A 700-Year-Long Record of Glacial Surging and Associated Flooding

kramer_b.pdf

kramer_b_Page_32.txt

kramer_b_Page_27.txt

kramer_b_Page_46.txt

kramer_b_Page_01.txt

kramer_b_Page_47.txt

kramer_b_Page_49.txt

kramer_b_Page_37.txt

kramer_b_Page_25.txt

kramer_b_Page_17.txt

kramer_b_Page_23.txt

kramer_b_Page_11.txt

kramer_b_Page_50.txt

kramer_b_Page_03.txt

kramer_b_Page_26.txt

kramer_b_Page_19.txt

kramer_b_Page_55.txt

kramer_b_Page_56.txt

kramer_b_Page_39.txt

kramer_b_Page_33.txt

kramer_b_Page_08.txt

kramer_b_Page_28.txt

kramer_b_Page_42.txt

kramer_b_Page_30.txt

kramer_b_Page_36.txt

kramer_b_Page_41.txt

EORB03B4D_VZ5S5K_xml.txt

kramer_b_Page_12.txt

kramer_b_Page_24.txt

kramer_b_Page_29.txt

kramer_b_Page_40.txt

kramer_b_Page_02.txt

kramer_b_Page_05.txt

kramer_b_Page_13.txt

kramer_b_Page_52.txt

kramer_b_Page_07.txt

kramer_b_Page_10.txt

kramer_b_Page_16.txt

kramer_b_Page_35.txt

kramer_b_Page_04.txt

kramer_b_pdf.txt

kramer_b_Page_09.txt

kramer_b_Page_21.txt

kramer_b_Page_34.txt

kramer_b_Page_20.txt

kramer_b_Page_45.txt

kramer_b_Page_14.txt

kramer_b_Page_43.txt

kramer_b_Page_44.txt

kramer_b_Page_48.txt

kramer_b_Page_06.txt

kramer_b_Page_38.txt

kramer_b_Page_54.txt

kramer_b_Page_51.txt

kramer_b_Page_53.txt

kramer_b_Page_22.txt

kramer_b_Page_15.txt

kramer_b_Page_31.txt

kramer_b_Page_18.txt

cal yr BP


3 10


- -


-


1200 -


-~_


Figue 51. Anonsteay sate ge ode wasdevlope wih th us of210Pbadltooi


Age! MoIdel


400

E 600









700 800 900 1000 1100 1200 1300 1400 1500 1500 1700 1800 1900 2000

8-








700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000

Figure 2-3. A paleo-temperature record for the Gulf of Alaska is based on tree ring data.
Periods of multi-decadal variability in mean summer air surface temperature are
highlighted in blue. Also highlighted is the Little Ice Age (gray), which is not
associated with the multi-decadal shifts in temperature. The horizontal bars represent
overall mean temperature with no significant difference for the given time period.
[Adapted from Wilson, R., Wiles, G., D'Arrigo, R., and Zweck, C., 2006, Cycles and
shifts: 1,300 years of multi-decadal temperature variability in the Gulf of Alaska:
Climate Dynamics, v. 28(4), p. 425-440.]








A 700-YEAR-LONG RECORD OF GLACIAL SURGING AND ASSOCIATED FLOODING:


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

2008









CHAPTER 2
BACKGROUND

Gulf of Alaska Climate

Climatic changes in the Gulf of Alaska (GOA) are primarily controlled by the position of

the Aleutian Low (AL) pressure system. The AL dominates climate in spring and winter months

while during the summer months it is dominated by the North PaciHc High pressure cell (Wiles

et al., 1998). The AL is also known to experience decadal scale variability, which has been

observed in climate indices such as the North PaciHc Index (NPI) (Trenberth and Hurrell, 1994)

and the PaciHc Decadal Oscillation (PDO) (Mantua et al., 1997). The NPI measures the change

in sea level pressure to observe the position and intensity of the AL (Trenberth and Hurrell,

1994). At an eastern or strong position, the AL provides increased precipitation to the coast

(Anderson et al., 2005) as winds bring warmer surface waters into the GOA (Figure 2-la).

During the western or weak position, the coast experiences drier conditions as winds are directed

along shore (Figure 2-1b). Along with the shifts in moisture, the air surface temperature is also

observed to fluctuate along with the phase shifts of the AL. For the past 80 yrs a strong phase in

the AL can be associated with warmer air surface temperatures as the weak phase is associated

with cooler temperatures (Figure 2-2). A multi-decadal shift between the strong and weak phase

of the PDO can also be observed for the past 80 yrs (Figure 2-2). Wilson et al. (2006) developed

a 1300 yr record of coastal temperatures for the GOA based on tree ring data, which revealed the

multi-decadal variability also occurred during ~800-950, ~1080-1100, and ~1300-1400 yr AD

(Figure 2-3). One notable period when the multi-decadal variability is not present is during the

little ice age (LIA; Figure 2-3), which occurred in the GOA region ~1650-1850 yr AD (Wiles et

al., 1999).









the thickness of the low-density beds observed in x-radiographs decrease to 1-3 cm, and they

become more frequent. The lithology at a depth in core of 2-10 m is characterized by

interlaminated-to-interbedded low- and high-density beds with occasional bioturbation (Figure 4-

5). The alternating low- and high-density beds produce a smaller range in GRA bulk density (0.3

g cm-3), but show more small scale variability from 2-6 m depth in core than in the upper 2 m.

Below 10 m, the laminations become rare and are separated by thick (5-20 cm) bioturbated beds

that grade into mottled and massive mud beds (Figure 4-6). The lack of pronounced density

changes in the x-radiographs still creates the same range of values as from 2-10 m (0.3 g cm-3

but the amount of small scale variability is less apparent (Figure 4-3). Common throughout the

core are low-density, low-susceptibility beds, but they are thick (5-20 cm) in the upper 2 m, but

are thinner (3-10 cm) below 2 m (Figure 4-3).

Grain Size

Throughout the core, sequences of low- and high-density beds are observed, and these

sequences can be divided in three distinct lithologic sections of the core as described in the

section above. The differences in the grain size distribution of a low- and high-density bed have

been compared from each of the three lithologies. The low-density beds observed in any of the

sections were finer grained then the high-density bed (Figure 4-7), and the low-density bed in the

upper 2 m of core was finer than low-density beds below 2 m (Figure 4-8).

The low density beds found throughout the core are clay-rich with greater than 60% clay

by mass while the dense beds have lower clay percentages (< 60%) and increased silt and sand

percentages (Figure 4-9a). The percentage of sand for the majority of the core remains relatively

low (< 10%), although at 4.5, 6.0, 6.3, 6.8, 9.1, and 11 m depth sand percentages were > 20%.

The percentage of silt shows the least amount of variability (20-40%) as it parallels sand for the

length of the core. The distribution of modal sand diameter shows little variation as it ranges












BIOGRAPHICAL SKETCH .............. ...............56....





























ITNE 141.=E IfoC irE 11.E 180 17.1. 1.... 1=,a 14% 17*@ l_.n ilow
Sea Level Presssue (rnb) Composite Mean
Jan to Mar: 1977

992 996 1000 1004 1008 1012 1016 1020 1024
a


















130E 140E 150E 160E 170E 180 170W 160W 150W 140W 130W 120W 110W
Sea Level Pressaue (rnb) Composite Mean
Jan to Mar: 1972


b 992 996 1000 1004 100B 1012 1016 1020 1024
Figure 2-1. The position of the Aleutian Low (AL) pressure system is one of the dominate
controls on climate in the Gulf of Alaska. A) The eastern/strong phase produces a
warm and moist climate. B) The west/weak phase produces cooler and drier
conditions.









CHAPTER 5
DISCUSSION

Chronology

To provide age constraint on the lithologic data within the core, an age model was

developed from 210Pb radioisotopes, five radiocarbon dates, and interpretations of the lithology

from x-radiographs. The 210Pb activity was used to calculate a first appearance accumulation

rate by identifying the depth at which 210Pb activities reach supported levels of 226Ra activity.

When supported levels of 226Ra activity have been reached, the 210Pb activity is produced solely

by the decay of 226Ra. At this depth, excess 210Pb that scavenged onto the sediment prior to

deposition has decayed by 4-5 half lives, which represents ~100 yr. In the trigger core supported

levels were never reached, but were trending towards supported near which the deepest

measurement (148 cm) was assumed to be at supported levels. The average accumulation rate

for the upper 2 m of core is ~1.5 cm yr- This rate is associated with episodic deposition and a

non-steady state system. Episodic deposition is evident as a decrease in 210Pb activity at 28 and

98 cm (Figure 4-1) suggesting rapid deposition (Dukat and Kuehl, 1995; Jaeger et al., 1998;

Nittrouer et al., 1979). When a particle is rapidly deposited it spends less time in the water

column were excess 210Pb can scavenge onto that particle, reducing the amount of activity

(Nittrouer et al., 1979). Both decreases in 210Pb activity correlate with the base of a low-density

bed, which corresponds with an outburst flood (Jaeger and Nittrouer, 1999).

The radiocarbon ages were calibrated with CALIB 5.0 (Stuvier and Reimer et al., 1993) for

a 700 yr reservoir correction. The 700 yr reservoir correction was derived using paired shell-

wood ages in the GOA (Tom Ager, pers. com.). Calibration to calendar ages of the five 14C ages

produced large ranges for each respective age. These large ranges limit the age model to broad










4-8 Grain size distribution of the three low-density beds from Figure 16 showing the
upper 2 m of core has a finer signature than the remainder of the core.............................37

4-9 The distribution of grain-size down core for 82JC is shown by a) relative percent
mass of sand, silt, and clay, and b) modal sand and mean portable silt.............................38

5-1 A non-steady state age model was developed with the use of 210Pb and lithologic
based accumulation rates. The increased accumulation rates during 300 500 cal yr
BP correspond to the timing of the LIA............... ...............47

5-2 Sediment lithologies are correlated with known changes in regional climate...................48

5-3 An aerial photograph shows the Malaspina Glacier, which is located east of the
Being Glacier at the mouth of Yakutat Bay. ............. ...............49













800



200 -


400 -



2 son -
O








1200 -


1400 -

Figure 4-2. AMS


1000


120@


1300


1400


radiocarbon ages for 82JC show a decrease of age with depth.


140: yr BP











GI~Bk Clensi~ 1I~br;j


o
O


e
0


1D-


Figure 4-6. The lithology below 10 m depth in core has rare laminations and is dominated by


1020


1025
I
I
1030
I







I




1050


1055









consistently warmer temperature. Instead, extreme warm (>?oC) to cold (< 40C) temperature

changes are observed to occur on less then decadal time scales causing an overall constant mean

air surface temperature (1850-1700 yr AD). This would allow for more frequent flooding during

years with warmer temperatures while years with colder temperatures would be associated with

less discharge and lower sediment accumulation rates. From 10-13.6 m depth in core (> 500 yr

BP), the Bering Glacier terminated at a more advance position than modern day as established by

previous investigators. The lack of preserved physical stratigraphy observed in this unit suggests

bioturbation coupled with low sediment accumulation rates was the dominant condition and

correlates to an overall cooler and possibly drier climate as established by tree-ring records.

Under these drier conditions, meltwater production would be lower, and the conditions favoring

Being Glacier surging were most likely not present.

This study indicates that the behavior of the Bering Glacier has varied during the late

Holocene. The record however did not extend back long enough to encompass another period of

multi-decadal variability (1300-1400 yr AD) in climate to accurately test the hypothesis.

Because of this reason it was difficult to make any interpretation or comparison on the influence

of the multi-decadal shifts in the Aleutian Low and their affect on the behavior of the Bering

Glacier. The collection of a deeper core at this site would allow for multiple periods of multi-

decadal shifts in the climate to be compared. The ability to compare these conditions could

provide greater insight into the potential influence of climate on glacial surging.






























Year
. The correlation of the 1977-1988 Wolverine Glacier positive mass balance
(Josberger et al., 2007) with an eastern/strong AL depicted as negative values for the
Aleutian Low Pressure Index and positive values for the Pacific Decadal Oscillation
Index.


r


O


-1-
-2
1920


Figure 2-6


1930 14 90 16 90 18 90 20









CHAPTER 3
METHODS

During the 2004 R/V Ewing cruise, a 0.5 m-long multicore, 2 m-long trigger core, and 14

m-long jumbo piston core (e.g., EWO408 81MC, 82TC, and 82JC) were respectively collected

~16 km southwest from the mouth of the Seal River on the adjacent continental shelf in 150

meters of water. The collection of three cores allowed for complete recovery with minimum loss

of sediment from the sediment water interface to depth down core. Duplicate Ocean Drilling

Program (ODP)-style u-channels (2 x 2 x 150 cm) were extracted in parallel from the core

archived at Oregon State University. The physical properties and sedimentary structures of the

core were examined using a multi-sensor core logger (MSCL), x-ray radiographs, and grain size

analysis. An age model for the core was constructed using 210Pb geochronology and radiocarbon

dates.

Core Logging Data

After collection at sea, the cores were sectioned into 150 cm lengths and immediately

processed on a GeoTek Multi-Senor Core Logger (MSCL) for the following physical properties;

gamma-ray attenuation (GRA) bulk density, and volume magnetic susceptibility. These data

were processed at a coarse-scale resolution (1 cm). After extraction, ODP-style u-channels were

re-analyzed with an MSCL, at UF, at a finer-scale (0.5 cm) for GRA bulk density and at 1-cm

resolution for magnetic properties.

The depth scale from the shipboard MSCL data was determined by the shipboard party to

represent the most accurate depth scale for all cores, so the u-channel depths were corrected to

the shipboard data. Each section's depth was corrected by adding an additional 0.5 cm to the

core logging data of the u-channel to match that of the shipboard data. Depth correction to the

shipboard data allows for uniform depths among all cores collected during the 2004 cruise.










Jaeger, J. M., and Nittrouer, C. A., 1999, Marine record of surge-induced outburst floods from
the Bering Glacier, Alaska: Geology, v. 27, p. 847-850.

Jaeger, J. M., and Nittrouer, C. A., 2006, A quantitative examination of modern sedimentary
lithofacies formation on the glacially influenced Gulf of Alaska continental shelf:
Continental Shelf Research, v. 26, p. 2187-2204.

Jaeger, J. M., Nittrouer, C. A., Scott, N. D., and Milliman, J. D., 1998, Sediment accumulation
along a glacially impacted mountainous coastline: north-east Gulf of Alaska: Basin
Research, v. 10, p. 155-173.

Joughin, I., Howat, I., Alley, R. B., Ekstrom, G., Fahnestock, M., Moon, T., Nettles, M., Truffer,
M., and Tsai, V. C., 2008, Ice-front variations and tidewater behavior on Helheim and
Kangerdlugssuaq Glaciers, Greenland: Journal of Geophysical Research, v. 113, p.
F01004.

Josberger, E.G., Bidlake, W.R., March, R.S., and Kennedy, B.W., 2007, Glacier Mass-Balance
Fluctuations in the Pacific Northwest and Alaska, USA: Annals of Glaciology, v. 46, p.
291-296.

Kumar, K., Dumka, R. K., Miral, M. S., Satyal, G. S., and Pant, M., 2008, Estimations of retreat
rate of Gangotri glacier using rapid static and cinematic GPS survey: Current Science, v.
94, p. 258-262.

Single, C. S., and Fatland, D. R., 2003, Does englacial water storage drive temperate glacier
surges?: Annals Of Glaciology, v. 36, p.14-20.

Mantua, N. J., Hare, S. R., Zhang, Y., Wallace, J. M., and Francis, R. C., 1997, A Pacific
interdecadal climate oscillation with impacts on salmon production: Bulletin of the
American Meteoroical Society, v. 78, p. 1069-1079.

McCave, N., Manighetti, B., and Robinson, S. G., 1995, Sortable silt and fine sediment
size/composition slicing: Parameters for palaeocurrent speed and palaeoceanography:
Paleoceanography, v. 10 (3), p. 593-610.

Meier, M. F., and Post, A., 1969, What Are Glacier Surges: Canadian Journal Of Earth Sciences,
v. 6(4P2), p. 807.

Merrand, Y., and Hallet, B., 1996, Water and sediment discharge from a large surging glacier:
Being Glacier, Alaska, USA, summer 1994: Annals of Glaciology, v. 22, p. 233-40.

Molnia, B. F., and Post, A., 1995, Holocene history of Bering Glacier, Alaska: A prelude to the
1993-1994 surge: Physical Geography, v. 16, p. 87-117.

Molnia, B. F., Post, A., and Carlson, P. R., 1996, 20th-century glacial-marine sedimentation in
Vitus Lake, Bering Glacier, Alaska, U.S.A.: Annals of Glaciology, v. 22, p. 205-210.









warm and moist conditions during the past 1300 yrs, suggesting that surging could be a persistent

process of the Bering Glacier.

In this study, a record of shelf sedimentation was examined to provide an extended record

of glacial surging. I hypothesize that Bering Glacier surging will be more frequent when positive

annual mass balance (i.e., higher temperature /moisture delivery) periods persist for several

decades coupled with a lack of colder periods that promote freezing. To test this hypothesis an

age model for a 14 m-long core, EWO408 82JC, was developed using 210Pb geochronology and

five radiocarbon dates. X-radiographs of the core were used to identify sediment faces down

core. The physical properties of the core including gamma ray attenuation (GRA) bulk density,

volume magnetic susceptibility, and grain size were also used to identify and interpret

lithofacies. The interpreted climatic conditions for the GOA from Wilson et al. (2006) were used

in conjunction with the age model to compare the timing of climatic events to lithologic

properties to test the hypothesis.









sedimentary fabric suggests a decrease in Ls (< 3 cm) to less than Lb Of in Sediment accumulation

rate. A thicker Lb WOuld allow for episodic deposition to still occur, but not be preserved

because of lower accumulation rates (Bentley et al., 2006).

Coincident with the deposition of the lower unit, the Bering Glacier terminus was at a

more advanced position than modern day, but not on the coast (Figure 5-2b). The position of

this terminus caused glacial drainage to be diverted to a more eastern position as the Kaliakh,

Tsivat, and Tsiu Rivers built an outwash apron between 1300-200 yr BP (Muller and Fleisher,

1995). Similar to the middle unit, the dispersal of a constant volume of sediment over a larger

area, results in a deceased Ls. The decrease in Ls is evident in the x-radiographs as the number of

preserved event layers have decreased and the preservation of biological traces has increased.

The first 50 cm of the lower unit consist of low-density laminations interbedded with thick (> 15

cm) bioturbated beds, which suggests episodic deposition (Jaeger and Nittrouer, 2006). This

type of low-density deposit is similar to deposits observed in the middle unit having a coarser

grain size distribution, but they are less common. Below 11 m, the core is composed of a

mottled to massive mud faces indicating complete mixing from bioturbation and lower

accumulation rates (< 1.0 cm yr Figure 18c) (Jaeger and Nittrouer, 2006).

Lithofacies Interpretations

The three units of the core have notable differences in thickness of the low-density event

beds and bioturbated intervals. The change in thickness observed down core is most likely

caused by the dispersal pathways of glacial drainage in relation to the position of the glacial

terminus, as described above. An understanding of glacial behavior with respect to temporal

climatic change can also be utilized to interpret the deposits observed within the core. This can

be done by calculating a recurrence interval for the low-density event beds from each unit. The

recurrence interval represents the amount of time that has elapsed between the deposition of low-




































Figure 5-3. An aerial photograph shows the Malaspina Glacier, which is located east of the
Being Glacier at the mouth of Yakutat Bay. A modern analog for the terminus
position of the Bering Glacier during the LIA, the Malaspina has seven main
tributaries draining into the GOA.









transport energy during the process of deposition, which can span over a few months (Jaeger and

Nittrouer, 1999). After a surge, the glacier experiences a 20-30 yr-long quiescent period.

During this period, annual meltwater floods produce plumes, but flood faces are not evident in

the sediment record. The majority of sediment from annual melt water floods is probably

deposited in Vitus Lake, which has recorded a maximum accumulation rate from seismic profiles

of 10 m yr- since 1967 (Molina et al., 1996). The annual meltwater events most likely produce a

smaller Ls (< 3 cm) on the shelf, which is reworked by bioturbation and not preserved. The

bioturbated dense layers can be found overlying outburst flood faces as organisms begin to re-

establish the bioturbated layer (Lb).

Middle Unit

Strata at these depths in core are interlaminated to interbedded high- and low-density beds

that are separated by bioturbated beds (Figure 5-2d). The high- and low-density deposits differ

from the upper unit because they are thinner (1-3 cm) and more frequent. Although Ls is reduced

the preservation of these deposits suggests that there was an increase in the overall sediment

accumulation rate (3 cm yr- Figure 5-2c) (Bentley et al., 2006; Jaeger and Nittrouer, 2006).

This is evident in the thin or non-existent bioturbated layers. The increase in sedimentation rates

and lack of bioturbation suggests sedimentation has changed from modern deposition, which is

dominated by thick low-density beds interbedded with bioturbated beds.

The middle unit was deposited around the time when the Bering Glacier reached its

Neoglacial maximum during the LIA and terminated on the coastline (Figure 5-2b). From this

position glacial drainage and floods would likely flow directly onto the continental shelf. When

the Being Glacier was at the coast it was possibly similar to the modern piedmont Malaspina

Glacier, in southern Alaska, with numerous discharge points to the ocean (Figure 5-3). A coastal

terminus would suggest that there was not a single outlet present to focus flow toward one









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

A 700-YEAR-LONG RECORD OF GLACIAL SURGING AND ASSOCIATED FLOODING:
BEING GLACIER ALASKA

By
Branden James Kramer

August 2008
Chair: John M. Jaeger
Major: Geology

Being Glacier, Alaska is one of the largest glaciers in North America, and is largely

known for its dramatic surging events, five of which have occurred in the last century. A primary

late Holocene history of the Bering has been previously established from on-shore studies of

glacial termini position and evidence of glacial advances, but the Little Ice Age (LIA) record of

glacial surging and associated flooding has not been examined. A 14 m-long jumbo core

collected on the adjacent continental shelf reveals a 700-year-long record of flood deposition.

The core was dated using 210Pb chronology and five radiocarbon dates, and can be separated into

three distinct lithologic units based on examination of x-radiographs and physical properties: 1)

the uppermost unit dates from ~120 yr BP to the present and is characterized by bioturbated mud

interbedded with faintly laminated, thick (5-20 cm) low-density beds, 2) the middle unit dates

from ~120-500 yr BP and includes abundant laminated-to-interbedded low- and high-density

beds with some evidence of bioturbation, and 3) a lowermost unit post dates 500 yr BP and is

composed of rare laminated beds grading into mottled to massive mud. In each of these units,

the laminated lithofacies from this oxic mid-shelf location indicates both flood and gravity flow

deposition. Based on previous terrestrial studies, from 850-400 yr BP, the terminus was at a

slightly advanced position relative to the present, and it was at its Holocene Neoglacial












GIHBa~k mesiy lrr


btasseM~cmaEJc9.#ri~rqe (S Ulf)


1.B


e


490-


Figure 4-5. A characteristic section of lithology with in 2-10 m section showing interbedded
low- and high-density beds with some bioturbation (483-490 cm). Ls is typically 1 -
3 cm thick compared to 5-20 cm in the upper 2 m, but events beds are still preserved.
The magnetic susceptibility data is not resolved well enough to show the small scale
variations associated with the thin alternating event beds although they are present in
the GRA bulk density data.









X-ray Radiographs

X-radiographs of each u-channel (~150 cm) were taken in 30 cm segments; 15 cm-long

segments were scanned onto a computer and were spliced into continuous sections using Adobe

Photoshop. To verify that segments were placed together properly, the core logging data were

used to match low density beds from the logging data with low density beds in the x-ray.

Correcting the x-rays with the core logging data also ensures that x-rays have the same depth

scale as the shipboard data. The x-rays were then inverted from negatives to positives so dense

objects appear opaque and less dense objects appear brighter. The contrast of each section was

enhanced to make fainter contacts between beds more apparent. The lithologic properties of the

core were then described and interpreted from the x-radiographs. The x-radiographs were also

utilized in the extraction of grain size samples by printing out true scale versions of the x-

radiographs and placing them next to the actual u-channel to ensure proper sample selection.

Grain Size Data

Numerous grain-size samples were collected from the entire core to provide a detailed

distribution of sediment size. The core logging data were used to determine the relative grain

size of the sediment (e.g., low-density and magnetic susceptibility = fine grain, high-density and

magnetic susceptibility = coarse grain) and to pick samples based on the following properties;

low-density and low-susceptibility, low-density and high-susceptibility, high-density and low-

susceptibility, or high-density and high-susceptibility. The core logging properties were used to

provide a variety of different lithologic attributes found throughout the core, but still have the

ability to compare those samples. Subsamples were extracted at 1.5 cm intervals from the u-

channel, homogenized, and disaggregated so they could be wet sieved at 53 microns. The mud

fraction was then analyzed on a 5100 Micrometrics Sedigraph (Coakley and Syvistski, 1991) and

the sand portion with a settling column (Syvitski et al., 1991). The separate sedigraph and









CHAPTER 1
INTRODUCTION

The dynamic behavior of ice and its relationship to global climate change have become a

major environmental concern. Glaciers in particular are experiencing increased ice discharge

and unprecedented rates of retreat (Joughlin et al., 2008; Nesje et al., 2008; Throst and Truffer,

2008). The retreat is likely climate driven (Joughlin et al., 2008; Nesje et al., 2008; Throst and

Truffer, 2008; Wiles et al., 2008), but there is a poor understanding of glacial response to climate

change over millennial time scales (Brown and Rivera, 2007; Wiles et al., 2008; Zumbithl et al.,

2008). An understanding of dynamic glacial behavior especially the time scale between melting

and/or advance events will provide insight to changes observed in modern day. The study of

modern glacial behavior has been enhanced with the establishment of new techniques for

determining rates of retreat using high-precision GPS (Kumar et al., 2008), ice thickness using

laser altimetry (Arendt et al., 2002; Kumar et al., 2008), and changes in ice mass using Gravity

Recovery and Climate Experiment (GRACE) satellites (Arendt et al., 2007). Observations at this

scale allow for the production of high resolution, temporal and spatial data, but the comparison

of modern data to paleo-data are needed to determine whether the modern observations reveal

anomalous activity or are within the range of previous fluctuations.

The history of Holocene glaciation, in southern Alaska, has been constrained by glacial

termini positions (Calkin et al., 2001; Wiles et al., 2008) and summer air surface reconstruction

for the GOA from tree-ring data (Wilson et al., 2006). Interpreting climatic change from the

terminus position of glaciers is best done using a composite record of glacial terminus histories

to reduce the affects of reworking by the most recent glacier advance (Wilson et al., 2008). The

collection of multiple incomplete or fragmented records could be eliminated by the recovery of a

more complete record. One such environment with the potential to preserve a more complete









state system produces an Ls << Lb where as episodic deposition can create a situation where Ls >

Lb. The preservation of episodic deposition is also related to the residence time (Tr= [Lb -

Ls/2]/t) or amount of time required to bury the bed below Lb (Wheatcroft and Drake, 2003).

Residence time can be decreased if multiple event beds are deposited consecutively essentially

making Ls >> Lb.

GOA Shelf Sedimentation

The southeastern coastline of Alaska is tectonically active as the Pacific plate subducts

underneath of the North American plate. The combination of high coastal relief and large

amounts of precipitation contribute to one of the largest annual shelf sediment accumulation

rates (250 x 106 tons yr ) in North America (Jaeger et al., 1998). The majority of the sediment

entering the GOA occurs during the late summer and fall months as elevated glacial melting and

increased precipitation increase the amount of meltwater and runoff (Stabeno et al., 2004). The

input to the GOA of large volumes of fresh water runoff and along-shore winds produces the

Alaskan Coastal Current (ACC) (Weingartner et al., 2002; Stabeno et al., 2004). The ACC flows

in a western direction causing buoyant surface plumes entering the shelf to be re-directed to the

west and remain inshore of the 50 m isobath (Jaeger and Nittrouer, 2006). Also inshore of the 50

m isobath, bottom wave orbital velocities are the highest. The combination of the ACC and

wave orbital velocities inshore of the 50 m isobath creates a high energy environment with

minimal mud accumulation (Carlson et al., 1977). Other currents available to transport sediment

across shelf are downwelling currents produced by eastern alongshore winds during the winter

months (Stabeno et al., 2004). Also during the winter, strong storms produce higher wave

energies than in the summer months causing the resuspension of muddy sediment in water depths

of < 100 m (Jaeger and Nittrouer, 2006).









BIOGRAPHICAL SKETCH

Branden Kramer grew up in the little town of Highland Heights, KY, and aspired to be a

marine biologist so he could be the best dolphin trainer in the world. To obtain his dream, he

consistently watched the Discovery channel to soak in the wonderful opportunities the ocean had

to offer. Once he graduated high school, he enrolled into the Marine Science program at Coastal

Carolina University. While at Coastal, Branden realized that the job market for dolphin trainers

was few and far between, so he decided to search for other opportunities within the marine

realm. After a few weeks of searching, he stumbled upon a marine geologist by the name of Eric

Wright that needed help collecting samples one summer. As summer ended, Branden acquired

an interest for the geosciences. Eric allowed him to continue working on various projects and

eventually led to a senior project where Branden used ground penetrating radar (GPR) to study

the stratigraphy and evolution of a barrier island system. Branden had finally found his calling

as a geologist. Knowing that he would want to continue his education, he applied to graduate

schools and was accepted to the University of Florida. During his three years at UF he studied

glacially influenced, continental shelf deposits in the Gulf of Alaska. He received his master's

degree in August of 2008 and has moved onto the next step of his life.









Upper Unit

Episodic deposition is noted in the strata of the upper 2 m by interbedded high- and low-

density beds (Figure 5-2d). The high-density beds are characterized by bioturbated mud that

typically lack sedimentary fabric. The low-density beds correspond to thick clay-rich deposits

with sharp basal contacts and internal laminations. The preservation of the low-density beds (Ls

= 5-20 cm) is a product of their thickness compared to the thickness of the zone of active

bioturbation (Lb ~3-7 cm; Jaeger and Nittrouer, 2006). These low-density beds have been

correlated to the historic surges and associated outburst floods of the Bering Glacier, and the

bioturbated intervals correlate to the quiescent periods between surges (Jaeger and Nittrouer,

1999).

For the past century, the Bering Glacier has terminated into proglacial Vitus Lake (Figure

5-2b). The lake acts as a sediment trap for coarse grains only allowing silt and clay sized

particles to be transported via the Seal River to the GOA (Merrand and Hallet, 1996). The

sedimentary evidence of this are the low-density beds observed in the upper unit that have a finer

grain size distribution compared to low-density beds in the lower two units (Figure 4-8), a period

when Vitus Lake is not present. This increase can also be observed in the percentage of clay by

mass (> 60%) in the upper unit (Figure 4-9a). As an outburst flood propagates from the Bering

Glacier, the Seal River acts to focus the flood plume out onto the shelf. The focusing of the

plume along with the ACC (Stabeno et al., 2004) cause deposition to be concentrated within the

50 m isobath. The large supply of suspended sediment could create a high concentration regime

where bottom wave orbital velocities allow for the sediment to remain suspended at the seafloor

until a critical concentration (10 gl ) is reached producing a fluid mud (Fan et al., 2004), and the

resulting deposits produce thick low-density, clay-rich outburst flood faces (Figure 5-2d).

Observation of dense internal laminations within the low-density bed suggests fluctuations in









The observation of more frequent low-density beds suggests that the glacier was flooding more

frequently, which would continually flush water and sediment from underneath the glacier. Not

allowing large volumes of water and sediment to build underneath the glacier would produce

smaller magnitude floods. This combined with the dispersal of sediment discharge over a larger

area would create a thinner deposit as observed. The recurrence interval between interbedded to

interlaminated low- and high-density sequences was 3-10 yrs. This recurrence interval is not as

prolonged as in the upper unit and may be related to the shorter time periods between shifts from

warm to cold conditions during the LIA (1850-1650 AD, Figure 5-2c). The bioturbated intervals

could represent a colder drier period when meltwater flooding was at a minimum, resulting in

small scale events not preserved in the strata. The interbedded low-density sequence on the other

hand would be produced during a warmer, moist period when the glacier was actively flooding

sub glacial water was not being trapped underneath the glacier as observed in modern conditions.

This would suggest that during this time period the Bering Glacier was not surging on the

modern multi-decadal time scale.

Similar to the middle unit, the lower unit represents a period of time with multi-year long

periods between shifts from warm to cool temperatures, but the shifts were not as dramatic as

during the LIA (1700-1450 AD, Figure 5-2a). This would have produced less intense

precipitation during warmer periods compared to the LIA. Overall the lower unit was deposited

when the average climate was cooler and most likely drier causing smaller magnitude flood

events. The combination of a cooler climate and a lack of preserved event imply that this period

of time was not favorable for glacial surging.














TABLE OF CONTENTS


page


ACKNOWLEDGMENTS .............. ...............3


LIST OF FIGURES .............. ...............6


LI ST OF AB BREVIAT IONS ............... ...............8


AB ST RAC T ............... ...............9


CHAPTERS


1 INTRODUCTION ............... ...............11


2 BACKGROUND .............. ...............14


Gulf of Alaska Climate............... ...............14

Being Glacier............... ...............15

Preservation of Event Layers............... ...............16

GOA Shelf Sedimentation .............. ............... 17


3 METHODS .............. ...............25


Core Logging Data .............. ...............25

X-ray Radiographs............... ...............26
Grain Size Data............... ...............26

Age Model .............. ...............27


4 RE SULT S .............. ...............28


Chronology .............. ...............28

Lithologic Description............... ...............28

Grain Size .............. ...............29


5 DI S CU SS ION ............... ...............3 9


Chronology .............. ...............39

Lithofacies .............. ...............40

Upper Unit ............... ...............41
Middle Unit .............. ...............42

Lower Unit............... ...............43

Lithofacies Interpretations.............. ...............44


6 CONCLUSIONS .............. ...............50


LIST OF REFERENCES............... ...............52





4


































-- ------ --~


1 10


100


1 10
Microns


100


Figure 4-7. The grain size distributions for low- and high-density beds show low-density beds










Mass Percent


Grain size (Ip)


*
*
, *


*
*
. *
*


*
,*
* **
*


Figure 4-9. The distribution of grain-size down core for 82JC is shown by a) relative percent
mass of sand, silt, and clay, and b) modal sand and mean portable silt. Clay-rich (>
60%) low-density, low-susceptibility beds are represented by the highlighted gray
areas. The modal sand diameter remains relatively constant down core, but the mean
portable silt size does show some variation although there is no discernable down core
trends.


Sand 9
O
-SiR o
.5
Clay
-Ei
e


. O Modal sand
Mean
*
Sortable sil










Throst, D. E., and Truffer, M., 2008, Glacier recession on Heard Island, southern Indian Ocean:
Artic, Antarctic, and Alpine Research, v. 40(1), p. 199-214.

Trenberth, K. E., and Hurrell, J. W., 1994, Decadal atmosphere-ocean variations in the PaciHc:
Climate Dynamics, v. 9, p. 303-319.

Weingartner, T.J., Coyle, K., Finney, B., Hoperoft, R., Whitledge, T., Brodeur, R., Dagg, M.,
Farley, E., Haidvogel, D., Haldorson, L., Hermann, A., Hinckley, S., Napp, J., Stabeno, P.,
Kline, T., Lee, C., Lessard, E., Royer, T., and Strom, S., 2002, The Northeast PaciHc
Globec Program: Coastal Gulf of Alaska: Oceanography, v. 15 (2), p. 10-35.

Wheatcroft, R. A., 2000, Oceanic flood sedimentation: a new perspective: Continental Shelf
Research, v. 20 (16), p. 2059-2066.

Wheatcroft, R. A., and Borgeld, J. C., 2000, Oceanic flood deposits on the northem Califomia
shelf: large scale distribution and small-scale physical properties: Continental Shelf
Research, v. 20, p. 2163-2190.

Wheatcroft, R.A., and Drake, D.E., 2003, Post-depositional alteration and preservation of
sedimentary event layers on continental margins, I. The role of episodic sedimentation:
Marine Geology, v. 199 (1-2), p. 123-137.

Wheatcroft, R. A., SommerHeld, C. K., Drake, D. E., Borgeld, J. C., and Nittrouer, C. A., 1997,
Rapid and widespread dispersal of flood sediment on the northern Califomia margin:
Geology, v. 25, p. 163-166.

Wiles, G. C., Barclay, D. J., Calkin, P. E., and Lowell, T. V., 2008, Century to millennial-scale
temperature variations for the last two thousand years indicated from glacial geologic
records of Southem Alaska: Global and Planetary Change, v. 60, p. 115-125.

Wiles, G. C., D'Arrigo, R. D., and Jacoby, G. C., 1998, Gulf of Alaska atmosphere-ocean
variability over recent centuries inferred from coastal tree-ring records: Climate Change, v.
38, p. 289-306.

Wiles, G. C., Post, A., Muller, E. H., and Molnia, B. F., 1999, Dendrochronology and late
Holocene history of Bering Piedmont Glacier, Alaska: Quaternary Research, v. 52, p. 185-
195.

Wilson, R., Wiles, G., D'Arrigo, R., and Zweck, C., 2006, Cycles and shifts: 1,300 years of
multi-decadal temperature variability in the Gulf of Alaska: Climate Dynamics, v. 28(4), p.
425-440.

Zumbithl, H. J., Steiner, D., and Nussbaumer, S. U., 2008, 19th century glacier representations
and fluctuations in the central and westem European Alps: An interdisciplinary approach:
Global and Planetary Change, v. 60, p. 42-57.













GFtWllkhmrily@ cr~


ac


DD
-




05
-



O
100 -
O
O



e
Cl

110
-



115
-


Fiue44 -aigrp oiie fpaallrgtadletucanl ee xrce ln h


Volme MusplelicanwplihMy(S islile)
21 2 a a a









a



100








150



its














Supper


1 10


110


Low-density


Microne







































O 2008 Branden James Kramer









ACKNOWLEDGMENTS

A special thanks to my family, Barbry, Danny-boy, Justbry, and Kel-Kel, for all of the

support they have shown. Thanks to my girl friend, Erin, for keeping me sane during the stress

caused by writing this masterpiece. I thank my advisor, John Jaeger, and the rest of my

committee (i.e. Ellen Martin and Joann Mossa) for all of the help and guidance that they have

provided me in the last few years. I would like to say thanks to all of my friends in the

department, especially Derrick, David, and Mike. Last but not least I would also like to thank

Gilly for her time spent teaching me how things worked in the lab.

















200lal -na 400 yr B


0~, 0 y P


400 850 yr BP









positive mass balance for the Wolverine Glacier has been observed to correlate with an

eastern/strong phase of the AL (Figure 2-6). The periods of an eastern/strong AL for the past 80

yrs can be correlated to the five historic surges observed for the Bering Glacier (Figure 2-2).

The Bering Glacier experiences two different types of flooding; typical summer meltwater

floods and outburst floods. Associated with Bering Glacier surges are large outburst floods that

discharge ~70% more then a typical summer seasonal ablation (Merrand and Hallet, 1996). Both

types of floods transport large volumes of suspended sediment, most of which is deposited into

Vitus Lake (Merrand and Hallet, 1996; Molnia et al., 1996). The finer grained sediment not

deposited in Vitus Lake is transported to the Gulf of Alaska in a buoyant surface plume. A study

by Jaeger and Nittrouer (1999) recognized that outburst floods are deposited as thick low-density

beds that are internally laminated and were rapidly deposited, based on 234Th data. Jaeger and

Nittrouer (1999) were able to correlate outburst faces to a historical surge event using 210Pb

geochronology on the core. In contrast, the meltwater floods are not well preserved in the

sedimentary record.

Preservation of Event Layers

The preservation of an outburst faces or any event bed in the sedimentary record is

controlled by three factors (1) sediment accumulation rate (t), (2) thickness of the event layer

(Ls), and (3) the thickness of the bioturbation layer (Lb) (Figure 2-7). To preserve the

sedimentary fabric of an event layer in a steady state system, Ls needs to be thicker than Lb

(Nittrouer and Sternberg, 1981). A thicker Ls produces rapid burial of the event layer base and

allows complete preservation. High preservation is also found from the rapid accumulation of

sediment (Nittrouer and Sternberg, 1981). It can be assumed that areas experiencing episodic

deposition have a larger potential to preserve sedimentary fabric (Wheatcroft and Drake, 2003)

than in a steady state system (Bentley et al., 2006). The constant flux of sediment for a steady










within 1 cp for the entire length of core (Figure 4-9b). The mean sortable silt is the average size


0.10


10.00


m


40
.


60

o
80
-


a.
e 100





140
-


a I#gh-denehy EHMI


+ 81MC210Fb Total Activity


160 -' 160

Figure 4-1. 210Pb and 226Ra activity from core 81MC and 82TC and a schematic of core 82TC
show decreases in 210Pb activity correlate to low-density beds. Accumulation rates
calculated from the 210Pb data were ~1.5 cm yr .


Activity (dpmig)


so=









Being Glacier

The Bering Glacier is one of the largest temperate glaciers in North America with an area

of 5,200 km2. It extends down from the Bagley Ice Field in the St. Elias Mountains where its

lower piedmont lobe is currently composed of the Stellar and Bering lobes and is more than 40

km wide. During the mid-Holocene however the terminus of the Bering Glacier was located 50

km or greater inland, creating a broad bay and narrow fjord where the Bering currently

terminates (Mueller and Fleisher, 1995). It began to advance from this inland position around

5,000 BP building the Neoglacial foreplain observed today. Wiles et al., (1999) were able to

identify three major termini positions of the Bering Glacier during the late-Holocene (850-0 yr

BP). These positions include a more advanced terminus than modern day (850-400 yr BP), its

Neoglacial maximum that terminated at the coast (400-200 yr BP), and its modern position (200-

0 yr BP) (Figure 2-4). The Bering lobe is currently terminating into proglacial Vitus Lake which

drains into the GOA via the Seal River (Figure 2-5).

The Bering Glacier is known to experience quasi-periodic increase in ice flow, known as

glacial surging. The first indications of glacial surging were observed from aerial observations

of folding in the medial moraines of the Bering Glacier (Post, 1972). For many glaciers, surges

appear to follow decadal-long periods of slow flow and thickening. Eventually, an instability

threshold is reached leading to sudden acceleration of flow and rapid ice thinning. The system

then returns to slow, quiescent flow (Meier and Post, 1969; Raymond, 1987). Some surging

glaciers in Alaska (e.g., Bering Glacier, Variegated Glacier) appear to follow this cycle. Eisen et

al. (2001) reconstructed the annual mass balance for the Variegated Glacier, and found that the

length of time between surges was due to the variability in ice accumulation rates prior to the

surge. Also, there is evidence that Bering Glacier surges are triggered by several consecutive

years of above-normal winter balances (Tangborn, 2002; Lingle and Fatland, 2003). A period of









CHAPTER 4
RESULTS

Chronology

A chronology for the trigger core was determined using 210Pb radioisotope geochronology

while AMS radiocarbon dates were derived from the jumbo core samples. The profile of 210Pb

and 226Ra activity shows that 210Pb activity decreases from 10-1 dpm g- while supported levels

of226Ra remain close to 1 dpm g- with depth in the core (Figure 4-1). The 210Pb profile can be

characterized by a non-steady state profile (Jaeger et al., 1998). This is evident as activity

decreases to supported levels at 28 and 98 cm depth. Both of these depths are associated with

low density beds, which are interpreted to represent flood beds (Jaeger and Nittrouer, 1998). It

should be noted that the 210Pb activities have not yet reached supported levels at 148 cm depth.

Based on the first appearance approach an average accumulation rate of 1.5 cm yr- was

calculated (Goodbred and Kuehl, 1998). The AMS radiocarbon dates increase with depth from

~900-1300 14C yr BP (Figure 4-2).

Lithologic Description

The core logging data show large variations in the upper 2 m of core (Figure 4-3). The

GRA bulk density has a range of 0.6 g cm-3 in the upper 2 m while it typically only ranges by 0.3

g cm-3 FOr the remainder of the core. X-radiographs show that the strata of the upper 2 m are

composed of thick (5-20 cm), faintly laminated low-density beds interbedded with dense

bioturbated beds (Figure 4-4). The low-density beds in the x-radiographs correlate with low

values in the GRA bulk density and volume magnetic susceptibility, and the dense bioturbated

beds correlate with higher GRA bulk density and magnetic susceptibility values (Figure 4-4). It

is also evident that the susceptibility data only show large scale variations in the core while the

density data correlate well with changes observed in the x-radiographs (Figure 4-4). Down core










LIST OF FIGURES


FM Dage

2-1 The position of the Aleutian Low (AL) pressure system is one of the dominate
controls on climate in the Gulf of Alaska. .............. ...............18

2-2 Records of the Pacific Decadal Oscillation (PDO) index and air surface temperatures
from Cordova for the Gulf of Alaska appear to be correlated, which relates a strong
or eastern Aleutian Low (AL) to warm temperatures............... ...............19

2-3 A paleo-temperature record for the Gulf of Alaska is based on tree ring data. .................20

2-4 A reconstruction of three termini positions of the Bering Glacier for the past 850 yrs
were derived from radiocarbon dates of buried forest............... ...............21

2-5 Map of the Bering Glacier, showing its 2001 terminus position that terminates into
proglacial Vitus Lake............... ...............22

2-6 The correlation of the 1977 1988 Wolverine Glacier positive mass balance with an
eastern/strong AL depicted as negative values for the Aleutian Low Pressure Index
and positive values for the Pacific Decadal Oscillation Index. ............. ...............23

2-7 The variables affecting the preservation potential of an event bed include the
sediment accumulation rate (t), event bed thickness (Ls), and depth of bioturbation
(Lb). ............. ...............24

4-1 210Pb and 226Ra activity from core 81MC and 82TC and a schematic of core 82TC
show decreases in 210Pb activity correlate to low-density beds. ............. ...............30

4-2 AMS radiocarbon ages for 82JC show a decrease of age with depth............... ...............31

4-3 The shipboard GRA bulk density and volume magnetic susceptibility for core 82JC
are illustrating the physical properties of each lithologic unit............... ...............32

4-4 X-radiograph positives of parallel right and left u-channels were extracted along the
center of sections from core 82JC............... ...............33

4-5 A characteristic section of lithology with in 2 10 m section showing interbedded
low- and high-density beds with some bioturbation (483 490 cm). ............. ...............34

4-6 The lithology below 10 m depth in core has rare laminations and is dominated by
bioturbated intervals (1027 1048 cm)............... ...............35

4-7 The grain size distributions for low- and high-density beds show low-density beds
have a finer signature then high-density beds............... ...............36










LIST OF REFERENCES


Anderson, L., Abbott, M. B., Finney, B. P., and Bums, S. J., 2005, Regional atmospheric
circulation change in the North Pacific during the Holocene inferred from lacustrine
carbonate oxygen isotopes, Yukon Territory, Canada: Quatemary Research, v. 64, p. 21-
35.

Aren't, A. A., Echelmeyer, K. A., Harrison, W. D., Lingle, C. S., and Valentine, V. B., 2002,
Rapid wastage of Alaska Glaciers and their contribution to rising sea level: Science, v.
287, p. 382-386.

Aren't, A., Luthake, S., Abdelati, W., Larsen, C., Lingle, C., Echelmeyer, K., Rowlands, D., and
Krabill, W., 2007, Estimates of glacier mass change in the St. ~Elias Mountians of Alaska,
USA and Yukon Territory, Canada: a strategy for combining GRACE and aircraft laser
altimetry data: Eos Trans. AGU, v. 88(52), Fall Meet. Suppl., Abstract, C12A-07.

Bentley, S. J., Sheremet, A., and Jaeger, J. M., 2006, Event sedimentation, bioturbation, and
preserved sedimentary fabric: Field and model comparisons in three contrasting marine
settings: Continental Shelf Sedimentation, v. 26, p. 2108-2124.

Brown, F., and Rivera, A., 2007, Climate changes and recent glacier behaviour in the Chilean
Lake District: Global and Planetary Change, v. 59, p. 79-86.

Calkin, P. E., Wiles, G. C., and Barclay, D. J., 2001, Holocene coastal glaciation of Alaska:
Quaternary Science Reviews, v. 20, p. 449-461.

Carlson, P. R., Molina, B. F., Kittlson, S. C., and Hampson, J. C., 1977, Distribution of bottom
sediments on the continental shelf, northern Gulf of Alaska, Scale 1:500,000: US
Geological Survey Misc. Field Study Map MF-876, p. 1-13.

Coakley, J. P., and Syvitski, J. P. M., Syvitski, J. P. M., (ed) 1991, Sedigraph Technique,
Methods and Applications ofParticle Size Analysis: Cambridge University Press, New
York, p. 129-142.

Dukat, D. A., and Kuehl, S. A., 1995, Non steady state 210Pb flux and the use of 228Ra/226Ra as a
geochronometer on the Amazon continental shelf: Marine Geology, v. 125, p. 329-350.

Eisen, O., Harrison, W. D., and Raymond, C. F., 2001, The surges of Variegated Glacier, Alaska,
USA, and their connection to climate and mass balance: Journal Of Glaciology, v. 47(158),
p. 351-358.

Fan, S., Swift, D., Traykovski, P., Bentley, S., Borgeld, J., Reed, C., and Niedoroda, A., 2004,
River flooding, storm resuspension, and event stratigraphy on the northem Califomia shelf:
observations compared with simulations: Marine Geology, v. 210, p. 17-41.

Goodbred, S. L., and Kuehl, S. A., 1998, Floodplain processes in the Bengal Basin and the
storage of Ganges-Brahmaputra river sediment: an accretion study using 137C and 210Pb
geochronology: Sedimentary Geology, v. 121, p. 239-258.








- ~PDO Index Teprur


2.5
2.0
1.5
1.0
0.5
0.0
-0.5
-1.0


-2.0


8

70
e
6 L..

5e
a.
E
4 e

3 m

2

1 2
0


-2.5 -C-


Figure 2-2. Records of the Pacific Decadal Oscillation (PDO) index and air surface temperatures
from Cordova for the Gulf of Alaska appear to be correlated, which relates a strong or
eastern Aleutian Low (AL) to warm temperatures. The five shaded regions represent
the five historical surges for the Bering Glacier, which occur within 5 yr after a multi-
year period of increased warming or strong AL. The climate record and the surges
both exhibit a multi-decadal periodicity.


1930 1940 1950 1960 1970 1980 1990 2000









portion of the shelf rather a flood would be dispersed across a larger portion of the shelf, but still

directed westward by the ACC. The dispersal of the same volume of sediment across a larger

area would produce a thinner Ls (1-3 cm), as observed from x-radiographs. The large amounts of

sediment associated with modern outburst flooding and the absence of Vitus Lake to trap

sediment should produce thick flood deposits during this interval, but they are not observed in x-

radiographs. This could imply that the area of flood dispersal increased or that more frequent

floods resulted in lower sediment yields per event.

The absence of Vitus Lake would also allow for the distribution of a coarser range of grain

sizes to the shelf. This is evident in a coarser signature for the low-density beds of the middle

unit compared to the upper unit (Figure 4-8). It is also apparent in the increase of sand beds

observed from x-radiographs, but the modal sand size of these sand beds is similar to other sand

beds found throughout the core (Figure 4-9b). The range of modal sand varies between 3.1-3.89

and implies that similar shear stresses were available during transport to the core location. The

increase in the number of sand beds could be a result of increased storm activity causing the

frequency of sand deposition to increase. Another possibility could be related to the volume of

sediment the Bering is delivering onto the shelf while positioned on the coastline. This would

supply a large amount of sand, which can then be transported across shelf by normal winter

storms. With more sediment available for transport, Ls would increase allowing for the

preservation of more sand beds. The sand beds are less common in the upper and lower units,

indicating that these are periods of time when large volumes of sand were not being deposited at

the core site.

Lower Unit

The laminated deposits in this unit become less abundant and are interbedded with thick (>

20 cm) beds of mottled mud (Figure 5-2d). A decrease in the preservation of physical










































*75m





-270m


Figure 2-5 Map of the Bering Glacier, showing its 2001 terminus position that terminates into
proglacial Vitus Lake. Coloring of the terrestrial landscape represents uncovered ice
(blue), debris covered ice (red), and ice free areas (green). The jumbo piston core
82JC is located ~16 km from the mouth of the Seal River in 150 m of water. The
location of other cores collected during a 1994-95 cruise are also shown. Bathymetric
data at the core location shows surface morphology of varying depth of the shelf and
trough. We gratefully acknowledge the USGS for the Landsat image of the Bering
Glacier, and Larry Mayer and the staff at the Center for Coastal and Ocean Mapping
(CCOM)/ Joint Hydrographic Center (JHC) for collecting, processing, and sharing
these swath bathymetry data.









Molina, B. F., Post, A., Trabant, D. C., and Krimmel, R. M., 1994, The 1993-1994 surge of the
Being Glacier, Alaska: An overview: Eos (Transactions, American Geophysical Union),
v. 75, p. F62.

Muller, E. H., and Fleisher, P. J., 1995, Surging history and potential for renewed retreat: Bering
Glacier, Alaska, U.S.A.: Arctic and Alpine Research, v. 27, p. 81-88.

Mulder, T., and Syvitski, J. P. M., 1995, Turbidity currents generated at river mouths during
exceptional discharges to the world oceans: Joumal of Geology, v. 103, p. 285-299.

Nesje, A., Bakke, J., Dahl, S. O., Lie, 0., and Matthews, J. A., 2008, Norwegian moutian
glaciers in the past, present, and future: Global and Planetary Change, v. 60, p. 10-27.

Nittrouer, C.A., and Sternberg, R.W., 1981, The formation of sedimentary strata in an
allochthonous shelf environment: The Washington Continental Shelf: Marine Geology, v.
42, p. 210-232.

Nittrouer, C. A., Sternberg, R. W., Carpenter, R., and Bennett, J. T., 1979, The use of 210Pb
geochronology as a sedimentological tool, application to the Washington continental shelf:
Marine Geology, v. 31, p. 297-316.

Post, A., 1972, Periodic surge origin of folded medial moraines on Bering Piedmont Glacier,
Alaska: Joumal of Glaciology, v. 11, p. 219-226.

Raymond, C. F., 1987., How Do Glaciers Surge A Review: Joumal Of Geophysical Research-
Solid Earth And Planets, v. 92(B9), p. 9121-9134.

Rodionov, S., Bering Climate and Ecosystems: NOAA, July, 14 2006, April, 18 2008,
http://www.beringelimate.noaa.gov/data/BCresult.php.

Sommerfield, C. K., and Wheatcroft, R. A., 2007, Late Holocene sediment accumulation on the
northem California shelf: ocean, fluvial, and anthropogenic influences: GSA Bulletin, v.
119, p. 1120-1134.

Stabeno, P.J., Bond, N.A., Hermann, A.J., Kachel, N.B., Mordy, C.W., and Overland, J.E., 2004,
Meteorology and oceanography of the northern Gulf of Alaska: Continental Shelf
Research, v. 24 (7-8), p. 859-897.

Stuiver, M., and Reimer, P. J., 1993, Extended 14C database and revised CALIB radiocarbon
calibration program: Radiocarbon 35:215-230.

Syvitski, J. P. M., Asprey, K. W., and Clattenburg, D. A., Syvitski, J. P. M., (ed) 1991,
Principles, design, and calibration of settling columns, Methods and Applications of
Particle Size Analysis: Cambridge University Press, New York, p. 45-63.

Tangborn, W., 2002, Connecting Winter Balance and Runoff to Surges of the Bering Glacier,
Alaska: http://www.hymet.com/docs/yakuart4.pdf, p. 1-17.








Bio~turba~tion


bL


1111Historical Zone:11111111


Figure 2-7. The variables affecting the preservation potential of an event bed include the
sediment accumulation rate (t), event bed thickness (Ls), and depth of bioturbation
(Lb). An exponential curve shows the rate of biological mixing (2) is the most
intense at the sediment water interface and decreases with depth. [Adapted from
Bentley, S. J., Sheremet, A., and Jaeger, J. M., 2006, Event sedimentation,
bioturbation, and preserved sedimentary fabric: Field and model comparisons in three
contrasting marine settings: Continental Shelf Sedimentation, v. 26, p. 2108-2124.]


zone of











Vokre


AmumumanRale










I
I



I
12


Temperature (OC)


0-200yrBP


200- 40 y B


"~~ __


g 1 Ig 400 850 yr B
P it *,

Figure 5-2. Sediment lithologies are correlated with known changes in regional climate. The three lithologic units have been
separated by distinct lithologic properties, but can also be described in terms of A) climate and B) glacial terminus
position. The change in air surface temperature for the last 1300 yrs was derived from tree ring data collected from the
Gulf of Alaska (Wilson et al., 2006). Three terminus positions for the Bering Glacier during the past 850 yr are shown
youngest to oldest from top to bottom (Wiles et al., 1999). Also shown in this figure are C) core logging and accumulation
rate data and D) x-radiographs of characteristic lithologies from each unit.









interpretations, but when coupled with the 210Pb data they provide an adequate chronology to

develop an age model.

A curve was fit to the 14C chronology by estimating ages based on accumulation rates

(Figure 5-1). The ages were calculated by dividing a constant depth interval by the 210Pb or

lithologic based accumulation rate. The 210Pb derived accumulation rate (1.5 cm yr ) was used

for the upper 2 m of the core while the accumulation rates for the remainder of the core were

assumed based on lithologic properties. The lack of bioturbation and preservation of physical

strata between 2-10.4 m depth indicate accumulation rates of > 2-3 cm yr (Jaeger and Nittrouer,

2006). Below 10.4 m depth, bioturbated beds become more dominant while physical strata is not

well preserved suggesting that accumulation rates decrease to ~1 cm yr (Jaeger and Nittrouer,

2006). The resulting age model indicates that the upper 2 m of thicker interbedded deposits

formed between 120 cal yr BP and the present. The age model also indicates that the middle

interval (2-10 m) of thinner but more frequent interbeds formed from 120-500 cal yr BP, and that

the lower bioturbated beds formed between 500-700 cal yr BP.

Lithofacies

The distinct lithologies of the core that were used to determine the accumulation rates for

the age model can be used to separate the core into three distinct lithologic units. The age ranges

for each lithologic unit, determined from the age model, can be correlated with three major

changes in the position of the Bering Glacier terminus (Figure 2-4) (Wiles et al., 1999). The

position of the terminus likely affects the routing of glaciofluvial drainage and ultimately the

transport of sediment to the GOA. In this section, each unit will be discussed with respect to the

positions of the terminus, and how sedimentation is affected at the core site.









Vhanme


GRA But Danisty


Figure 4-3. The shipboard GRA bulk density and volume magnetic susceptibility for core 82JC
are illustrating the physical properties of each lithologic unit. Depths experiencing a
low in density and susceptibility have been highlighted in gray. It should be noted
that large shift in the density and susceptibility data in the upper 2 m are not observed
with depth downcore. Section breaks in the core are delineated by dashed lines.









maximum extent, which was close to the modern coastline, during the LIA (200-350 yr BP). The

thick low-density, clay-rich beds in the uppermost unit correlate with historic outburst floods

associated with known surge events. During the LIA, bioturbated intervals are rare and thin

while laminated intervals are common. Given average radiocarbon-supported sedimentation

rates of 2-3 cm yr this style of interbedding indicates frequent, if not annual, deposition at the

core's location. This would suggest that melt water plumes and redistribution by winter storms

are more prevalent during the LIA rather than outburst flood deposits typical of the past century.

The infrequent deposition of event layers in the lowermost unit could be attributed to the

enhanced diversion of glacial drainage to the eastern Kaliakh, Tsiu, and Tsivat Rivers instead of

present day Seal River. The observation of thinner flood beds during the LIA differs appreciably

from the past century, suggesting that the Bering Glacier was not surging at a multi-decadal

interval, and its position during this period allowed for increased accumulation rates and

preservation of annual melt water flood deposits.









density event beds, and was calculated by dividing an average thickness of bioturbated strata

between low-density event beds by an assumed accumulation rate. In the upper unit, the average

thickness of bioturbated strata was 20 cm which are associated with an average accumulation

rate of 1.0 cm yr- which corresponds to a recurrence interval of20 yrs. The middle unit has

thinner bioturbated intervals (5-15 cm) and higher sediment accumulation rate (1.5 cm yr- )

giving it a recurrence interval of 3-10 yrs. The difference in sediment accumulation rates

between the upper and middle unit was determined based on the amount of preserved physical

strata observed in the bioturbated intervals. The upper unit displayed more prevalent biological

structures (e.g., burrows) with little to no preservation of physical strata while the middle unit

exhibits physical strata that were only partially destroyed.

The 20-30 yr recurrence interval corresponds to the quiescent period before the Bering

Glacier surges. During the quiescent period volumes of water and sediment are stored

underneath the glacier and if enough water is stored over time could produce the buoyant force

that is thought to trigger a surge event. This excess water is possibly produced during the multi-

decadal periods of increased temperature and moisture associated with a strong AL (Figure 5-

2a). The excess water and the positive mass balance created during a strong AL could produce

the instability required to allow a surge to occur. The stored water and sediment is then released

during and after the surge in an outburst flood. The large amounts of sediment and the focusing

of the outburst flood by Vitus Lake and the Seal River combine to help create the thick, low-

density deposits observed on the shelf.

It is evident from the lithology and grain size of the core that glacier surging, as manifested

by thick, low-density beds, is not as pronounced in the lower two units. The low-density beds in

the middle unit are thinner but more frequent while bioturbated intervals were less abundant.









CHAPTER 6
CONCLUSIONS

The results of this study reveal that over the past 500 years changes in sedimentation on

the continental shelf seaward of the Bering Glacier correlate with historical records of surging

events, outburst floods, and with previously established paleo-temperature record based on

coastal tree-ring data. These correlations suggest that large infrequent outburst floods typical of

the past century are a result of multi-decadal periods of higher precipitation coupled with higher

mean summer temperatures. This study also reveals that the Bering Glacier terminus position

influences the delivery of sediment to the adjacent continental shelf.

Modern day sedimentation consisting of thick interbedded mottled and laminated mud

beds is present in the upper 2 m of core (0-120 yr BP), which was deposited during a period of

time when the Bering Glacier terminates in proglacial Vitus Lake. The lithofacies that dominate

this unit are thick, laminated outburst flood faces, which correlate with glacial surge events.

These low-density beds representing outburst flood faces have a ~20 yr recurrence interval,

which correlates with the prolonged multi-decadal warm trends (~70C) in mean summer air

surface temperature for the past century. During these prolonged warmer climate periods excess

meltwater could be stored sub-glacially, potentially creating the buoyant force needed to trigger a

surge. Between 2-10 m depth in core (120-500 yr BP), a period when other investigators have

determined that the Bering Glacier terminated on the coast, thin interbedded to interlaminated

faces were preserved. These interbedded faces represent a period of more frequent flooding,

but this flooding is not likely related to modern multi-decadal scale glacial surging as flood beds

have a short recurrence interval (3-10 yr). The more frequent flooding could reduce sub-glacial

meltwater storage, potentially making it less likely to trigger a surge. Unlike climatic conditions

during the deposition of the upper unit, the middle unit did not have multi-decadal periods of









record is the continental shelf. A continental shelf environment proximal to a glacier with (1)

high sediment accumulation rates, (2) episodic deposition, and (3) reduced rates of bioturbation

provide an ideal location to preserve a high resolution stratigraphic record. Studies like

STRATAFORM on the northern California shelf have shown that sediment transport processes

(Wheatcroft, 2000), oceanic storms (Fan et al., 2004), fluvial floods (Wheatcroft et al., 1997;

Wheatcroft and Borgeld, 2000), and anthropogenic influences (Sommerfield and Wheatcroft,

2007) can all be recorded within the marine shelf strata.

For this study, the sediment record on the shelf adjacent to the Bering Glacier in the Gulf

of Alaska (GOA) was used to identify glacial dynamics. The Bering Glacier is known to

experience quasi-periodic abrupt increases in ice-flow velocity, known as surging. The most

recent surge event occurred from 1993 to 1995 (Molnia and Post, 1995) and caused the Bering

Glacier to advance ~9 km within several months into Vitus Lake (Molina et al., 1994). Although

the exact triggering mechanism is unknown, it has been suggested that the timing between surges

is climatically driven, and the events occur at a period of enhanced ice accumulation (Tangborn,

2002; Lingle and Fatland, 2003). Large outburst floods that transport sediment and water in the

form of buoyant surface plumes to the Gulf of Alaska (GOA) are associated with glacial surging.

Jaeger and Nittrouer (1999) were able to identify outburst flood faces and correlate each deposit

to known surges for the last century. The 20-30 yr periodicity of historic surges and shifts in the

Aleutian Low (AL) pressure system could influence the conditions favorable for producing a

positive mass balance in ice accumulation. As the AL shifts from strongest to weak/west, the

GOA experiences changes in air surface temperature and precipitation. This suggests that a 20-

30 yr period of warm and moist conditions could produce the positive mass balance needed to

trigger a surge, and Wilson et al. (2006) have identified other such multi-decadal periods of









LIST OF ABBREVIATIONS
BP Before Present

GOA Gulf of Alaska

LIA Little Ice Age

MSCL Multi-Sensor Core Logger

GRA Gamma Ray Attenuation

GPS Global Positioning System

dpm disintegration per minute
AL Aleutian Low

NPI North Pacific Index

PDO Pacific Decadal Oscillation









settling column data were then combined by normalizing the mud and sand fraction to their

relative masses to form a complete distribution from sand to clay size particles.

Age Model

A chronology for the core was developed using 210Pb radioisotopes for the trigger core and

AMS radiocarbon dates for the jumbo core. The 210Pb samples were analyzed by gamma

spectroscopy using a low intrinsic germanium detector. Samples were powered and placed into

plastic counting jars to 'age' for three weeks to capture 222Rn gas and ensure equilibrium

between 226Ra and daughters 214Pb and 214Bi (Goodbred and Kuehl, 1998). The samples were

then counted for a 24 hr period. 210Pb activities were determined from the isolated gamma rays

at 46.6 keV while 226Ra was based from the decay of daughters 214Pb (295 keV and 351 keV) and

214Bi (609 keV) assuming secular equilibrium. The 210Pb activity was used to develop an

accumulation rate for the past ~100 years (4-5 half lives) by using the first appearance approach

(Goodbred and Kuehl, 1998).

Whole articulated bivalves were collected from the jumbo core and from within the u-

channels for the AMS radiocarbon dating. After extraction they were rinsed with de-ionized

water and dried in a 500F oven for 24 hours. The samples were analyzed at the National Ocean

Science AMS (NOSAMS) facility for radiocarbon ages.

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 EORB03B4D_VZ5S5K INGEST_TIME 2011-04-22T19:29:44Z PACKAGE UFE0022345_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES



PAGE 1

1 A 700-YEAR-LONG RECORD OF GLACIAL SURGING AND ASSOCIATED FLOODING: BERING GLACIER, ALASKA By BRANDEN JAMES KRAMER 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 2008

PAGE 2

2 2008 Branden James Kramer

PAGE 3

3 ACKNOWLEDGMENTS A special thanks to m y family, Barbry, Da nny-boy, Justbry, and Kel-Kel, for all of the support they have shown. Thanks to my girl fr iend, Erin, for keeping me sane during the stress caused by writing this masterpiece. I thank my advisor, John Jaeger, and the rest of my committee (i.e. Ellen Martin and Joann Mossa) for all of the help and gui dance that they have provided me in the last few years. I would lik e to say thanks to all of my friends in the department, especially Derrick, David, and Mike. Last but not least I would also like to thank Gilly for her time spent teaching me how things worked in the lab.

PAGE 4

4 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................3 LIST OF FIGURES.........................................................................................................................6 LIST OF ABBREVIATIONS.......................................................................................................... 8 ABSTRACT.....................................................................................................................................9 CHAP TERS 1 INTRODUCTION..................................................................................................................11 2 BACKGROUND.................................................................................................................... 14 Gulf of Alaska Climate......................................................................................................... ..14 Bering Glacier.........................................................................................................................15 Preservation of Event Layers..................................................................................................16 GOA Shelf Sedimentation...................................................................................................... 17 3 METHODS.............................................................................................................................25 Core Logging Data.................................................................................................................25 X-ray Radiographs..................................................................................................................26 Grain Size Data................................................................................................................ .......26 Age Model..............................................................................................................................27 4 RESULTS...............................................................................................................................28 Chronology.............................................................................................................................28 Lithologic Description......................................................................................................... ...28 Grain Size..................................................................................................................... ..........29 5 DISCUSSION.........................................................................................................................39 Chronology.............................................................................................................................39 Lithofacies.................................................................................................................... ..........40 Upper Unit.......................................................................................................................41 Middle Unit.....................................................................................................................42 Lower Unit.......................................................................................................................43 Lithofacies Interpretations.................................................................................................... ..44 6 CONCLUSIONS.................................................................................................................... 50 LIST OF REFERENCES...............................................................................................................52

PAGE 5

5 BIOGRAPHICAL SKETCH.........................................................................................................56

PAGE 6

6 LIST OF FIGURES Figure page 2-1 The position of the Aleutian Low (AL) pressure system is one of the dominate controls on climate in the Gulf of Alaska.......................................................................... 18 2-2 Records of the Pacific Decadal Oscillat ion (PDO) index and air su rface temperatures from Cordova for the Gulf of Alaska appear to be correlated, wh ich relates a strong or eastern Aleutian Low (AL) to warm temperatures........................................................ 19 2-3 A paleo-temperature record for the Gulf of Alaska is based on tree ring data. ................. 20 2-4 A reconstruction of three termini positions of th e Bering Glacier for the past 850 yrs were derived from radiocarbon dates of buried forest.......................................................21 2-5 Map of the Bering Glacier, showing its 2001 term inus position that terminates into proglacial Vitus Lake......................................................................................................... 22 2-6 The correlation of the 1977 1988 Wolverin e Glacier positive mass balance with an eastern/strong AL depicted as negative values for the Aleutian Low Pressure Index and positive values for the Paci fic Decadal Oscillation Index.......................................... 23 2-7 The variables affecting the preservati on potential of an ev ent bed includ e the sediment accumulation rate (t), event bed thickness (Ls), and depth of bioturbation (Lb)....................................................................................................................................24 4-1 210Pb and 226Ra activity from core 81MC and 82TC and a schematic of core 82TC show decreases in 210Pb activity corr elate to low-density beds.........................................30 4-2 AMS radiocarbon ages for 82JC show a decrease of age with depth................................ 31 4-3 The shipboard GRA bulk density and vol um e magnetic susceptibility for core 82JC are illustrating the physical properties of each lithologic unit........................................... 32 4-4 X-radiograph postivies of parallel right a nd left u-channels were extracted along the center of sections from core 82JC......................................................................................33 4-5 A characteristic section of lithology with in 2 10 m section showing interbedded lowand high-density beds with some bioturbation (483 490 cm)................................ 34 4-6 The lithology below 10 m depth in core has r are laminations and is dominated by bioturbated intervals (1027 1048 cm)............................................................................. 35 4-7 The grain size distributions for lowa nd high-density beds show low-density beds have a finer signature then high-density beds. ................................................................... 36

PAGE 7

7 4-8 Grain size distribution of the three lo w-density beds from Figure 16 showing the upper 2 m of core has a finer signature than the remainder of the core............................. 37 4-9 The distribution of grain-size down core for 82JC is shown by a) relative percent m ass of sand, silt, and clay, and b) modal sand and mean sortable silt............................. 38 5-1 A non-steady state age model wa s developed with the use of 210Pb and lithologic based accumulation rates. The increased accumulation rates during 300 500 cal yr BP correspond to the timing of the LIA.............................................................................47 5-2 Sediment lithologies are correlated w ith known changes in regional clim ate................... 48 5-3 An aerial photograph shows the Malaspina Glacier, which is located eas t of the Bering Glacier at the mouth of Yakutat Bay..................................................................... 49

PAGE 8

8 LIST OF ABBREVIATIONS BP Before Present GOA Gulf of Alaska LIA Little Ice Age MSCL Multi-Sensor Core Logger GRA Gamma Ray Attenuation GPS Global Positioning System dpm disintegrations per minute AL Aleutian Low NPI North Pacific Index PDO Pacific Deca dal Oscillation

PAGE 9

9 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 A 700-YEAR-LONG RECORD OF GLACIAL SURGING AND ASSOCIATED FLOODING: BERING GLACIER ALASKA By Branden James Kramer August 2008 Chair: John M. Jaeger Major: Geology Bering Glacier, Alaska is one of the larges t glaciers in North Amer ica, and is largely known for its dramatic surging events, five of wh ich have occurred in the last century. A primary late Holocene history of the Bering has been pr eviously established from on-shore studies of glacial termini position and eviden ce of glacial advances, but the L ittle Ice Age (L IA) record of glacial surging and associated flooding has not been examined. A 14 m-long jumbo core collected on the adjacent contin ental shelf reveals a 700-year-l ong record of flood deposition. The core was dated using 210Pb chronology and five radiocarbon dates, and can be separated into three distinct lithologic units based on examina tion of x-radiographs and physical properties: 1) the uppermost unit dates from ~120 yr BP to the present and is characterized by bioturbated mud interbedded with faintly laminated, thick (5-20 cm) low-density beds, 2) the middle unit dates from ~120-500 yr BP and includes abundant la minated-to-interbedded lowand high-density beds with some evidence of bioturbation, and 3) a lowermost unit post dates 500 yr BP and is composed of rare laminated beds grading into mottled to massive mud. In each of these units, the laminated lithofacies from this oxic mid-shel f location indicates both flood and gravity flow deposition. Based on previous te rrestrial studies, from 850-400 yr BP, the terminus was at a slightly advanced position relative to the present, and it was at its Holocene Neoglacial

PAGE 10

10 maximum extent, which was close to the modern coastline, during the LIA (200-350 yr BP). The thick low-density, clay-rich beds in the uppermos t unit correlate with hi storic outburst floods associated with known surge events. During the LI A, bioturbated interval s are rare and thin while laminated intervals are common. Given average radiocarbon-supported sedimentation rates of 2-3 cm yr-1, this style of interbedding indicates frequent, if not annual, deposition at the cores location. This would suggest that melt wa ter plumes and redistribution by winter storms are more prevalent during the LIA rather than outburst flood deposits typica l of the past century. The infrequent deposition of event layers in th e lowermost unit could be attributed to the enhanced diversion of glacial drai nage to the eastern Kaliakh, Tsi u, and Tsivat Rivers instead of present day Seal River. The observation of thin ner flood beds during the LIA differs appreciably from the past century, suggesting that the Bering Glacier was not surging at a multi-decadal interval, and its position duri ng this period allowed for incr eased accumulation rates and preservation of annual melt water flood deposits.

PAGE 11

11 CHAPTER 1 INTRODUCTION The dynam ic behavior of ice and its relationshi p to global climate change have become a major environmental concern. Glaciers in par ticular are experiencing increased ice discharge and unprecedented rates of retreat (Joughlin et al., 2008; Nesje et al., 2008; Throst and Truffer, 2008). The retreat is likely climate driven (Joug hlin et al., 2008; Nesje et al., 2008; Throst and Truffer, 2008; Wiles et al., 2008), but there is a poor understanding of glac ial response to climate change over millennial time scales (Brown and Rivera, 2007; Wiles et al., 2008; Zumbhl et al., 2008). An understanding of dynamic glacial behavi or especially the time scale between melting and/or advance events will provide insight to changes observed in modern day. The study of modern glacial behavior has been enhanced w ith the establishment of new techniques for determining rates of retreat using high-precisi on GPS (Kumar et al., 2008 ), ice thickness using laser altimetry (Arendt et al., 2002; Kumar et al., 2008), and cha nges in ice mass using Gravity Recovery and Climate Experiment (GRACE) satellite s (Arendt et al., 2007). Observations at this scale allow for the production of high resolution, temporal and spatial data, but the comparison of modern data to paleo-data are needed to determine whether the modern observations reveal anomalous activity or are within th e range of previous fluctuations. The history of Holocene glaciation, in southe rn Alaska, has been constrained by glacial termini positions (Calkin et al., 2001; Wiles et al., 2008) and summer air surface reconstruction for the GOA from tree-ring data (Wilson et al., 20 06). Interpreting climatic change from the terminus position of glaciers is best done using a composite record of glacial terminus histories to reduce the affects of reworking by the most recent glacier advance (Wilson et al., 2008). The collection of multiple incomplete or fragmented records could be eliminated by the recovery of a more complete record. One such environment w ith the potential to preserve a more complete

PAGE 12

12 record is the continental shelf. A continental shelf environment proximal to a glacier with (1) high sediment accumulation rates, (2) episodic deposition, and (3) reduced rates of bioturbation provide an ideal location to preserve a high resolution stratigraphic r ecord. Studies like STRATAFORM on the northern Calif ornia shelf have shown that sediment transport processes (Wheatcroft, 2000), oceanic storms (Fan et al ., 2004), fluvial floods (W heatcroft et al., 1997; Wheatcroft and Borgeld, 2000), and anthropogenic influences (Sommerfield and Wheatcroft, 2007) can all be recorded within the marine shelf strata. For this study, the sediment record on the shelf adjacent to the Bering Glacier in the Gulf of Alaska (GOA) was used to identify glaci al dynamics. The Bering Glacier is known to experience quasi-periodic abrupt increases in ice-flow velocity, known as surging. The most recent surge event occurred from 1993 to 1995 (Molnia and Post, 1995) and caused the Bering Glacier to advance ~9 km within several months into Vitus Lake (Molina et al., 1994). Although the exact triggering mechanism is unknown, it has b een suggested that the timing between surges is climatically driven, and the events occur at a period of e nhanced ice accumulation (Tangborn, 2002; Lingle and Fatland, 2003). Larg e outburst floods that transpor t sediment and water in the form of buoyant surface plumes to the Gulf of Al aska (GOA) are associated with glacial surging. Jaeger and Nittrouer (1999) were able to identi fy outburst flood facies a nd correlate each deposit to known surges for the last century. The 20-30 yr pe riodicity of historic su rges and shifts in the Aleutian Low (AL) pressure system could infl uence the conditions favo rable for producing a positive mass balance in ice accumulation. As the AL shifts from strong/east to weak/west, the GOA experiences changes in air surface temperature and precipitation. This suggests that a 2030 yr period of warm and moist conditions c ould produce the positive ma ss balance needed to trigger a surge, and Wilson et al. (2006) have identified other such multi-decadal periods of

PAGE 13

13 warm and moist conditions during the past 1300 yrs, suggesting that surging could be a persistent process of the Bering Glacier. In this study, a record of sh elf sedimentation was examined to provide an extended record of glacial surging. I hypothesize that Bering Glac ier surging will be more frequent when positive annual mass balance (i.e., higher temperature /m oisture delivery) periods persist for several decades coupled with a lack of colder periods th at promote freezing. To test this hypothesis an age model for a 14 m-long core, EW0408 82JC, was developed using 210Pb geochronology and five radiocarbon dates. X-radiographs of the core were used to identify sediment facies down core. The physical properties of the core including gamma ray attenuation (GRA) bulk density, volume magnetic susceptibility, and grain size were also used to identify and interpret lithofacies. The interpre ted climatic conditions for the GOA fr om Wilson et al. (2006) were used in conjunction with the age model to compare th e timing of climatic events to lithologic properties to test the hypothesis.

PAGE 14

14 CHAPTER 2 BACKGROUND Gulf of Alaska Climate Clim atic changes in th e Gulf of Alaska (GOA) are primar ily controlled by the position of the Aleutian Low (AL) pressure system. The AL dominates climate in spring and winter months while during the summer months it is dominated by the North Pacific High pressure cell (Wiles et al., 1998). The AL is also known to experi ence decadal scale vari ability, which has been observed in climate indices such as the North Pa cific Index (NPI) (Trenberth and Hurrell, 1994) and the Pacific Decadal Oscillation (PDO) (Mantua et al., 1997). The NPI measures the change in sea level pressure to observe the position a nd intensity of the AL (Trenberth and Hurrell, 1994). At an eastern or strong position, the AL provides increased prec ipitation to the coast (Anderson et al., 2005) as winds bring warmer surface waters into the GOA (Figure 2-1a). During the western or weak positio n, the coast experiences drier conditions as winds are directed along shore (Figure 2-1b). Along wi th the shifts in moisture, the air surface temperature is also observed to fluctuate along with the phase shifts of the AL. For the past 80 yrs a strong phase in the AL can be associated with wa rmer air surface temperatures as the weak phase is associated with cooler temperatures (Figure 2-2). A multidecadal shift between the strong and weak phase of the PDO can also be observed for the past 80 yrs (Figure 2-2). Wilson et al. (2006) developed a 1300 yr record of coastal temp eratures for the GOA based on tree ring data, which revealed the multi-decadal variability also occurre d during ~800-950, ~1080-1100, and ~1300-1400 yr AD (Figure 2-3). One notable period when the multi-decadal variability is not present is during the little ice age (LIA; Figure 2-3) which occurred in the GOA region ~1650-1850 yr AD (Wiles et al., 1999).

PAGE 15

15 Bering Glacier The Bering Glacier is on e of the largest temper ate glaciers in North America with an area of 5,200 km2. It extends down from the Bagley Ice Field in the St Elias Mountains where its lower piedmont lobe is currently composed of the Stellar and Bering lobes and is more than 40 km wide. During the mid-Holocene however the terminus of the Bering Glacier was located 50 km or greater inland, creating a broad bay and narrow fjord where the Bering currently terminates (Mueller and Fleisher, 1995). It began to advance from this inland position around 5,000 BP building the Neoglacial foreplain observed today. Wiles et al., (1999) were able to identify three major termini positions of the Be ring Glacier during the late-Holocene (850-0 yr BP). These positions include a more advanced terminus than modern day (850-400 yr BP), its Neoglacial maximum that terminat ed at the coast (400-200 yr BP), and its modern position (2000 yr BP) (Figure 2-4). The Bering lobe is curren tly terminating into proglacial Vitus Lake which drains into the GOA via the S eal River (Figure 2-5). The Bering Glacier is known to experience quasi-periodic increase in ice flow, known as glacial surging. The first indicat ions of glacial surg ing were observed from aerial observations of folding in the medial moraines of the Beri ng Glacier (Post, 1972). Fo r many glaciers, surges appear to follow decadal-long periods of slow fl ow and thickening. Eventually, an instability threshold is reached leading to sudden acceleration of flow and rapid ice thinning. The system then returns to slow, quiescent flow (Meier and Post, 1969; Raymond, 1987). Some surging glaciers in Alaska (e.g., Bering Glacie r, Variegated Glacier) appear to follow this cycle. Eisen et al. (2001) reconstructed the annua l mass balance for the Variegated Glacier, and found that the length of time between surges was due to the va riability in ice accumulation rates prior to the surge. Also, there is evidence that Bering Gl acier surges are triggered by several consecutive years of above-normal winter balances (Ta ngborn, 2002; Lingle and Fa tland, 2003). A period of

PAGE 16

16 positive mass balance for the Wolverine Glacier has been observed to correlate with an eastern/strong phase of the AL (F igure 2-6). The periods of an eastern/strong AL for the past 80 yrs can be correlated to the five historic surges observed for the Bering Glacier (Figure 2-2). The Bering Glacier experiences two different types of floodi ng; typical summer meltwater floods and outburst floods. Associated with Beri ng Glacier surges are large outburst floods that discharge ~70% more then a typical summer seas onal ablation (Merrand and Hallet, 1996). Both types of floods transport large volumes of suspe nded sediment, most of which is deposited into Vitus Lake (Merrand and Hallet, 1996; Molnia et al., 1996). The finer grained sediment not deposited in Vitus Lake is transported to the Gu lf of Alaska in a buoyant surface plume. A study by Jaeger and Nittrouer (1999) recognized that ou tburst floods are deposite d as thick low-density beds that are internally laminated and were rapidly deposited, based on 234Th data. Jaeger and Nittrouer (1999) were able to correlate outburs t facies to a historical surge event using 210Pb geochronology on the core. In contrast, the me ltwater floods are not well preserved in the sedimentary record. Preservation of Event Layers The preservation of an outburst facies or any event bed in the sedim entary record is controlled by three factors (1) sediment accumula tion rate (t), (2) thic kness of the event layer (Ls), and (3) the thickness of the bioturbation layer (Lb) (Figure 2-7). To preserve the sedimentary fabric of an event layer in a steady state system, Ls needs to be thicker than Lb (Nittrouer and Sternber g, 1981). A thicker Ls produces rapid burial of the event layer base and allows complete preservation. High preservation is also found from the rapid accumulation of sediment (Nittrouer and Sternber g, 1981). It can be assumed th at areas experiencing episodic deposition have a larger potential to preserve sedime ntary fabric (Wheatcroft and Drake, 2003) than in a steady state system (Bentley et al., 200 6). The constant flux of sediment for a steady

PAGE 17

17 state system produces an Ls << Lb where as episodic deposition can create a situation where Ls > Lb. The preservation of episodic deposition is also related to the residence time (Tr = [Lb Ls/2]/t) or amount of time required to bury the bed below Lb (Wheatcroft and Drake, 2003). Residence time can be decreased if multiple even t beds are deposited consecutively essentially making Ls >> Lb. GOA Shelf Sedimentation The southeastern coastline of Alaska is te ctonically active as the Pacific plate subducts underneath of the North Am erican plate. Th e combination of high co astal relief and large amounts of precipitation contribute to one of th e largest annual shelf sediment accumulation rates (250 x 106 tons yr-1) in North America (Jaeger et al., 199 8). The majority of the sediment entering the GOA occurs during the late summer a nd fall months as elevated glacial melting and increased precipitation increase the amount of meltwater and runo ff (Stabeno et al., 2004). The input to the GOA of large volumes of fresh water runoff and along-shore winds produces the Alaskan Coastal Current (ACC) (Weingartner et al., 2002; Stabeno et al., 2004). The ACC flows in a western direction causing buoyant surface plumes entering the shelf to be re-directed to the west and remain inshore of the 50 m isobath (Jaeger and Nittrouer, 2006). Also inshore of the 50 m isobath, bottom wave orbital velocities are the highest. The combination of the ACC and wave orbital velocities inshor e of the 50 m isobath creates a high energy environment with minimal mud accumulation (Carlson et al., 1977). Other currents available to transport sediment across shelf are downwelling currents produced by eastern alongshore winds during the winter months (Stabeno et al., 2004). Also during th e winter, strong storms produce higher wave energies than in the summer months causing the resuspension of muddy sediment in water depths of < 100 m (Jaeger and Nittrouer, 2006).

PAGE 18

18 a b Figure 2-1. The position of the Aleutian Low (AL) pressure system is one of the dominate controls on climate in the Gulf of Alaska. A) The eastern/strong phase produces a warm and moist climate. B) The wes t/weak phase produces cooler and drier conditions.

PAGE 19

19 Figure 2-2. Records of the Pacific Decadal Osci llation (PDO) index and air surface temperatures from Cordova for the Gulf of Alaska appear to be correlated, which relates a strong or eastern Aleutian Low (AL) to warm temperat ures. The five shaded regions represent the five historical surges for the Bering Gl acier, which occur within 5 yr after a multiyear period of increased warming or strong AL. The climate record and the surges both exhibit a multi-decadal periodicity.

PAGE 20

20 Figure 2-3. A paleo-temperature record for the Gulf of Alaska is based on tree ring data. Periods of multi-decadal variability in mean summer air surface temperature are highlighted in blue. Also highlighted is the Little Ice Age (gray), which is not associated with the multi-decadal shifts in temperature. The horizontal bars represent overall mean temperature with no significan t difference for the given time period. [Adapted from Wilson, R., Wiles, G., D Arrigo, R., and Zweck, C., 2006, Cycles and shifts: 1,300 years of multi-decadal temperat ure variability in the Gulf of Alaska: Climate Dynamics, v. 28(4), p. 425-440.]

PAGE 21

21 Figure 2-4. A reconstruction of three termini pos itions of the Bering Glacier for the past 850 yrs were derived from radiocarbon dates of buried forest.

PAGE 22

22 Figure 2-5 Map of the Bering Glacier, showing its 2001 terminus position that terminates into proglacial Vitus Lake. Coloring of the te rrestrial landscape represents uncovered ice (blue), debris covered ice (red), and ice fr ee areas (green). The jumbo piston core 82JC is located ~16 km from the mouth of the Seal River in 1 50 m of water. The location of other cores colle cted during a 1994-95 cruise ar e also shown. Bathymetric data at the core location shows surface mo rphology of varying depth of the shelf and trough. We gratefully acknowledge the USGS for the Landsat image of the Bering Glacier, and Larry Mayer and the staff at the Center for Coastal and Ocean Mapping (CCOM)/ Joint Hydrographic Center (JHC ) for collecting, processing, and sharing these swath bathymetry data.

PAGE 23

23 Figure 2-6. The correlation of the 1977-1988 Wolverine Glacier pos itive mass balance (Josberger et al., 2007) with an eastern/strong AL depicted as negative values for the Aleutian Low Pressure Index and positive values for the Pacific Decadal Oscillation Index.

PAGE 24

24 Figure 2-7. The variables aff ecting the preservation potential of an event bed include the sediment accumulation rate (t), event bed thickness (Ls), and depth of bioturbation (Lb). An exponential curve shows the rate of biological mixing ( ) is the most intense at the sediment water interface a nd decreases with depth. [Adapted from Bentley, S. J., Sheremet, A., and Jaeg er, J. M., 2006, Event sedimentation, bioturbation, and preserved sedimentary fabr ic: Field and model comparisons in three contrasting marine settings: Continenta l Shelf Sedimentation, v. 26, p. 2108-2124.]

PAGE 25

25 CHAPTER 3 METHODS During the 2 004 R/V Ewing cruise, a 0.5 m-l ong multicore, 2 m-long trigger core, and 14 m-long jumbo piston core (e.g., EW0408 81MC, 82TC, and 82JC) were respectively collected ~16 km southwest from the mouth of the Seal River on the adjacent continental shelf in 150 meters of water. The collecti on of three cores allowed for comp lete recovery with minimum loss of sediment from the sediment water interface to depth down core. Duplicate Ocean Drilling Program (ODP)-style u-channels (2 x 2 x 150 cm ) were extracted in parallel from the core archived at Oregon State Universi ty. The physical properties and sedimentary structures of the core were examined using a multi-sensor core logger (MSCL), x-ray radiographs, and grain size analysis. An age model for the core was constructed using 210Pb geochronology and radiocarbon dates. Core Logging Data After collection at sea, the cores were sectioned into 150 cm lengths and immediately processed on a GeoTek Multi-Senor Core Logger (MSCL) for the following physical properties; gamma-ray attenuation (GRA) bulk density, and volume magnetic susceptibility. These data were processed at a coarse-scale resolution (1 cm). After extr action, ODP-style u-channels were re-analyzed with an MSCL, at UF, at a finerscale (0.5 cm) for GRA bulk density and at 1-cm resolution for magnetic properties. The depth scale from the shipboard MSCL data was determined by the shipboard party to represent the most accurate depth scale for all core s, so the u-channel depths were corrected to the shipboard data. Each sections depth was corrected by adding an additional 0.5 cm to the core logging data of the u-channe l to match that of the shipboar d data. Depth correction to the shipboard data allows for uniform depths am ong all cores collected during the 2004 cruise.

PAGE 26

26 X-ray Radiographs X-radiograp hs of each u-channel (~150 cm) were taken in 30 cm segments; 15 cm-long segments were scanned onto a computer and were spliced into continuous sections using Adobe Photoshop. To verify that segments were placed together properly, the co re logging data were used to match low density beds from the loggi ng data with low density beds in the x-ray. Correcting the x-rays with the core logging data also ensures that x-rays have the same depth scale as the shipboard data. The x-rays were th en inverted from negatives to positives so dense objects appear opaque and less dens e objects appear brighter. The contrast of each section was enhanced to make fainter contacts between beds more apparent. The lithologic properties of the core were then described and interpreted from th e x-radiographs. The x-radiographs were also utilized in the extraction of grain size sample s by printing out true scale versions of the xradiographs and placing them next to the actual u-channel to en sure proper sample selection. Grain Size Data Num erous grain-size samples we re collected from the entire core to provide a detailed distribution of sediment size. Th e core logging data were used to determine the relative grain size of the sediment (e.g., low-density and magnetic susceptibility = fine grain, high-density and magnetic susceptibility = coarse grain) and to pick samples based on the following properties; low-density and low-susceptibility, low-density and high-susceptibilit y, high-density and lowsusceptibility, or high-density and high-susceptibility. The core l ogging properties were used to provide a variety of different lithologic attri butes found throughout the core, but still have the ability to compare those samples. Subsamples were extracted at 1.5 cm intervals from the uchannel, homogenized, and disaggr egated so they could be wet sieved at 53 microns. The mud fraction was then analyzed on a 5100 Micrometrics Sedigraph (Coakley and Syvistski, 1991) and the sand portion with a settling co lumn (Syvitski et al., 1991). The separate sedigraph and

PAGE 27

27 settling column data were then combined by normalizing the mud and sand fraction to their relative masses to form a complete distri bution from sand to clay size particles. Age Model A chronology for the core was developed using 210Pb radioisotopes for the trigger core and AMS radiocarbon dates for the jumbo core. The 210Pb samples were analyzed by gamma spectroscopy using a low intrinsic germanium dete ctor. Samples were powered and placed into plastic counting jars to age for three weeks to capture 222Rn gas and ensure equilibrium between 226Ra and daughters 214Pb and 214Bi (Goodbred and Kuehl, 1998). The samples were then counted for a 24 hr period. 210Pb activities were determined from the isolated gamma rays at 46.6 keV while 226Ra was based from the decay of daughters 214Pb (295 keV and 351 keV) and 214Bi (609 keV) assuming secular equilibrium. The 210Pb activity was used to develop an accumulation rate for the past ~100 years (4-5 half lives) by using the first appearance approach (Goodbred and Kuehl, 1998). Whole articulated bivalves were collected from the jumbo core and from within the uchannels for the AMS radiocarbon dating. After extraction they were rinsed with de-ionized water and dried in a 50F oven for 24 hours. The samples were analyzed at the National Ocean Science AMS (NOSAMS) facility for radiocarbon ages.

PAGE 28

28 CHAPTER 4 RESULTS Chronology A chronology for the trigger core was determ ined using 210Pb radioisotope geochronology while AMS radiocarbon dates were derived from the jumbo core samples. The profile of 210Pb and 226Ra activity shows that 210Pb activity decreases from 10-1 dpm g-1 while supported levels of 226Ra remain close to 1 dpm g-1 with depth in the core (Figure 4-1). The 210Pb profile can be characterized by a non-steady state profile (Jaeger et al., 1998). This is evident as activity decreases to supported levels at 28 and 98 cm depth. Both of thes e depths are associated with low density beds, which are interpreted to repr esent flood beds (Jaeger and Nittrouer, 1998). It should be noted that the 210Pb activities have not yet reached supported levels at 148 cm depth. Based on the first appearance approach an average accumulation rate of 1.5 cm yr-1 was calculated (Goodbred and Kuehl, 1998). The AMS radiocarbon dates increas e with depth from ~900-1300 14C yr BP (Figure 4-2). Lithologic Description The core logging data show large variations in the upper 2 m of core (Figure 4-3). The GRA bulk density has a range of 0.6 g cm-3 in the upper 2 m while it typically only ranges by 0.3 g cm-3 for the remainder of the core X-radiographs show that th e strata of the upper 2 m are composed of thick (5-20 cm), faintly lamina ted low-density beds interbedded with dense bioturbated beds (Figure 4-4). The low-density beds in the x-radiographs correlate with low values in the GRA bulk density and volume magnetic susceptibil ity, and the dense bioturbated beds correlate with higher GRA bulk density and magnetic susceptibi lity values (Figure 4-4). It is also evident that the susceptibility data only show large scale variations in the core while the density data correlate well with changes observed in the x-radiogra phs (Figure 4-4). Down core

PAGE 29

29 the thickness of the low-density beds observed in x-radiographs decrease to 1-3 cm, and they become more frequent. The lithology at a depth in core of 2-10 m is characterized by interlaminated-to-interbedded lowand high-density beds with oc casional bioturbation (Figure 45). The alternating lowand high-density beds produce a smaller range in GRA bulk density (0.3 g cm-3), but show more small scale variability from 2-6 m depth in core than in the upper 2 m. Below 10 m, the laminations become rare and are separated by thick (5-20 cm) bioturbated beds that grade into mottled and massive mud beds (Figure 4-6). The lack of pronounced density changes in the x-radiographs still creates the same range of values as from 2-10 m (0.3 g cm-3), but the amount of small scale va riability is less apparent (Fig ure 4-3). Common throughout the core are low-density, low-susceptibility beds, but they are thick (5-20 cm) in the upper 2 m, but are thinner (3-10 cm) below 2 m (Figure 4-3). Grain Size Throughout the core, sequences of lowand high-density beds are observed, and these sequences can be divided in three distinct lithologic sections of the core as described in the section above. The differences in the grain size distribution of a lowa nd high-density bed have been com pared from each of the three lithologies The low-density beds observed in any of the sections were finer grained then the high-density bed (Figure 4-7) and the low-density bed in the upper 2 m of core was finer than low-de nsity beds below 2 m (Figure 4-8). The low density beds found throughout the core are clay-rich with greater than 60% clay by mass while the dense beds have lower clay percentages (< 60%) and increased silt and sand percentages (Figure 4-9a). The pe rcentage of sand for the majority of the core remains relatively low (< 10%), although at 4.5, 6.0, 6.3, 6.8, 9.1, and 11 m depth sand percentages were > 20%. The percentage of silt shows the least amount of variability (20-40%) as it parallels sand for the length of the core. The distribution of modal sand diameter shows little variation as it ranges

PAGE 30

30 within 1 for the entire length of core (Figure 4-9b). The mean sortable silt is the average size fraction that ranges from 10-63 m (4-6.5 ) and is an indicator of flow speed for the depositing current (McCave et al., 1995). The mean sortab le silt does show vari ation down core, but no distinct patterns or trends we re observed (Figure 4-9b). Figure 4-1. 210Pb and 226Ra activity from core 81MC and 82TC and a schematic of core 82TC show decreases in 210Pb activity correlate to low-density beds. Accumulation rates calculated from the 210Pb data were ~1.5 cm yr-1.

PAGE 31

31 Figure 4-2. AMS radiocarbon ages for 82JC show a decrease of age with depth.

PAGE 32

32 Figure 4-3. The shipboard GRA bulk density and volume magnetic susceptibility for core 82JC are illustrating the physical properties of each lithologic unit. Depths experiencing a low in density and susceptibility have been highlighted in gray. It should be noted that large shift in the density and susceptib ility data in the upper 2 m are not observed with depth downcore. Section breaks in the core are delinea ted by dashed lines.

PAGE 33

33 Figure 4-4. X-radiograph positives of parallel righ t and left u-channels were extracted along the center of sections from core 82JC. The x-radiographs were stretched horizontally to accentuate bedding features. Dark beds represent high density and light beds are low density. The SI units for magnetic susceptibility are times 10-5.

PAGE 34

34 Figure 4-5. A characteristic sect ion of lithology with in 2-10 m section showing interbedded lowand high-density beds with some bioturbation (483-490 cm). Ls is typically 1 3 cm thick compared to 5-20 cm in the upper 2 m, but events beds are still preserved. The magnetic susceptibility data is not re solved well enough to show the small scale variations associated with the thin alternat ing event beds although they are present in the GRA bulk density data.

PAGE 35

35 Figure 4-6. The lithology below 10 m depth in co re has rare laminations and is dominated by bioturbated intervals (1027-1048 cm). Si milar to the middle core interval, the magnetic susceptibility does not record ch ange of small scale variation in the lithology.

PAGE 36

36 Figure 4-7. The grain size distri butions for lowand high-density beds show low-density beds have a finer signature then high-density beds.

PAGE 37

37 Figure 4-8. Grain size distribution of the three low-density beds from Figure 16 showing the upper 2 m of core has a finer signature than the remainder of the core.

PAGE 38

38 Figure 4-9. The distribution of grain-size down core for 82JC is shown by a) relative percent mass of sand, silt, and clay, and b) modal sand and mean sortable silt. Clay-rich (> 60%) low-density, low-susceptibility beds are represented by the highlighted gray areas. The modal sand diameter remains rela tively constant down core, but the mean sortable silt size does show some variation although there is no discernable down core trends.

PAGE 39

39 CHAPTER 5 DISCUSSION Chronology To provide a ge constraint on the lithologic data within the core, an age model was developed from 210Pb radioisotopes, five radiocarbon date s, and interpretati ons of the lithology from x-radiographs. The 210Pb activity was used to calcula te a first appearance accumulation rate by identifying the depth at which 210Pb activities reach supported levels of 226Ra activity. When supported levels of 226Ra activity have been reached, the 210Pb activity is produced solely by the decay of 226Ra. At this depth, excess 210Pb that scavenged onto the sediment prior to deposition has decayed by 4-5 half lives, which represents ~100 yr. In the trigger core supported levels were never reached, but were trendi ng towards supported near which the deepest measurement (148 cm) was assumed to be at su pported levels. The average accumulation rate for the upper 2 m of core is ~1.5 cm yr-1. This rate is associated with episodic deposition and a non-steady state system. Episodic depos ition is evident as a decrease in 210Pb activity at 28 and 98 cm (Figure 4-1) suggesting rapid deposition (Dukat and Kuehl, 1995; Jaeger et al., 1998; Nittrouer et al., 1979). When a particle is rapidly deposited it spends less time in the water column were excess 210Pb can scavenge onto that partic le, reducing the amount of activity (Nittrouer et al., 1979). Both decreases in 210Pb activity correlate with the base of a low-density bed, which corresponds with an outburst flood (Jaeger and Nittrouer, 1999). The radiocarbon ages were calibrated with CALI B 5.0 (Stuvier and Reimer et al., 1993) for a 700 yr reservoir correction. The 700 yr reserv oir correction was derived using paired shellwood ages in the GOA (Tom Ager, pers. com.). Calibration to calendar ages of the five 14C ages produced large ranges for each respective age. These large ranges limit the age model to broad

PAGE 40

40 interpretations, but when coupled with the 210Pb data they provide an adequate chronology to develop an age model. A curve was fit to the 14C chronology by estimating ages based on accumulation rates (Figure 5-1). The ages were calculated by dividing a constant depth interval by the 210Pb or lithologic based accumulation rate. The 210Pb derived accumulation rate (1.5 cm yr-1) was used for the upper 2 m of the core while the accumulati on rates for the remainder of the core were assumed based on lithologic properties. The l ack of bioturbation and preservation of physical strata between 2-10.4 m depth indicate accumulation rates of > 2-3 cm yr-1 (Jaeger and Nittrouer, 2006). Below 10.4 m depth, bioturbate d beds become more dominant while physical strata is not well preserved suggesting that accumu lation rates decrease to ~1 cm yr-1 (Jaeger and Nittrouer, 2006). The resulting age model indicates that the upper 2 m of thicker interbedded deposits formed between 120 cal yr BP and the present. The age model also indicates that the middle interval (2-10 m) of thinner but more frequent interbeds formed from 120-500 cal yr BP, and that the lower bioturbated beds formed between 500-700 cal yr BP. Lithofacies The distinct litho logies of the core that were used to determine the accumulation rates for the age model can be used to separate the core in to three distinct lithologic units. The age ranges for each lithologic unit, determined from the ag e model, can be correl ated with three major changes in the position of the Be ring Glacier terminus (Figure 2-4) (Wiles et al., 1999). The position of the terminus likely affects the routing of glaciofluvial draina ge and ultimately the transport of sediment to the GOA. In this sect ion, each unit will be discu ssed with respect to the positions of the terminus, and how sedimentation is affected at the core site.

PAGE 41

41 Upper Unit Episodic deposition is noted in the strata of the upper 2 m by inte rbedded highand lowdensity beds (Figure 5-2d). The high-density beds are characterized by bioturbated mud that typically lack sedimentary fabric. The low-de nsity beds correspond to thick clay-rich deposits with sharp basal contacts and internal laminati ons. The preservation of the low-density beds (Ls = 5-20 cm) is a product of their thickness compar ed to the thickness of the zone of active bioturbation (Lb ~3-7 cm; Jaeger and Nittrouer, 2006). These low-density beds have been correlated to the historic surges and associated outburst floods of the Bering Glacier, and the bioturbated intervals correlate to the quiescent periods between surges (Jaeger and Nittrouer, 1999). For the past century, the Bering Glacier has te rminated into proglacial Vitus Lake (Figure 5-2b). The lake acts as a sediment trap for co arse grains only allowing silt and clay sized particles to be transported via the Seal Ri ver to the GOA (Merrand a nd Hallet, 1996). The sedimentary evidence of this are the low-density beds observed in the upper unit that have a finer grain size distribution compared to low-density beds in the lower two units (Figure 4-8), a period when Vitus Lake is not present. This increase can also be observed in th e percentage of clay by mass (> 60%) in the upper unit (F igure 4-9a). As an outburst flood propagates from the Bering Glacier, the Seal River acts to focus the flood plume out onto the shelf. The focusing of the plume along with the ACC (Stabeno et al., 2004) cau se deposition to be concentrated within the 50 m isobath. The large supply of suspended sedi ment could create a high concentration regime where bottom wave orbital velocities allow for the sediment to remain suspended at the seafloor until a critical concentration (10 g l-1) is reached producing a fluid m ud (Fan et al., 2004), and the resulting deposits produce thick low-density, clay -rich outburst flood facies (Figure 5-2d). Observation of dense internal laminations within the low-density bed suggests fluctuations in

PAGE 42

42 transport energy during the process of deposition, which can span over a few months (Jaeger and Nittrouer, 1999). After a surge, the glacier experiences a 20-30 yr-long quiescent period. During this period, annual meltwater floods produ ce plumes, but flood facies are not evident in the sediment record. The majority of sediment from annual melt water floods is probably deposited in Vitus Lake, which has recorded a maximum accumulation rate from seismic profiles of 10 m yr-1 since 1967 (Molina et al., 1996). The annua l meltwater events most likely produce a smaller Ls (< 3 cm) on the shelf, which is rework ed by bioturbation and not preserved. The bioturbated dense layers can be found overlying outburst flood faci es as organisms begin to reestablish the bioturbated layer (Lb). Middle Unit Strata at these depths in core are in terlamin ated to interbedded highand low-density beds that are separated by bioturbated beds (Figure 5-2d). The high and low-density deposits differ from the upper unit because they are thinne r (1-3 cm) and more frequent. Although Ls is reduced the preservation of these deposits suggests that there was an in crease in the overall sediment accumulation rate (3 cm yr-1, Figure 5-2c) (Bentley et al., 20 06; Jaeger and Nittrouer, 2006). This is evident in the thin or non-existent biotur bated layers. The increas e in sedimentation rates and lack of bioturbation suggest s sedimentation has changed from modern deposition, which is dominated by thick low-density beds interbedded with bioturbated beds. The middle unit was deposited around the time when the Bering Glacier reached its Neoglacial maximum during the LIA and terminated on the coastline (Figure 5-2b). From this position glacial drainage and floods would likely flow directly onto the continental shelf. When the Being Glacier was at the coast it was possi bly similar to the modern piedmont Malaspina Glacier, in southern Alaska, with numerous discha rge points to the ocean (Figure 5-3). A coastal terminus would suggest that there was not a si ngle outlet present to focus flow toward one

PAGE 43

43 portion of the shelf rather a flood would be dispersed across a larger portion of the shelf, but still directed westward by the ACC. The dispersal of the same volume of sediment across a larger area would produce a thinner Ls (1-3 cm), as observed from x-ra diographs. The large amounts of sediment associated with modern outburst fl ooding and the absence of Vitus Lake to trap sediment should produce thick flood deposits during this interval, but they are not observed in xradiographs. This could imply that the area of fl ood dispersal increased or that more frequent floods resulted in lower sediment yields per event. The absence of Vitus Lake would also allow fo r the distribution of a coarser range of grain sizes to the shelf. This is evident in a coarse r signature for the low-density beds of the middle unit compared to the upper unit (Fig ure 4-8). It is also apparent in the increase of sand beds observed from x-radiographs, but the modal sand size of these sand beds is similar to other sand beds found throughout the core (Figure 4-9b). The range of modal sand varies between 3.1-3.8 and implies that similar shear stresses were avai lable during transport to the core location. The increase in the number of sand beds could be a result of increased storm activity causing the frequency of sand deposition to increase. Another possibility could be related to the volume of sediment the Bering is delivering onto the shelf while positioned on the coastline. This would supply a large amount of sand, which can then be transported across shelf by normal winter storms. With more sediment available for transport, Ls would increase allowing for the preservation of more sand beds. The sand beds are less common in the upper and lower units, indicating that these are periods of time when large volumes of sa nd were not being deposited at the core site. Lower Unit The lam inated deposits in this unit become le ss abundant and are inte rbedded with thick ( 20 cm) beds of mottled mud (Figure 5-2d). A decrease in the preservation of physical

PAGE 44

44 sedimentary fabric suggests a decrease in Ls (< 3 cm) to less than Lb or in sediment accumulation rate. A thicker Lb would allow for episodic deposition to still occur, but not be preserved because of lower accumulation rates (Bentley et al., 2006). Coincident with the deposition of the lower unit, the Bering Glacier terminus was at a more advanced position than modern day, but not on the coast (Figure 5-2b). The position of this terminus caused glacial drainage to be di verted to a more eastern position as the Kaliakh, Tsivat, and Tsiu Rivers built an outwash apron between 1300-200 yr BP (Muller and Fleisher, 1995). Similar to the middle unit, the dispersal of a constant volume of sediment over a larger area, results in a deceased Ls. The decrease in Ls is evident in the x-radiographs as the number of preserved event layers have decreased and the pr eservation of biological traces has increased. The first 50 cm of the lower unit consist of lowdensity laminations interbedded with thick (> 15 cm) bioturbated beds, which suggests episodic deposition (Jaeger and Nittrouer, 2006). This type of low-density deposit is similar to de posits observed in the middle unit having a coarser grain size distribution, but they are less common. Below 11 m, the core is composed of a mottled to massive mud facies indicating co mplete mixing from bioturbation and lower accumulation rates (< 1.0 cm yr-1, Figure 18c) (Jaeger and Nittrouer, 2006). Lithofacies Interpretations The three units of the core have notable diffe rences in thickness of the low-density event beds and bioturbated intervals. The change in thickness observed down core is most likely caused by the dispersal pathways of glacial draina ge in relation to the position of the glacial terminus, as described above. An understanding of glacial behavi or with respect to temporal climatic change can also be uti lized to interpret the deposits obser ved within the core. This can be done by calculating a recurrence interval for the low-density event beds from each unit. The recurrence interval represents the amount of time that has elapsed between the deposition of low-

PAGE 45

45 density event beds, and was calcu lated by dividing an average thic kness of bioturbated strata between low-density event beds by an assumed a ccumulation rate. In the upper unit, the average thickness of bioturbated strata was 20 cm which are associated with an average accumulation rate of 1.0 cm yr-1, which corresponds to a recurrence inte rval of 20 yrs. The middle unit has thinner bioturbated intervals (5-15 cm) and higher sediment accumulation rate (1.5 cm yr-1) giving it a recurrence interval of 3-10 yrs. The difference in sediment accumulation rates between the upper and middle unit was determined based on the amount of preserved physical strata observed in the bioturbated intervals. The upper unit displayed mo re prevalent biological structures (e.g., burrows) with little to no pres ervation of physical stra ta while the middle unit exhibits physical strata that we re only partially destroyed. The 20-30 yr recurrence interval corresponds to the quiescent period before the Bering Glacier surges. During the quiescent period volumes of water and sediment are stored underneath the glacier and if e nough water is stored over time could produce the buoyant force that is thought to trigger a surge event. This excess water is possibly produced during the multidecadal periods of increased temperature and mo isture associated with a strong AL (Figure 52a). The excess water and the positive mass ba lance created during a strong AL could produce the instability required to allow a surge to occur. The stored water and sediment is then released during and after the surge in an outburst flood. The large amounts of sediment and the focusing of the outburst flood by Vitus Lake and the Seal River combine to help create the thick, lowdensity deposits observed on the shelf. It is evident from the lithology a nd grain size of the core that glacier surging, as manifested by thick, low-density beds, is not as pronounced in the lower two un its. The low-density beds in the middle unit are thinner but more frequent wh ile bioturbated interval s were less abundant.

PAGE 46

46 The observation of more frequent low-density beds suggests that the glacier was flooding more frequently, which would continually flush water and sediment from under neath the glacier. Not allowing large volumes of water and sediment to build underneath the glacier would produce smaller magnitude floods. This combined with the dispersal of sediment discharge over a larger area would create a thinner deposit as observed. The recurrence interval between interbedded to interlaminated lowand high-density sequences was 3-10 yrs. This recurrence interval is not as prolonged as in the upper unit and may be related to the shorter time periods between shifts from warm to cold conditions during the LIA (1850-1650 AD, Figure 5-2c ). The bioturbated intervals could represent a colder drier period when meltw ater flooding was at a minimum, resulting in small scale events not preserved in the strata. The interbedded low-dens ity sequence on the other hand would be produced during a warmer, moist period when the glacier was actively flooding subglacial water was not being trapped underneath the glacier as observed in modern conditions. This would suggest that during this time pe riod the Bering Glacier was not surging on the modern multi-decadal time scale. Similar to the middle unit, the lower unit repr esents a period of time with multi-year long periods between shifts from warm to cool temper atures, but the shifts were not as dramatic as during the LIA (1700-1450 AD, Figure 5-2a). This would have produced less intense precipitation during warmer periods compared to the LIA. Overall the lower unit was deposited when the average climate was cooler and mo st likely drier causing smaller magnitude flood events. The combination of a cooler climate and a lack of preserved event imply that this period of time was not favorable for glacial surging.

PAGE 47

47 Figure 5-1. A non-steady state age mode l was developed with the use of 210Pb and lithologic based accumulation rates. The increased accumulation rates during 300-500 cal yr BP correspond to the timing of the LIA.

PAGE 48

48 Figure 5-2. Sediment lithologies are correlated with known changes in regional climate. The thre e lithologic units have been separated by distinct lithologic properties, but can also be described in terms of A) climate and B) glacial terminus position. The change in air surface temperature for the last 1300 yrs was derived from tree ring data collected from the Gulf of Alaska (Wilson et al., 2006). Three terminus positions for the Bering Glacier during the past 850 yr are shown youngest to oldest from top to bottom (Wile s et al., 1999). Also shown in this fi gure are C) core logging and accumulation rate data and D) x-radiographs of characteristic lithologies from each unit.

PAGE 49

49 Figure 5-3. An aerial photograph shows the Malaspina Glacier, which is located east of the Bering Glacier at the mouth of Yakutat Ba y. A modern analog for the terminus position of the Bering Glacier during the LIA, the Malaspina has seven main tributaries draining into the GOA.

PAGE 50

50 CHAPTER 6 CONCLUSIONS The results of this study reveal that over the past 500 years changes in sedim entation on the continental shelf seaward of the Bering Glacier correlate with historical records of surging events, outburst floods, and with previously established paleo-temper ature record based on coastal tree-ring data. These corr elations suggest that large infr equent outburst floods typical of the past century are a result of multi-decadal peri ods of higher precipita tion coupled with higher mean summer temperatures. This study also reveals that the Bering Glacier terminus position influences the delivery of sediment to the adjacent continental shelf. Modern day sedimentation consisting of th ick interbedded mottled and laminated mud beds is present in the upper 2 m of core (0120 yr BP), which was depos ited during a period of time when the Bering Glacier terminates in proglaci al Vitus Lake. The lithofacies that dominate this unit are thick, laminated out burst flood facies, which correlate with glacial surge events. These low-density beds represen ting outburst flood facies have a ~20 yr recurrence interval, which correlates with the prolonged multi-decadal warm trends (~7C) in mean summer air surface temperature for the past century. During these prolonged warmer climate periods excess meltwater could be stored sub-gl acially, potentially creating the buoya nt force needed to trigger a surge. Between 2-10 m depth in core (120-500 yr BP), a period when other investigators have determined that the Bering Glacier terminated on the coast, thin interbedded to interlaminated facies were preserved. These interbedded facies represent a period of more frequent flooding, but this flooding is not likely related to modern multi-decadal scale glacial surging as flood beds have a short recurrence interval (3-10 yr). The more frequent flooding could reduce sub-glacial meltwater storage, potentially making it less likely to trigger a surg e. Unlike climatic conditions during the deposition of the upper unit, the middle unit did not have multi-decadal periods of

PAGE 51

51 consistently warmer temperature. Instead, extr eme warm (>7C) to cold (< 4C) temperature changes are observed to occur on less then decada l time scales causing an overall constant mean air surface temperature (1850-1700 yr AD). This would allow for more frequent flooding during years with warmer temperatures while years with colder temperatures would be associated with less discharge and lower sediment accumulation ra tes. From 10-13.6 m depth in core (> 500 yr BP), the Bering Glacier terminated at a more a dvance position than modern day as established by previous investigators. The l ack of preserved physical stratigraphy observed in this unit suggests bioturbation coupled with low sediment accumulation rates was the dominant condition and correlates to an overall cooler a nd possibly drier climate as esta blished by tree-ring records. Under these drier conditions, me ltwater production would be lowe r, and the conditions favoring Bering Glacier surging were most likely not present. This study indicates that the behavior of th e Bering Glacier has va ried during the late Holocene. The record however did not extend b ack long enough to encompass another period of multi-decadal variability (1300-1400 yr AD) in climate to accurately test the hypothesis. Because of this reason it was difficult to make any interpretation or comparison on the influence of the multi-decadal shifts in the Aleutian Low and their affect on the behavior of the Bering Glacier. The collection of a deeper core at th is site would allow for multiple periods of multidecadal shifts in the climate to be compared. The ability to compare these conditions could provide greater insigh t into the potential influence of climate on glacial surging.

PAGE 52

52 LIST OF REFERENCES Anderson, L., Abbott, M. B., Finney, B. P., and Burns, S. J., 2005, Regional atm ospheric circulation change in the North Pacific dur ing the Holocene inferred from lacustrine carbonate oxygen isotopes, Yukon Territory, Ca nada: Quaternary Research, v. 64, p. 2135. Arendt, A. A., Echelmeyer, K. A., Harrison, W. D., Lingle, C. S., and Valentine, V. B., 2002, Rapid wastage of Alaska Glaciers and their c ontribution to rising sea level: Science, v. 287, p. 382-386. Arendt, A., Luthake, S., Abdelati, W., Larsen, C., Lingle, C., Echelmeyer, K., Rowlands, D., and Krabill, W., 2007, Estimates of glacier mass cha nge in the St. ~Elias Mountians of Alaska, USA and Yukon Territory, Canada: a strategy for combining GRACE and aircraft laser altimetry data: Eos Trans. AGU, v. 88(52) Fall Meet. Suppl., Abstract, C12A-07. Bentley, S. J., Sheremet, A., and Jaeger, J. M., 2006, Event sedimentation, bioturbation, and preserved sedimentary fabric: Field and model comparisons in three contrasting marine settings: Continental Shel f Sedimentation, v. 26, p. 2108-2124. Brown, F., and Rivera, A., 2007, Climate changes and recent glacier behaviour in the Chilean Lake District: Global and Pl anetary Change, v. 59, p. 79-86. Calkin, P. E., Wiles, G. C., and Barclay, D. J., 2001, Holocene coastal glaciation of Alaska: Quaternary Science Reviews, v. 20, p. 449-461. Carlson, P. R., Molina, B. F., Kittlson, S. C., and Hampson, J. C., 1977, Distribution of bottom sediments on the continental shelf, north ern Gulf of Alaska, Scale 1:500,000: US Geological Survey Misc. Field Study Map MF-876, p. 1-13. Coakley, J. P., and Syvitski, J. P. M., Syvits ki, J. P. M., (ed) 1991, Sedigraph Technique, Methods and Applications of Particle Size Analysis : Cambridge University Press, New York, p. 129-142. Dukat, D. A., and Kuehl, S. A., 1995, Non steady state 210Pb flux and the use of 228Ra/226Ra as a geochronometer on the Amazon continenta l shelf: Marine Geology, v. 125, p. 329-350. Eisen, O., Harrison, W. D., and Raymond, C. F., 2001, The surges of Variegated Glacier, Alaska, USA, and their connection to climate and ma ss balance: Journal Of Glaciology, v. 47(158), p. 351-358. Fan, S., Swift, D., Traykovski, P., Bentley, S ., Borgeld, J., Reed, C., and Niedoroda, A., 2004, River flooding, storm resuspension, and event st ratigraphy on the northern California shelf: observations compared with simulations: Marine Geology, v. 210, p. 17-41. Goodbred, S. L., and Kuehl, S. A., 1998, Floodplai n processes in the Bengal Basin and the storage of Ganges-Brahmaputra rive r sediment: an accretion study using 137C and 210Pb geochronology: Sedimentary Geology, v. 121, p. 239-258.

PAGE 53

53 Jaeger, J. M., and Nittrouer, C. A., 1999, Marine record of surge-induced outburst floods from the Bering Glacier, Alas ka: Geology, v. 27, p. 847-850. Jaeger, J. M., and Nittrouer, C. A., 2006, A qua ntitative examination of modern sedimentary lithofacies formation on the glacially influen ced Gulf of Alaska continental shelf: Continental Shelf Research, v. 26, p. 2187-2204. Jaeger, J. M., Nittrouer, C. A., Scott, N. D ., and Milliman, J. D., 1998, Sediment accumulation along a glacially impacted mountainous coas tline: north-east Gulf of Alaska: Basin Research, v. 10, p. 155-173. Joughin, I., Howat, I., Alley, R. B., Ekstrom, G. Fahnestock, M., Moon, T., Nettles, M., Truffer, M., and Tsai, V. C., 2008, Ice-front variati ons and tidewater beha vior on Helheim and Kangerdlugssuaq Glaciers, Greenland: J ournal of Geophysical Research, v. 113, p. F01004. Josberger, E.G., Bidlake, W.R., March, R. S., and Kennedy, B.W., 2007, Glacier Mass-Balance Fluctuations in the Pacific Northwest and Alaska, USA: Annals of Glaciology, v. 46, p. 291-296. Kumar, K., Dumka, R. K., Miral, M. S., Satyal, G. S., and Pant, M., 2008, Estimations of retreat rate of Gangotri glacier using rapid static and kinematic GPS survey: Current Science, v. 94, p. 258-262. Lingle, C. S., and Fatland, D. R., 2003, Does englacial water storage dr ive temperate glacier surges?: Annals Of Glaciology, v. 36, p.14-20. Mantua, N. J., Hare, S. R., Zhang, Y., Wallace, J. M., and Francis, R. C., 1997, A Pacific interdecadal climate oscilla tion with impacts on salmon production: Bulletin of the American Meteoroical Society, v. 78, p. 1069-1079. McCave, N., Manighetti, B., and Robinson, S. G., 1995, Sortable silt and fine sediment size/composition slicing: Parameters for pa laeocurrent speed and palaeoceanography: Paleoceanography, v. 10 (3), p. 593-610. Meier, M. F., and Post, A., 1969, What Are Glacie r Surges: Canadian Journal Of Earth Sciences, v. 6(4P2), p. 807. Merrand, Y., and Hallet, B., 1996, Wate r and sediment discharge from a large surging glacier: Bering Glacier, Alaska, USA, summer 1994: Annals of Glaciology, v. 22, p. 233-40. Molnia, B. F., and Post, A., 1995, Holocene histor y of Bering Glacier, Alaska: A prelude to the 1993-1994 surge: Physical Geography, v. 16, p. 87-117. Molnia, B. F., Post, A., and Carlson, P. R., 1996, 20th-century glacial-marine sedimentation in Vitus Lake, Bering Glacier, Alaska, U. S.A.: Annals of Glaciology, v. 22, p. 205-210.

PAGE 54

54 Molina, B. F., Post, A., Trabant, D. C., and Krimmel, R. M., 1994, The 1993-1994 surge of the Bering Glacier, Alaska: An overview: Eos (Transactions, American Geophysical Union), v. 75, p. F62. Muller, E. H., and Fleisher, P. J., 1995, Surging history and potential for renewed retreat: Bering Glacier, Alaska, U.S.A.: Arctic and Alpine Research, v. 27, p. 81-88. Mulder, T., and Syvitski, J. P. M., 1995, Turbid ity currents generated at river mouths during exceptional discharges to the world oceans: Journal of Geology, v. 103, p. 285-299. Nesje, A., Bakke, J., Dahl, S. O., Lie, ., and Matthews, J. A., 2008, Norwegian moutian glaciers in the past, presen t, and future: Global and Planetary Change, v. 60, p. 10-27. Nittrouer, C.A., and Sternberg, R.W., 1981, The formation of sedimentary strata in an allochthonous shelf environment: The Washingt on Continental Shelf: Marine Geology, v. 42, p. 210. Nittrouer, C. A., Sternberg, R. W., Carpente r, R., and Bennett, J. T., 1979, The use of 210Pb geochronology as a sedimentological tool, application to the Washington continental shelf: Marine Geology, v. 31, p. 297-316. Post, A., 1972, Periodic surge orig in of folded medial morain es on Bering Piedmont Glacier, Alaska: Journal of Glaciology, v. 11, p. 219-226. Raymond, C. F., 1987., How Do Glaciers Surge A Review: Journal Of Geophysical ResearchSolid Earth And Planets, v. 92(B9), p. 9121-9134. Rodionov, S., Bering Climate and Ecosys tems: NOAA, July, 14 2006, April, 18 2008, http://www.beringclimate.noaa.gov/data/BCresult.php. Sommerfield, C. K., and Wheatcroft, R. A., 2007, Late Holocene sediment accumulation on the northern California shelf: ocean, fluvial, and anthropogenic influences: GSA Bulletin, v. 119, p. 1120-1134. Stabeno, P.J., Bond, N.A., Hermann, A.J., Kach el, N.B., Mordy, C.W., and Overland, J.E., 2004, Meteorology and oceanography of the norther n Gulf of Alaska: Continental Shelf Research, v. 24 (7), p. 859. Stuiver, M., and Reimer, P. J., 1993, Extended 14C database and revised CALIB radiocarbon calibration program: Radiocarbon 35:215-230. Syvitski, J. P. M., Asprey, K. W., and Clatte nburg, D. A., Syvitski, J. P. M., (ed) 1991, Principles, design, and calibration of settling columns, Methods and Applications of Particle Size Analysis : Cambridge University Press, New York, p. 45-63. Tangborn, W., 2002, Connecting Winter Balance and Runoff to Surges of the Bering Glacier, Alaska: http://www.hymet.com/docs/yakuart4.pdf, p. 1-17.

PAGE 55

55 Throst, D. E., and Truffer, M., 2008, Glacier re cession on Heard Island, southern Indian Ocean: Artic, Antarctic, and Alpine Research, v. 40(1), p. 199-214. Trenberth, K. E., and Hurrell, J. W., 1994, Decadal atmosphere-ocean variations in the Pacific: Climate Dynamics, v. 9, p. 303-319. Weingartner, T.J., Coyle, K., Finney, B., H opcroft, R., Whitledge, T., Brodeur, R., Dagg, M., Farley, E., Haidvogel, D., Hal dorson, L., Hermann, A., Hinckley, S., Napp, J., Stabeno, P., Kline, T., Lee, C., Lessard, E., Royer, T., and Strom, S., 2002, The Northeast Pacific Globec Program: Coastal Gulf of Alaska: Oceanography, v. 15 (2), p. 1035. Wheatcroft, R. A., 2000, Oceanic flood sedimentation: a new perspective: Continental Shelf Research, v. 20 (16), p. 2059-2066. Wheatcroft, R. A., and Borgeld, J. C., 2000, Oceanic flood deposits on the northern California shelf: large scale distribution and small-s cale physical properties: Continental Shelf Research, v. 20, p. 2163-2190. Wheatcroft, R.A., and Drake, D.E., 2003, Post-d epositional alteration and preservation of sedimentary event layers on continental margin s, I. The role of episodic sedimentation: Marine Geology, v. 199 (1), p. 123. Wheatcroft, R. A., Sommerfield, C. K., Drake, D. E., Borgeld, J. C., and Nittrouer, C. A., 1997, Rapid and widespread dispersal of flood se diment on the northern California margin: Geology, v. 25, p. 163-166. Wiles, G. C., Barclay, D. J., Calkin, P. E., and Lowell, T. V., 2008, Century to millennial-scale temperature variations for th e last two thousand years indi cated from glacial geologic records of Southern Alaska: Globa l and Planetary Change, v. 60, p. 115-125. Wiles, G. C., DArrigo, R. D., and Jacoby, G. C., 1998, Gulf of Alaska atmosphere-ocean variability over recent centuries inferred from coastal tree-ring records: Climate Change, v. 38, p. 289-306. Wiles, G. C., Post, A., Muller, E. H., a nd Molnia, B. F., 1999, Dendrochronology and late Holocene history of Bering Piedmont Glacier Alaska: Quaternary Research, v. 52, p. 185195. Wilson, R., Wiles, G., DArrigo, R., and Zwec k, C., 2006, Cycles and shifts: 1,300 years of multi-decadal temperature variability in the Gulf of Alaska: Climate Dynamics, v. 28(4), p. 425-440. Zumbhl, H. J., Steiner, D., and Nussbaumer, S. U., 2008, 19th century glacier representations and fluctuations in the central and western Eu ropean Alps: An interdisciplinary approach: Global and Planetary Change, v. 60, p. 42-57.

PAGE 56

56 BIOGRAPHICAL SKETCH Branden Kram er grew up in the little town of Highland Heights, KY, and aspired to be a marine biologist so he could be the best dolphin trainer in the world. To obtain his dream, he consistently watched the Discovery channel to soak in the wonderful opportunities the ocean had to offer. Once he graduated high school, he enro lled into the Marine Sc ience program at Coastal Carolina University. While at Coastal, Branden r ealized that the job mark et for dolphin trainers was few and far between, so he decided to search for other opportunitie s within the marine realm. After a few weeks of searching, he stum bled upon a marine geologist by the name of Eric Wright that needed help collecting samples one summer. As summer ended, Branden acquired an interest for the geosciences. Eric allowed him to continue worki ng on various projects and eventually led to a senior project where Bra nden used ground penetra ting radar (GPR) to study the stratigraphy and evolution of a barrier island system. Branden had finally found his calling as a geologist. Knowing that he would want to continue his education, he applied to graduate schools and was accepted to the University of Florid a. During his three years at UF he studied glacially influenced, continental shelf deposits in the Gulf of Alaska. He received his masters degree in August of 2008 and has moved ont o the next step of his life.