Download Integrating satellite observations and modern climate measurements

JOURNAL OF GEOPHYSICAL RESEARCH: OCEANS, VOL. 118, 3444–3461, doi:10.1002/jgrc.20243, 2013
Integrating satellite observations and modern climate measurements
with the recent sedimentary record: An example from Southeast
Alaska
Jason A. Addison,1 Bruce P. Finney,2 John M. Jaeger,3 Joseph S. Stoner,4 Richard D. Norris,5
and Alexandra Hangsterfer5
Received 9 January 2013; revised 9 May 2013; accepted 14 May 2013; published 17 July 2013.
[1] Assessments of climate change over time scales that exceed the last 100 years require
robust integration of high-quality instrument records with high-resolution paleoclimate
proxy data. In this study, we show that the recent biogenic sediments accumulating in two
temperate ice-free fjords in Southeast Alaska preserve evidence of North Pacific Ocean
climate variability as recorded by both instrument networks and satellite observations.
Multicore samples EW0408-32MC and EW0408-43MC were investigated with 137Cs and
excess 210Pb geochronometry, three-dimensional computed tomography, high-resolution
scanning XRF geochemistry, and organic stable isotope analyses. EW0408-32MC
(57.162 N, 135.357 W, 146 m depth) is a moderately bioturbated continuous record that
spans AD 1930–2004. EW0408-43MC (56.965 N, 135.268 W, 91 m depth) is composed
of laminated diatom oozes, a turbidite, and a hypopycnal plume (river flood) deposit. A
discontinuous event-based varve chronology indicates 43MC spans AD 1940–1981.
Decadal-scale fluctuations in sedimentary Br/Cl ratios accurately reflect changes in marine
organic matter accumulation that display the same temporal pattern as that of the Pacific
Decadal Oscillation. An estimated Sitka summer productivity parameter calibrated using
SeaWiFS satellite observations support these relationships. The correlation of North Pacific
climate regime states, primary productivity, and sediment geochemistry indicate the
accumulation of biogenic sediment in Southeast Alaska temperate fjords can be used as a
sensitive recorder of past productivity variability, and by inference, past climate conditions
in the high-latitude Gulf of Alaska.
Citation: Addison, J. A., B. P. Finney, J. M. Jaeger, J. S. Stoner, R. D. Norris, and A. Hangsterfer (2013), Integrating satellite
observations and modern climate measurements with the recent sedimentary record: An example from Southeast Alaska, J. Geophys.
Res. Oceans, 118, 3444–3461, doi:10.1002/jgrc.20243.
1.
Introduction
[2] High-latitude regions of the Northern Hemisphere
are currently experiencing rapid environmental change as a
result of combined forcing by natural and anthropogenic
Additional supporting information may be found in the online version of
this article.
1
Climate and Land Use Change R&D Program, United States Geological Survey, Menlo Park, California, USA.
2
Departments of Biological and Geological Sciences, Idaho State University, Pocatello, Idaho, USA.
3
Department of Geological Sciences, University of Florida, Gainesville,
Florida, USA.
4
College of Ocean and Atmospheric Sciences, Oregon State University,
Corvallis, Oregon, USA.
5
Scripps Institution of Oceanography, University of California, San
Diego, California, USA.
Corresponding author: J. A. Addison, United States Geological Survey,
345 Middlefield Road, Mail Stop 910, Menlo Park, CA 94025, USA.
([email protected])
©2013. American Geophysical Union. All Rights Reserved.
2169-9275/13/10.1002/jgrc.20243
processes. In many regions of Alaska, observations include
the rapid retreat of formerly advancing glaciers in both alpine and coastal settings [Molnia, 2007], thawing permafrost [Romanovsky et al., 2007], and widespread shifts in
terrestrial [De Valpine and Harte, 2001] and marine [Mueter et al., 2009] biogeographic distributions. From a historical perspective, many of these changes are unprecedented
and their future impacts are largely unknown due to relatively short instrumental records that span only the last 100
years. High-quality paleoclimate data sets are, therefore,
essential for reconstructing earlier climate analogues to
provide further insight into future conditions, yet the spatial
coverage of such records is sparse.
[3] Long-term instrument observations are rare in the
Gulf of Alaska sector of the Subarctic North Pacific Ocean,
yet conditions here are an important component of Northern Hemisphere climate patterns with atmospheric teleconnections reaching as far east as the U.S. Atlantic Coast
[Trenberth and Hurrell, 1994] and as far south as the Sonora Desert in northern Mexico [Latif and Barnett, 1994].
Gulf of Alaska precipitation, sea surface temperatures
(SST), and fluvial discharge also show strong correlations
3444
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 1. SeaWiFS-derived net primary productivity measurements [Westberry et al., 2008] in the
Gulf of Alaska during June 2007 (a), with black box indicating area over which values were averaged to
generate the productivity time series in Figure 1g. Locations of the PFEL upwelling index site and the
NOAA NOMAD buoys are indicated. Inset map (b) depicts location of cores EW0408-32MC and
EW0408-43MC in Katlian Bay and Deep Inlet, respectively, along with locations of USGS stream gage
stations and epicenter of M7.6 Sitka Earthquake.
to the Pacific Inter-Decadal Oscillation (PDO) index
[Ebbesmeyer et al., 1991; Mantua et al., 1997; Trenberth,
1990]. There is strong evidence that the PDO strongly regulates the Gulf of Alaska marine ecosystem at multiple
trophic levels [Beamish, 1993; Hollowed et al., 2001; King
et al., 2000], likely through bottom-up environmental forcing [Gargett, 1997]. This decadal-scale North Pacific climate variability has further been extended to historic and
ancient Pacific sockeye salmon abundances [Beamish and
Bouillon, 1993; Finney et al., 2000]. Understanding highfrequency climate shifts in the context of high-latitude
change is therefore critical to predicting future impacts on
Gulf of Alaska resources and requires long paleoclimate
data sets to supplement existing instrument networks.
[4] In this paper, the last 100 years of marine sediment
from two coastal temperate (ice-free) fjords in Southeast
Alaska are compared to modern climate observations to
determine if these fjords are high-fidelity recorders of regional Gulf of Alaska climate dynamics. Fjord basins can
preserve exceptional records of marine productivity [Gilbert, 2000], and as primary productivity appears to be
forced by environmental conditions in the Gulf of Alaska
[Gargett, 1997], temperate fjord sediment records thus
offer a rare opportunity to reconstruct past high-latitude climate conditions if they are indeed sensitive to regional climate variability. To test this hypothesis, we examine two
multicore-type sediment cores from Deep Inlet (56.965 N,
135.268 W, 91 m water depth) and Katlian Bay (57.162 N,
135.357 W, 146 m water depth), a pair of temperate fjords
(Figure 1). Sediment accumulation rates in both fjords are
very high (>250 cm/kyr; Addison [2009]), and offer an opportunity to investigate past Gulf of Alaska oceanography
at exceedingly high temporal resolution. However, the
preservation of highly resolved paleoclimate information is
dependent upon the nature of sedimentation and postdepositional preservation. To understand the quality of the
records preserved in Deep Inlet and Katlian Bay, we
employ computed tomography (CT) scanning, scanning
XRF geochemistry, linescan color imaging, and gammaray attenuated (GRA) bulk density analyses to describe the
most recent 100 years of accumulation. These results are
then related to the modern cycle of sedimentation at work
along the southeastern Alaska coast.
2.
Methods
2.1. Site Overview
[5] Deep Inlet and Katlian Bay are two temperate icefree fjords located on Baranof Island in the Alexander Archipelago of southeastern Alaska (Figure 1). Deep Inlet
was carved from the local Cretaceous Sitka Graywacke
bedrock [Loney et al., 1975] by the extinct Cordilleran Ice
Sheet [Kaufman and Manley, 2004], while Katlian Bay is
located within the meta-sedimentary Kelp Bay Group
3445
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 2. (a) Sketch of ‘‘reverse-estuary’’ circulation mode in temperate ice-free fjords, (b) temperature, (c) salinity, and (d) potential density data collected in August 2004. Offshore Baranof Island hydrocast taken at 56.8 N, 136.3 W, 1784 m water depth. Hydrocast plotted with Ocean Data View (Schlitzer,
2013, Ocean Data View, http://odv.awi.de).
[Loney et al., 1975]. Both fjords are connected to the Gulf
of Alaska through Sitka Sound, a larger temperate fjord
with a sill (likely a submerged Pleistocene terminal moraine) at approximately 130 m water depth. There are no
modern tidewater glaciers along the coast of Baranof
Island, nor are there any alpine glaciers in the high relief
regions surrounding these fjords. Deep Inlet is 6 km long,
0.75 km wide, has a maximum depth of 91 m and is confined by a shallow sill at its mouth that is only 24 m deep.
Several small unnamed streams drain directly into Deep
Inlet, with the largest two forming a delta complex at the
head of the fjord that have a combined drainage area of
approximately 25 km2. Katlian Bay is larger than Deep
Inlet and is 7.5 km long, 1.5 km wide, and approximately
143 m deep. The sill at the mouth of Katlian Bay is also
deeper than at Deep Inlet, at a water depth of 47 m, with a
deep scour pool lying just behind the sill. Katlian Bay also
has a much larger delta at its head, generated by the confluence of four rivers that drain approximately 148 km2
[Haight et al., 2006].
[6] Temperature and salinity hydrocast data from Deep
Inlet and Katlian Bay show a three-layer water column
(Figure 2), which is characteristic of the ‘‘reverse estuary’’
circulation pattern common to temperate fjords [Syvitski et
al., 1987]. Both fjords contain brackish surface mixed
layers derived from fluvial discharge and surface runoff,
which overlie intermediate zones that are freely exchangeable with waters from the open Gulf of Alaska (Figure 2d).
The bottom waters below the sills in southeastern Alaskan
ice-free fjords tend to exhibit dissolved oxygen depletion
(either seasonally or permanently), depending on the rate
of oxygen-renewal events [Burrell, 1989]. These dysoxicto-anoxic bottom waters limit sedimentary bioturbation and
enhance organic matter preservation [Skei, 1983].
3446
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
2.2. Modern Climate Observations
[7] A collection of regional instrument-derived data sets
is used to assess modern climate parameters and marine
primary productivity, as neither Katlian Bay nor Deep Inlet
contain sediment traps or instrumented moorings. The location and intensity of the Aleutian Low atmospheric pressure cell is described by the North Pacific Index, which is
the monthly area-weighted sea level pressure (SLP) over
the region 30–65 N latitude and 160 E to 140 W longitude
[Trenberth and Hurrell, 1994] with coverage beginning in
AD 1899. The Aleutian Low is of special significance in
this study, as it has been identified as one of the major drivers of North Pacific regional environmental change [Trenberth, 1990]. The coastal upwelling index for 57 N, 137 W
from the Pacific Fisheries Environmental Laboratory
(PFEL; [Schwing et al., 1996]) has been calculated for AD
1946–2010 and is based on estimates of Ekman transport
driven by regional geostrophic wind stresses. Monthly
mean surface air temperature (SAT) and monthly total precipitation observations for Sitka, Alaska (AD 1948–2010 at
the Rocky Gutierrez Airport; AD 1922–1989 at the Sitka
Magnetic Observatory) are from the NOAA National Climate Data Center. SST are from two different sources, a
pair of instrumented NOMAD buoys at Cape Edgecumbe
(56.612 N, 136.065 W) and the Fairweather Grounds
(58.237 N, 137.986 W) maintained by the National Data
Buoy Center with coverage from AD 2001 to 2010, and the
satellite-based 1 1 NOAA Optimum Interpolation v.2
SST data set with coverage since November 1981 [Reynolds, 1988; Reynolds et al., 2002]. Monthly USGS stream
discharge rate data from the Baranof Island area are sparse
and discontinuous between AD 1915 to present, with a
large data gap between 1957 and 1976. Most discharge
measurements are from Sawmill Creek (site 15088000),
but this record is discontinuous with several breaks in data
coverage (Table 1). A combination of gage stations at Starrigavan Creek (site 15087618), the lower Indian River (site
15087700), and Bear Cove (site 15088200) contain data
coverage between AD 1999 and 2008. Discharge data from
the USGS stations are used in this study after normalization
by drainage area.
[8] Net primary productivity (NPP) estimates are from
SeaWiFS ocean color reflectance spectra and the carbonbased productivity model (CbPM) of Westberry et al.
[2008]. The CbPM algorithm differs from traditional
SeaWiFS-based estimates of net productivity that use a
combination of chlorophyll concentrations and parameterizations to account for changing light, nutrient, and temperature conditions [Behrenfeld and Falkowski, 1997].
Instead, the CbPM approach relates NPP to (i) phytoplankton carbon biomass measured using SeaWiFS particulate
backscattering coefficients, and (ii) phytoplankton growth
rates derived from SeaWiFS chlorophyll:C ratios
[McClain, 2009]. For this study, NPP values are averaged
at monthly time-steps between October 1997 and October
2007 using the SeaDAS 6.1 image processing software
(SeaDAS Development Group, NASA) for the grid box
shown in Figure 1a, which includes 181 individual NPP
nodes. The grid box covers approximately 31,000 km2 of
ocean surface area and is intentionally large to minimize
the contributions from localized anomalies related to high
turbidities in near-shore waters [International OceanColour Coordinating Group, 2000]. Due to extensive cloud
coverage and/or low light levels during the winter in Southeast Alaska, there are no data for the months November to
January.
2.3. Sediment Core Collection and Description
[9] The cores considered in this study were collected by
the R/V Maurice Ewing in September 2004 (Figure 1).
Both EW0408-32MC and EW0408-43MC were collected
using the Oregon State University (OSU) multicore rig in
Katlian Bay and Deep Inlet, respectively. EW0408-32MC
is 38 cm long, and EW0408-43MC is 61.5 cm long. Following retrieval, both cores were analyzed shipboard using
a GEOTEK Multi-Sensor Core Logger (MSCL) at 1 cm
intervals; data from GRA wet bulk density measurements
are considered in this paper. The cores were then split, lithologies described, and linescan images were collected
using the MSCL GEOSCAN 3-CCD camera, which yielded
RGB images with a pixel resolution of 100 pixels/cm.
137
Cs and excess 210Pb data for both EW0408-32MC and
EW0408-43MC have been presented previously [Walinsky
et al., 2009], with minimum detectable activity limits of
0.15 dpmg1 for 137Cs. All EW0408 cores are archived at
the OSU Core Lab.
2.4. CT Scans
[10] Computed tomography (CT) provides rapid nondestructive three-dimensional mapping of a material’s X-ray
attenuation variability at submillimeter spatial scales that
can be used to identify changes in sedimentary facies
[Duliu, 1999; Ketcham and Carlson, 2001; St-Onge and
Long, 2009]. CT imaging was performed at the OSU Veterinary College using a Toshiba Aquilion 64-Slice instrument. Axial (0.5 mm slices), coronal (longitudinal; 0.5 mm
slices), and sagittal (transverse ; 1.0 mm slices) images
were collected using a source radiation of 120 keV and 400
mA. CT measurements are expressed as Hounsfield units
(HU), which are a measure of a material’s X-ray mean
absorption coefficient relative to that of water, which in
turn reflects a combination of bulk density, effective atomic
number, mineralogy, and porosity [Boespflug et al., 1994;
Boespflug et al., 1995] and were extracted from the CT
scans using the software program ImageJ [Abramoff et al.,
2004]. HU numbers were then calibrated to wet bulk density using the low-resolution (e.g., every 1 cm) GEOTEK
GRA density measurements.
[11] Semiautomated lamination thickness measurements
were made on contrast-enhanced linescan and CT images
of EW0408-43MC using the WinGeol lamination software
[Meyer et al., 2006]. After initial processing of the
extracted RGB transect data by the WinGeol algorithm and
boundaries between sublamina were automatically
assigned, each sublamina boundary was evaluated and corrected by hand when necessary. Three separate downcore
transects were measured and the mean depth of each
boundary was recorded.
2.5. Scanning XRF Geochemistry
[12] Semiquantitative geochemical data were collected on
u-channel samples from EW0408-32MC and EW040843MC using an Avaatech XRF core scanner [Richter et al.,
2006] at Scripps Institution of Oceanography. The u-channel
3447
3448
–
0.93
92
0.99
59
–
0.47
90
0.54
94
0.89
322
0.99
65
0.99
94
0.52
346
0.53
350
b
Bold face indicates significant correlation at the 0.01 level. Italic numbers are population size.
USGS gage station at Sawmill Creek (site 15088000).
c
Discontinuous with coverage between AD 1920–1922, 1928–1936, 1937–1942, 1945–1957, and 2001–2009.
a
–
–
Cape Edgecumbe
NDBC Buoy SST
Fairweather Grounds
NDBC Buoy SST
NOAA Optimum
Interpolation SST
Sitka Precipitation
(Spliced)
USGS Fluvial
Dischargeb
SeaWiFS Net
Primary Productivity
0.55
62
0.55
65
0.89
64
–
0.54
1035
0.63
752
–
0.76
776
Sitka SAT (Spliced)
PFEL Coastal
Upwelling Index
North Pacific Index
0.32
436
0.25
241
0.57
418
0.44
54
0.37
79
0.41
100
0.43
420
0.18
1037
20.22
752
20.09
1038
0.16
64
0.10
92
0.07
322
0.69
86
0.67
34
0.53
50
0.56
96
20.13
86
0.08
64
0.66
96
0.63
96
http://www.emc.ncep.noaa.gov/
research/cmb/sst_analysis
http://cdo.ncdc.noaa.gov
AD 1981
AD 1997
http://www.science.oregonstate.
edu/ocean.productivity
AD 1920c http://waterdata.usgs.gov
AD 1922
http//www.ndbc.noaa.gov
http://www.ndbc.noaa.gov
http://www.cgd.ucar.edu/cas/
jurrell/npindex.html
http://www.pfeg.noaa.gov/products/
PFEL/modeled/indices/upwelling/
upwelling.html
http://cdo.ncdc.noaa.gov
Data
Source
AD 2001
AD 2002
AD 1922
AD 1946
AD 1899
PFEL Coastal
Cape
Fairweather
NOAA Optimum
Sitka
USGS Fluvial
SeaWiFS Beginning
Upwelling
Sitka SAT Edgecumbe NDBC Grounds NDBC
Interpolation
Precipitation
Discharge
Net Primary Date of
Index
(Spliced)
Buoy SST
Buoy SST
SST
(Spliced)
(Sawmill Creek) Productivity
Record
Table 1. Pearson Correlation Coefficients for Sitka Regional Climate and SeaWiFS Productivity Measurementsa
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
surfaces were lightly scraped to expose a less oxidized layer
of sediment and then covered with 4 m thick SPEXCerti
R foil to prevent excess drying or contaminaPrep UltraleneV
tion of the sediments during analysis as the XRF sensor
moved downcore. Both cores were analyzed at 10, 30, and
50 keV energies for 10, 20, and 30 s, respectively, where the
10 keV energy level was used to detect the elements Al
through Fe, 30 keV was used for Cu to Bi, and 50 keV measured Cd to Ba. A thick Cu filter and a Pd filter were placed
in front of the Rh X-ray source for the 30 and 50 keV scans,
respectively, to limit the lower energy bremsstrahlung and
improve the signal-to-noise ratio. Measurement slit size for
EW0408-32MC was 10 2 mm, while EW0408-43MC was
analyzed using a smaller 10 0.2 mm slit size.
[13] X-ray spectra were quantified using the WinAxil
software package (Canberra Industries, Inc.), where an iterative nonlinear least square fit (based on a Gaussian function) was used to model both K and K peak areas. The
postprocessing also removed the bremsstrahlung background and corrected for escape- and sum peaks. The
resulting data were then converted into total counts for
each detected element, and a 2 goodness-of-fit parameter
(D-area) was calculated. All scanning XRF element intensity measurements were then normalized by count times
and converted to centered natural-log ratios (clr) [Aitchison, 1982; Pawlowsky-Glahn and Egozcue, 2006; Welte
and Tjallingii, 2008] prior to compositional analysis. This
data treatment helps to eliminate the effects of physical
changes in the sediment matrix and scanner setup (e.g.,
grain size, water content, core surface topography, water,
and/or air bubbles that accumulate under the foil). Lighter
elements, generally those obtained in the 10 keV scan, are
most affected by the presence of water beneath the foil surface while elements with higher atomic masses (such as
those obtained in the 30 and 50 keV scans) are not as
affected, and can more reliably be reported without much
mathematical manipulation [Kido et al., 2006; Tjallingii et
al., 2007; Ziegler et al., 2008] (see also supporting information). Clr-transformation of X-ray intensities also have the
added benefit of correcting the constant-sum issue for compositional data [Aitchison, 1986] and are linearly proportional
to concentrations for most elements [Welte and Tjallingii,
2008]. All scanning XRF element intensities considered
exceed the 3 lower limit of detection (LLD), calculated as
LLDj ¼ 3 冑(cj/tj), where c ¼ total number of counts for element j and t is the count time in seconds [Potts, 1987]. Visual
inspection of representative X-ray spectra from selected samples in both cores was also used to confirm high signal-tobackground intensities for key elements.
2.6. Biogenic Silica, Total Organic Carbon,
Molar N/C, and Organic Matter d13C
[14] Bulk sediment samples (1 cm thick 2 cm wide)
were collected every 5 cm from EW0408-32MC and
EW0408-43MC, and biogenic silica (opal), total organic
carbon (TOC), total nitrogen, and organic matter 13C compositions were measured on homogenized carbonate-free
splits according to the methods described in Addison et al.
[2012]. Opal was measured using molybdate-blue spectrophotometry and a 0.1 M Na2CO3 extraction technique
[Mortlock and Froelich, 1989]. TOC, molar ratios of
TN:TOC (hereafter referred to as molar N/C ratios), and
stable isotope analyses were performed at the Alaska Stable
Isotope Facility (ASIF) using a Costech 4010 HCNS elemental analyzer coupled to a Finnigan DeltaplusXP isotope
ratio mass spectrometer (IRMS) on EW0408-32MC. A
similar set of analyses were performed on EW0408-43MC
by the USGS Isotope Tracers Project Lab using a Carlo
Erba NA 1500 elemental analyzer and an Elementar (formerly Micromass) Optima IRMS. Several samples measured by ASIF were also measured by USGS to ensure data
were of similar quality and that results would be directly
comparable. Organic matter 13C values are reported in
permil units (%) relative to the Vienna Peedee Belemnite
(V-PDB),
where
13C ¼ [{(13C/12C)sample/(13C/12C)V15
PDB} 1] 1000%. Sedimentary N data follow the
same permil convention relative to atmospheric nitrogen.
Replicate analyses of internal standards indicate standard
deviations (1 ) of approximately 1.0% for opal for both
cores, and for EW0408-32MC (measured at ASIF; n ¼ 5) 1
deviations for TOC were 0.88%, 0.012 for molar N/C
ratios, 0.10% for organic matter 13C, and 0.06% for sedimentary 15N measurements, while the 1 deviations for
EW0408-43MC (measured at USGS; n ¼ 17) were 0.98%
for TOC, 0.010 for molar N/C, 0.09% for organic matter
13C, and 0.14% for sedimentary 15N.
3.
Results
3.1. Modern Climate Data
[15] The monthly instrument data from Southeast Alaska
compiled for the years AD 1998 to 2008 show evidence of
strong seasonality in continental shelf waters (Figure 3).
Low values of the North Pacific Index during the winter
months indicate intensified Aleutian Low atmospheric dynamics in the eastern Gulf of Alaska (Figure 3a), which
contributes to strong onshore winds and coastal downwelling (Figure 3b). Winter is also the period of minimum
SAT and SST (Figures 3c and 3d). The peak in NPP occurs
during the spring and summer (Figure 3g), due to a combination of the seasonal increase in solar insolation (Figure
3d), euphotic zone thermal stratification, and a reduction in
downwelling intensity. The maxima in precipitation and
fluvial discharge follow in the autumn, with the peak flow
season tending to occur between the months of September
and December (Figures 3e and 3f).
[16] The full suite of climate time-series data exhibit
several significant correlations. All measurements of SST
correlate at the 99% significance level with the longer Sitka
SAT data (Table 1), and there is a significant correlation
between Sitka SAT and the SeaWiFS-derived NPP record.
A somewhat weaker correlation exists between Sitka precipitation and the Sawmill Creek discharge data, though it
still exceeds the 99% significance level. The relatively
small mountainous drainage basins adjacent to Katlian Bay
and Deep Inlet offer little surface water storage area or
groundwater discharge, thus favoring a rapid fluvial
response to seasonal peaks in precipitation. Collectively,
these results indicate that (i) the long Sitka SAT record is
representative of local ocean SST conditions, particularly
with regard to seasonality (Figure 4a) and (ii) the long Sitka
precipitation record is likely a good indicator of regional
precipitation that can be cross checked against the discontinuous USGS fluvial discharge network (Figure 5).
3449
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 3. Instrument data from the Baranof Island region for the years AD 1998–2008: (a) SLPderived North Pacific Index; (b) PFEL wind-derived upwelling index (negative values indicate downwelling); (c) monthly mean air temperature at the Sitka Airport; (d) SST from two NOMAD buoy stations
and the 1 1 NOAA Optimum Interpolation data set (extracted from the grid box location in Figure
1a); (e) total monthly precipitation at the Sitka Airport; (f) USGS fluvial discharge rates normalized by
drainage basin area ; and (g) SeaWiFS net primary productivity measured offshore Baranof Island, where
individual data points represent the mean value calculated for the grid area in Figure 1a, with 1 error
bars. Note reversed axes for both the North Pacific Index (a) and the PFEL upwelling index (b) data.
[17] Because of the short duration of SeaWiFS observations, a multiple (stepwise) linear regression model [Davis,
2002] was derived to extend the NPP record to the beginning of the Sitka weather observations in AD 1922 (Figure
4b). Using the available climate data from Table 1, data
sets were selected for inclusion in the regression model if
the probability of the resulting F-statistic was less than a
significance level of 0.05, while data sets were excluded if
the probability of the resulting model F-statistic exceeded
the 0.10 significance level. For the period between AD
1922 and 1981, the stepwise approach identified both the
Sitka SAT record and the North Pacific Index as key variables, such that
SeaWiFS net primary productivity ¼ ½SitkaSAT 41:579
þ ½North Pacific Index 44:673 þ k1
ð1Þ
where k1 is a constant of 387.529. When compared to the
measured NPP data, equation (1) accounts for 60.8% of the
total variance (p < 0.01). For the period between AD 1981
to 2009, the stepwise regression included the NOAA Optimum Interpolation SST data, where
SeaWiFS net primary productivity ¼ ½SitkaSAT 87:017
þ ½North Pacific Index 45:719 þ ½NOAA SST 68:534
þ k2
ð2Þ
where k2 ¼ 136.864. This equation improved the fit
between the modeled and measured productivity data,
increasing the variance descriptor to 66.3% which is a significant improvement (p < 0.01) over the model derived
from equation (1) (Figure 4b).
3.2. EW0408-32MC
[18] This multicore sample recovered the sediment-water
interface, and consists of very dark greenish-gray clay that
contains diffuse centimeter-scale laminations visible in the
3450
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 4. (a) Monthly climatology for Sitka surface air
temperature (SAT), the NOAA Optimum Interpolation SST
(mean for grid box in Figure 1a), and the SeaWiFS derived
net primary productivity (NPP; mean for grid box in Figure
1a). All error bars are 1. (b) Time series of multiple (stepwise) regression model of NPP from equation (2) (see text)
during period of overlap with recent SeaWiFS-measured
NPP observations. Correlation between model and data is
highly significant.
linescan imagery (Figure 6), but which appear to be moderately bioturbated in the CT images. Calibration of the
HU units from the CT scan against the low-resolution
GEOTEK GRA wet bulk density measurements (n ¼ 40,
r ¼ 0.59, p < 0.01) yields a mean bulk density of
1.08 6 0.01 g/cm3.
[19] The carbonate-free organic element and 13C data
(supporting information Table S1) indicate high concentrations of organic material, including mean opal concentrations of 27.4 6 2.0% and TOC values of 4.9 6 0.5%. Mean
molar N/C ratios are 0.094 6 0.013, organic matter 13C
values are 21.9 6 0.4%, and sedimentary 15N is
6.4 6 0.2%. Negative Pearson correlations exist between
natural-log-transformed values of opal and TOC (n ¼ 8,
r ¼ 0.90, p < 0.01), and natural-log-transformed values of
organic 13C and molar N/C ratios (n ¼ 8, r ¼ 0.894,
p < 0.01). Applying a linear mixing model to estimate the
fraction of marine-derived organic matter that uses a terrestrial end-member 13C composition of 26% and a marine
end-member 13C ratio of 21% [Meyers, 1994; McQuoid
et al., 2001; Walinsky et al., 2009; Addison et al., 2012]
yields a range of marine-derived organic matter fractions of
between 0.71 and 0.93 (Figure 7).
[20] Comparisons between clr element intensities that
exceeded the detection limits of the Avaatech scanning
XRF instrument show several coherent trends (Figure 8).
Clr Al, Si, Ti, K, Fe, Ca, and Mn intensities exhibit rapid
increases with increasing core depth between 0 and 12 cm
that transitions to a more gradual increase that persists to
the base. These similarities are reflected in the low covariance between most of these elements (supporting information Table S2). The profile of clr Ba is similar, but with a
subsurface maxima between 7 and 17 cm depth. These
trends generally contrast with those of clr Zn, Cu, Pb, Y,
Zr, Rb, and Sr, which show an overall decrease toward the
base of 32MC. Between 0 and 21 cm depth, both clr Cl
and S follow similar trends. Below 21 cm, S increases toward the base of the core, while Cl decreases. The intensity
pattern of clr Br is distinct from any other element in
32MC, in that it oscillates between intervals of high and
low intensities, superimposed on an overall linear decrease
with depth.
[21] The radioisotope data for EW0408-32MC (Figure
6a) indicates several important features. The firstappearance datum (FAD) for 137Cs activity occurs between
25 and 30 cm, and the FAD for excess 210Pb activity occurs
below 35 cm (e.g., below the base of EW0408-32MC). A
subsurface peak in 137Cs activity occurs at 25 cm. The
activities of the two uppermost excess 210Pb samples are
approximately equal, as are the two uppermost 137Cs samples, implying a surface mixed layer 5 cm thick. Depth
trends in both excess 210Pb and 137Cs activity profiles suggest steady-state accumulation, with a maximum apparent
accumulation rate of between 5 and 6 mm/yr for 137Cs, and
3–6 mm/yr for excess 210Pb based on the constant initial
concentration accumulation model [Nittrouer et al., 1979].
A linear composite sedimentation model was developed
from these data using a rate of 5 6 1 mm/yr that places the
core top at AD 2004 (date of collection), and the base at
AD 1934. Combined with the CT-scan evidence of bioturbation (e.g., mottled fabric between 0 and 10 cm), these
geochronology data indicate EW0408-32MC proxy data
can be resolved at subdecadal but not annual time
resolutions.
3.3. EW0408-43MC
[22] The sediment-water interface was not recovered in
multicore EW0408-43MC, indicating an unknown volume
of material was lost during core recovery. Both linescan
and CT images of EW0408-43MC show parallel beds that
vary both in color and HU value intensity (Figure 9). When
the HU data are empirically converted into density values
using the GEOTEK MSCL low-resolution bulk density
measurements (n ¼ 60, r ¼ 0.68, p < 0.01), EW0408-43MC
has a mean bulk density of 0.92 6 0.03 g/cm3.
[23] Using a combination of color, HU, and relative bulk
density, the stratigraphy of EW0408-43MC can be classified into four broad categories : (i) alternating thin light
brown low-density (0.88 g/cm3 ; CT-dark) sublaminae
and dark brown/black medium-density (0.92 g/cm3 ; CTlight) sublaminae ; (ii) a thick normally graded unit with a
coarse very high-density (1.28 g/cm3) base between 12.4
and 14.7 cm depth, interpreted as a turbidite; (iii) a thick
high-density (0.96 g/cm3) unit between 32.8 and 42.0 cm
depth with gradational contacts ; and (iv) a bioturbated horizon between 18.7 and 24.7 cm. The repeating nature of
the light brown and dark brown/black sublaminae suggests
3451
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 5. Spliced (a) annual total (black) and monthly total (gray) precipitation record for the Sitka
area for AD 1922 to 2010, from combining observations from the Sitka Airport and the Sitka Magnetic
Observatory (period of overlap 1948–1989; n ¼ 503, r ¼ 0.966, p < 0.01). Gap in data coverage between
1996 and 1999. White triangle on left axis indicates mean annual total precipitation, with 1 (black) and
2 (red) levels indicated by dashed lines. (b) Monthly climatology derived from Figure 5a, indicating the
wettest months are September to December, which alone account for 50% of the mean total annual precipitation. Error bars are 1 from monthly mean. (c) Standardized wet season precipitation anomalies
from Figure 5a, and discharge for Sawmill Creek (USGS station 15088000). The two driest wet seasons
occur consecutively during 1950 and 1951.
these are cyclical couplets whereas the higher density
deposits appear to be episodic.
[24] Organic geochemistry and isotope results for
EW0408-43MC (supporting information Table S1) indicate
extremely high concentrations of opal (mean 50.7 6 8.0%)
and TOC (10.8 6 1.4%). Molar N/C ratios are also high
(0.085 6 0.005), as are organic matter 13C (22.1 6
0.5%) and sedimentary 15N (6.1 6 0.3%). Negative Pearson correlations exist between natural-log-transformed values of opal and TOC (n ¼ 12, r ¼ 0.86, p < 0.01), and
natural-log-transformed values of organic 13C and molar
N/C ratios (n ¼ 12, r ¼ 0.97, p < 0.01). Using the same
parameters to calculate the proportion of marine contribu-
tions as for EW0408-32MC (see section 3.2) yields a range
between 0.63 and >1.0 of marine-derived organic matter
(Figure 8), with the high value indicating that some samples had 13C values more enriched than the marine endmember value of 21%.
[25] The lithological complexity in 43MC is also
reflected in the geochemical scanning XRF data (Figure 9).
The clr-transformed intensities for Ti, Ca, Sr, and Fe follow
similar patterns, with the highest intensities associated with
the turbidite deposit and the other high-density units. These
trends contrast sharply with those of Si, Al, Cr, Mn, Mo,
Cu, Ba, and Zn, all of which have minimum intensities
associated with the turbidite and high-density intervals, and
3452
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 6. Geochronology results for (a) EW0408-32MC show a maximum apparent steady-state linear
sedimentation rate of 5 6 1 mm/yr by excess 210Pb, and an 137Cs subsurface maxima from the global fallout maxima during AD 1964 [Appleby, 2001]. These results indicate EW0408-32MC is a continuous
record from AD 1934 to 2004. (b) EW0408-43MC over-penetrated and lost the most recent core-top
sediments, and CT imagery reveals a non-steady-state lithostratigraphy (see text for discussion). The
FAD of 137Cs occurs in the subsurface, indicating the interval between 45 and 50 cm corresponds to the
early AD 1950s, supporting the inference that a hypopycnal plume deposit linked to cessation of lowflow conditions in Southeast Alaska occurred during the early 1950s (Figure 5). A relative varve count
age-depth model anchored using both the hypopycnal plume deposit and a seismically generated turbidite (30 July 1972) indicates that EW0408-43MC spans AD 1940 to 1981, with a hiatus between AD
1956 to 1972.
maxima associated with the low-density intervals. Both S
and Cl show some similarities to the Si group as well, with
maxima at 6, 17, and 28 cm depth. Clr Br is complex, as
the lowest values are associated with the turbidite deposit,
intermediate values occur within the low-density laminations, high values occur with the high-density unit between
33 and 42 cm, and the highest Br intensities are at the core
top. The heterogeneity of EW0408-43MC is also reflected
in the higher variance (supporting information Table S2),
with the elements Fe and Al showing the greatest variance
of any element in the scanning XRF data set.
[26] Results of the radioisotopic analyses for the
EW0408-43MC samples parallel the complex lithostratigraphy described above (Figure 6b). As the core barrel
for EW0408-43MC over-penetrated during collection,
there is no intact sediment-water interface. The presence
3453
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 7. Sedimentary organic matter relationships for EW0408-32MC and EW0408-43MC. (a) Negative correlations between natural-log-transformed opal and TOC concentrations indicate TOC is diluted
by the greater siliceous productivity flux. Dashed lines are intracore linear regressions, both of which are
significant. (b) Correlations between natural-log-transformed opal and (absolute) organic 13C show that
the EW0408-32MC data is not significant, but that the high correlation for EW0408-43MC suggests the
13C variability reflects changes in siliceous growth rates. Organic matter provenance diagrams using (c)
13C and N/C, and (d) 13C and 15N, indicate both cores are predominantly autochtonous marine organic matter, typical of southeastern Alaskan temperate fjords. Dashed lines are linear mixing models
between 13C and N/C or 15N compositions for marine phytoplankton and vascular plant debris, and
phytoplankton and soil organic matter, with percentages indicating proportion of marine-derived fraction. GoAK ¼ Gulf of Alaska. VPD ¼ vascular plant detritus. Compositional ranges are from regional
surface sediment samples [Walinsky et al., 2009].
of two episodic depositional units (e.g., the turbidite and
the high-density unit between 32.8 and 42.0 cm depth)
and the highly variable thicknesses of the laminated
horizons (mean 7.6 6 9.2 mm at 1 standard deviation)
preclude the use of steady-state accumulation rate models to calculate age-depth relationships for EW040843MC. The episodic depositional horizons noted above
tend to bracket anomalously low excess 210Pb and/or
137
Cs activities, further supporting the interpretation that
these deposits do not represent steady-state depositional
conditions. The FAD of anthropogenically derived 137Cs
occurs between 45 and 50 cm depth, which suggest
that this interval dates to the early AD 1950s. The
FAD of excess 210Pb occurs below the basal depth of
EW0408-43MC, which indicates the base is younger
than AD 1900.
4.
Discussion
4.1. Event Sedimentology and Geochemistry of
EW0408-43MC
4.1.1. Sublaminae Couplets and Bioturbation Horizon
[27] There is a high degree of geochemical overlap
between the CT-light and CT-dark sublaminae, as well as
the bioturbation horizon. This result is not unexpected as (i)
the bioturbation horizon likely represents a mixture of these
different sublaminae and (ii) the processes that lead to the
generation of the CT-dark (e.g., marine phytoplankton) and
3454
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 8. Linescan imagery, CT scanning results and calibrated bulk density, organic matter analysis,
and scanning XRF geochemistry for EW0408-32MC. Fraction of marine-derived organic matter calculated using a terrestrial end-member 13C of 26% and a marine end-member 13C ratio of 21%.
Additional XRF element intensity data contained within online supplement.
CT-light (e.g., terrigenous material) sublaminae operate
year-round in Southeast Alaska (Figure 3), though there is
also a strong seasonal difference in the periods of maximum
accumulation for these different components.
4.1.2. Turbidite (Seismite)
[28] Prior work by Walinsky et al. [2009] suggested that
the upper 10.0–18.0 cm of EW0408-43MC was in
nonsteady-state accumulation, possibly as a result of submarine gravity flows. The CT images support this interpretation, as indicated by the thick turbidite deposit between 12.4
and 14.7 cm (Figure 9). Using a combination of multivariate
discriminant function analysis and ternary diagrams of key
geochemical elements (Figure 10), the turbidite appears distinct from the other EW0408-43MC units, possibly as a
result of being composed of coarser terrigenous clasts that
were originally deposited upslope from the main fjord basin.
[29] While such deposits can result from aseismic processes [Syvitski et al., 1987], it seems more likely this turbidite is the result of slumping associated with a shallow
M7.6 earthquake that occurred 30 July 1972 with an epicenter less than 20 km distant from the EW0408-43MC
core site (Figure 1b) [Page, 1973; Schell and Ruff, 1989].
The 1972 earthquake also had an associated tsunami, suggesting the combination of slumping and tsunami backwash
would be very likely to cause turbidite deposition, as was
observed with the 1964 Good Friday earthquake (M9.2) in
Port Valdez [Sharma and Burbank, 1973]. The turbidite
also overlies the FAD of 137Cs in EW0408-43MC (Figure
6b), which supports the inference that the turbidite is
younger than the early AD 1950s. The presence of scourand-fill structures at the base of the turbidite indicates an
erosive contact with the underlying laminae.
4.1.3. Hypopycnal Plume Deposit
[30] The second high-density unit between 32.8 and 42.0
cm in EW0408-43MC lacks both a coarse-grained basal
unit and erosive contacts (Figure 9), which suggests this
unit is not a turbidite. Instead, this deposit has the characteristics of a fluvial hypopycnal plume deposit, as the high
density of the unit reflects a greater contribution of terrigenous sediment whereas the fine grain-size could be
accounted for by settling of the suspended fluvial load as
flocculated clastic mud [Syvitski et al., 1987; Ducassou et
al., 2008]. The hypopycnal plume deposit describes a geochemical population that differs from the mixed sublaminae/bioturbation populations (Figure 10), and both the CT
bulk density and scanning XRF geochemical data show it is
composed of three distinct subunits (Figures 9 and 10). The
upper and lower subunits both have similar densities and
geochemical compositions, while the middle subunit
appears more similar to the sublaminae/bioturbation compositions. Visual smear slide microscopy further indicates
that the upper and lower subunits are composed of 75%
terrigenous clays (<4 mm), while the middle subunit is
only 50% terrigenous clays. All three subunits are also
composed of biogenic components (predominantly diatoms
and amorphous organic matter), with the middle subunit
composed of much greater proportions of fragile Chaetoceros sp. diatom spores which are difficult to transport
without fragmentation (J. Barron, personal communication)
and suggests the middle subunit is a primary depositional
feature. The internal stratification, stratigraphy, and geochemistry of the hypopycnal deposit thus suggests: (i) generation of the overall deposit differs from the normal
sedimentation regime that generates the CT-dark/CT-light
sublaminae ; (ii) the thickness of the deposit (10 cm) is
due to rapid accumulation ; and (iii) the upper and lower
subunits are derived from allochthonous fine-grained lithic
clays, while the biogenic-rich middle subunit is derived
from autochthonous planktonic sources. Overall, these
observations suggest that the hypopycnal deposit was deposited during a single autumn/winter flow season, but in
detail, it appears to be composed of two different
3455
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 9. Linescan imagery, CT scanning results, calibrated bulk density, organic elemental and isotopic analyses, and scanning XRF geochemistry for EW0408-43MC. Additional XRF element intensity
data contained within online supplement. Note that the density record has been truncated.
‘‘flushing’’ events separated by a late-season phytoplankton
bloom. It is difficult to ascribe a single mechanism to the
cause of such a bloom, but perhaps it was stimulated by the
rapid influx of terrestrially derived nutrients to Deep Inlet
from the first hypopycnal event.
[31] Fluvial hypopycnal plume deposits require a strong
salinity contrast between a low-salinity surface water mass
and underlying saline waters, a condition that is common in
temperate fjords (Figure 2). Precipitation and fluvial discharge records from the Sitka area suggest the most likely
Figure 10. EW0408-43MC scanning XRF intensity data by lithology. Both (a) multivariate discriminant function analysis and (b) centered ternary Fe–Ca–Br diagrams [Thio-Henestrosa and Martin-Fernandez, 2005] indicate that the geochemical composition of the turbidite is distinctive from the other
units, while both sublaminae types and the bioturbation zone are geochemically similar. The hypopycnal
plume deposit has been subdivided into three subunits, and geochemical analysis indicates the upper and
lower subunits are geochemically similar, while the middle subunit is more similar to the biogenic-rich
sublaminae composition.
3456
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
period for such a hypopycnal deposit to accumulate in
Deep Inlet would be associated with the resumption of
high-flow conditions during the wet season of AD 1952
(Figure 5). This period is anomalous in that it received very
low precipitation (two standard deviations below the annual
mean) during the two preceding years, which resulted in an
accumulation of fine-grained detritus and sediment in local
stream beds. The return of normal high-flow conditions,
therefore, would facilitate transport of this multiyear accumulation of terrigenous material into coastal waters.
4.2. Construction of a Varve Chronology
for EW0408-43MC
[32] The assignment of the turbidite and the hypopycnal
deposit to specific periods allows the construction of a preliminary chronology for EW0408-43MC based on the hypothesis that the laminated couplets preserve the annual
cycle of deposition in southeastern Alaska (e.g., varves).
Since the CT-dark sublaminae are dominated by organic
matter, and the peak in NPP in Southeast Alaska occurs
between May and August (Figure 4a), the CT-dark sublaminae are designated as ‘‘0.5’’ year intervals (e.g. 1945.5,
1957.5, etc.). Similarly, as the CT-light sublaminae are
composed primarily of detrital material from precipitation
run-off and fluvial discharge which peaks in southeastern
Alaska between September and December (Figure 5b),
these units are designated as ‘‘0.0’’ year intervals (e.g.,
1945.0, 1958.0, etc.) by convention.
[33] The gradational contact between the turbidite and
the overlying CT-dark sublaminae supports the inference
that the episodic deposition of the turbidite occurred during
an interval of high organic matter accumulation (likely the
summer diatom bloom; Figure 4a) without disrupting the
continuity of the record following the deposition of the turbidite. To maintain continuity with the preliminary varve
chronology, the contact between the overlying CT-dark
sublaminae and the turbidite is assigned an age of AD
1972.5. The hypopycnal plume deposit is assigned an age
of AD 1952.0, and as it has gradational contacts both above
and below with adjacent CT-dark sublaminae, this unit is
part of the continuous record.
[34] A time discontinuity in the EW0408-43MC varve record occurs between 14.7 and 24.7 cm. The upper depth discontinuity is due to the erosional scour at the base of the
turbidite that removed an unknown number of underlying
sublaminae. The lower discontinuity depth is at the base of
the bioturbated horizon, which represents the maximum penetration depth of homogenization of several lamina couplets,
thus obscuring the varve chronology for this interval as well.
Combined, the turbidite scour and the bioturbation have
effectively removed the time constraints on the laminae couplets between 14.7 and 18.7 cm. This gap in the chronology
is constrained to between AD 1956 and 1972.5.
[35] The resulting chronology is presented in Figure 6
and represents the mean depth of each sublaminae measured from three downcore transects. This floating varvebased chronology places the core top age at AD 1981, the
base of EW0408-43MC at AD 1940, and includes a hiatus between AD 1956 and 1972. The resulting EW040843MC chronology also explains the absence of key 137Cs
horizons, including the lack of an AD 1963/1964 subsurface maxima (which would have been deposited during the
hiatus), as well as the lack of an AD 1986 maxima related
to the high-latitude deposition of radionuclide fallout from
the Chernobyl accident [e.g., Sugai et al., 1994] due to
over-penetration during core collection. Given the uncertainties associated with the generation of this age-depth
model, the chronology for EW0408-43MC is likely accurate only to subdecadal timescales.
4.3. Shifts in Recent Primary Productivity and
Sedimentary Br/Cl Ratios
[36] Few instrument data sets from Southeast Alaska are
long enough to compare directly against the multidecadal
record of sedimentation preserved in multicore samples
EW0408-32MC and EW0408-43MC. The Pacific Decadal
Oscillation record is based on statistical analysis of North
Pacific Ocean SST and extends to the beginning of the 20th
Century AD [Mantua et al., 1997]. It has been used in
many studies as an estimator of the state of North Pacific
climate regimes [e.g., Minobe, 1999]. Therefore, the PDO
index is used in this study as a measure of North Pacific
environmental variability, against which to compare the (i)
derived measures of Sitka primary productivity and (ii) the
geochemistry of the recent sedimentary record accumulating in Katlian Bay and Deep Inlet.
[37] Equations (1) and (2) allow an estimation of NPP
from the long Sitka SAT, NPI, and SST records. NPP is not
normally used as an estimator for export productivity, particularly in regions where particle export tends to be anticorrelated with biological activity such as in the Southern Ocean
[Sarmiento and Gruber, 2006]. However, sediment flux
studies in the high-latitude North Pacific Ocean [Francois et
al., 2002; Whitney et al., 2005] show high particle export
efficiencies, but low transfer efficiencies to the deep ocean
(PEeff and TEeff, respectively, as defined by Henson et al.
[2012]). Henson et al. [2012] use a global compilation of
open-ocean sediment trap data [Honjo et al., 2008] to calibrate satellite estimates of PEeff and TEeff and show that the
coastal Gulf of Alaska has (i) high levels of primary production; (ii) high surface particle export rates; (iii) high PEeff ;
and (iv) low TEeff. The shallow depths of the core sites in
Katlian Bay (146 m depth) and Deep Inlet (91 m depth) minimize the impact of TEeff to this study, as a low TEeff would
result in a skewed and/or discontinuous sediment record.
The recognition that diatoms greatly increase rates of PEeff
in high-latitude settings especially during bloom seasons
[e.g., Buesseler, 1998], such as those that occur in the Gulf
of Alaska, further supports the inference that surface productivity is comparable to the export productivity that accumulates beneath the shallow waters of southeastern Alaska
temperate fjords. The elevated concentrations of diatombound opal in both Katlian Bay and Deep Inlet (mean 27
and 51%, respectively) also reflect this high PEeff. Sediment
trap studies performed in other Northeast Pacific temperate
fjords (Boca de Quadra: Burrell [1989]; Effingham Inlet:
Chang et al. [2013]) show that the highest total biogenic
fluxes occur during the late spring/early summer seasons,
which is consistent with the SeaWiFS NPP observations for
the Sitka region (Figure 2). The Chang et al. [2013] study
measured opal, TOC, and CaCO3 fluxes, with the spring/
summer opal flux nearly four times higher than during the
subsequent autumn and winter seasons. Collectively, these
observations suggest that SeaWiFS NPP can be used to
3457
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 11. Modern (a) monthly Pacific Decadal Oscillation (PDO) index, (b) estimated Sitka net primary productivity from equation (2), and the clr-transformed Br/Cl ratio
from (c) EW0408-32MC and (d) EW0408-43MC. PDO
index from Mantua et al. [1997]. Thick dashed line indicates timing of North Pacific regime shift that occurred during the winter of AD 1976/1977.
estimate export productivity in Katlian Bay and Deep Inlet
in lieu of having direct sediment trap data, though these estimates should probably be considered only semiquantitative.
[38] SeaWiFS observations between AD 1997 and 2009
(Figure 3) indicate the dominant peak in NPP in Southeast
Alaska occurs between June and August (Figure 4a). Comparing deviations from the overall summer mean NPP
shows multidecadal variability with periodicity similar to
that of the PDO (Figures 11a and 11b). This result is not
surprising as both the PDO and Sitka productivity anomaly
data sets include an SST term, though there are subtle differences (e.g., the PDO is based on a North Pacific basinwide SST mean, while the Sitka record is based on local
SST conditions, and includes SLP and SAT terms). Therefore, this SeaWiFS-based estimator of NPP is likely a
semi-independent estimator of regional Southeast Alaska
productivity through the early 20th century. Together, the
PDO index and the Sitka productivity estimate yield a modern data comparison of North Pacific climate regime state
and export productivity, respectively, that correspond to
the same time intervals that are preserved in EW040832MC and EW0408-43MC.
[39] The bromine (Br) X-ray intensities from both multicore samples show large variations since the early AD
1930s (Figures 8 and 9). Several studies have shown strong
coupling exists between sediment-hosted halides and organic carbon accumulation [Calvert et al., 2001; Mahn and
Gieskes, 2001; Price and Calvert, 1977], though Br is also
a conservative inorganic component of seawater [Morris
and Riley, 1966]. Recent synchrotron-radiation analyses of
organic-rich North Pacific marine sediments indicate an organic Br pool [Leri et al., 2010], suggesting that measurements of Br could potentially be used as a proxy of organic
matter accumulation [Ziegler et al., 2008]. The stable organic matter 13C ratios suggest that marine organic matter
accounts for >80% of biogenic accumulation in both
EW0408-32MC and EW0408-43MC. Therefore, variations
in Br likely reflect changes in marine organic matter accumulation which is ultimately driven by export productivity
in southeast Alaskan fjords. Bromine is also uniquely wellsuited to XRF core scanning analysis due to its high excitation energy (K1 ¼ 11.9 keV), which limits X-ray attenuation due to (1) high porewater content and (2) water films
that form between the sediment surface and X-ray transparent foils (see supporting information) [Kido et al., 2006;
Ziegler et al., 2008]. To account for inorganic seawater Br,
a normalization using clr-transformed Cl was applied to the
clr Br data, which yielded trends similar to the nonnormalized clr Br data that confirms organic-bound Br as the predominant phase in these sediments.
[40] The continuous Br/Cl ratio data from EW040832MC contains decadal-scale shifts superimposed over a
decreasing linear trend from the core top toward the base
(Figure 11). The high-resolution discontinuous data in
EW0408-43MC also shows similar behavior. In particular,
both cores show a shift in the Br/Cl ratio during the late AD
1970s. This period corresponds to the major North Pacific
climate regime shift that occurred rapidly during the winter
of AD 1976/1977 [Ebbesmeyer et al., 1991], which resulted
in a significant shift in Gulf of Alaska marine ecosystem dynamics toward a higher productivity baseline [Beamish,
1993; Hollowed et al., 2001]. Both cores also show evidence of an earlier shift in their respective Br/Cl ratios during the early AD 1950s, another period of noted change in
North Pacific climate and ecosystems [Mantua et al., 1997].
4.4. Reconciling Discrepancies Between Traditional
Productivity Proxies and the Br/Cl Ratio
[41] The organic element concentrations (opal and TOC)
and stable isotope ratio (13C and 15N) data sets indicate
that EW0408-32MC and EW0408-43MC have different
sensitivities to sediment accumulation regimes. When the
biogenic fractions are calculated for both cores using the
equation
Biogenic fraction ¼ opal þ ð2 TOC Þ þ CaCO 3
ð3Þ
and treating CaCO3 as a constant of 6.0% based on a mean
concentration of surface samples from 10 temperate Southeast Alaska fjords [Walinsky et al., 2009]), the mean biogenic content of EW0408-32MC is 43.2 6 1.2% and
EW0408-43MC is 77.2 6 5.4%. Assuming authigenic precipitation of other sediment phases is negligible and that
the lithic fraction represents the remainder of the bulk
3458
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
Figure 12. Time-series analysis of the EW0408-32MC Br/Cl data set. (a) After detrending, a sequential
Student’s t test algorithm [Rodionov, 2004, 2006] was used to identify distinct regimes (b) that exceeded
a 99% confidence level. Thick line is the weighted mean value of each regime calculated with a 10-yr
minimum duration and an IP4 prewhitened red-noise estimator (subsample size ¼ 16). Regime shift dating errors determined from radioisotopic age-model errors, which increase in the downcore direction. (c)
REDFIT Lomb–Scargle Fourier transform frequency analysis of the EW0408-32MC detrended Br/Cl
data [Schulz and Mudelsee, 2002]. Significance of spectral amplitude evaluated against the theoretical
red-noise spectrum (thin dotted line) and a Monte Carlo test at the 95% significance level (thick line).
sediment phases, there is a much greater lithic fraction
present in EW0408-32MC than EW0408-43MC.
[42] An unexpected result from EW0408-32MC is the
opposite downcore trends between opal and the Br/Cl ratio
(Figure 7). Downcore changes in opal are also similar to
the low-frequency background for the lithogenic elements
(e.g., Ti, Fe, and K) and wet bulk density, which is unexpected as the lithogenic and biogenic fractions are nearly
equal in EW0408-32MC. The downcore shift in all solidphase components can best be explained by changes in
sediment porosity, which does not necessarily require
changes in the relative concentrations of either biogenic or
lithogenic phases (see supporting information for porewater
model results). The opposite trend between opal and TOC
is, therefore, assumed to be due to a combination of dilution by porewaters, as well as dilution of the remaining biogenic sediment phases by the greater contribution of opal
(Figure 7a). Since Br/Cl is proportional to the TOC concentration of the sediment, the secular decline in the Br/Cl data
downcore thus also reflects a combination of dilution by
porewaters and opal.
[43] The downcore plots of EW0408-43MC biogenic and
lithogenic data (Figure 9) do not exhibit the same porositydriven secular decrease as that found in EW0408-32MC.
This result is likely due to (i) the lack of recovery of the
sediment-water interface of EW0408-43MC where the exponential decline in porosity would be most pronounced
and (ii) the much higher concentrations of opal and organic
matter may create a gel-like rheology [Di Giuseppe et al.,
2009] that differs in its response to burial and compaction
from that of the clastic clays of EW0408-32MC.
4.5. Time-Series Analysis of EW0408-32MC
[44] The continuous and steady-state accumulation of
sediments in EW0408-32MC suggests this record is ideal for
detailed time-series analysis. After removing the porosityinduced trend from the Br/Cl ratio data using linear regression, the resulting detrended Br/Cl ratio was analyzed using
a modified sequential Student’s t-test algorithm [Rodionov,
2004, 2006] to test the occurrence of coherent ‘‘regime’’
states in the data (Figures 12a and 12b). Six regimes were
identified, with shifts exceeding the 99% confidence threshold occurring in AD 1991 (63 years), 1975 (66 years),
1965 (68 years), 1954 (611 years), and 1941 (613 years),
all of which have been previously identified as regime shifts
in the PDO. Analysis of the frequency domain of the
detrended EW0408-32MC Br/Cl data shows a significant
multidecadal peak with a period of 23 years/cycle (Figure
12c) which corresponds with known periodicities of North
Pacific climate regime shifts [Mantua et al., 1997]. It should
be noted that the short timescale of the EW0408-32MC record precludes the identification of longer PDO periods,
such as the 50–70 yr periodicities found by Minobe [1999].
5.
Conclusions
[45] Modern instrument and paleoceanographic proxy data
demonstrate the prevalence of low-frequency oscillations that
dominate Gulf of Alaska climate. Bottom-up forcing of the
marine ecosystem by changes in atmospheric and ocean climate conditions, particularly associated with the North Pacific regime shift in AD 1976/1977, is also expressed in the
accumulation record of marine-derived organic matter in the
3459
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
nearshore temperate fjords of Southeast Alaska. The characteristics of these ice-free fjords (low bottom water oxygen
concentrations and high primary productivity) make them
ideal depocenters for preserving high-quality paleoclimate archives, though they can also introduce complications, such as
turbidite erosion and deposition, and episodic oxygenation of
bottom waters. Despite significant differences in interfjord
geomorphology such as sill depth and upland drainage basin
physiography, reliable paleoclimate records can be extracted
and cross-validated at a regional scale.
[46] The complex sediment deposition mechanisms that
occur in temperate fjords necessarily require highresolution scanning and imaging techniques that are normally beyond the scope of studies focused on traditional
deep-water marine depocenters. Cryptic episodic deposition events discovered in surface sediments using such
scanning techniques (e.g., turbidites, hypopycnal plume
deposits, etc.) can be linked to modern records of seismicity, precipitation, or other instrument-derived data sets, and
thus be utilized as chronostratigraphic horizons to supplement other chronometric techniques.
[47] Using a combination of ultra-high-resolution sediment core CT scanning and scanning XRF techniques, as
well as radioisotopic nuclide and organic elemental analyses, a pair of short multicore samples from Baranof Island
was found to record decadal-scale fluctuations in Br/Cl
ratios, which reflect changes in marine organic matter accumulation. The pattern of changes in Br/Cl follows the same
temporal pattern as that of the Pacific Decadal Oscillation,
as well as an estimated Sitka summer productivity anomaly
parameter calibrated with recent SeaWiFS observations.
The correlation of North Pacific climate regime states, primary productivity, and sedimentary Br/Cl ratios indicate
the accumulation of biogenic sediment in Southeast Alaska
temperate ice-free fjords can be used as a sensitive recorder
of past productivity variability, and by inference, climate
conditions in the high-latitude Gulf of Alaska.
[48] Acknowledgments. The authors wish to thank the crew of the R/
V Maurice Ewing and members of science party EW0408, as well as the
Oregon State University core curatorial staff (Bobbi Conard, Mysti Weber,
and Maizet Cheseby). Analytical support from Andrea Krumhardt, the Oregon State University Veterinary Hospital CT Facility (Jason Wiest), the
Alaska Stable Isotope Facility at the University of Alaska Fairbanks, and the
USGS Isotope Tracers program (Carol Kendall, Steve Silva and Doug Choy)
were critical for this study. Rama Kotra and the USGS Mendenhall Postdoctoral Fellowship Program are acknowledged for funding to the lead author.
Discussions with John Barron, Lesleigh Anderson, George Waldbusser, and
Alan Mix greatly improved an early version of this manuscript, as did the
comments of two anonymous reviewers and the editorial guidance of S.
Bradley Moran. Any use of trade, product, or firm names is for descriptive
purposes only and does not imply endorsement by the U.S. Government.
References
Abramoff, M. D., P. J. Magelhaes, and S. J. Ram (2004), Image processing
with ImageJ, Biophotonics Int., 11(7), 36–42.
Addison, J. A. (2009), High-resolution paleoceanography from the Gulf of
Alaska, Subarctic Northeast Pacific Ocean, since the last glacial maximum: Insights into a dynamic atmosphere-ocean-ecosystem linkage at
decadal timescales, Dissertation Thesis, 248 pp., Univ. of Alaska Fairbanks, Fairbanks, Alaska.
Addison, J. A., B. P. Finney, W. E. Dean, M. H. Davies, A. C. Mix, J. S.
Stoner, and J. M. Jaeger (2012), Productivity maxima and sedimentary
15N for the last 17,000 years in the Gulf of Alaska, Paleoceanography,
27, PA1206, doi:10.1029/2011PA002161.
Aitchison, J. (1982), The statistical analysis of compositional data, J. R.
Stat. Soc. Ser. B (Statistical Methodology), 44(2), 139–177.
Aitchison, J. (1986), The Statistical Analysis of Compositional Data, 416
pp., Chapman and Hall, London, U. K.
Appleby, P. G. (2001), Chronostratigraphic techniques in recent sediments,
in Tracking Environmental Change Using Lake Sediments, vol. 1, Basin
Analysis, Coring, and Chronological Techniques, edited by W. M. Last
and J. P. Smol, pp. 171–203, Kluwer, Dordrecht, Netherlands.
Beamish, R. J. (1993), Climate and exceptional fish production off the
West Coast of North America, Can. J. Fish. Aquat. Sci., 50(10), 2270–
2291.
Beamish, R. J., and D. R. Bouillon (1993), Pacific salmon production trends
in relation to climate, Can. J. Fish. Aquat. Sci., 50(5), 1002–1016.
Behrenfeld, M. J., and P. G. Falkowski (1997), Photosynthetic rates
derived from satellite-based chlorophyll concentration, Limnol. Oceanogr., 42, 1–20.
Boespflug, X., N. Ross, B. Long, and J. F. Dumais (1994), Tomodensitometrie axiale: Relation entre l’intensite tomographique et al densite de
la matiere, Can. J. Earth Sci., 31, 426–434.
Boespflug, X., B. F. N. Long, and S. Occhietti (1995), CAT-scan in marine
stratigraphy: A quantitative approach, Mar. Geol., 122, 281–301.
Buesseler, K. O. (1998), The decoupling of production and particulate
export in the surface ocean, Global Biogeochem. Cycles, 12(2), 297–310.
Burrell, D. C. (1989), Biogeochemistry of a Boreal Fjord: Boca de Quadra,
Southeast Alaska, 454 pp., Inst. of Mar. Sci., Univ. of Alaska Fairbanks,
Fairbanks, Alaska.
Calvert, S. E., T. F. Pedersen, and R. E. Karlin (2001), Geochemical and
isotopic evidence for post-glacial palaeoceanographic changes in Saanich Inlet, British Columbia, Mar. Geol., 174(1–4), 287–305.
Chang, A. S., M. A. Bertram, T. Ivanochko, S. E. Calvert, A. Dallimore,
and R. E. Thomson (2013), Annual record of particle fluxes, geochemistry and diatoms in Effingham Inlet, British Columbia, Canada, and the
impact of the 1999 La Nina event, Mar. Geol., 337, 20–34, doi:
10.1016/j.margeo.2013.01.003.
Davis, J. C. (2002), Statistics and Data Analysis in Geology, 3rd ed., 638
pp., John Wiley, N. Y.
De Valpine, P., and J. Harte (2001), Plant responses to experimental
warming in a montane meadow, Ecology, 82(3), 637–648.
Di Giuseppe, E., F. Funiciello, F. Corbi, G. Ranalli, and G. Mojoli (2009),
Gelatins as rock analogs: A systematic study of their rheological and
physical properties, Tectonophysics, 473, 391–403.
Ducassou, E., T. Mulder, S. Migeon, E. Gonthier, A. Murat, M. Revel, L.
Capotondi, S. M. Bernasconi, J. Mascle, and S. Zaragosi (2008), Nile floods
recorded in deep Mediterranean sediments, Quat. Res., 70, 382–391.
Duliu, O. (1999), Computer axial tomography in geosciences: An overview, Earth Sci. Rev., 48, 265–281.
Ebbesmeyer, C. C., D. R. Cayan, D. R. McLain, F. H. Nichols, D. H. Peterson, and K. T. Redmond (1991), 1976 step in Pacific climate: Forty environmental changes between 1968–1975 and 1977–1984, in Proceedings
of the Seventh Annual Pacific Climate (PACLIM) Workshop, April 1990,
edited by J. L. Betancourt and V. L. Tharp, pp. 115–126, Calif. Dep. of
Water Resour., Pacific Grove.
Finney, B. P., I. Gregory-Eaves, J. Sweetman, M. S. V. Douglas, and J. P.
Smol (2000), Impacts of climatic change and fishing on Pacific salmon
abundance over the past 300 years, Science, 290(5492), 795–799.
Francois, R., S. Honjo, R. Krishfield, and S. Manganini (2002), Factors controlling the flux of organic carbon to the bathypelagic zone of the ocean,
Global Biogeochem. Cycles, 16(4), 1087, doi:10.1029/2001GB001722.
Gargett, A. E. (1997), The optimal stability ‘‘window’’: A mechanism
underlying decadal fluctuations in North Pacific salmon stocks?, Fish.
Oceanogr., 6(2), 109–117.
Gilbert, R. (2000), Environmental assessment from the sedimentary record
of high-latitude fiords, Geomorphology, 32, 295–314.
Haight, R. E., G. M. Reid, and N. Weemes (2006), Distribution and Habitats of Marine Fish and Invertebrates in Katlian Bay, Southeastern
Alaska, 1967 and 1968, 60 pp., Alaska Fish. Sci. Center, Natl. Mar. Fish.
Serv., Natl. Oceanic and Atmos. Admin., Juneau, Alaska.
Henson, S. A., R. Sanders, and E. Madsen (2012), Global patterns in efficiency of particulate organic carbon export and transfer to the deep ocean,
Global Biogeochem. Cycles, 26, GB1028, doi:10.1029/2011GB004099.
Hollowed, A. B., S. R. Hare, and W. S. Wooster (2001), Pacific Basin climate variability and patterns of Northeast Pacific marine fish production, Prog. Oceanogr., 49(1–4), 257–282.
Honjo, S., S. J. Manganini, R. A. Krishfield, and R. Francois (2008), Particulate organic carbon fluxes to the ocean interior and factors controlling the biological pump: A synthesis of global sediment trap programs
since 1983, Prog. Oceanogr., 76, 217–285.
3460
ADDISON ET AL.: MODERN SE ALASKA FJORD SEDIMENT RECORDS
International Ocean-Colour Coordinating Group (2000), Remote sensing of
ocean colour in coastal, and other optically-complex, waters, in Reports
of the International Ocean-Colour Coordinating Group, No. 3, edited by
S. Sathyendranath, 140 pp., Dartmouth, Canada.
Kaufman, D., and W. F. Manley (2004), Pleistocene maximum and late
Wisconsinan glacier extents across Alaska, U.S.A., in Quaternary Glaciations—Extent and Chronology, Part II: North America, edited by J.
Ehlers and P. L. Gibbard, pp. 9–27, Elsevier, Amsterdam.
Ketcham, R. A., and W. D. Carlson (2001), Acquisition, optimization and
interpretation of X-ray computed tomographic imagery: Applications to
the geosciences, Comput. Geosci., 27, 381–400.
Kido, Y., T. Koshikawa, and R. Tada (2006), Rapid and quantitative major
element analysis method for wet fine-grained sediments using an XRF
microscanner, Mar. Geol., 229, 209–225.
King, J. R., G. A. McFarlane, and R. J. Beamish (2000), Decadal-scale
patterns in the relative year class success of sablefish (Anoplopoma fimbria), Fish. Oceanogr., 9(1), 62–70.
Latif, M., and T. P. Barnett (1994), Causes of decadal climate variability
over the North Pacific and North America, Science, 266(5185), 634–
637.
Leri, A. C., J. A. Hakala, M. A. Marcus, A. Lanzirotti, C. M. Reddy, and
S. C. B. Myneni (2010), Natural organobromine in marine sediments:
New evidence of biogeochemical Br cycling, Global Biogeochem.
Cycles, 24, GB4017, doi:10.1029/2010GB003794.
Loney, R. A., D. A. Brew, L. J. P. Muffler, and J. S. Pomeroy (1975), Reconnaissance Geologic Map of Chichagof, Baranof, and Kruzof Islands,
Southeastern Alaska, U. S. Gov. Print. Off., Wash., D. C.
Mahn, C. L., and J. M. Gieskes (2001), Halide systematics in comparison
with nutrient distributions in sites 1033B and 1034B, Saanich Inlet:
ODP Leg 169S, Mar. Geol., 174, 323–339.
Mantua, N. J., S. R. Hare, Y. Zhang, J. M. Wallace, and R. C. Francis
(1997), A Pacific interdecadal climate oscillation with impacts on
salmon production, Bull. Am. Meteorol. Soc., 78(6), 1069–1079.
McClain, C. R. (2009), A decade of satellite ocean color observations, Ann.
Rev. Mar. Sci., 1, 19–42, doi:10.1146/annurev.marine.010908.163650.
McQuoid, M. R., M. J. Whiticar, S. E. Calvert, and T. F. Pedersen (2001),
A post-glacial isotope record of primary production and accumulation
in the organic sediments of Saanich Inlet, ODP Leg 169S, Mar. Geol.,
174(1–4), 273–286.
Meyer, M. C., R. Faber, and C. Spotl (2006), The WinGeol lamination
tool: New software for rapid, semi-automated analysis of laminated climate archives, The Holocene, 16(5), 753–761.
Meyers, P. A. (1994), Preservation of elemental and isotopic source identification of sedimentary organic matter, Chem. Geol., 114(3–4), 289–302.
Minobe, S. (1999), Resonance in bidecadal and pentadecadal climate
oscillations over the North Pacific: Role in climatic regime shifts, Geophys. Res. Lett., 26, 855–858.
Molnia, B. F. (2007), Late nineteenth to early twenty-first century behavior
of Alaskan glaciers as indicators of changing regional climate, Global
Planet. Change, 56(1–2), 23–56.
Morris, A. W., and J. P. Riley (1966), The bromide/chlorinity and sulphate/chlorinity ratio in sea water, Deep Sea Res., 13, 699–705.
Mortlock, R. A., and P. N. Froelich (1989), A simple method for the rapid
determination of biogenic opal in pelagic marine sediments, Deep Sea
Res., 36(9), 1415–1426.
Mueter, F. J., C. Broms, K. F. Drinkwater, K. D. Friedland, J. A. Hare, G.
L. Hunt, W. Melle, and M. Taylor (2009), Ecosystem responses to
recent oceanographic variability in high-latitude Northern Hemisphere
ecosystems, Prog. Oceanogr., 81(1–4), 93–110.
Nittrouer, C. A., R. W. Sternberg, R. Carpenter, and J. T. Bennett (1979),
The use of Pb-210 geochronology as a sedimentological tool: Application to the Washington continental shelf, Mar. Geol., 31, 297–316.
Page, R. A. (1973), The Sitka, Alaska, earthquake of 1972: An unexpected
visitor, Earthquake Inform. Bull., US Geol. Survey, 5(5), 4–9.
Pawlowsky-Glahn, V., and J. J. Egozcue (2006), Compositional data and
their analysis: An introduction, in Compositional Data Anlaysis in the
Geosciences: From Theory to Practice, edited by A. Buccianti, G.
Mateu-Figueras, and V. Pawlowsky-Glahn, pp. 1–10, Geol. Soc. of London, London, U. K.
Potts, P. J. (1987), A Handbook of Silicate Rock Analysis, 622 pp., Chapman and Hall, N. Y.
Price, N. B., and S. E. Calvert (1977), The contrasting geochemical behaviours of iodine and bromine in recent sediments from the Namibian
shelf, Geochim. Cosmochim. Acta, 41, 1769–1775.
Reynolds, R. W. (1988), A real-time global sea surface temperature analysis, J. Clim., 1, 75–86.
Reynolds, R. W., N. A. Rayner, T. M. Smith, D. C. Stokes, and W. Q.
Wang (2002), An improved in situ and satellite SST analysis for climate, J. Clim., 15(13), 1609–1625.
Richter, T. O., S. Van Der Gaast, B. Koster, A. Vaars, R. Gieles, H. De
Stigter, H. De Haas, and T. C. E. Van Weering (2006), The Avaatech
XRF Core Scanner: Technical description and applications to NE Atlantic sediments, in New Techniques in Sediment Core Analysis, edited by
R. G. Rothwell, pp. 39–50, Geol. Soc. of London Special Publications,
London, U. K.
Rodionov, S. (2004), A sequential algorithm for testing climate regime
shifts, Geophys. Res. Lett., 31, L09204, doi:10.1029/2004GL019448.
Rodionov, S. (2006), Use of prewhitening in climate regime shift detection, Geophys. Res. Lett., 33, L12707, doi:10.1029/2006GL025904.
Romanovsky, V. E., T. S. Sazonova, V. T. Balobaev, N. I. Shender, and D.
O. Sergueev (2007), Past and recent changes in air and permafrost temperatures in eastern Siberia, Global Planet. Change, 56(3–4), 399–413.
Sarmiento, J. L., and N. Gruber (2006), Ocean Biogeochemical Dynamics,
503 pp., Princeton Univ. Press, Princeton, N. J.
Schell, M. M., and L. J. Ruff (1989), Rupture of a seismic gap in southeastern Alaska: the 1972 Sitka earthquake (Ms 7.6), Phys. Earth Planet.
Inter., 54, 241–257.
Schulz, M., and M. Mudelsee (2002), REDFIT: Estimating red-noise spectra directly from unevenly spaced paleoclimatic time series, Comput.
Geosci., 28, 421–426.
Schwing, F. B., M. O’Farrell, J. M. Steger, and K. Baltz (1996), Coastal
Upwelling Indices West Coast of North America 1946–1995, 32 pp.,
Southwest Fish. Sci. Center, Natl. Mar. Fish. Serv., Natl. Oceanic and
Atmos. Admin., Pacific Grove, Calif.
Sharma, G. D., and B. C. Burbank (1973), Geological oceanography, in
Environmental Studies of Port Valdez, edited by D. W. Hood, W. E.
Shiels, and E. J. Kelley, pp. 15–100, Inst. of Mar. Sci., Univ. of Alaska,
Fairbanks, Alaska.
Skei, J. (1983), Geochemical and sedimentological considerations of a permanently anoxic fjord—Framvaren, South Norway, Sediment. Geol.,
36(2–4), 131–145.
St-Onge, G., and B. F. Long (2009), CAT-scan analysis of sedimentary
sequences: An ultrahigh-resolution paleoclimatic tool, Eng. Geol., 103,
127–133.
Sugai, S. F., M. J. Alperin, and W. S. Reeburgh (1994), Episodic deposition and 137Cs immobility in Skan Bay sediments: A ten-year 210Pb
and 137Cs time series, Mar. Geol., 116, 351–372.
Syvitski, J. P. M., D. C. Burrell, and J. M. Skei (1987), Fjords: Processes
and Products, 379 pp., Springer, N. Y.
Thio-Henestrosa, S., and J. A. Martin-Fernandez (2005), Dealing with compositional data: The freeware CoDaPack, Math. Geol., 37(7), 773–793.
Tjallingii, R., U. Rohl, M. Kolling, and T. Bickert (2007), Influence of the
water content on X-ray fluorescence core-scanning measurements in
soft marine sediments, Geochem. Geophys. Geosyst., 8(2), Q02004,
doi:10.1029/2006gc001393.
Trenberth, K. E. (1990), Recent observed interdecadal climate changes in
the Northern Hemisphere, Bull. Am. Meteorol. Soc., 71(7), 988–993.
Trenberth, K. E., and J. W. Hurrell (1994), Decadal atmosphere-ocean variations in the Pacific, Clim. Dyn., 9(6), 303–319.
Walinsky, S. E., F. G. Prahl, A. C. Mix, B. P. Finney, J. M. Jaeger, and G.
P. Rosen (2009), Distribution and composition of organic matter in surface sediments of coastal southeast Alaska, Cont. Shelf Res., 29(13),
1565–1579, doi:10.1016/j.csr.2009.04.006.
Welte, G. J., and R. Tjallingii (2008), Calibration of XRF core scanners
for quantitative geochemical logging of sediment cores: Theory and
application, Earth Planet. Sci. Lett., 274, 423–438.
Westberry, T., M. J. Behrenfeld, D. A. Siegel, and E. Boss (2008), Carbon-based primary productivity modeling with vertically resolved photoacclimation, Global Biogeochem. Cycles, 22, GB2024, doi:10.1029/
2007GB003078.
Whitney, F. A., W. R. Crawford, and P. Harrison (2005), Physical processes that enhance nutrient transport and primary productivity in the
coastal and open ocean of the subarctic NE Pacific, Deep Sea Res. Part
II, 52(5–6), 681–706.
Ziegler, M., T. Jilbert, G. J. de Lange, L. J. Lourens, and G.-J. Reichart
(2008), Bromine counts from XRF scanning as an estimate of the marine organic carbon content of sediment cores, Geochem. Geophys. Geosyst., 9(5), Q05009, doi:10.1029/2007GC001932.
3461