georgia and north Florida. The Penman-Monte ith model incorporates evapot ranspiration only from the land surface and was applied only to the area around OLeno State Park. Monthly potential ET (PET) was calcula ted using the Thornthwaite method for comparison (Figure 4-3). For the three year pe riod during this study, th e average annual PET calculated with the Thornthwaite method is 102 cm/yr. This value agrees with the values calculated by Bush and Johnston (1988). This is mo st likely a result of both methods being based on mean temperatures. Monthly ca lculations of PET and ET usi ng the respective methods show distinct seasonal trends, with increased values in the summ er months and less in the winter months (Figure 4-3). Comparing both methods shows that PET calcu lated with the Thornthwaite method is greater than calculated ET with the PenmanMonteith method during the summer months, by up to 5 cm in July and August. Result s also show that during the winter months the ET calculated by the Penman-Monteith method is sometimes estimated to be higher than the PET calculated with the Thornthwai te method. This is likely a resu lt of the differences in the method. Xu and Chen (2005) worked with three me thods that calculate ET and compared them 60

PAGE 61

to four methods that calculate PET, one of which was the Thornthwaite method. Although the Penman-Monteith method was not included in their research, results indicated that ET calculated with other methods can sometimes be greater th an PET calculations made with the Thorthwaite method. Lemur and Zhang (1990) evaluate three methods of calculating ET, and their results demonstrate that the evapotranspiration calcu lated with Penman-Mont eith reflects values calculated from a water balance better than th e other two methods. The Penman-Montieth model has become the most widely used approach at estimating evapotranspiration from land surfaces (Dingman, 2002), and therefore would likely represen t the best estimate of evapotranspiration in this region. Water Levels River Rise data have been used in various prior studies since 2001 (S creaton et al, 2004; Martin et al, 2006) (Figure 4-4). Water levels were also recorded at the River Sink and Sweetwater Lake locations during this time; however, records were interrupted at various points due to logger malfunction and/or lack of logger ava ilability, and therefore th e record is much less complete (Table 3-2). Water levels recorded sp ecifically during this study were between April 2004 and April 2007 (Figure 4-5). During this reco rding period, there were three storm events distinctly displayed by three larg e peaks in the River Rise hydr ograph (Figure 4-5). These storm events were also observed in fl uctuations of water levels at various monitoring wells (Figure 45). Increase in head at the monitoring wells ra nged from as high as 3.5 m during the largest September 2004 event, to 1 m during the smalle r December 2005 event. Similar increases in head were observed at the River Rise. Water le vels at Well 1 are greater than the River Rise. This is because Well 1 is influenced by the change in head at the River Sink, which is upgradient of the River Rise and thus has higher water leve ls. The maximum water level at the River Rise 61

PAGE 62

occurs slightly before the maximum water level at most of the monitoring wells, indicating that there is a small time-lag for the head change to reach the wells. Sink and Rise Discharge Stage at the River Sink and the River Rise were converted to daily discharge rates based on the appropriate rating curve. Daily discharge rates are plotted for this study between April 2004 and April 2007 (Figure 4-6). The blue highli ghted area of the fi gure indicates the time period during this study where th e gauge at the River Rise wa s becoming increasingly warped. Discharge rates were also corrected for a possibl e 0.3 ft error and are also plotted in figure 4-6. The corrected water level values shift the Rise discharge curve slightly lower. During this time period, there were three stor m events which increased discharge (Figure 4-6). During these storm events, the discharge in to the River Sink is equal or greater than the discharge out of the River Rise. This indicates that some of the storm water which entered the Sink is leaving the conduit, presumably to the matrix. The idea that water moves into the matrix during these large events is also supported by the fact that change in head at the monitoring wells during storm events is much larger than th e amount of calculated r echarge divided by the specific yield. In contrast, during pe riods of low river stage, the disc harge out of the River Rise is noticeably higher than the discharge into the Ri ver Sink. This indicates that the conduit is gaining water along its flow path. Recharge Diffuse Recharge Calculated From Water Budget Excess precipitation was calculated daily from 2002 through April 2007 by subtracting evapotranspiration calculated with the Penman -Monteith method from precipitation. Because runoff is minimal in the unconfined Floridan aqui fer, recharge was calculated using this excess precipitation and accounting for soil moisture storag e (Table 4-1). Daily recharges values are 62

PAGE 63

plotted for the study period of April 2004 to April 2007 (Figure 4-7). Most of the recharge during this time occurred within three time periods, wh ich are associated with the three large storm events (Figure 4-7). The largest single recharge peak (~20 cm) occurred during the September 2004 storm event. During 2006, there are very few in stances where recharge occurs (Figure 4-7). This is caused by less precipitation occurring dur ing this time (Figure 4-1). Comparing the daily recharge values with the precipitation record also indicates that much of the precipitation never actually makes it to recharge. Instead, this preci pitation is evapotranspired or occupies soil moisture storage and is later evapotranspired. Annual recharge relies heavily on the amount of precipitation that the area receives over the course of a year. Diffuse Recharge during Large Storm Events During large storm events, precipitation is si gnificantly greater than evapotranspiration, and soil moisture storage is likely to reach capacity quickly. The combination of a large diffuse recharge component combined with an input from the conduit, results in the largest increases in observation well hydrographs in response to these events (Figure 45). Diffuse recharge is shown for each of three events occurring after Ap ril 2004 (September 2004, March 2005, and December 2005), as well as the March 2003 even t analyzed by Martin et al. ( 2006) (Table 4-2). Each storm event varied in the amount of precipitation a nd ET, based on the time of the year and the magnitude of the storm. The calculated rechar ge during the September 2004 hurricane events was 41 cm, which occurred over a relatively sh ort period of time (24 days). In contrast, calculated recharge for the March 2003 storm event was 27 cm and for the December 2005 storm event was 24 cm; both of these results are nearly half the amount of r echarge spread out over double the amount of time (50 days). The smalle st event was in March and April 2005, when only 10 cm was recharged over 28 days. The durati on of the storm events were chosen based on 63

PAGE 64

the water level records, with a at least three to four days included before the increase in water levels and at least five to ten days included during the recession. Diffuse Recharge Calculated Fr om Chloride Concentrations Over the course of 2006, hydrographs from conduit locations and well locations both show a consistent recessional to flat trend (Figure 45). During these recessions, water may flow from the shallow well locations towards the conduit, and t hus it is likely that the shallow wells are not receiving any conduit water. Therefore, concentrat ions of chloride in the various shallow well locations should primarily reflect the evapotranspiration process, and can be used to make an estimate of the evapotranspirati on occurring at the well locations. Water samples were available from six sampling trips beginning in April 2006 and ending January 2007 (Moore, unpublished data). Shallow well bulk water chemistry was anal yzed after each sample period, and chloride concentrations are available for the four shallow wells (Table 4-3). The average chloride concentration for the six sampling periods was used to calculate a concentration factor for each shallow well loca tion (Table 4-3). Concentration factors were calculated assuming a chloride conc entration value of 1 mg/L for th e precipitation that occurs at Oleno State Park. This value was chosen from a contoured map created by Junge and Werby (1958) with data collected at va rious locations across the United States between July 1955 and July 1956. Chloride concentrations in Florida from Junge and Werbys (1958) study range from 0.66 mg/L over the Florida panhandle, to 2.44 mg/L near Miami. The closest location to OLeno State Park was near St. Augustine, with a ch loride concentration of 1.02 mg/L. The 1.0 mg/L value was chosen based on the 1.0 mg/L contour that falls over the OLeno State Park study area. The highest concentration factor determined fr om the chloride concentrations was .917, which occurred at well 5a. Concentration factors of well 4a and 5a are almost identical, and all of the concentration factors are within 6% of each other. If the chloride concentration is actually lower 64

PAGE 65

than is estimated from the contour map, then the resulting concentration factors would be higher. Differences in concentration factors may be a re sult of differences in the local vegetation near the respective well locati on, which directly affects the amounts of evapotranspiration that would occur. Differences may also be a result of the thickness of the vadose zo ne and thus the amount of water that can evapotranspire from the top of water table. Evapotranspiration was calculated for 2006 by comparing the chloride concentration factors determined at each well location to the total amount of observed precipitation (Table 44). Average evapotranspiration for the four we lls was 75 cm, which is consistent with the average calculated value of 76 cm/yr using th e Penman-Montieth method for the five-year period. This value is slightly higher than the 66 cm calculated for 2006 with the PenmanMonteith model. Annual recharge was then calc ulated for 2006 by taking the difference between total precipitation during 2006 and the calculated evapotranspira tion using the concentration factor. Calculated recharge valu es range between 7 cm and 12 cm, which is lower than the 17 cm calculated with the water budget method. Both re sults support that 2006 diffuse recharge was well below the average of 45 to 60 cm/yr calculated by Grubbs (1998). Diffuse Recharge Estimated Using the Water Table Fluctuation Method The water table fluctuation method is ofte n applied as a comparative method to other recharge calculations (Healy and Cook, 2002). This method is commonly used in regions where change in water level is attributed solely to wa ter arriving at the water ta ble via diffuse recharge. The water table fluctuation method assumes that th e change in water level observed at the water table is related to the amount of precipitation that reaches the water table divided by the aquifer storage (specific yield). During lo w-flow periods in the Oleno study area, discharge out of the River Rise is larger than disc harge into the River Sink. We expe ct that there is little to no movement of water from the conduit to the ma trix. During these times, the overall hydrographs 65

PAGE 66

are experiencing a recession. It is common that a sm all rain event will trigger a minor blip in the hydrograph, but will not change the long term recessional trend. Water levels at shallow well locations 5a, 6a, and 7a, which are screened at the water table, were monitored from the time of their installation on February 21, 2006 through April 2007. Because most of 2006 was dry, the overall trend of the hydrographs at the wells was a recession. However, over the summer of 2006 there were two rainfall events on record that caused small perturbations to water levels. The first event occurred in June 2006 when 12.3 cm of precipitation was recorded over three days. The diffuse recharge calculated w ith the water budget method for this event was 3.2 cm. The second event occurred in July of 2006 when 7.8 cm of precipitation was recorded over three days. This event produced 0 cm of recharge calculated with the water b udget method (Figure 4-9). During the time when these events occur, the River Ri se water level continued to experience a recession (Figure 4-10). During the recessi on, the River Rise water level in creases very slightly (<.05 cm) after the June event, which may suggest that the change in water levels observed in the matrix wells might be influenced by the conduit. For each of these small events, the raw 10 mi nute data were smoothed using an interval of 50 points (Figures 4-11 through 4-16). The recession curve was then projected from the smoothed line, and the difference was calculated between the projected curve and the maximum spike in water level. Recharge was estimated by taking the maximum change in head and dividing by the estimated specific yield of 0.2. Recharge calculated using this method is compared to recharge calculations using the water budget method (Table 4-5). For the June 2006 event, the water table fluctuati on method predicts significantly less recharge at a ll three shallow well locations than the 3.18 cm predicted by th e water budget method. In contrast, for the July 2006 event, the water table fluctu ation calculations show a small amount of recharge (between 66

PAGE 67

0.38 and 0.43 cm), while the water budget calculati on indicates that no recharge should have occurred (0 cm) (Table 4-5). Transmissivity Analytical Method without Recharge Martin et al. (2006) investig ated transmissivity in this study area using an analytical curve-matching method described by Pinder et al. (1969) that estimated monitoring well response to head changes in the conduit system; diffuse rechar ge was not included. This method was applied to monitoring well hydrographs from a storm event in March 2003 (Figure 4-4). Results showed that transmissivity between the conduit and various monitoring wells ranged by orders of magnitude, and was hi ghly dependent on the distance of the monitoring well to the conduit. For example, calculated transmissiv ity to Tower well (approximately 3700 m from the conduit; Figure 2-2) and Well 1 (approximately 500m from the conduit; Figure 2-2), were 160,000 m2/day and 97,000 m2/day respectively, while calculated transmissivity to Well 4 (approximately 115 m from the conduit; Figure 2-2) and Well 6 (approximately 140 m from the conduit; Figure 2-2) were 950 m2/day and 900 m2/day respectively (Marti n et al, 2006). The difference in transmissivity between the various well locations was attributed to the fact that water traveling a further distance through a hi ghly karstic system will be more likely to encounter preferential flow paths along the way. Martin (2003) also observed that calculated transmissivity could vary at a single well location depending on whether the calculated da ta was matched to th e observed data by correlating the timing of the hydrograph peak or by correlating the amplitude of the hydrograph peak. For example, when using the time lag as th e matching point, the estimat ed transmissivity at Tower Well was 120,000 m2/day, which is significan tly less than the 550,000 m2/day calculated when matching the curve based on amplitude. 67

PAGE 68

In this study, the Pinder et al (1969) method was applied to the three storm events that occurred after March 2003. Transmissivity was calculated between the River Rise and Wells 4, 6, and 7 (Table 4-6). Transmissivity could not be calculated between th e River Sink and Tower Well or Well 1 because the data logger was not operating at the River Sink during these times. However, Well 7 is located at a distance of 1025 me ters from the conduit, which is intermediate in distance between Tower Well and Well 1. Results from the curve matching between actual head change and calculated head change for th e three storm events are shown in the figures: September 2004 (Figures 4-17, 4-18, 4-19), March 2005 (Figures 4-20, 4-21, 4-22), and December 2005 (Figures 4-23, 4-24, 4-25). In these figures, head cha nge refers to the difference between water level on a given day and the water level on the first day of the time period being analyzed. For all three events, at least 1-3 days of small head change (< .02 m) exist prior to the start of simulations. For events that have two peaks in the hydrograph, the amplitude match was made with the first peak because the first peak on the hydrograph shoul d represent the initial pressure head change in the conduit that reaches the well after the onset of the event from a base flow condition. The following peaks occur as a re sult of precipitation that happens after the hydrograph has already increased sign ificantly, and thus is not starti ng from the base flow initial conditions. Transmissivity calculated for these events varied by orders of magnitude depending on the distance of the observation well from the conduit, as previously observed by Martin et al. (2006). For example, using the time lag match, the September 2004 event produced transmissivity values from Wells 4 and 6 (115 m and 140 m from the conduit) of 800 m2/day and 1200 m2/day respectively. These values are signifi cantly less than the transmissivity of 150,000 m2/day determined from Well 7 (1025 m from the conduit). Transmissivity also varied 68

PAGE 69

significantly when using a time-lag match as co mpared to an amplitude match. For the same September event, the amplitude-match produced a transmissivity of 350,000 m2/day from well 7 as compared to the time-lag calculation of 150,000 m2/day. 1-Dimensional Model with and without Recharge in MODFLOW The Pinder et al. (1969) analytical me thod allows for prediction of ground water fluctuations at monitoring wells based on water level changes in the nearby conduit, but does not include diffuse recharge. To j udge how much diffuse recharge contributes to storm event hydrographs and in the calculation of transmi ssivity, a one-dimensional, transient finitedifference model with a recharge component was created in MODFLOW. The model grid has 50 columns, with an initial spacing of 10 m clos e to the conduit and an increasing cell size by a multiplier of 1.1 away from the conduit (Figure 4-26). The spacing is refined near the conduit because this is the region where the model is most affected by changes in the conduit water level (Figure 4-26). The total length of the model is 11638 m. The total thickne ss of the aquifer was assigned to be 100 m. The stress period was set at 1 day with 10 uniform time steps applied to each stress period. The total time of the model was altered to work w ith each of the storm events, with the longest model duration being 70 days for the September 2004 event, and the shortest model being 29 days for the March 2005 event. The conduit was simulated as a transient constant-head boundary with water level values from the River Rise imported for each day of the event. In order to account for the differences between the conduit head and the head at the observation wells at the beginning of the simulation, change in head from the onset of the storm event was used at both locations. Thus, the initial head throughout the model at the beginning of all the simulations was set at zero. The boundary c ondition at the other end of the model (furthest point from conduit) was assigne d to be a no flow boundary. 69

PAGE 70

The specific yield of the unconfined aquifer was set at 0.2 for the model runs. This value was chosen as a best estimate of specific yield and to also to be consistent with all other calculations; sensitivity to the selected specific yi eld was tested as part of the modeling. In the model, three observation points were located at 110 m, 140 m, and 1020 m from the constant head boundary to represent Wells 4, 6, and 7 respectively. Observed data from each well for each storm event were compared with simulation results. The curve match that produced the lowest root mean square error was used to determine th e best fit hydraulic cond uctivity. Transmissivity was then calculated by multiplying the best fit h ydraulic conductivity determined in the model by the assigned aquifer thickness of 100 m (Table 4-6). The 1D model was first simulated for all f our storm events without including calculated recharge. Except for effects from the discreti zation of space and time, and the no flow boundary on one side, simulating the model without rechar ge should produce results identical to solving the Pinder et al (1969) analytical solution without rech arge. This is verified by results of a hypothetical simulation of the Pinder et al (1969 ) analytical solution and the numerical model without recharge (Figure 4-27). In the hypothetical simulation, bot h methods were simulated for a hypothetical observation well 1500 m from the c onduit location, with a th ree meter change in head in the conduit and a transmissivity of 100,000 m2/day. Figure 4-27 shows that predicted curves with both methods produce re sults that are very similar. Estimated transmissivity values calculated from the storm events with the analytical solution without recharge and the model without rech arge differ (Table 4-6) This is a result of the curves being matched differently between th e two methods. The model was calibrated to find a best fit hydraulic conductivity by using the lowest root mean s quare error between the observed 70

PAGE 71

and calculated heads at the observation well. Th e analytical solution was calibrated using a visual match of amplitude or time-lag. After simulating the system w ithout recharge and determining the best fit transmissivity, the model was then rerun and recalibrated with the calculated recharges included. To simplify the modeling, recharge was assumed to reach the aquifer on the day of occurrence and thus was applied to the time step when recharge was reco rded. The surficial sediments that make up the unsaturated zone are mostly sands, and are less th an 6 m thick at all well locations except Well 1 (~17 m thick). The maximum depth to the water table at the peak of the smallest event (December 2005) was 6.38 m, while during the larg er events the water table came extremely close to the surface (<.1 m during the September 2004 event). The small thickness of the surficial sediments and the proximity of the water table to the surface support th e idea that water should move through the unsaturated region and quickly become recharge This is also supported by the small June 2006 event where water level changes occur on the same day as estimated recharge occurs (Figure 4-8). However, model results (Fi gure 4-28 through 4-40) indica te that for three of the four storm events, including the diffuse r echarge component on the same day results in predicted curves that are pointed and jagged. Th ese results are not cons istent with the smooth curves that are seen in the observed data, and su ggest that the timing of the recharge input may be more spread out than is applied in the mode l simulation. The assumption that recharge arrives at the water table on the same day as precipitation occurs maximizes the impact of diffuse recharge in the calculation because the recharge is arriving in a lump sum rather than over a few days. The addition of recharge to the model simu lations results in predicted curves at Well 7 during the September 2004 and December 2005 events that plot higher than the observed data for 71

PAGE 72

these respective events (Figure 4-34, 4-40). A lthough these predicted curves produce the lowest root mean square error in the curve match, it seems unlikely that the predicted curve should vary so significantly from the observed data. Two possible explanations are availa ble. If the recharge is overpredicted, then the mode l would produce an unreasonable shift of the curve above the observed data. This is unlikely, considering that the same recharge has been applied to other well locations and this result is not observed. Th e second possible explanation may be that the influence of the conduit is overpredicted at We ll 7. Based on the location of Well 7 with respect to the River Rise (Figure 2-2), th is is possible if Well 7 is actua lly being influenced by the Santa Fe River at a location downstream of the River Rise; at this locat ion head changes in response to storms may be less and thus we would expect the calculated transm issivity to also be less. Comparing 1D model simulations with and wi thout recharge indicate s that the best fit transmissivity values are higher when the recharge is neglected. This result is seen in 10 out of 12 model simulations, with the only exception be ing the matches possibly in error at Well 7 (Table 4-6). These results suggest that in or der to match observed and simulated curves, the calculated transmissivity must adjust for the lack of recharge by increasing. Therefore, the curve matching techniques which neglect the recharge component will overp redict transmissivity from storm events. For the storm events in this study, the transmissivity required to make up for the lack of recharge can be up to four times greater than when recharge is included in the calculation (e.g. Well 6 during the September 2004 event). Consistent with results of the analytical method, results from both the 1D model with and without recharge indicate that calculated transmissivity can be one to two orders of magnitude higher for Well 1, Well 7, and Tower Well, than fo r wells closer to the conduit (Table 4-6). For example, model results at Well 7 fo r the September 2004 event are 300,000 m2/day, while the 72

PAGE 73

highest transmissivity calculated from We lls 4 and 6 during all simulations was 3,000 m2/day. Both the model with and without recharge also s how that the storm event with the highest change in head and most diffuse recharge (September 2004) produces higher calc ulated transmissivity values than the smaller events. For example, We lls 4 and 6 have a calculated transmissivity of 3000 m2/day from the September 2004 event, while the transmissivity values calculated at these locations from the other events does not exceed 1000 m2/day. Sensitivity to Specific Yield For all the calculations performed to date, a specific yield of 0.2 was used as a best estimate for the highly karstic, unconfined Floridan Aquifer based on porosity estimates from Palmer (2002). If this is not an accurate estimate, then the calculated transmissivity values would also be in error. To test the effect of changi ng specific yield on the tran smissivity calibration and to observe whether the specific yield value affect s the quality of the cu rve match, the 1D model with recharge was simulated with a smaller specific yield of 0.1, and a larg er specific yield of 0.3. Including recharge in the solution allows us to separate transmissivity and specific yield. The best fit transmissivity was once again chosen from the root mean square error of the curve matching (Table 4-7). Decreasing the specific yield to a value of 0.1 results in signifi cant changes to the calculated transmissivity from all four storm events (Table 4-7). For example, calculated transmissivity between the conduit and Well 4 during the March 2003 storm event is three times as much with the lower specific yield, while dur ing the March 2005 event the transmissivity to Well 4 is about four times less. However, more im portant than the change in transmissivity is that decreasing the specific yield to 0.1 significantly increases the ro ot mean square errors for all model simulations. This indicates that the specific yield of 0.2 yields a better fit to the data than a lower specific yield of 0.1. Results are perhaps more interesting when increasing the specific 73

PAGE 74

yield to a value of 0.3. For almost half of the simu lations, the root mean square errors decrease as specific yield is increased from 0.2 to 0.3. These results suggest that specific yield might vary between the conduit and the various well locations. Fo r the purpose of calcula tions at this scale, the model results generally confirm that the origin al assumption of an average specific yield of 0.2 is reasonable. Ideally, a sa mple set of more well locati ons spaced at various points throughout the matrix would provide a better esti mate of how specific yield varies across the study area. Finer sensitivity analys is would help pin down a more exact estimate of the specific yield at these locations. Recession Curve Analysis Unlike the Pinder et al. (1969) analyti cal curve matching technique and the 1dimensional model with recharge, the recession curve analysis method does not require a conduit signal as an input. Therefore, the benefit of this method is that it can be applied to calculate transmissivity at all well locations where an uninterrupted well hydrograph recession curve is available. However, it is also important to note that this method is used to calculate matrix transmissivity or what is described by Sheven ell (1996) as the slower hydrologic response of the intergranular porosity. This transmissivity is mainly derived from the third, and broadest sloped, segment of the recession curve. Therefore, these values do not necessarily indicate flow through all of the preferential flow paths, and therefore do not represen t transmissivity on the formation scale. For this study, 14 segments from various recession curves were available for analysis. These recession curves spanned all four storm ev ents since 2003 (Table 4-6). Length of recession varied for the different storm events, with th e shortest recessions coming after the March 2003 event (~40 days), and the longest recessions coming after the December 2005 event (~55 days). A specific yield of 0.2 was used for all of the calc ulations. Inflection points were chosen visually 74

PAGE 75

on the recession curve at the firs t and second noticeable change of slope (Figures 4-41 4-44). Finding the inflection points was difficult for th e March 2005 event, where a few small rainfall events caused slight perturbations in the recession curve (Figure 4-43). Results show that calculated transmissivity varies per storm event, with the highest transmissivity values being calculated for th e larger 2003 and 2004 events (Table 4-6). The transmissivity calculated for the larger events is on average twice the transmissivity calculated for the two events in 2005. This is similar to wh at is observed with the 1D model with recharge, where transmissivity values calculated from the September 2004 storm event were greater than the other three events. Perhaps more interesting is th at transmissivity is generally consistent at all the well locations on a per storm ev ent basis. For example, results do not show a transmissivity at Well 7 that is orders of magnitude larger th an at Wells 4 and 6, as observed in the previous analyses. This is likely a result of the met hod used, as the recession curve values are only indicative of the matrix transmissivity, as opposed to the effective transmi ssivity of the entire formation between the conduit and well. Basin Area Calculations Recession curves were also used in this study to estimate the aquifer surface area upgradient of the River Rise whic h contributes to the discharge of the Santa Fe River Sink-Rise system. This has previously been a difficult valu e to assess because groundwater divides are not clearly defined by water table mapping in this area. Knowing the surface area of the groundwater basin that discharges to the River Rise is valuable because it allo ws quantification of a volumetric input to the system via the diffuse recharge component. These values can then be compared to the volume that the aquifer receives from conduit losses during storm events. For these calculations, recharge to the confined porti on of the Floridan aquifer within the basin is neglected because all calculations are made using the assumed specific yield value of 0.2 for the 75

PAGE 76

unconfined aquifer. Making this assumption undere stimates overall area because the confined region will have a specific yield that is much smaller than the unconfined region. However, this assumption is acceptable in this study because we are only concerned with recharge to the unconfined portion of the aquifer. Recession curves following the storm events in March 2005 and December 2005 were used for the calculations. The time periods used for the analyses are November through December 2005 and March through December 2006. The recession curves following the September 2004 event were not used because of the possible error caused by the warpage of the River Rise water level gauge. The slopes of the recession curves were determined by finding the change in head per number of days of recession. Fo r the recession curves occurring as a result of the December 2005 storm event, the curves was br oken into segments which clearly varied in slope (Figure 4-45). The average slope of all the recession segments was 0.0078 m/day, and ranged from 0.0053 m/day to 0.0116 m/day depending on the storm event and the segment of the curve. In total there were 14 recession segments available to calculate area. Change in volume per day of the recession was calcu lated by taking the daily difference between the Rise and Sink discharge. Basin area was cal culated using the equation: Basin area = V / Sy h V = Change in volume (m3) Sy = Specific Yield h = Change in head (m) The surface area draining to the River Rise calculated from the average of 14 recession segments is 4.44 x 108 m2 or 444 km2 (Table 4-8). For the recession curves resulting from the December 2005 event, the calculated area varied for the upper and lower portions of the recession curves. The first portion (upper half be fore first inflection point) of the recession 76

PAGE 77

77 curves yielded calculated areas that were less th an the second portion (l ower half after first inflection point). The average area for the first portion of the curves was 356 km2, while the average area for the second por tion of the curves was 547 km2. A possible explanation for this result is that the specific yield is changing as the saturated thickness of the aquifer decreases during recession. During storm events, the water le vel reaches well into the surficial sediments. If these sediments have a different specific yield than the carbonate aquifer, then using a value of 0.2 for specific yield for both the upper and lowe r portions of the curve would result in the apparent difference in area. If the inferred change in area is due to change s in specific yield, we should see the curve inflection point occur at the same water level for each respective well location. However, this is not th e case, as the inflection points occur at different water levels after each storm event. Another pos sible explanation for the changing area is that there is a nonstationary groundwater divide. If the groundwater divide moves aw ay from the discharge point as the curve recedes, then we would see the increased area from the lower portion of the recession curves. The area calculations are sensitive to the accu racy of the discharge measurement. If the stage at the River Rise is incorrect, the resul ting discharge and calculate d area would change. To estimate a possible range of error in the area calculation due to su spected problems with the rise gauge, a calculation was made to see how the area changes if th e stage at the River Rise was overestimated by 0.3 ft. Calculations show that a decrease of 0.3 ft decreases the average discharge at the Rise and thus increases the difference between the Sink and Rise discharge. This only has a small effect on the calculated area, which would be 3.35 x 108 m2 or 335 km2. If some of the error is due to changes at the River Sink, we would expect a similar affect. Increasing the River Sink stage by .3 ft would result in a calculated area of 3.25 x 108 m2, or 325 km2.

PAGE 78

0 5 10 15 20 254/23/04 6/7/04 7/22/04 9/5/04 10/20/04 12/4/04 1/18/05 3/4/05 4/18/05 6/2/05 7/17/05 8/31/05 10/15/05 11/29/05 1/13/06 2/27/06 4/13/06 5/28/06 7/12/06 8/26/06 10/10/06 11/24/06 1/8/07 2/22/07 4/8/07Precipitation (cm)Date 78 Figure 4-1. Daily precipitation records fr om OLeno State Park during this study. Three storm events are highlighted.

PAGE 79

0 0.1 0.2 0.3 0.4 0.5 0.64/23/04 6/7/04 7/22/04 9/5/04 10/20/04 12/4/04 1/18/05 3/4/05 4/18/05 6/2/05 7/17/05 8/31/05 10/15/05 11/29/05 1/13/06 2/27/06 4/13/06 5/28/06 7/12/06 8/26/06 10/10/06 11/24/06 1/8/07 2/22/07 4/8/07Evapotranspiration (cm)Date 79 Figure 4-2. Daily calculated evapotranspira tion during this study using the Penman-Montei th Model. Spaces in the data indicate a value of zero.

PAGE 80

0 5 10 15 20Apr-04 Jun-04 Aug-04 Oct-04 Dec-04 Feb-05 Apr-05 Jun-05 Aug-05 Oct-05 Dec-05 Feb-06 Apr-06 Jun-06 Aug-06 Oct-06 Dec-06 Feb-07 Apr-07 Actual ET Potential ETEvapotranspiration (cm)Month 80 Figure 4-3. Monthly calculated potential eva potranspiration using the Thor nthwaite method and monthly actual evapotranspiration using the Penman-Monteith method

PAGE 81

81 Figure 4-4. Long term record of River Rise wa ter levels. Storm events that have been cited in previous work and in this study a re highlighted.

PAGE 82

82 Figure 4-5. Water levels of vari ous monitoring wells during this study and River Rise water level for co mparison. Three storm e vents are highlighted.

PAGE 83

0 1005 1061 1071.5 1072 1074/1/0410/1/044/1/0510/1/054/1/0610/1/064/1/07 Sink Rise Rise (-0.09 m)Discharge (m3/day)Date 83 Figure 4-6. River Sink and River Rise discharge during this study. Th ree storm events resulting in increased discharge are high lighted. The time period where the River Rise gauge was warp ed is represented with the blue highlight.

PAGE 84

0 5 10 15 20 254/23/04 6/7/04 7/22/04 9/5/04 10/20/04 12/4/04 1/18/05 3/4/05 4/18/05 6/2/05 7/17/05 8/31/05 10/15/05 11/29/05 1/13/06 2/27/06 4/13/06 5/28/06 7/12/06 8/26/06 10/10/06 11/24/06 1/8/07 2/22/07 4/8/07Recharge (cm)Date 84 Figure 4-7. Daily calculated diffuse recharge during this study using the water budget method. Rech arges associated with the th ree major storm events are highlighted.

PAGE 85

0 2 4 6 8 106/6/06 6/7/06 6/8/06 6/9/06 6/10/06 6/11/06 6/12/06 6/13/06 6/14/06 6/15/06 6/16/06 6/17/06 6/18/06 6/19/06 6/20/06 6/21/06 6/22/06 6/23/06 6/24/06 Precipitation (cm) Recharge (cm) Date 85 Figure 4-8. Precipitation and calcu lated diffuse recharge with the water budge t method for the small event in June 2006.

PAGE 86

86 0 2 4 6 8 107/16/06 7/17/06 7/18/06 7/19/06 7/20/06 7/21/06 7/22/06 7/23/06 7/24/06 7/25/06 7/26/06 7/27/06 7/28/06 7/29/06 7/30/06 7/31/06 8/1/06 Precipitation (cm) Recharge (cm) date Figure 4-9. Precipitation and calculated di ffuse recharge with the water budget method for the small event in July 2006.

PAGE 87

9.75 9.8 9.85 9.9 9.95 10 10.05 5/15/066/15/067/15/068/15/069/15/0610/15/06Rise Water Level (masl)Date Figure 4-10. River Rise water level during summer 2006. Small events in June and July are highlighted. 87

PAGE 88

9.86 9.88 9.9 9.92 9.94 9.96 9.98 10 6/6/066/9/066/12/066/15/066/18/066/21/066/24/06Well 5a June 2006Water Level (masl)Date Figure 4-11. Water level perturbation at Well 5a after a small rain event in June 2006 with recession curve projected. 88

PAGE 89

9.86 9.88 9.9 9.92 9.94 9.96 9.98 10 6/6/066/9/066/12/066/15/066/18/066/21/066/24/06Well 6a June 2006Water Level (masl)Date Figure 4-12. Water level perturbation at Well 6a after a small rain event in June 2006 with recession curve predicted. 89

PAGE 90

9.82 9.84 9.86 9.88 9.9 9.92 9.94 9.96 9.98 10 6/6/066/9/066/12/066/15/066/18/066/21/066/24/06Well 7a June 2006Water Level (masl)Date Figure 4-13. Water level perturbation at Well 7a after a small rain event in June 2006 with recession curve projected. 90

PAGE 91

9.82 9.84 9.86 9.88 9.9 9.92 7/16/067/20/067/24/067/28/068/1/06Well 5a July 2006Water Level (masl)Date Figure 4-14. Water level perturbation at Well 5a after a small rain event in July 2006 with recession curve projected. 91

PAGE 92

9.76 9.78 9.8 9.82 9.84 9.86 7/16/067/20/067/24/067/28/068/1/06Well 6a July 2006Water Level (masl)Date Figure 4-15. Water level perturbation at Well 6a after a small rain event in July 2006 with recession curve projected. 92

PAGE 93

9.74 9.76 9.78 9.8 9.82 9.84 7/16/067/20/067/24/067/28/068/1/06Well 7a July 2006Water Level (masl)Date Figure 4-16. Water level perturbation at Well 7a after a small rain event in July 2006 with recession curve projected. 93

PAGE 94

0 0.5 1 1.5 2 2.5 3 3.5 4 9/2/049/7/049/12/049/17/049/22/049/27/04 Well 4 Observed Well 4 Amplitude match Well 4 time-lag matchChange in head (m)Date Figure 4-17. Curve matching results at Well 4 for the September 2004 storm event using the Pinder et al. (1969) method. 0 0.5 1 1.5 2 2.5 3 3.5 4 9/2/049/7/049/12/049/17/049/22/049/27/04 well 6 observed Well 6 amplitude match Well 6 time-lag matchChange in head (m)Date Figure 4-18. Curve matching results at Well 6 for the September 2004 storm event using the Pinder et al. (1969) method. 94

PAGE 95

0 0.5 1 1.5 2 2.5 3 3.5 4 9/2/049/7/049/12/049/17/049/22/049/27/04 well 7 observed Well 7 amplitude match Well 7 time-lag matchChange in head (m)Date Figure 4-19. Curve matching results at Well 7 for the September 2004 storm event using the Pinder et al. (1969) method. 0 0.5 1 1.5 2 3/21/053/26/053/31/054/5/054/10/054/15/054/20/054/25/05 Well 4 observed Well 4 amplitude match Well 4 time-lag matchChange in head (m)Date Figure 4-20. Curve matching results at Well 4 fo r the March 2005 storm event using the Pinder et al. (1969) method. 95

PAGE 96

0 0.5 1 1.5 2 3/21/053/26/053/31/054/5/054/10/054/15/054/20/054/25/05 Well 6 observed well 6 amplitude match Well 6 time-lag matchChange in head (m)Date Figure 4-21. Curve matching results at Well 6 fo r the March 2005 storm event using the Pinder et al. (1969) method. 0 0.5 1 1.5 2 3/21/053/26/053/31/054/5/054/10/054/15/054/20/054/25/05 well 7 observed Well 7 amplitude match Well 7 time-lag matchwell 7 observedDate Figure 4-22. Curve matching results at Well 7 fo r the March 2005 storm event using the Pinder et al. (1969) method. 96

PAGE 97

0 0.5 1 1.5 2 12/12/0512/22/051/1/061/11/061/21/061/31/06 Well 4 observed Well 4 amplitude match Well 4 time-lag matchChange in head (m)Date Figure 4-23. Curve matching results at Well 4 for the December 2005 storm event using the Pinder et al. (1969) method. 0 0.5 1 1.5 2 12/12/0512/22/051/1/061/11/061/21/061/31/06 Well 6 observed Well 6 amplitude match Well 6 time-lag matchChange in head (m)Date Figure 4-24. Curve matching results at Well 6 for the December 2005 storm event using the Pinder et al. (1969) method. 97

PAGE 98

98 Figure 4-25. Curve matching results at Well 7 for the December 2005 storm event using the Pinder et al. (1969) method. 0 0.5 12/12/0512/22/051/1/061/11/061/21 1 1.5 2 /061/31/06 ChanDate Well 7 observed Well 7 amplitude Well 7 time-lagge in Head (m)

PAGE 99

99 Observation Wells: tower 7 1 6 4 Conduit water level No flow boundary Figure 4-26. 1-Dimensional m odel with recharge grid

PAGE 100

9.5 10 10.5 11 11.5 12 12.5 051015202530 1D Model without recharge Analytical solution without rechargeHypothetical Water Level (m)Day Figure 4-27. Results from of the Pinder et al (1969) analytical solution without recharge and the 1D model without recharge for a hypothetical storm event 100

PAGE 101

0 0.5 1 1.5 2 2.5 01 02 03 04 05 0 Well 4 observed Well 4 with recharge (T=150) Well 4 without recharge (T=450)Change in Head (m)day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-28. Curve matching results at Well 4 for the March 2003 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference. 101

PAGE 102

0 0.5 1 1.5 2 2.5 01020304050 Well 6 observed Well 6 with recharge (T=250) Well 6 without recharge (T=1300)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-29. Curve matching results at Well 6 for the March 2003 storm event using the 1D model with measured precipitation and calcu lated diffuse recharge for reference 102

PAGE 103

0 0.5 1 1.5 2 2.5 01020304050 Tower well observed Tower well with recharge (T=60000) Tower well without recharge (T=330000)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-30. Curve matching results at Tower We ll for the March 2003 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 103

PAGE 104

0 0.5 1 1.5 2 2.5 3 3.5 4 01020304050 Well 1 observed Well 1 with recharge (T=50000) Well 1 without recharge (T=90000)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-31. Curve matching results at Well 1 for the March 2003 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 104

PAGE 105

0 0.5 1 1.5 2 2.5 3 3.5 4 01020304050607080 Well 4 Observed Well 4 with recharge (T=3000) Well 4 without recharge (T=9000)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-32. Curve matching results at Well 4 fo r the September 2004 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 105

PAGE 106

0 0.5 1 1.5 2 2.5 3 3.5 4 01020304050607080 Well 6 observed Well 6 with recharge (T=3000) Well 6 without recharge (T=12000)Change in Head (m)day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-33. Curve matching results at Well 6 fo r the September 2004 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 106

PAGE 107

0 0.5 1 1.5 2 2.5 3 3.5 4 01020304050607080 Well 7 Observed Well 7 with recharge (T=300000) Well 7 without recharge (T=300000)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-34. Curve matching results at Well 7 fo r the September 2004 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 107

PAGE 108

0 0.25 0.5 0.75 1 1.25 1.5 051015202530 Well 4 observed Well 4 with Recharge (T=450) Well 4 without recharge (T=900)Change in Head (m)day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-35. Curve matching results at Well 4 for the March 2005 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 108

PAGE 109

0 0.25 0.5 0.75 1 1.25 1.5 051015202530 Well 6 observed Well 6 with recharge (T=650) Well 6 without recharge (T=1300)Change in Head (m)day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-36. Curve matching results at Well 6 for the March 2005 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 109

PAGE 110

0 0.25 0.5 0.75 1 1.25 1.5 051015202530 Well 7 observed Well 7 with recharge (T=13500) Well 7 without recharge (T=30500)Change in Head (m)day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated recharge (m) Figure 4-37. Curve matching results at Well 7 for the March 2005 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 110

PAGE 111

0 0.25 0.5 0.75 1 1.25 1.5 01020304050 Well 4 observed Well 4 with recharge (T=200) Well 4 without recharge (T=2500)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated Recharge (m) Figure 4-38. Curve matching results at Well 4 for the December 2005 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 111

PAGE 112

0 0.25 0.5 0.75 1 1.25 1.5 01020304050 well 6 observed Well 6 with recharge (T=1000) Well 6 without recharge (T=4000)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated Recharge (m) Figure 4-39. Curve matching results at Well 6 for the December 2005 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 112

PAGE 113

0 0.25 0.5 0.75 1 1.25 1.5 01020304050 Well 7 observed Well 7 with recharge (T=125000) Well 7 without recharge (T=62500)Change in Head (m)Day 0 0.05 0.1 0.15 0.2 0.25 Precipitation (m) Calculated Recharge (m) Figure 4-40. Curve matching results at Well 7 for the December 2005 storm event using the 1D model with measured precipitation and cal culated diffuse recharge for reference 113

PAGE 114

10 10.5 11 11.5 12 12.5 3/2/033/16/033/30/034/13/034/27/035/11/035/25/03Water Level (masl)date Figure 4-41. Recession curve breakdown at We ll 4 for the hydrograph produced by the March 2003 storm event 10 10.5 11 11.5 12 12.5 13 13.5 9/27/0410/7/0410/17/0410/27/0411/6/0411/16/0411/26/04Water Level (masl)date Figure 4-42. Recession curve breakdown at Well 4 for the hydrograph produced by the September 2004 storm event 114

PAGE 115

9.8 10 10.2 10.4 10.6 10.8 11 11.2 11.4 6/30/057/30/058/30/059/30/0510/30/05Water Level (masl)date Figure 4-43. Recession curve breakdown at We ll 4 for the hydrograph produced by the March 2005 storm event 10 10.2 10.4 10.6 10.8 11 11.2 11.4 2/5/062/19/063/5/063/19/064/2/064/16/064/30/065/14/06Water Level (masl)date Figure 4-44. Recession curve breakdown at Well 4 for the hydrograph produced by the December 2005 storm event 115

PAGE 116

9.8 10 10.2 10.4 10.6 10.8 3/12/063/26/064/9/064/23/065/7/065/21/066/4/066/18/06Water Level (masl)Date Figure 4-45. Example of using two portions of the recession curve that results from the December 2005 event at Well 6 to calculate basin area 116

PAGE 117

117 Table 4-1. Annual precipitation, annual evapotranspiration, and annual calculated recharge from 2002 through 2006 Year Precipitation (cm) Evapotranspiration (cm) Diffuse Recharge (cm) 2002 124.7 81.6 38.8 2003 142.1 65.9 81.2 2004 187.4 81.0 102.3 2005 132.7 84.8 47.8 2006 84.1 66.7 17.8 Table 4-2. Precipitation, potential evapotranspiration, and calculat ed recharge for three large storm events Storm Event Dates Precipitation (cm) Evapotranspiration (cm) Diffuse Recharge (cm) 2/5/2003 3/26/2003 31.5 3.3 26.9 9/2/2004 9/25/2004 44.5 6.2 41.0 3/24/2005 4/21/2005 15.4 8.4 9.6 12/14/2005 1/24/2006 28.6 4.6 24.2 Table 4-3. Chloride concentrations in sh allow wells from various sampling trips Sample Date Clconcentration Well 4a (mg/L) Clconcentration Well 5a (mg/L) Clconcentration Well 6a (mg/L) Clconcentration Well 7a (mg/L) 4/11/2006 11.7 12.8 6.5 9.7 6/15/2006 12.1 12.7 6.2 9.2 7/12/2006 10.0 11.0 6.0 7.0 8/28/2006 12.2 12.0 7.4 8.0 10/12/2006 12.0 12.0 8.0 8.0 1/17/2007 12.0 12.0 7.0 7.0 Average 11.7 12.1 6.9 8.2 Table 4-4. Recharge results for 2006 us ing chloride concentration factors Well 4a Well 5a Well 6a Well 7a Concentration Factor .914 .917 .854 .877 Evapotranspiration (cm) 76.9 77.1 71.8 73.8 Recharge (cm) 7.2 7.0 12.3 10.3

PAGE 118

Table 4-5. Calculated diffuse recharge fo r two non-conduit influenced events using th e water budget and water table fluctuation methods Date Well # Precipitation (cm) Calculated ET (cm) Amount that enters soil moisture (cm) Recharge water budget method (cm) Change in head (cm) Specific yield Recharge water table fluctuation (cm) 6/12/2006 6/14/2006 5a 12.3 0.4 8.7 3.2 2.2 0.2 0.4 6/12/2006 6/14/2006 6a 12.3 0.4 8.7 3.2 2.8 0.2 0.6 6/12/2006 6/14/2006 7a 12.3 0.4 8.7 3.2 7.4 0.2 1.5 7/22/2006 7/25/2006 5a 7.8 1.4 6.4 0.0 2.2 0.2 0.4 7/22/2006 7/25/2006 6a 7.8 1.4 6.4 0.0 1.9 0.2 0.4 7/22/2006 7/25/2006 7a 7.8 1.4 6.4 0.0 2.4 0.2 0.5 118

PAGE 119

Table 4-6. Summary of calculated transmissivities (m2/day) with different methods in the Sa nta Fe Sink-Rise System. Root mean square errors associated with model simulations are in cluded. A value of (-) indicates no data was available for calculations. Results for March 2003 analytical method are from Martin (2003). Storm event Well # Distance to well (m) Pinder et al. (1969) analytical time lag match Pinder et al. (1969) analytical amplitude match 1D model without recharge in MODFLOW Root mean square error for model w/o recharge 1D model with recharge in MODFLOW Root mean square error for model w/ recharge Recession curve analysis March 2003 1 470 97,000 97,000 90,000 50,000 March 2003 2 1050 1,898 March 2003 4 110 950 950 450 0.060 150 0.173 2,646 March 2003 6 140 900 1,300 250 March 2003 Tower 3750 120,000 550, 000 330,000 0.069 60,000 0.187 September 2004 4 110 800 10, 000 9,000 0.330 3,000 0.205 2,102 September 2004 6 140 1,200 16, 000 12,000 0.432 3,000 0.256 2,233 September 2004 7 1020 150,000 350, 000 300,000 0.176 300,000 0.356 2,013 March 2005 2 1050 991 March 2005 4 110 500 800 900 0.124 450 0.090 1,234 March 2005 6 140 600 800 1,300 0.123 650 0.089 966 March 2005 7 1020 25,000 26,000 30,500 0.056 13,500 0.076 616 December 2005 1 470 1,635 December 2005 2 1050 1,035 December 2005 4 110 700 2,200 2,500 0.274 200 0.114 1,211 December 2005 6 140 1,100 2, 200 4,000 0.183 1,000 0.062 1,011 December 2005 7 1020 40,000 40, 000 62,500 0.074 125,000 0.266 1,034 119

PAGE 120

Table 4.7. Sensitivity of 1D mode l results to specific yield Sy = .1 Sy = .2 Sy = .3 Best match Transmissivity (m2/day) root mean square error Best match Transmissivity (m2/day) root mean square error Best match Transmissivity (m2/day) root mean square error M a r c h 2 0 0 3 a r c h 2 0 0 5 Well 4 450 0.472 150 0.173 75 0.105 Tower Well 540,000 0.603 60,000 0.187 195,000 0.115 September 2004 Well 4 9,000 0.324 3,000 0.205 4,500 0.216 Well 6 7,500 0.338 3,000 0.256 6,000 0.305 Well 7 1,500,000 0.440 300,000 0.356 300,000 0.270 M Well 4 90 0.140 450 0.090 900 0.097 Well 6 130 0.143 650 0.089 1,300 0.096 Well 7 675 0.137 13,500 0.076 27,000 0.061 December 2005 Well 4 1,600 0.262 200 0.114 600 0.158 Well 6 10,000 0.244 1,000 0.062 1,100 0.062 Well 7 1,500,000 0.323 125,000 0.266 37,500 0.157 120

PAGE 121

121 Table 4-8. Area calculations using recession curves Date range Well # # of Days Head (max) Head (min) h (m) Slope ( h/day) Sy Average Q (m3/day) Area (m2) Average Q (-0.09m) correction (m3/day) Area (m2) 10/17/05 11/19/05 1 33 10.40 10.18 0.22 0.0067 0.2 5.99 x 105 4.50 x 108 4.42 x 105 3.31 x 108 10/12/05 11/20/05 2 40 10.52 10.28 0.24 0.006 0.2 5.84 x 105 4.87 x 108 4.24 x 105 3.53 x 108 10/12/05 11/25/05 4 45 10.42 9.99 0.43 0.0096 0.2 5.86 x 105 4.55 x 108 4.24 x 105 3.29 x 108 10/12/05 11/25/05 6 45 10.36 10.07 0.29 0.0064 0.2 5.86 x 105 3.07 x 108 4.24 x 105 2.21 x 108 10/12/05 11/25/05 7 45 10.20 9.96 0.24 0.0053 0.2 5.86 x 105 5.49 x 108 4.24 x 105 3.97 x 108 3/22/06 4/12/06 1 22 10.50 10.31 0.19 0.0086 0.2 6.92 x 105 4.00 x 108 5.24 x 105 3.04 x 108 3/17/06 4/23/06 2 38 10.82 10.42 0.40 0.0105 0.2 6.98 x 105 3.31 x 108 5.31 x 105 2.52 x 108 3/17/06 4/23/06 4 38 10.59 10.23 0.36 0.0095 0.2 6.98 x 105 3.02 x 108 5.31 x 105 2.29 x 108 3/17/06 4/23/06 6 38 10.63 10.19 0.44 0.0116 0.2 6. 98 x 105 3.69 x 108 5.31 x 105 2.80 x 108 3/17/06 4/23/06 7 38 10.43 10.08 0.35 0.0092 0.2 6. 98 x 105 3.79 x 108 5.31 x 105 2.88 x 108 4/23/06 5/15/06 2 23 10.42 10.27 0.15 0.0065 0.2 6. 80 x 105 5.22 x 108 5.26 x 105 4.03 x 108 4/23/06 5/11/06 4 19 10.23 10.11 0.12 0.0063 0.2 6.77 x 105 5.29 x 108 5.53 x 105 4.15 x 108 4/23/06 6/9/06 6 48 10.19 9.87 0.32 0.0067 0.2 7.05 x 105 5.36 x 108 5.23 x 105 4.14 x 108 4/23/06 5/15/06 7 23 10.08 9.95 0.13 0.0057 0.2 6.80 x 105 6.02 x 108 5.26 x 105 4.65 x 108 Average = 4.44 x 108 m2 Average = 3.35 x 108 m2

PAGE 122

CHAPTER 5 DISCUSSION Applicability of Different Rech arge Calculation Methods and E ffect of Conduit Boundaries Calculated recharge values for the small J une 2006 event using the water table fluctuation method at wells 5a, 6a, and 7a are 0.4 cm, 0.6 cm, and 1.5 cm respectively. Water table fluctuation calculations at all three locations pr edict less recharge than the 3.2 cm calculated with the water budget method. These results may have occurred if the water budget method is overpredicting the amount of water that actually makes it to the water table. In a highly karstic aquifer, it is possible that water making it past the root zone a nd soil moisture storage may flow laterally in the unsaturated zone be fore recharging the water table. If water were to flow laterally at a similar time scale to which recharge enters the vadose zone, then pe rhaps we are not seeing the full recharge pulse in the water table fluctu ation at the wells. Instead the water will flow towards the conduit location and rechar ge at some point along the way. It is also possible that since the volume of recharge is small, the loss of water to the conduit may occur in the saturated zone on a sim ilar time scale as the r echarge is raising the water table. The effect that fluctuations of water level in the conduit boundary has on the water level change at matrix monitori ng wells has been difficult to asse ss. The most apparent effect from the conduit boundary comes after significant ch anges in conduit water le vel related to large storm events. It is during these times that the water table fluctu ation method is least applicable; this is because it does not differentiate between sources of water and therefore during large storm events the water table method would overestim ate the amount of diffuse recharge. However during small events, where the changes in head at both the wells and the boundary are small, the effect of the conduit is less obvious. 122

PAGE 123

In the case of the June 2006 small event, th e River Rise water leve l does increase very slightly (~0.005 m), suggesting that there may be some conduit influence contributing to the fluctuation of the water table. To test the condui t influence, the analytical solution was used by substituting the 0.005 m change in head at the conduit over a three day period and observing if this can cause changes in head in the observation wells. Transmissivity values determined from model results at the different wells were used for their respective well lo cations. The calculated effect of the conduit is less th an .00001 m at all three observa tion well locations, which is significantly less than the change in head observed at these locations during the small event. Therefore, we are confident that the change at the observation well during the small events is primarily related to diffuse rechar ge arriving at the water table. The change in head at the observation we ll during the small events may actually be underpredicted if water is flowing away from the well towards the conduit at the same rate as water is arriving at the water table. We can assess this situation in the same way as we approached the effect of the conduit on the we lls. An average head change of 0.022 m was observed at shallow Wells 4 and 6 during these two storm events. Applying this head change over the same three day period, and us ing a transmissivity value of 500 m2/day, the distance the conduit would need to be away from the well befo re it would stop having an effect is ~200 m. This result suggests that Wells 4 and 6 can potentially be losi ng some amount of recharge to lateral flow towards the conduit. This would explain why the water table fluctuation method predicts less recharge during the small events when compared with the water budget method. Risser et al (2005) discusse d the effect of proximity to a stream boundary on the water table fluctuation method applied in a small wate rshed of eastern Pennsylvania. Risser et al (2005) used a 1D model in MODFLOW to simulate changes in head at observation wells as a 123

PAGE 124

result of diffuse recharge. Three observation we lls were simulated which are located at three different distances away from a stream bounda ry. The result was th at the observation well furthest from the stream boundary produced the greate st change in head after the diffuse recharge was applied. The observation well that was cl osest to the stream boundary produced a head change that was only 20% of what was observed at the furthest well. Results from the study by Risser et al (2005) and from this study indicate that the accuracy of the water table fluctuation method will be compromised in regions where hydrologic boundaries may influence the results. The small event that occurred in July of 2006 was similar to the June event in that it lasted only four days and the soil moisture st orage was minimal at the beginning of the storm. However, in contrast to the June 2006 event, we see the opposite results for the two recharge methods applied to the July 2006 event. For this storm, the water budget method predicts zero recharge, while the water table fl uctuation predicts a small amount of recharge (<0.5 cm) at all of the well locations. We can be confident that wate r is arriving to the water table, based on the increase in the well hydrograph. We can also be confident that water is not arriving from the conduit because there is no increase in the Rive r Rise water level. A possible explanation for why the water budget method does not predict any r echarge is that the soil moisture storage capacity may have been over predicted. When the soil moisture has reached its wilting point (i.e. the point when soil moisture content is so low th at the surface tension of the soil-water interface exceeds the osmotic pressure of th e roots and water will no longer enter the roots (Fetter, 1994)), then the estimated 10 cm of precipitation is needed before any recharge can occur. The July 2006 event resulted in a little less than 8 cm of r echarge, thus was not able to account for the soil moisture storage. If the storage capacity is actually closer to 8 cm, then we would see the small recharge observed in the water le vel. Another possible explanation to consider is that since the 124

PAGE 125

amount of recharge calculated with the water table fluctuation is so small, it is possible that a small amount of water might bypass the soil moisture storage. This could happen if some recharge were to hit a fast flow path (fracture) which allows it to move quickly through the unsaturated zone, and thus avoiding the soil moisture storage. This amount of water could be affected by the intensity of the rainfall. For exam ple, if all the rain falls within a ten minute period, it is more likely that some might bypass the storage as opposed to a slow and steady rainfall over the course of a three day period. Be cause precipitation was recorded on a daily basis during this study, the differences in intensity can not be evaluated. In the case of these small summer events, it is difficult to assess the accuracy of the water table fluctuation method. If the conduit is truly not contributing any water, as demonstrated by the receding trend in water level at the River Ri se (Figure 4-10) and the calculations of conduit effect, then we know that the slight change in wa ter level must be a result of water arriving from the diffuse component. If water is arriving from any other sources then the diffuse recharge portion of the head change will be over predicte d using the water table fluctuation. However, if water is flowing away from the well location to the conduit at the same rate that it is recharging the water table, then the water ta ble fluctuation results may be an under estimate. Fluctuations in the water table can be potentially valuable in this system for seeing the timing of recharge; however, a calculation of the effect of a nearby boundary (such as a stream or a conduit) should be considered before applying the water table fl uctuation method to estimate volumes of diffuse recharge. In this study area, there are no we lls that are far enough away from a boundary condition that we can be totally confiden t in the water table fluctuation method. The chloride method provides an alternat ive long term method fo r quantifying recharge. This method is valuable in systems where th ere is only once source of chloride and where 125

PAGE 126

evapotranspiration is abundant. This method works well in dry years (e.g. 2006) in the unconfined Santa Fe River basin because diffuse recharge and the conduit component are not constantly diluting the chlorine concentrations Since the concentrations are dependent on the precipitation and evapotranspiration, the more measurements of chlo ride concentrations that can be made will result in the most accurate calcula tion of recharge. The chloride concentration is also good as a method to determine the long term differences in recharge at the various well locations. Daily Versus Monthly Recharge and Evapotranspiration Calculations Because precipitation records are readily av ailable and usually easily accessible, they are often used as a substitute to approximate rechar ge, rather than calculating recharge on a daily basis. Daily recharge calculati ons require determining soil moisture storage and adjusting for daily evapotranspiration, and thus require signi ficantly more time and effort. However, results from this study indicate that in this study area da ily precipitation (Figure 4-1) and daily recharge records (Figure 4-7) differ trem endously. It is not uncommon fo r an afternoon thunderstorm in the summer to supply more than 5 cm of precipi tation, yet none of that water is calculated to become recharge. For example, this happens after the small storm in July of 2006 when there was 8 cm of precipitation, but no calculated recharge. The recharge does not exactly mimic precipitation because so much water is either stored in soil moistu re, or is evapotranspired back to the atmosphere; this result would be expected in any region with a high soil moisture capacity and high evapotranspiration rates. The recharge values calculated on a daily basi s were especially impor tant to this study because they were calculated on the same time s cale as the 1-dimensional model. Daily recharge calculations allowed for calculation of recharge fo r any time-step of the model. If the recharge 126

PAGE 127

could be calculated on even a finer time scale, we could reduce the time step of the model, and this may improve the calculated transmissivity results. Potential evapotranspiration is commonly calculated on a monthly basis to emphasize the seasonal differences throughout a year. This is a relatively easy calculation with very few inputs using the standard Thornthwaite method (Schwartz and Zhang, 2003). In this study, evapotranspiration was calculated on a daily basis with the much more detailed PenmanMonteith equation. The monthly cal culated values of evapotranspi ration tend to be consistent with the summed daily calculations (Figures 4-2 and 4-3). Although this PenmanMonteith method requires more data input and some assump tions, the daily evapotranspiration calculations are very valuable. For example, more than half of the precipitation that fa lls in this region every year is lost to evapotranspiration, and therefore daily values of evapotranspiration are necessary to produce an accurate daily recharge calculation. Daily values can also be used to indicate dry times in the record, as can be seen by the peri ods of zero calculated evapotranspiration in figure 4-2. In this study area, this commonly happens during the month of May, when there is not enough precipitation to produce recharge. Transmissivity Influence of Diffuse Recharge on Stor m Event Hydrographs and Transmissivity Calculations Comparing results of the 1D model with a nd without recharge s howed that including diffuse recharge in the solution will significantly change the predicted transmissivity. These results suggest that transmissivity may be overpre dicted with the methods that do not include the diffuse recharge. This happens because an artificia lly high transmissivity is needed to create a good match in methods that disregard diffuse input This was the case in Martin et al. (2006), where transmissivities were calculated using only the Pinder et al. (1969) analytical solution 127

PAGE 128

which disregards the diffuse component. Calculated transmissivity values from the same storm were three times more when calculated with the 1D model without recharge. This suggests that choosing a method which does not include recharge is a crucia l omission when using a large scale natural event to estimate transmissivity through passive monitoring. In local karst regions, dissolution of the landscape lead s to high susceptibility to pol lution via groundwater recharge. Overpredicted transmissivity values suggest fa ster movement of pollutants through the system than what may actually be observed. Although the inclusion of diffuse recharge changes the calculated transmissivity values, it is important to point out the di fficulties that go along with incl uding the recharge. Three of the four model simulations including the diffuse rechar ge component result in predicted curves that are unrealistically pointed and jagged when compar ed to the smooth curves that are produced in the observed data (Figures 27-30, 35-40) This effect is most noticeable during the small December 2005 event, and is not seen in the large September 2004 event. Assuming the total amount of calculated recharge per storm event is correct, then the jagged predicted curves may be displaying a lag time effect. This happens if the full amount of the calculated recharge per day did not fully reach the well during that same calculated day. Rather, the recharge actually reaches the aquifer slowly over the course of a few da ys after the rain event. To correct for this in the model, recharge would have to be spr ead across the time steps in varying amounts. Determining how to spread the recharge ac ross the model is difficult because the excess precipitation calculation can not be refined to better than a daily time scale. It is also difficult because the influence that the diffuse recharge has on the hydrograph varies per storm event. For example, the jaggedness of the predicted curv es during the March 2005 storm event is much more apparent than during the September 2004 storm event. This may be a result of most of the 128

PAGE 129

recharge (>30 cm) occuring over a short two da y period. In contrast, th e other storms have recharge occur sporadically thr oughout the course of the event. The heavily focused recharge coincides with the large storm pulse arriving via the conduit during th e September 2004 event, and results in a quick, sharp increase in the st orm hydrograph (3.5 m in 10 days). This helps mask the effect of the diffuse recharge co mponent, as a smooth hydrograph is predicted. The time lag of diffuse recharge to the water table is influenced by the location of the water table as compared to the surface during the various storm ev ents. During larger events like in September 2004, the water table came extremely close to the su rface (Table 5-1), and thus it is expected that recharge occurs quickly. In smaller events like March 2005 and December 2005 the water table starts low and does not increase too far into the su rficial sediment (Table 5-1); the greater depths to the water table would result in greater amount of time for recharge to reach the water table as compared to the September event. The time lag of the diffuse recharge is also affected by the storage that occurs in the unsatur ated zone. The ability to temporarily store more water in this area will increase the time it takes for r echarge to reach the water table. Choosing one pattern of water table response to rainfall would also be difficult in this region because of the varying flow that may occu r in the unsaturated zone produced by different magnitude storms. When soil is not saturated, soil moisture flows dow nward by gravity flow through interconnected pores (Fetter, 1994). With increasing water content, more pores will fill and the rate of downward water movement increases (Fetter, 1994). Because the vertical unsaturated hydraulic conductivity is not constant, th e rate of flow will be affected by the amount of diffuse recharge entering the uns aturated zone. This amount varies with intensity of the storm. Influence of Scale Effects on Transmissivity Calculations Both the analytical method without recharge a nd the 1D model simulations show the effect that scale has on the transmissivity calculation. It has been noted in previous work on karst 129

PAGE 130

aquifers (e.g., Bradbury and Muldoon, 1990; Rovey, 1994; Martin et al, 2006) that when increasing scale, preferential pathways tend to dominate a larger percentage of groundwater flow, thus increasing transmissivity. This seem s to be the case around the Santa Fe Sink-Rise system, where transmissivity to the wells furt hest from the mapped or inferred conduit (Well 1, Tower Well, and possibly Well 7) are two to three orders of magn itude greater than transmissivity of wells near the mapped conduit (Well 4 and Well 6). The scale effect is seen with all four storm events since 2003. The fact that transmissivity in creases so much with increasing distance away from the river supports the idea that the preferential flow paths exist, and thus the karstic limestone of the Floridan aquifer is highly heterogeneous in nature. In contrast to both of the curve matching t echniques, the recession curve analysis does not show the scaling effect. However, the recession curv e analysis results should be interpreted in a different manner than results produced by the two curve matching methods. The recession analysis method yields insight into the transmissi vity of the intergranular matrix rock. This does not include the conduit and fracture zones that make up the preferenti al flow paths in the highly karstic parts of the Floridan aqui fer. This is most obvious for lo cations at a far distance away from the conduit (Well 1, Well 2, and Well 7) wh ere the recession curve transmissivities are significantly lower than those calculated by the analytical solution a nd model (Table 4-6). Transmissivity calculated with th e recession curve analysis at well locations that are close to the conduit (Well 4 and Well 6) yield very similar resu lts to those calculated with the other methods. This is most likely because the amount of pref erential flow paths encountered in the short distance between the wells and c onduit is much less than over th e whole formation, so therefore the transmissivity of the formation between th ese points mimic the matrix transmissivity around the wells. These results suggest that the analytical solution wit hout recharge and 1-dimensional 130

PAGE 131

model with recharge are more us eful for predicting how the system will respond to a storm pulse arriving from the conduit location, while the recession curv e analysis is bette r for understanding how a particular well responds after the storm pulse. For example, if the goal was to understand how a contaminant will move from the conduit to the matrix during a storm event, then one would be more concerned with the understanding the maximum transmissiv ity of the whole rock formation. In contrast, if a contaminant was dumpe d at a point source (e.g. near an observation well), one might be concerned with how that cont aminant will move near the region in which it was dumped. Prior to this study, air permeab ility tests were performed on co res recovered from Well 5 in the OLeno State Park study area. In regions where at least 50% of the core was recovered, permeabilities averaged about .4 darcys, which is equivalent to a hydrauli c conductivity of about .297 m/day (Moore and Florea, unpublished data). Sl ug tests were also preformed in this study area at well locations 3, 4, 5, 6 and 7 by Ham ilton (unpublished data) and Langston (unpublished data). Slug test results analyzed using the B ouwer and Rice Slug Test Method yielded average hydraulic conductivity values between 0.864 m/day and 3.89 m/day at the six well locations (Hamilton, unpublished data). Average hydraulic c onductivity values from this study ranged from 1.5 m/day at Well 4 during the March 2003 event to 3000 m/day at Well 7 during the September 2004 event. These values are greater than those calculated from permeability and slug tests, which is a common result in hydraulic conductivity determined from laboratory and well tests as compared to those determined from pa ssive monitoring. These va lues further support the scale effect in this region. Influence of Storm Event Magnitude on Transmissivity Calculations The September 2004 storm event was the larges t event recorded dur ing this study, with head changes greater than 3.5 m at the monitoring wells. Calculated transmissivity values from 131

PAGE 132

the September 2004 storm pulse are significantly gr eater than calcu lated transmissivity values from the other three storm events. For example, transmissivity values from Wells 4 and 6 using the September event were 3000 m2/day, while calculated transmissivity did not exceed 1000 m2/day from the other storm events. Transmissi vity for the September 2004 event at Well 7 was 300,000 m2/day, which is about 20 times greater than the transmissivity of 13,500 m2/day calculated from Well 7 during the March 2005 event. In an unconfined aquifer, the level of saturati on rises or falls with th e amount of water that is in storage (Fetter, 1994). One possible explan ation for the higher calculated transmissivity values is that the amount of preferential flow paths encountered by the storm pulse is increasing as the saturated thickness of the unconfined aquifer increases into the surficial sediments during a storm event. This can also happen if the permeab ility of surficial sediments in this region are higher than the limestone. Based on descriptive analysis of a core taken during the installation of well location 5, surficial sediment to depths of a bout 2 m are lightly consolidated fine to coarse sands with interspersed chunks of gravel. Duri ng the September 2004 storm event, water levels were less than 2 m below the surf ace on average at wells 4, 6, and 7, and were as little as 0.1 m from the surface at Well 6 specifically (Table 5-1). This indicates that water was almost completely saturating the surficial sediments in some areas. In contrast, water levels at Wells 4, 6, and 7 were greater than 3 m below the surface on average during the March 2003 event and greater than 4 m below the surface during the other two storm events (Table 5-1). Since the boundary between the limestone bedrock and surficial sediments lies at about 3 4 m below the surface on average, during the small events th e groundwater levels rarely reach above the limestone. If the saturated thic kness of the aquifer increases above the limestone/surficial 132

PAGE 133

sediment boundary, then the calculated transmissivity may increase as a result of flow through this zone of higher porosity. If it is true that transmissivity can change relative to the increasing storm size, then determining the maximum transmissivity of th is system requires a storm event that fully saturates the system. The storm event in Septem ber 2004 is represents th is scenario, as water levels were at or near the surface. One intere sting point to note is that overland flow was observed in some areas at the OLeno fiel d site during the Sept ember 2004 event (Moore, personal communication). Overla nd flow might induce an artif icially high calculation of transmissivity if water were to flow faster ov erland and infiltrate downward into the wells. This happens because overland flow decreases the distance from the conduit to the monitoring wells. Overland flow may have occurred near Well 6 dur ing this storm, as can be seen by the water level being extremely close to the surface. However, overland flow is not likely at Wells 4 and 7, which still had water levels about 2 m below the surface at the peak of the event. Overland flow could potentially be a possible source of error in any pa ssive monitoring calculation of transmissivity, and should be noted during large events where it may occur. Choosing the Appropriate Method to Calculate Transmissivity Three methods were used to calculate tran smissivity throughout this study. The most important differences between th e three methods are whether the recharge is included in the method and the type of transmissivity that is being calculated (i.e. matrix versus whole rock formation). Because of these differences, comp arison between methods should be approached with caution. As previously discussed, the 1-dimensional m odel without recharge and the Pinder et al. (1969) analytical method without recharge essentially produce th e same results (Figure 4-27). The only real difference between the two methods in this study is the way in which curve 133

PAGE 134

matching was performed. With the analytical met hod, curves are matched visually by time-lag or by amplitude. In the model, the curves are cal ibrated numerically from deviation between predicted and observed curves via root mean square error. The difference that this has on the calculation of transmissivity in some cases can be significant (Table 4-6). The model is most likely more accurate because the best fit curve is determined numerically as opposed to making a visual estimate. Methods like thes e that neglect the diffuse rechar ge component are adequate to use in regions where little diffuse recharge occurs, such as a confined karst aquifer. The 1-dimensional model with recharge is usef ul in regions similar to the unconfined Santa Fe River basin, where diffuse recharge is signific ant. The model used in this study maximizes the effect of incorporating recharge by assuming that all of the calcul ated excess precipitation arrives as recharge to the water table on the same day it was calculated for. Results show that disregarding the diffuse component in such regions can lead to overpredicted transmissivity values. Creation of the very basic 1-dimensional numerical model used in this study is a similar amount of work compared to the computing iter ations involved with so lving the analytical solution. Therefore, it would be hard to argue that the analytical solution is any easier to apply. Also, the modeling allows for simple calibration of various parameters with numerical based sensitivity analysis of those parameters. Diffuse Recharge Component and Possible Implications for Dissolution An important part of understanding the potenti al karstification of the unconfined Floridan aquifer region in response to storm events is being able to accurately make water budget calculations. This requires quantif ication of the volumetric inputs and outputs of the system; most importantly the volume contributed by diffuse recharge in comparison to the volume of water received from the conduit. Previously, quan tifying the diffuse recharge input had been a challenge in this study area because the basin area contributing to the system had been difficult 134

PAGE 135

to assess. However, using the area estimates made from recession curves in this study (Table 48) we now have the ability to make some co mparisons between these two components. Possible sources of error to the overall budget include if there are other inputs and outputs to the system. Besides the Sink and the diffuse recharge, a third possible input to the system may be from a feeder conduit that has been ma pped to the east of the OLeno St ate Park study area (Figure 2-2). Other possible inputs include at Vinzants landing just north of the River Sink, and any recharge that might occur to the confined Floridan aquife r. Other possible outputs include springs that are located downstream of the River Rise. Screaton et al. (2004) discuss possible local karstification in the Santa Fe River system resulting from inflow of conduit water to the aq uifer after a storm event. A small storm event from late September 2001 was used (Figure 4-4), wh ich produced a little mo re than seven cm of precipitation over 16 days. The amount of water lo st to the matrix during this time was 2.2 x 106 m3, calculated from the difference between Rive r Sink and River Rise daily discharges. The degree of undersaturation was estimated from ch emistry data collected in 1998 at the River Sink, for a stage similar to the stage recorded during the 2001 event. Because the residence time of the conduit water in the matrix was not known, the calculation was simplified by assuming all the water lost from the conduit remains in the matrix rock sufficient time to reach equilibrium. It was calculated that the volume of rock needed to dissolve in order to bring the calcium concentration back to equilibrium was 1.06 x 104 m3. An estimated area of di ssolution surroundi ng the conduit of 3.0 x 107 m2 was used, which implied regional dissolution of 4.6 x 10-4 m. This regional dissolution rate was an order of magn itude greater than rates of 2.63 x 10-5 m reported by Opdyke et al (1984) from measurements of dissolved calcium in Floridas springs. 135

PAGE 136

The calculations of possible dissolution in Screaton et al. (2004) do not account for dissolution occurring from diffuse recharge. Using the calculated basin area of 4.44 x 108 m2 from above, we can now calculate the total volume of water that enters the matrix via diffuse recharge. For the same 2001 storm event, assuming that precipitation rechar ges evenly across the entire basin area, this volu me would be equal to 3.20 x 107 m3. If we use the low end of the calculated basin area of 3.35 x 108 m2, than this volume of wate r would be equal to 2.38 x 107 m3. These calculated volumes of diffuse recharge are an order of magnitude higher than the volume contributed to the system by the conduit. Determining the dissolution caused by this volume will be possible once chemistry data from precipitation in the study area is determined. However, the fact that the total volume of water input from the diffuse recharge during this small event is much larger than the conduit input, suggest s that dissolution will actually be greater than calculated by Screaton et al. (2004). With the addition of more storm event data since the original calcu lations in Screaton et al. (2004) were performed for the 2001 event, it is now possible to make a similar calculation of dissolution for a larger event. The March 2003 event discussed in Martin et al (2006) had a head increase of near three meters, and a full record of daily discharge values. During the 19 days of this storm event, the conduit lost 2.83 x 107 m3 of water to the matrix based on the Sink and Rise daily discharges. This volume is an order of ma gnitude higher than the amount of water lost to the matrix during the smaller 2001 event. Using the same chemical data from the River Sink collected in 1998 (Screaton et al 2004), .3 moles of calcite per liter of water is needed to dissolve in order to reach equi librium at a River Sink stage of 13.43 meters. This would equal a dissolved rock volume of 3.13 x 105 m3 from the conduit component. These results can be converted to regional denuda tion rates using the same estimated area of 3.0 x 107 m2. The 136

PAGE 137

137 regional denudation resulting from this storm would be 1.04 x 10-2 m, which is two to three orders of magnitude higher than the original reports of 4 x 10-5 m/yr by Opdyke et. al (1984) and 4.60 x 10-4 m by Screaton et al (2004) for the smaller 2001 event. During this same large 2003 storm event there was 12 cm of diffuse recharge. This would equal a total diffuse r echarge input of 5.33 x 107 m3 of water into the matrix using an area of 4.44 x 108 m2, and an diffuse recharge input of 4.02 x 107 m3 using an area of 3.35 x 108 m2. Even if we assume the lower calculated value for diffuse input is correct, this is still much greater than the total amount contributed by th e conduit during the storm. This would increase the estimated dissolution over the entire basin area.

PAGE 138

138 Table 5-1. Elevations of the ground surface, top of limestone bedr ock, and water table before and at the peak of various large storm events at different well locations (in meters above sea level) Well # Surface limestone bedrock Water Table before March 2003 event Water Table at peak of March 2003 event Water Table before September 2004 event Water Table at peak of September 2004 event Water Table before March 2005 event Water Table at peak of March 2005 event Water Table before December 2005 event Water Table at peak of December 2005 event 1 14.45 -2.62 10.17 13.81 10.50 13.25 10.25 12.10 4 17.89 13.32 10.17 12.23 9.85 13.13 10.56 11.89 9.94 11.28 6 13.51 8.64 10.11 13.38 10.57 11.89 10.07 11.28 7 15.22 9.73 9.97 13.05 10.41 11.46 9.98 10.95

PAGE 139

CHAPTER 6 SUMMARY Understanding aquifer parameters in ka rst regions is an important concern because of the reliance on these regions for potable water and because of the possibility for contamination across these landscapes. Th ere are many methods available to evaluate aquifer parameters, including laboratory tests and aquifer tests. However, these methods are difficult to apply in karst regions because the small scale often neglects to incorporate the heterogeneous nature of the whole rock formation. In this study, passive monitoring of water levels was used as a method to evaluate transmissivity between a conduit location and observation wells at varyi ng distances away from the conduit. The Pinder et al. (1969) an alytical solution was first applied in this study area by Martin et al. (2006) to storm event data from 2003. This method predic ts water levels at observation wells based on response in a nearby boundary (conduit) location, but does not include a recharge component in the solu tion. Results indicated that transmissivity values range based on vicinity to the condu it location, with wells at a further distance away having the highest calcu lated transmissivity values. These scale effects were consistent with results in this study using the Pinder et al. (1969) analytical method applied to three more storm events occurring after 2003. Three methods including a water budget, chloride concentrations, and water table fluctuation were used to calcula te diffuse recharge in the study area. Results indicate that diffuse recharge is quantifiab le and especially significant during large storm events, and thus should not be neglected in the transmi ssivity calculation. A 1-dimensional model in MODFLOW was created to test the influence of diffuse recharge on the transmissivity calculation. Like the analytical solution, the model predicts water levels in observation 139

PAGE 140

wells based on change in nearby boundary cond itions; however, the model also allows for recharge to be entered for any time ste p. Calculated recharge s using the water budget were entered into the mode l on the day of occurrence. Models were calibrated to observed data and a best fit transmissivity wa s determined using root mean square error as the criteria for fit. Results indicated that neglecting to include diffuse recharge will result in overpredicted calculated transmissi vity. Results also showed some difficulties with adding recharge to the model. These in clude the possibility that a lag time may exist from when the initial portion of diffuse rechar ge reaches the water ta ble to the final total amount of diffuse recharge, and also that the effect of recharge on the observed hydrograph may be influenced by the storm event magnitude. Calculation of the basin ar ea contributing to the discha rge of the Santa Fe River Basin upstream of the River Rise, in combination with quantifying the diffuse recharge, allows for comparing the components of th e water budget in the system. Results show that the total input of diffuse recharge during a storm event, as compared to input from the conduit boundary, will vary based on the magn itude of the storm. Diffuse recharge should be an important part of the potentia l dissolution of limestone in this study area, and should be included in future calculations when chemistry data become available. 140

PAGE 141

LIST OF REFERENCES Allen, R.G., Jensen, M.E., Wright, J.L. and Burman, R.D., 1989. Operational Estimates of Reference Evapotranspiration. Agronomy Journal, 81: 650-662. Anderson, M.P. and Woessner, W.W., 2002. Applied Groundwater Modeling. Academic Press, San Diego, 381 pp. Andres, A.S. and Klingbeil, A.D., 2006. Thickne ss and Transmissivity of the Unconfined Aquifer of Eastern Sussex County, Delawa re. In: D.G. Survey (Editor), Newark. Atkinson, T.C., 1977. Diffuse Flow and Conduit Flow in Limestone Terrain in the Mendip Hills, Somerset. Jour nal of Hydrology, 35: 93-110. Batiot, C., Linan, C., Andreo, B. and Embl anch, C., 2003. Use of Total Organic Carbon (TOC) as a Tracer of Diffuse Infiltrati on in a Dolomitic Karstic System: The Nerja Cave (Andalusia, Spain). Geop hysical Research Letters, 30(22). Bradbury, K.R. and Muldoon, M.A., 1990. Hydraulic Conductivity Determinations in Unlithified Glacial and Fluvi al Materials. In: D.M. Neilson and A.I. Johnson (Editors), Hydraulic Conduc tivity and Waste Contamin ant Transport in Soils. American Society for Testing and Materials, Philadelphia, pp. 138-151. Budd, D.A. and Vacher, H.L., 2004. Matrix Pe rmeability of the Confined Floridan Aquifer, Florida, USA. Hydr ogeology Journal, 12: 531-549. Bush, P.W. and Johnston, R.H., 1988. GroundWater Hydraulics, Regional Flow, and Ground-Water Development of the Floridan Aquifer System in Florida and in Parts of Georgia, South Carolina, an d Alabama. U.S. Geological Survey Professional Paper, 1403(C). Dingman, S.L., 2002. Physical Hydrology. Pr entice Hall, Upper Saddle River, New Jersey, 646 pp. Federer, C.A., Vorosmarty, C. and Fekete B., 1996. Intercomparison of Methods for Calculating Evaporation in Regional a nd Global Water Balance Models. Water Resources Research, 32: 2315-2321. Fetter, C.W., 1994. Applied Hydrogeology. Pr entice Hall, Upper Saddle River, New Jersey. Florea, L.J. and Vacher, H.L., 2005. Springflo w Hydrographs: Eogenetic vs. Telogenetic Karst. Groundwater: 1-10. Gilman, E.F., 1991. Community Tree Care. University of Floridan IFAS extension, Gainesville, Florida. 141

PAGE 142

Grubbs, J.W., 1998. Recharge Rates to the U pper Floridan Aquifer in the Suwannee River Water Management District, Florida. In: U.S.G. Survey (Editor). WaterResources Investigations Report 97-4283, Tallahassee. Healy, R.W. and Cook, P.G., 2002. Using Groundwater Levels to Estimate Recharge. Hydrogeology Journal, 10: 91-109. Hisert, R.A., 1994. A Multiple Tracer Appr oach to Determine the Ground and Surface Water Relationships in the Western Santa Fe River, Columbia county, Florida., University of Florida, Gainesville, 212 pp. Hunn, J.D. and Slack, L.J., 1983. Water Resource s of the Santa Fe River Basin, Florida. In: U.S.G. Survey (Editor). Water Re sources Investigations Report 83-4075, Tallahassee. Jones, I.C., Banner, J.L. and Humphrey, J.D., 2000. Estimating Recharge in a Tropical Karst Aquifer. Water Resources Research, 36(5): 1289-1299. Jones, W.K., 1999. Pump Tests of Wells at the National Training Center Near Sheperdstown, West Virginia. Karst Wate rs Institute Special Publication, 5: 259262. Junge, C.E. and Werby, R.T., 1958. The Concen tration of Chloride, Sodium, Potassium, Calcium, and Sulfate in Rain Water over the United States. Journal of Meteorology, 15: 417-425. Katz, B.G., Coplen, T.B., Bullen, T.D. and Davis, J.H., 1997. Use of Chemical and Isotopic Tracers to Characterize the Interactions between Ground Water and Surface Water in Mantled Karst. Groundwater, 35(6): 1014-1028. Lemeur, R., and Zhang, L., 1990. Evaluation of Three Evapotranspiration Models in Terms of their Applicability for an Arid Region. Journal of Hydrology, 114:395411 Maher, K. and DePaolo, D.J., 2003. Vadose Zone Infiltration Rate at Hanford, Washington, Inferred from Sr Isotope M easurements. Water Resources Research, 39(8). Martin, J.B. and Dean, R.W., 1999. Temperatur e and a Natural Tracer of Short Residence Times for Groundwater in Karst Aquife rs. Karst Waters Institute Special Publication, 5: 236-242. Martin, J.B. and Dean, R.W., 2001. Exchange of Water Between Conduits and Matrix in the Floridan Aquifer. Chemical Geology, 179: 145-165. 142

PAGE 143

Martin, J.M., 2003. Quantification of the Matr ix Hydraulic Conductivity in the Santa Fe River Sink/Rise system with Implications for the Exchange of Water between the Matrix and Conduits, University of Florida, Gainesville, 80 pp. Martin, J.M., Screaton, E.J. and Martin, J.B., 2006. Monitoring Well Responses to Karst Conduit Head Flucuations: Implicatio ns for Fluid Exchange and Matrix Transmissivity in the Floridan Aquifer. Geological Society of America Special Paper, 404. Monteith, J.L., 1965. Evaporation and Environment. Proceedings of the 19th Symposium of the Society for Experimental Biology. Cambridge University Press, New York, 205-233 pp. Mylroie, J.E., 1984. Hydrologic Classification of Caves and Karst. In: R.G. LaFleur (Editor), Groundwater as a Geomorphic Agent. Rensselaer Polytechnic Institute, Troy, New York. Palmer, A.N., 2002. Karst in Paleozoic Rocks: How Does it Differ from Florida? Karst Waters Institute Specia l Publication, 7: 185-190. Pinder, G.F., Bredehoeft, J.D. and JR, H.H.C., 1969. Determination of Aquifer Diffusivity From Aquifer Response to Flucuations in River Stage. Water Resources Research, 5(4): 850-855. Powers, J.G. and Shevenell, L., 2000. Transm issivity Estimates from Well Hydrographs in Karst and Fractured Aquifers Groundwater, 38(3): 361-369. Rasmussen, W.C. and Andreasen, G.E., 1959. Hydrologic Budget of the Beaverdam Creek Basin, Maryland. In: U.S.G. Survey (Editor). Water Supply Paper. Rindels, S., 1992. Tree Root Systems, ww w.ipm.iastate.edu/ipm/hortnews/1992/4-11992/treeroot.html. Iowa State University Department of Horticulture, Ames, Iowa, pp. 43-44. Risser, D.W., Gburek, W.J. and Folmar, J ., 2005. Comparison of Methods for Estimating Ground-Water Recharge and Base Flow at a Small Watershed Underlain by Fractured Bedrock in the Eastern United St ates. In: U.S.G. Survey (Editor). U.S. Department of the Interior, Reston, Virginia. Rovey, C.W. and Cherkauer, D.S., 1995. S cale Dependency of Hydraulic Conductivity Measurements. Groundwater, 33(5): 769-780. Rushton, K.R., Eilers, V.H.M. and Carter, R.C., 2005. Improved Soil Moisture Balance Methodology for Recharge Estimatio n. Journal of Hydrology, 318: 379-399. Schwartz, F.W. and Zhang, H., 2003. Fundame ntals of Ground Water. John Wiley and Sons, Inc. New York, 583 pp. 143

PAGE 144

Screaton, E., Martin, J.B., Ginn, B. and Smith, L., 2004. Conduit Properties and Karstification in the Unconfined Floridan Aquifer. Groundwater, 42(3): 338-346. Shevenell, L., 1996. Analysis of Well Hydrogr aphs in a Karst Aquifer: Estimates of Specific Yields and Continuum Transmi ssivities. Journal of Hydrology, 174: 331355. Stewart, J.B., 1988. Modelling Surface Conducta nce of Pine Forest. Agricultural and Forest Meteorology, 43: 19-35. Vacher, H.L. and Ayers, J.F., 1980. Hydrology of Small Oceanic Islands Utility of an Estimate of Recharge Inferred from the Ch loride Concentration of the Freshwater Lenses. Journal of Hydrology, 45: 21-37. Vacher, H.L. and Mylroie, J.E., 2002. Eogenetic Karst from the Perspective of an Equivalent Porous Medium. Car bonates and Evaporites, 17(2): 182-196. White, W.B., 1999. Conceptual Models for Ka rstic Aquifers. Karst Waters Institute Special Publication, 5: 11-15. White, W.B., 2002. Karst Hydrology: Recent Developments and Open Questions. Engineering Geology, 65: 85-10. XU, C.Y. and Chen, D., 2005. Comparison of Seven Models for Estimation of Evapotranspiration and Groundwater Rech arge Using Lysimeter Measurement Data in Germany. Hydrologi cal Processess, 19: 3717-3734. 144

PAGE 145

BIOGRAPHICAL SKETCH Michael Ritorto was born in Long Island, NY and lived there for 18 years. His family includes his parents Frank Ritorto a nd Denise Ritorto, a nd his sister Brittany Ritorto. After graduating from Commack Hi gh School in 2001, Michael moved to Ann Arbor, Michigan to attend the University of Michigan (UM). Michael received his bachelors degree in environmental geosci ences from UM in December 2004. Michael went on to pursue his masters degree in geol ogy at the University of Florida, starting in September 2005. Michael is cu rrently seeking employment in the environmental and geological science field. 145