Hydrography of Chromophoric Dissolved
Organic Matter in the North Atlantic
Norman B. Nelson,*1 David A. Siegel,1,2 Craig. A. Carlson,1,3
Chantal Swan,1 and William M. Smethie, Jr.4
* Corresponding Author: norm@icess.ucsb.edu
1: Institute for Computational Earth System Science, Mail Code 3060, University of
California, Santa Barbara, CA 93106
2: Also Department of Geography, University of California, Santa Barbara, CA 93106
3: Also Department of Ecology, Evolution, and Marine Biology, University of California,
Santa Barbara, CA 93106
4: Lamont-Doherty Earth Observatory of Columbia University, Palisades, NY
Draft: February 28, 2006 - DV
Abstract
The distribution and optical absorption characteristics of chromophoric dissolved
organic matter (CDOM) were systematically investigated along three meridional
transects in the North Atlantic Ocean and Caribbean Sea conducted with the 2003
CLIVAR Repeat Hydrography survey. Hydrographic transects covered in aggregate a
latitude range of 5 to 62 degrees north along longitudes 20°W (line A16N, Leg 1), 52°W
(A20), and 66°W (A22). Absorption spectra of samples of filtered seawater were
collected at depth ranges from the surface to ~6000 m, sampling all the ocean water
masses in the western basin of the subtropical North Atlantic and several stations on the
North and South American continental slopes. The lowest surface values of CDOM (<
0.1 m-1 absorption coefficient at 325 nm) were found in the central subtropical gyres
while the highest surface values (~0.7 m-1 absorption coefficient at 325 nm) were found
along the continental shelves and the subpolar gyre, confirming recent satellite-based
assessments of CDOM absorption coefficients. Characteristics of the CDOM absorption
spectrum and the DOC-mass-specific CDOM absorption coefficient show the roles of
solar bleaching and heterotrophic processing on changes in the chemical and optical
characteristics of CDOM. In the ocean interior, CDOM values were relatively high (0.10.2 m-1 absorption coefficient at 325 nm) except in the subtropical mode water, where a
local minimum exists due to the subduction of low CDOM surface waters during
ventilation. In the sub-thermocline water masses of the western basin (Antarctic
Intermediate Water and part of the North Atlantic Deep Water), decreases in CDOM are
correlated with increasing ventilation age as assessed using chlorofluorocarbon (CFC)
concentrations and the atmospheric CFC history. Extrapolation of the age of the overflow
water component of the NADW yields an estimated CDOM value at the time of
ventilation similar to that observed for the subpolar waters where the water mass was
formed. The overall CDOM distribution in the North Atlantic reflects the rapid advection
and mixing processes of the basin and demonstrates that remineralization in the ocean
interior is not a significant sink of CDOM. This supports the potential of CDOM as an
ocean circulation tracer.
Nelson et al. North Atlantic CDOM
Introduction
The chromophoric fraction of dissolved organic matter (CDOM) is ubiquitous in
natural waters (e.g., Kalle, 1938; Jerlov, 1953; Siegel et al., 2002). CDOM is
operationally defined as material that passes 0.2 m filters and absorbs light in the visible
and UV-A regions and is generally quantified by using the optical absorption coefficient
(m-1) at a selected wavelength, such as 325 nm (Nelson and Siegel, 2002). The absorption
spectrum of CDOM in natural waters increases exponentially with decreasing
wavelength, a characteristic of the absorption spectrum that allows it to be distinguished
from the other predominant light absorbing materials in the water column, phytoplankton
pigments (e.g., Siegel and Michaels, 1996; Nelson et al. 1998). This in turn allows the
global surface distribution of CDOM to be assessed from space by ocean color sensors
(Siegel et al. 2002). CDOM regulates the penetration of UV light into the ocean and
mediates photochemical reactions, therefore playing an important role in many
biogeochemical processes in the surface ocean including primary productivity and the
air-sea exchange of radiatively important trace gases (e.g., Mopper et al. 1991; Arrigo
and Brown, 1996; Zepp et al. 1998; Toole and Siegel, 2004). CDOM is also of interest as
a component of the dissolved organic matter (DOM) pool, which represents a significant
ocean carbon reservoir (e.g., Hansell, 2002). Understanding the dynamics of CDOM in
oceanic waters may assist in our understanding of the dynamics of the overall DOM pool,
including its photoremineralization and diagenesis.
It is well known that the abundance and distribution of CDOM in the coastal ocean
are dominated by terrestrial inputs from rivers and runoff, as decomposition of terrestrial
organic matter yields light absorbing compounds such as humic and fulvic acids (Kalle,
1938; Højerslev, 1982; Carder et al. 1989; Blough and Del Vecchio 2002, Del Vecchio
and Blough, 2004), but the open ocean has autochthonous sources and sinks (Nelson et
al. 1998; Nelson and Siegel 2002). Previous research has demonstrated that CDOM is
produced in the open ocean as a result of heterotrophic processes near the surface (Nelson
et al. 1998; 2004, Steinberg et al. 2004) and is destroyed by solar bleaching in stratified
1
Nelson et al. North Atlantic CDOM
surface waters (Vodacek et al. 1997; Nelson et al. 1998; Siegel et al. 2002; 2005; Nelson
et al. 2004; Del Vecchio and Blough 2004). These local sources and sinks are sufficient
to account for seasonal cycles in CDOM observed in satellite data by Siegel et al. (2002).
The optical activity of CDOM is [almost] never completely eliminated by solar bleaching
or other natural processes, suggesting a pool of CDOM that is at least partially resistant to
solar bleaching and microbial degradation. It is logical to assume that this material has
been carried into the intermediate and deep waters by convective processes and may act
as a tracer of ocean circulation and of diagenetic transformations of DOC.
CDOM has been detected in the deep ocean (> 1000m) by fluorescence methods
(Chen and Bada, 1992; Determann et al. 1996) and by absorption spectroscopy from
samples collected at BATS site in the northwestern Sargasso Sea (Nelson unpubl. data).
The "oceanographic" distribution of CDOM in Bermuda Atlantic Time-series Study
(BATS) profiles and in global ocean color imagery (Nelson et al. 1998; Siegel et al. 2002,
2005) suggests that CDOM is transported by large scale processes such as upwelling and
subduction of water masses. Thus CDOM shows potential as a semi-conservative
oceanographic tracer which can be bounded at the surface by remote sensing and
measured in situ using autonomous sensor systems. This distinguishes CDOM from other
oceanographic tracers which must be quantified, even at the surface, by analysis of
discrete water samples.
The overall goal of our study is to test the hypothesis that CDOM can act as a semiconservative tracer. Our main approach is to describe the global distribution of CDOM in
the major ocean basins in conjunction with hydrography and other established tracers.
This work covers much of the North Atlantic on meridional sections conducted in concert
with the CLIVAR/CARBON repeat hydrography section program. Here, we describe the
distribution of CDOM in surface waters of the North Atlantic and its principal deep water
masses, and assess this distribution in the context of hydrographic properties,
chlorofluorocarbon tracers, and optical properties of the CDOM itself. We find patterns
and concentrations of CDOM in surface waters are consistent with their assessment from
2
Nelson et al. North Atlantic CDOM
satellite ocean color imagery. Within the oceanic interior, we find that water mass
ventilation processes and diagenetic processes regulate CDOM concentrations in a
consistent manner throughout the North Atlantic. In all, this is the first consistent
hydrographic survey of CDOM absorption conducted in the open sea and provides new
insights for CDOM cycling and for the cycling of organic matter in general.
Methods and Data
Hydrographic Data
CDOM observations were made in concert with hydrographic measurements on three
transects (Figure 1) covering a latitude range of 5o to 62oN along longitudes of 20oW
(line A16N, Leg 1), 52oW (line A20), and 66oW (line A22). Hydrographic data were
taken with a nominal station spacing of every 50 km and CTD and 36 place bottle trips
were spread throughout the entire water column (Feely et al. 2005). Hydrographic
parameters sampled include temperature, salinity, dissolved oxygen, primary nutrients
(NO3, PO4 & SiO4), inorganic carbon concentrations (nominally pCO2 & DIC),
chlorofluorocarbon species (CFC-11, CFC-12 and CFC113) and dissolved organic carbon
(DOC) concentrations. WOCE standard protocols are used for all hydrographic
measurements. Details of the measurements protocols, cruise narratives and data sets are
available at the Repeat Hydrography Program website (http://ushydro.ucsd.edu).
Computations of neutral density, potential vorticity, partial pressure of CFC-12 (pCFC12), and AOU were performed using Ocean Data View (Schlitzer, 2004).
CDOM Absorption Data
Water samples were prepared for spectrophotometric analysis according to
established methods (Nelson et al. 1998, Nelson et al. 2004). Samples were drawn from
Niskin bottles into acid-washed and Milli-Q (Millipore) rinsed amber glass vials with
Teflon liner caps. The samples were then subjected to filtration using 0.2 m Nuclepore
membrane filters which were pre-conditioned by extraction with Milli-Q water to remove
any possible absorbing contaminants. On the A16N line, the samples were kept
3
Nelson et al. North Atlantic CDOM
refrigerated (not frozen) and shipped on ice to UCSB for subsequent analysis (within 4
weeks). We have found that CDOM samples from the open ocean (i.e. with low
concentrations of CDOM) prepared in this way remain stable, with repeatable absorbance
spectra, for months (Nelson et al. 2004, Nelson and Swan, unpubl. data). On the A20 and
A22 lines samples were analyzed immediately, usually within two hours of collection and
filtration.
Spectrophotometric analysis was performed using the UltraPath 200 cm liquid
waveguide cell (World Precision Instruments, Sarasota, FL, USA; Miller et al. 2003) and
the TIDAS-2 diode array spectrometer (J&M GmBH, FRG). This is a single beam
spectrophotometer, so assessment of absorbance of a sample required two scans: one of a
cell filled with blank solution (Milli-Q) and one of the sample itself. The liquid
waveguide absorption cell has several advantages over conventional spectrophotometers
using 10 cm cuvettes. Small sample volumes are smaller (ca. 15 ml) and analysis of an
individual sample is rapid (< 2 min/sample), allowing us to analyze many more samples
than is practical with a conventional spectrophotometer. Also, the 200 cm path can allow
for greater sensitivity, allowing us to better estimate absorption coefficients for low
CDOM open ocean surface waters (Nelson and Siegel 2002). The single beam design of
the instrument and the optical characteristics of the waveguide limit the accuracy and
precision of the instrument below what would be suggested by the pathlength and the
photometric accuracy of the spectrometer, which has some implications which we will
now describe.
The liquid waveguide cell has transmission properties which vary with seawater
refractive index (d’Sa et al. 1999; Byrne and Kaltenbacher, 2001; Miller et al. 2003).
Higher refractive index solutions such as seawater transmit more photons through the
waveguide than pure water, leading to an apparent optical absorbance less than zero. The
established procedure for correcting this refractive index effect is to prepare a blank
sodium chloride solution with a similar refractive index to the sample, using a
refractometer (Miller et al. 2003). This procedure was impractical for our purpose
because of the large number of samples to be collected and analyzed, so an empirical
method for correction was developed. This takes advantage of the fact that salinity is a
4
Nelson et al. North Atlantic CDOM
main factor controlling the refractive index of seawater as temperature effects are
eliminated by allowing samples to equilibrate before analysis. High-quality salinity
measurements were available for each sample bottle collected.
Artificial seawater medium (ASW) was prepared using HPLC-grade Optima™ water,
precombusted (450°C) MgSO4, NaCl, KBr, KCl, CaCl2, non-combusted MgCl2 and
sterile-filtered NaHCO3. This basal solution contained most of the diversity of ions and
their proportions found in natural seawater (Goldman and McCarthy 1978; McLachlan
1964). A range of ASW salinities was acquired through dilution and the salinity of each
solution determined from the ratio of conductivities measured by a Beckman induction
salinometer (model RS 7B). These solutions were analyzed in the UltraPath system using
Optima™ water as a reference blank, and are presented as optical density (negative base
10 logarithm of the light flux through the reference minus the light flux through the
sample). The refractive index effect increased in magnitude with increasing salinity and
was wavelength-dependent (Fig. 2). The overall curves were linear, and exhibited an
approximate optical density of 0.003 (dimensionless) per salinity unit dependence.
We determined the apparent optical absorbance spectra of filtered seawater samples
at sea. Typically, samples were analyzed in batches of 12 samples, with a blank
determination before and after each batch. This allowed us to track instrument drift (lamp
output or spectrometer sensitivity). The spectrometer collected data at 1 nm intervals with
an effective slit width of 2 nm, and five spectra were collected from each sample over a
two minute period. The five spectra were averaged unless there was a time trend in the
spectra over the two minute interval, in which case the spectrum was excluded from
analysis. The raw spectra were corrected at each wavelength by interpolating the
correction factor on the salinity of the sample (Fig. 2) and subtracting the corresponding
correction spectrum. The resulting dimensionless optical density spectra were converted
to absorption coefficient (m-1) by multiplying the spectrum by 2.303/l, where 2.303
converts decadal logarithmic absorbance to base e, and l is the effective optical
pathlength of the waveguide. For this waveguide the effective optical pathlength was
5
Nelson et al. North Atlantic CDOM
1.943 m and was determined at the factory through a linearity test using a Phenol Red
dye standard buffered solution with an absorbance of 1.0AU cm-1.
CDOM spectra, particularly in the visible waveband (400-700 nm) typically fit a
wavelength dependent exponential function with a single slope parameter, S (nm-1), such
that acdom() = acdom(o)*e-S(-o), where o is a reference wavelength. We computed the
logarithmic slope by fitting the spectrum from 320 to 650 nm to an exponential equation
using a least-squares nonlinear curve fitting approach (Twardowski et al. 2001; Blough
and Del Vecchio 2002). We did not include shorter or longer-wavelength portions of the
spectrum because the shorter wavelength portion of the spectrum does not fit a single
exponential well (Nelson et al. 2004) and the longer wavelength portion is strongly
affected by temperature – dependent absorption by water (Pegau et al. 1997). We
designate the parameter determined this way as Snlf.
Dissolved Organic Carbon Data
Concentrations of dissolved organic carbon (DOC, mol l-1) were determined from
the same Niskin bottle where CDOM samples were drawn. Samples were drawn into acid
leached high density polyethylene bottles and frozen at -20°C for later analysis at UCSB.
Samples collected at depths 1000m and shallower were subjected to inline gravity
filtration through a combusted GF/F filter (polycarbonate filter holders were attached
directly to the Niskin spigots so no additional water handling or apparatus were required).
DOC analyses were performed at shore laboratory at UCSB within 3-9 months of sample
collection. All samples were analyzed via the high temperature combustion technique
using a modified Shimadzu TOC-V analyzer. DOC analyses followed established
methods detailed in Carlson et al. (2004)
Results
Surface Waters
The meridional distribution of CDOM (defined here as dissolved absorption
coefficient at 325 nm, m-1) in samples taken at the surface (depths ≤ 10m) along the three
6
Nelson et al. North Atlantic CDOM
sections is shown in Figure 3. In this figure the triangles () represent measurements
taken in June 2003 along the WOCE A16N line near 20W (Figure 1), open circles (o)
represent measurements taken in September 2003 along the A20 line (52W) and the stars
(*) represent samples taken along the A22 line (66W) in October 2003. The highest
values were found on the continental shelves (N. and S. American) slope waters inshore
of the Gulf Stream, and in the Caribbean (presumably a relict of the Orinoco plume,
Muller-Karger et al. 1989, Siegel et al. 2002). These high CDOM values presumably
represent terrestrial influence, from coastal runoff and rivers, but may also reflect CDOM
resulting from high secondary productivity (Siegel et al. 2002, Nelson et al. 2004,
Steinberg et al. 2004). The lowest values were found in the Sargasso Sea (North Atlantic
subtropical gyre), where stratification and high solar radiation levels leading to bleaching
counters local production of CDOM (Nelson et al. 1998). Intermediate CDOM values
were found in the North Atlantic subpolar gyre waters, where absorption coefficients of
CDOM at the surface were approximately four-fold higher than in the Sargasso Sea
despite the absence of obvious terrestrial influence. This presumably reflects reduced
bleaching due to lower insolation, greater mixing, and higher biological productivity.
Where the sections overlapped latitudes in the subtropical gyre (between 30 and 40N for
all three transects, between 20 and 40N for A20 and A22), surface CDOM abundances
were similar (Figure 3). The basin-scale pattern of CDOM abundance shown in our in
situ measurements corresponds well to satellite views of the CDOM distribution (Siegel
et al. 2002, 2005).
Surface (top 200m) vertical profiles of CDOM absorption coefficient can be sorted
into several distinct patterns. In subpolar waters (62oN), values of CDOM are at their
maximum near the surface (Fig. 4A), declining to approximately half that value in the top
100m. In subtropical waters, the case is reversed: a steep increase in CDOM absorption
coefficient (up to 4x) is observed between the mixed layer and approximately 100m (Fig
4B). These profiles had relatively low resolution in surface waters (ca. 25m bottle
spacing) so the position of the subsurface maximum is approximate as data from BATS
(Nelson et al. 1998) indicate that the peak in the subsurface CDOM profile in subtropical
waters is actually shallower than 100m. Below 100m, values of CDOM decline. Within
7
Nelson et al. North Atlantic CDOM
continental slope and shelf waters the pattern is similar to the open ocean subpolar
profiles, with a larger surface concentration (Fig. 4C, D). Some stations on the
continental shelf and slope exhibited evidence of current or previous stratification and
bleaching (Fig. 4C, D). CDOM profiles in the surface waters are seen to be determined
by the interactions between bleaching and input (either allochthonous or autochthonous).
It is apparent from the subpolar and continental slope vertical profiles that production or
input processes likely occurs at the surface, and CDOM concentration is lower in the
main thermocline and the deep ocean (with the exception of the subtropical gyres).
Contours of CDOM in the top 1000 m of the A20 and A22 sections are shown in
color in Figures 5A and 5C. The corresponding 1000 to 6000 m sections are shown in
Figures 5B and 5D. The A16N section is not shown in contour form due to the sparseness
of the data set. At the surface, the surface lens of low CDOM in the subtropical gyre is
easily seen, as are the high CDOM values found on the continental shelves. In the
Caribbean portion of the A22 section (Fig 5C), surface to 100 m CDOM values are
higher than in the central gyre despite the fact that both have well-stratified water
columns. The feature we consider to be a relict of the Orinoco plume is visible near 15oN
and shallower than 100 m (Fig 4D). The subsurface maximum in subtropical waters is
seen spanning the central gyre in both sections. Below this lies a broad minimum in the
CDOM profile coincident with the subtropical mode water.
Subtropical Mode Water (STMW)
The subtropical mode water (STMW) lies between the seasonal or surface
thermocline and the main thermocline within the subtropical gyres (e.g., Worthington,
1959; Talley and Raymer, 1982). Circulation of STMW in the North Atlantic is presumed
to follow the Gulf Stream recirculation from northeast so southwest (Worthington 1976,
Joyce et al. 2000; Alfutis and Cornillon, 2001). In the Sargasso Sea, STMW is ventilated
during the winter on an occasional (mostly annual) basis in a region north of Bermuda
and south of the Gulf Stream (e.g., Talley and Raymer, 1982). Joyce et al. (2001) have
adopted the 26.4 and 26.6 kg m-3 isopycnals of neutral density (n) as the vertical
boundaries of the STMW in the Sargasso Sea. Figure 5 shows the neutral density
8
Nelson et al. North Atlantic CDOM
boundaries of the STMW overlain as contours on a field of the absorption coefficient of
CDOM at 325 nm (m-1). It can be clearly seen that the STMW represents an intermediate
minimum in the CDOM profile and defines the near-100m subsurface maxima observed
in the profiles (Fig. 4B). The STMW signature is most pronounced in the northern
latitudes of A20 (Fig. 5A) and A22 (Fig. 5C) sections. In both the A20 (Fig. 5A) and
A22 (Fig. 5C) sections the STMW is thickest in the region of 35 N (near where mode
water is ventilated) and thins to the south. Below the mode water the concentration of
CDOM in the main thermocline is roughly equal to that of the subsurface maximum.
Within the STMW the concentration of CDOM is variable, ranging from an
absorption coefficient of less than 0.075 m-1 at 325 nm to over 0.1 m-1. In general CDOM
abundance is greater to the south where the mode water thins, but in both the A20 and
A22 sections there are low CDOM features which span > 200 km interspersed with
higher CDOM features (Fig 5A, C, Fig. 6). We interpret this pattern (except for the local
maximum near 35N, see below) as reflecting variable amounts of CDOM in the mode
water at the time of formation and its subsequent southward advection in the Gulf Stream
recirculation, accompanied by diffusive effects which increase the CDOM in the mode
water (while decreasing PV) in older STMW to the south (e.g., Alfutis and Cornillon,
2001; Palter et al. 2005). This pattern was more pronounced along the A20 section at
52W (Fig. 6A) than it was along the A22 section at 66W (Fig. 6B). Of note is that no
significant correlation between the CDOM sampled in the STMW along these two
sections and the corresponding salinity, potential vorticity, and AOU (data not shown).
Along the A20 section at 35N just south of the Gulf Stream (Figure 5A) there is a
local maximum in CDOM found between 80 and 800 m (spanning the seasonal
thermocline, the STMW and the main thermocline), which is also reflected in the neutral
density contours. We provisionally identify this feature based on the isopycnal
deflections as a cyclonic eddy or cold core Gulf Stream ring, and the higher CDOM in
this location is a consequence of advection or local higher productivity (McNeil et al.
1999, McGillicuddy et al. 1999).
9
Nelson et al. North Atlantic CDOM
Deep Ocean Water Masses
For the purpose of our discussion we consider water masses lying at and below the
main thermocline to be the deep ocean. These include Antarctic Intermediate Water
(AAIW), North Atlantic Deep Water (NADW), and Antarctic Bottom Water (AABW)
(Joyce et al. 2001; Hall et al. 2004). Generally, CDOM (as 325 nm optical absorption
coefficient) increased with depth in the main thermocline and remained at a value of
0.125 to 0.2 m-1 throughout the deep North Atlantic (Figs. 5B and 5D). Deep water
masses have slightly higher concentrations of CDOM to the south which presumably
reflects the recirculation of NADW (Smethie et al. 2000). CDOM is also relatively high
(ca 0.15 m-1) in the deep (>1000m) waters of the Caribbean (Fig. 5D).
For the purpose of analyzing the distribution of CDOM by water mass in the North
Atlantic we have adopted the definitions of Joyce et al. (2001,2003) based upon neutral
density (n) isopycnals. We have condensed several of the layers to simplify the statistical
analyses (Table 1). For the Antarctic intermediate waters (AAIW) we consider two water
masses, corresponding to layer 6 of Joyce et al. (n range from 27 to 27.5 kg m-3) and
layer 7 and 8 together (n range 27.5 to 27.8). For the North Atlantic Deep Water we
consider the Upper Labrador Sea Water (ULSW, layer 10, n from 27.8 to 27.875 kg m-3)
and Labrador Sea Water (LSW, layers 11 and 12, n from 27.875 to 27.975 kg m-3). We
treat Overflow Water (OW) components of the NADW as one (layers 13-15, n from
27.975 to 28.1 kg m-3). Antarctic Bottom Water (AABW, layers 16 and 17) is considered
to be all water with neutral density greater than 28.1 kg m-3.
Within the subtropical gyre or between 15 and 35oN, mean values of CDOM in the
AAIW, NADW and AABW layers were similar on the two sections (Table 2). The deep
ocean mean CDOM values were also higher than the STMW and surface values.
Concentrations of CDOM in the AAIW fell between those of the STMW and the NADW
between 25 and 30oN along 66oW (Fig 5). In all cases the standard deviations were large
and overlapped values on the shallower and deeper layers.
To analyze the spatial distribution of CDOM within the deep ocean water masses, we
calculate estimates of water ventilation age using measurements of the CFC-12
10
Nelson et al. North Atlantic CDOM
concentration. Values of the partial pressure of CFC-12, pCFC-12, were computed as the
quotientof the CFC-12 concentrations measured in bottle samples where CDOM samples
were also collected and the solubility constant which as a function of temperature,
salinity and pressure (Warner and Weiss 1985). Time series of Northern Hemisphere
CFC-12 atmospheric mole fraction (Walker et al. 2000) were used, excluding all points
where the change in CFC concentration over time was zero or less. The ventilation year
was estimated by linear interpolation and the pCFC-12 age was found by subtracting the
ventilation year from the year of collection. We disregarded all values where CFC-12
concentration was less than or equal to 0.03 pmol/kg: this corresponds to a pCFC-12 near
5.5 atm in our data set. Using our lookup table this pCFC-12 yields a ventilation age
estimate of ~54.5 yr. The mole fraction of CFC-12 stopped increasing appreciably in the
atmosphere after 1999, so the ventilation ages of water less than approximately 4 years
are also not possible to estimate using this technique. The practical accuracy of the CFC12 method is approximately 0.03 pmol/kg (Smethie et al., 2000) but the extrapolated
error in pCFC-12 age then varies as a function of the slope of the pCFC-12 vs.
atmospheric concentration curve. Thus, the minimum estimated error in ventilation age is
approximately 0.5 years for pCFC-12 ages less than 40 years. For ages over 40 years the
error in pCFC-12 age based on the accuracy of CFC-12 determination increases to
between 3 and 6 years for ages over 50 years (not shown). CFC-12 measurement
accuracy is not the only source of error in estimating ventilation age from pCFC-12.
Undersaturation of CFC-12 in surface waters at time of ventilation (Fine et al. 2002) and
mixing of ‘newer,’ high-CFC water with ‘older’ low CFC layers above or below
(Smethie et al. 2000, Haine et al. 2003) also perturb age estimates from CFC
concentration. We will discuss the implications of these factors later in this paper.
Distribution of pCFC-12 age in the North Atlantic (Fig. 7) reflects presumed features
of the general circulation (Smethie et al. 2000; Smethie and Fine, 2001). The principal
features of interest to us in the present research are the gradients in pCFC-12 age vs.
latitude within the principal water masses. pCFC-12 ages within the AAIW are higher to
the southern end of the sections (Fig. 7B, 7D). The highest pCFC-12 ages within the
NADW elements are found in the mid-latitudes below the subtropical gyre (Fig. 7B, 7D).
11
Nelson et al. North Atlantic CDOM
Higher ages are also found in the Caribbean sector of the A22 section (Fig 7C, 7D). This
also corresponds to some of the highest deep ocean CDOM values observed.
At first glance these features do not appear to be reproduced in the distribution of
CDOM (Fig 5B, 5D). However, plotting individual points on the CDOM vs. pCFC-12
age axes within each deep ocean layer (excluding the Caribbean) reveals distinct trends
(Fig 8; corresponding regression statistics are found in Table 2). No significant trend is
found within the surface (not shown) and STMW layers (Fig 8A). Within each deep
ocean layer CDOM declines with increasing pCFC-12 age, and in the AAIW and LSW
layers there is a linear trend that is significant at a 95% level of confidence (Fig. 8B-D,
Table 2). The trend within the OW and AABW layers (Fig 8E-F) was not significant at
the 95% confidence interval but there is considerable scatter in the data and the leastsquares linear regression slope was close to zero. We estimated the specific rate of
CDOM decay in each layer (assessed as the regression slope of CDOM absorption
coefficient vs. age divided by the mean value of CDOM within the layer, Table 2) and the
mean absolute temperature within each layer (Fig. 9). The apparent rates of decay
decreased with depth and temperature but the relationship is not suggestive of an
Arrhenius plot (not shown), suggesting that processes other than biological or chemical
diagenesis cause the declines in CDOM concentration. Also, these rates are small: efolding times for CDOM degradation for the water masses shown in Fig. 9 range from
125 to 650 years. Finally, the linear model explains at best 22% of the variance in
CDOM even where the linear relationships are significant (Table 2).
DOC-specific absorption coefficient of CDOM
CDOM chromophores are part of the dissolved organic carbon (DOC) pool, though
are thought to be a rather small fraction (Nelson and Siegel, 2002). In the open ocean,
the relationship between CDOM and DOC is not constant and the surface layer CDOM
vs. DOC relationship will vary with the amount and quality of the terrestrial DOM input
and the integrated effects of solar bleaching for a water parcel (e.g., Vodacek et al. 1997;
Nelson et al. 1998; 2004; Nelson and Siegel 2002; Siegel et al. 2002; Del Vecchio and
Blough, 2004; Conmy et al. 2004). In the ocean interior, changes in this relationship may
reflect, for example, differential consumption of CDOM and DOC by microbes. We can
12
Nelson et al. North Atlantic CDOM
examine the changes in DOC quality induced by these processes by examining the
specific absorption coefficient of CDOM, or the quotient of CDOM absorption
coefficient and DOC concentration, which we will denote a*cdom. Values of DOC-specific
absorption coefficient of CDOM (a*cdom, units of m2g-1) are calculated by dividing the
measured CDOM absorption coefficient at 325 nm by the DOC concentration (g m-3)
measured from the same bottle.
The lowest values of a*cdom were found in surface waters in the subtropical gyre
where a*cdom values are 3 to 4 times lower than values found within the upper thermocline
(Fig. 10A, 10C). The highest a*cdom values near the surface were found on the South
American (Fig. 10A) and North American continental shelves (Fig 10C), but not on the
Grand Banks (Fig. 10A). There high values of a*cdom were found at depths of ~100m.
This suggests that terrestrial-origin CDOM has a higher a*cdom than that produced in situ,
but the station where the Orinoco plume was observed at the surface (Fig. 10C) had a
lower a*cdom value than nearby continental shelf stations. This may reflect partial
bleaching of the CDOM within the plume, assuming the transport time from the estuary
was significant (e.g., Del Castillo et al. 1999).
Within the intermediate and deep waters of the North Atlantic, values of a*cdom are
generally higher than they are in the upper 500 m (Figs. 10B and 10D). The mean value
of a*cdom within the AABW was the highest within the ocean interior (~0.26 m2g-1, Fig.
11A) but this value was not as high as values found on the continental shelves. The
rough pattern of the a*cdom distributions compares well to pCFC-12 age (Fig. 11B), but
scatter plots of a*cdom vs. pCFC-12 age did not reveal significant linear relationships
either within discrete water masses (as in Fig. 8) or for our entire data set (results not
shown). This suggests that a*cdom within a given water mass is determined by the source
of the CDOM and its bleaching state at its origin more than it is by processes that occur
over long periods of time.
Discussion
The observations presented here represent the first systematic survey of the colored
dissolved organic matter distribution of the North Atlantic Ocean and sampling of all
13
Nelson et al. North Atlantic CDOM
water masses from the surface ocean through the deep sea were made. As such, these
data along with the hydrographic observations available from the Repeat Hydrography
surveys provide clues of the processes controlling the oceanic CDOM distribution and its
implications. In the following, we use the present observations to address the processes
controlling the CDOM distribution in the North Atlantic. We then focus on the rates of
CDOM diagenesis that can be estimated from these observations and the transient tracer
age distributions. Last, we end with a discussion of the utility of CDOM as semiconserved tracer of biogeochemical process.
Controls on the CDOM distribution of the North Atlantic
The results of the present study confirm that the basin scale meridional distribution of
CDOM in North Atlantic surface waters (Fig. 3) is as estimated from satellite ocean color
imagery (Siegel et al. 2002, 2005). Highest values of CDOM abundance (absorption
coefficient) at the ocean surface are found along the continental shelves or in river
plumes, while the lowest are found in the permanently stratified subtropical gyres.
Intermediate values are found in surface waters in the more productive waters of the
North Atlantic subarctic gyre. This pattern supports the notion that solar bleaching is the
largest sink of CDOM in surface waters, and that open ocean CDOM distributions require
both terrestrial and open ocean sources. Our assertion that both terrestrial and in situ
sources of CDOM are required stems from our observation of high values of DOCspecific absorption coefficient, a*cdom, in shelf waters close to the continents but not off
the Grand Banks (Fig. 10), where riverine input is lower but biological productivity
remains high. The highest values of a*cdom are found along the continental shelves,
suggesting that terriginous DOM has the highest relative CDOM content, whereas
autochthonous sources result in CDOM with lower a*cdom and bleached CDOM found in
subtropical waters has the lowest.
Abundances of CDOM in the subtropical mode water appear to be governed, to first
order, by the concentration of CDOM in the source area at the time of formation, in the
northwestern Sargasso Sea south of the Gulf Stream. At the BATS site south of the
formation area, CDOM concentrations in the upper mixed layer are at or near the annual
minimum during the time of winter convective mixing (Nelson et al. 1998, 2004). If the
14
Nelson et al. North Atlantic CDOM
same pattern applies in the STMW formation area to the north of BATS, as seems likely,
then the appearance of STMW as a low in the CDOM profile is explained. However,
within the STMW the concentration of CDOM varies with latitude (Figs. 5 and 6),
indicating interannual differences in the concentration of CDOM at the time and place of
mode water formation (e.g., Ebbesmeyer and Lindstrom 1986; Alfutis and Cornillon,
2001).
Within the STMW, no consistent relationship between pCFC-12 age (Fig. 8A),
potential vorticity, remineralized nutrients, and AOU (not shown) is evident. This
suggests that the state of the mode water at the time of formation is more important than
diapycnal mixing (which would increase potential vorticity and CDOM in parallel, e.g.,
Palter et al. 2005), in situ production of CDOM through remineralization processes
(which would increase CDOM with age and correlate with AOU), nor diagenesis of
CDOM within the layer in determining the CDOM concentration within the STMW
along the transects. Environmental conditions which govern the formation of mode water
include the extent and severity of cold-air outbreaks and wind in the formation area
which controls the maximum depth of near-surface mixed layer (Ebbesmeyer and
Lindstrom 1986, Talley 1996, Joyce et al. 2000). Factors which would determine the
CDOM concentration at the time of mode water formation as well as the history of the
mixed layer and irradiance over the time scales appropriate for significant photobleaching
of CDOM (Nelson et al. 1998). It would appear from our present results that the
processes which result in lower PV water parcels during periods of formation are not
correlated with high or low CDOM abundance at the time of STMW ventilation. Further
no significant temporal relationship between indexes of the North Atlantic Oscillation,
ENSO, and CDOM within the STMW was found using pCFC-12 age (results not shown).
These results emphasize the decoupling between the convective processes that renew the
mode water and the biological plus physical processes (synthesis and bleaching) that
control the CDOM profile between renewal events.
CDOM and CFC-12 Age
Consistent slow decay of CDOM was correlated with increases in apparent
ventilation age in several deep water masses (Fig. 8, Table 2). The apparent decay rate of
15
Nelson et al. North Atlantic CDOM
CDOM decreased with depth and water temperature (Fig. 9). But this signature is
unlikely to be entirely due to biological or chemical diagenesis. Mixing of water masses
is a possible explanation for the apparent degradation of CDOM with age. Water masses
in the AAIW are expected to be low in CFC-12 concentration (and thus have large
apparent ages) because of the great distance from the North Atlantic to their source
waters, but in our data set this is not the case. Mixing with higher CFC-12 waters in the
well-ventilated North Atlantic decreases the apparent ventilation age to the range
observed, and may also introduce CDOM from higher CDOM source waters. Conversely,
in the LSW and OW, decreases in the CFC-12 concentration leading to higher age
estimates result from mixing with lower CFC-12 water masses such as AAIW and
AABW. The result in both cases would be a decrease in CDOM abundance with
increasing apparent water mass age, as we observed.
However, alternate explanations to mixing are also possible. Microbial activity may
be slowly remineralizing a portion of the CDOM along with other DOC in the deep
ocean. Another alternate explanation for the apparent temperature dependent degradation
of CDOM involves sorption of DOM onto sinking particles (Druffel et al., 1998). We
have no unambiguous way of discriminating between these mechanisms with our current
dataset. However, in the sub-thermocline water masses there is no significant relationship
between AOU and CDOM abundance (R2 = 0.0003 for the entire set, no significant linear
relationships in any layer). This in turn suggests remineralization is not playing a role in
the deep ocean dynamics of CDOM (production or removal) in the North Atlantic.
Diagenesis of CDOM
The present hydrographic data set available from the CLIVAR/CARBON Repeat
Hydrography program allows us to quantitatively assess changes in the optical and
chemical characteristics of the CDOM pool and how these changes occur relative to
DOC. The dominant transformation of CDOM abundances we observed was the
presumed bleaching of CDOM in the surface layer of the stratified subtropical central
gyre. Bleaching dramatically reduced the absorption coefficient of CDOM in the surface
waters of the subtropical gyres (Figs. 5A and C) and also reduced the DOC-specific
absorption coefficient a*cdom (Fig. 10A and C). This pattern reflects one of two things:
16
Nelson et al. North Atlantic CDOM
either bleaching of CDOM is not necessarily equivalent to photooxidation of DOC to
inorganic carbon species (Nelson et al. 1998, Del Vecchio and Blough, 2004) or the
amount of DOC in the CDOM fraction is so small that significant photooxidation of
CDOM chromophores does not affect DOC concentration in a measurable way (Nelson
and Siegel 2002). We also observed an increase in the spectral slope parameter Snlf that
correlated with decreases in acdom (not shown). High values of the CDOM spectral slope
parameter are often used to diagnose the past history of photobleaching (Green and
Blough, 1994). These reduced a*cdom, and increased Snlf are consistent with trends
observed in the Sargasso Sea (Nelson et al. 1998, 2004), on the North American
continental shelf (Vodacek et al. 1997), and in laboratory studies (Del Vecchio and
Blough 2002).
The pattern of changes in the value of a*cdom and its lack of a relationship with pCFC12 age provide more clues about the diagenetic status of ocean CDOM abundances. For
example, exposure to solar radiation greatly reduces the relative light absorption
properties of DOC while the consumption and/or decomposition of CDOM by aphotic
processes in the deep sea (Fig. 8) produces little change in the DOC-specific absorption
coefficient. The implication here is that CDOM in the deep North Atlantic originates
from surface water masses from higher latitudes that are typically not bleached.
Advection as viewed using the pCFC-12 age estimates appears to be fast enough to
overcome in situ CDOM decomposition (Fig. 8) or in situ sources from other processes,
such as the remineralization of the sinking rain of particulate organic carbon.
Is CDOM as a Semi-Conservative Tracer?
We suggest from our analysis of the distribution of CDOM in the intermediate and
deep waters of the central North Atlantic that CDOM is carried by the deep ocean
circulation into the ocean interior, as is DOC (Hansell and Carlson, 1998). Furthermore,
the relative homogeneity of CDOM in the deep North Atlantic suggests that advection
and mixing dominates the distribution, and subsurface sources of CDOM are not as
important. In the ocean interior (1000m – 3000m), CDOM declines with increasing age
after ventilation as assessed by pCFC-12, with the rates of decline related to temperature
(Fig. 8B-E), but in bottom water masses the rate of decline is negligible (Fig. 8F; see also
17
Nelson et al. North Atlantic CDOM
Table 2). Mixing or sorption onto particles may explain this effect. Despite this apparent
removal, in no water mass observed is CDOM completely eliminated. Can we
nevertheless use CDOM as a passive tracer of ocean circulation?
One test of this hypothesis is possible using the present data set. Portions of the
overflow water (OW) component of the NADW (specifically the Iceland-Scotland
Overflow Water, ISOW) were sampled on the A20 and A22 sections, but the source
water is also found near its sill at ca.1000m along the A16N section near 60N latitude.
Fine et al. (2002) have estimated the impact of mixing and undersaturation at the surface
for components of the NADW, and determined a ‘relic age’ for OW of 18 years. Fine et
al. suggest that the relic age should be subtracted from pCFC-11 ages to produce a more
accurate ventilation age. If we assume a similar relic age for pCFC-12 age and
extrapolate the OW CDOM vs. age line (Fig. 8E) back to zero (18) yr age, we find an
abundance (absorption coefficient at 325 nm) of ca. 0.14 m-1, which is similar to CDOM
found between within the neutral density surfaces delineating OW between 45N and 60N,
where the OW presumably passes through the section (0.134 m-1, n = 6). If processes
other than those parameterized by the linear regression of CDOM vs. age in this layer
were operating, than the concentration of CDOM in the OW along 52W and 66W within
would not necessarily be related to the concentration of CDOM in the OW along 20W by
the same relationship. These results suggest CDOM is in fact useful as a semiconservative tracer within a deep ocean water mass. Furthermore, CDOM is present in
analytical quantities everywhere in the deep ocean we have so far sampled, in contrast to
anthropogenic tracers. Finally, CDOM is quantifiable from space (Siegel et al. 2002,
2005b) so a surface boundary condition can be imposed on any model thus developed. It
remains to diagnose CDOM dynamics in the upper water column in the context of local
production and bleaching (Nelson et al. 1998).
Conclusion
In the present work we have presented a to-date-unique basin scale data set
encompassing CDOM, hydrographic data, and tracers of water mass ventilation at depths
18
Nelson et al. North Atlantic CDOM
ranging from the surface to the bottom. We have confirmed the presence of CDOM in the
deep sea and its circulation in the North Atlantic, and examined the correlation of CDOM
abundance with oceanographic variables. We have also documented diagenesis of
CDOM in the deep sea and have confirmed the distribution patterns of CDOM as
elucidated by satellite ocean color data. In our ongoing research we are extending these
analyses to include other ocean basins and continuing our work on source and sink
processes for CDOM in surface waters.
Acknowledgments
We acknowledge the support of NSF Chemical Oceanography (OCE-0241614) and
NASA Ocean Biology and Biogeochemistry (NAG5-13277) to DAS, NBN, and CAC.
We thank the Repeat Hydrography Program; Rik Wanninkof and Rana Fine, for support.
Jon Klamberg and Stu Goldberg provided essential assistance at sea and in the laboratory.
Thanks also to Chief Scientists John Bullister, John Toole, Terry Joyce, and the captains
and crew of the R/Vs Knorr and Ron Brown. Alexey Mishonov and Wilf Gardner,
TAMU, arranged collection and preparation of samples for us on A16N Leg 1.
Literature Cited
Alfutis, M.A., and P. Cornillon, Annual and interannual changes in the North Atlantic
STMW layer properties. Journal of Physical Oceanography 31, 2066–2086, 2001.
Arrigo, K.R. and C.W. Brown. The impact of chromophoric dissolved organic matter on
UV inhibition of primary productivity in the open ocean. Marine Ecology Progress
Series 140: 207-216, 1996.
Bricaud, A., A. Morel and L. Prieur, Absorption by dissolved organic matter of the sea
(yellow substance) in the UV and visible domains. Limnol. Oceanogr. 26, 43-53, 1981.
Byrne, R., and E. Kaltenbacher. Use of Liquid Core Waveguides For Long Pathlength
Absorbance Spectroscopy; Principles and Practice. Limnology and Oceanography 46
740-742, 2001.
Carder, K. L., Steward, R. G., Harvey, G. R., & Ortner, P. B. Marine humic and fulvic
acids: Their effects on remote sensing of ocean chlorophyll. Limnology and
Oceanography, 34, 68-81, 1989.
Carder, K.L., S.K. Hawes, K.A. Baker, R.C. Smith, R.G. Steward, and B.G. Mitchell,
Reflectance model for quantifying chlorophyll a in the presence of productivity
degradation products. J. Geophys. Res. 96, 20,599-20,611, 1991.
Carlson, C.A., S.J. Giovannoni, D.A. Hansell, S.J. Goldberg, R. Parsons, and K. Vergin.
2004. Interactions between DOC, microbial processes, and community structure in the
mesopelagic zone of the northwestern Sargasso Sea. Limnology and Oceanography 49:
1073-1083.
19
Nelson et al. North Atlantic CDOM
Chen, R.F. and Bada, J.L., The fluorescence of dissolved organic matter in seawater",
Marine Chemistry, 37: 191-221, 1992.
Conmy, R.N., P.G. Coble, R.F. Chen, G.B. Gardner, Optical properties of colored
dissolved organic matter in the Northern Gulf of Mexico. Marine Chemistry, 89, 127144, 2004.
D'Sa, E. J. R. G. Steward, A. Vodacek, N. V. Blough and D. Phinney, Optical absorption
of seawater colored dissolved organic matter determined using a liquid capillary
waveguide Limnology and Oceanography, 44, 1142-1148, 1999.
DeGrandpre, M.D., A. Vodacek, R.K. Nelson, E.J. Bruce and N.V. Blough, Seasonal
seawater optical properties of the U.S. Middle Atlantic Bight. J. Geophys. Res., 101,
22,727-22,736, 1996.
Del Castillo C.E., P.G. Coble, J.M. Morell, J.M. Lopez, J.E. Corredor, Analysis of the
optical properties of the Orinoco River plume by absorption and fluorescence
spectroscopy. Marine Chemistry, 66, 35-51, 1999.
Del Vecchio R., and N.V. Blough, On the origin of the optical properties of humic
substances. Environmental Science and Technology, 38, 3885-3891, 2004.
Del Vecchio R., and N.V. Blough, Spatial and seasonal distribution of chromophoric
dissolved organic matter and dissolved organic carbon in the Middle Atlantic Bight.
Mar. Chem., 89: 169-187, 2004.
Determann, S., Reuter, R., Willkomm, R., Fluorescent matter in the eastern Atlantic
Ocean. Part 2: vertical profiles and relation to water masses. Deep-Sea Research I, 43,
345-360, 1996.
Druffel E.R.M., S. Griffin, J.E. Bauer, D.M. Wolgast, and X.C. Wang, 1998. Distribution
of particulate organic carbon and radiocarbon in the water column from the upper slope
to the abyssal NE Pacific ocean. Deep-Sea Research II, 45: 667-687.
Feely, R. A., L. D. Talley, G. C. Johnson, C. L. Sabine, and R. Wanninkhof. Repeat
hydrography cruises reveal chemical changes in the North Atlantic. EOS, Transactions,
American Geophysical Union, 86,399,404-405, 2005.
Goldman, J. C. and McCarthy, J. J. 1978. Steady state growth and ammonium uptake of a
fast growing marine diatom. Limnol. Oceanogr. 23:695–703.
Green, S.A. and N.V. Blough, Optical absorption and fluorescence properties of
chromophoric dissolved organic matter in natural waters, Limnology and
Oceanography 39, 1903-1916, 1994.
Hall, M. M., T. M. Joyce, R. S. Pickart, W. M. Smethie Jr., and D. J. Torres, Zonal
circulation across 52oW in the North Atlantic, J. Geophys. Res., 109, C11008,
doi:10.1029/2003JC002103, 2004.
Hansell D.A. and C.A. Carlson. 1998. Deep ocean gradients in the concentration of
dissolved organic carbon. Nature 395: 263-266.
Hansell, D.A. 2002. DOC in the global ocean carbon cycle. In Biogeochemistry of
Marine Dissolved Organic Matter, eds. D.A. Hansell and C.A. Carlson, Academic
Press, San Diego. Pp. 685-715.
Herndl, G. J., G. Muller-Niklas, and J. Frick, Major role of ultraviolet-B in controlling
bacterioplankton growth in the surface layer of the ocean, Nature, 361, 717-719, 1993.
Højerslev, N. Yellow substance in the sea in The Role of Solar Ultraviolet Radiation in
Marine Ecosystems. (ed. J. Calkins) 263-281 Plenum Press, New York, 1982.
20
Nelson et al. North Atlantic CDOM
Jerlov, N.G., Influence of suspended and dissolved matter on the transparency of sea
water. Tellus, 5, 59-65, 1953.
Joyce, T.M., C. Deser and M.A. Spall, The relation between decadal variability of
subtropical mode water and the North Atlantic Oscillation. Journal of Climate 13,
2550–2569, 2000.
Joyce, T.M., R.S. Pickart and R.C. Millard Long-term hydrographic changes at 52 and
66oW in the North Atlantic Subtropical Gyre & Caribbean. Deep-Sea Research Part II
46, 245-278, 1999.
Joyce, T. M., Jr., Zonal circulation in the NW Atlantic and Caribbean from a meridional
World Ocean Circulation experiment hydrographic section at 66°W, J. Geophys. Res.,
106(C10), 22,095–22,114, 2001.
Kalle, K., Zum problem der meerwasserfarbe. Annalen der hydrologischen und marinen
mitteilungen 66, 1-13, 1938.
McGillicuddy, D.J., R. Johnson, D.A. Siegel, A.F. Michaels, N.R. Bates, and A.H. Knap.
1999. Mesoscale variations of biogeochemical properties in the Sargasso Sea. J.
Geophys. Res. 104: 13381-13394.
McLachlan, J. 1964. Some considerations of the growth of marine algae in artificial
media. Can. J. Microbiol. 10:769–82.
McNeil J.D., H.W. Jannasch, T. Dickey, D. McGillicuddy, M. Brzezinski, and C.M.
Sakamoto. 1999. New chemical, bio-optical and physical observations of upper ocean
response to the passage of a mesoscale eddy off Bermuda. J. Geophys. Res. 104:1553715548.
Mopper, K., X.L. Zhou, R.J. Kieber, D.J. Kieber, R.J. Sikorski and R.D. Jones,
Photochemical degradation of dissolved organic carbon and its impact on the oceanic
carbon cycle. Nature, 353, 60-62, 1991.
Moran, M.A., and R.G. Zepp, Role of photoreactions in the formation of biologically
labile compounds from dissolved organic matter. Limnol. Oceanogr. 42, 1307-1316,
1997.
Nelson, N.B., and D.A. Siegel, 2002: Chromophoric DOM in the Open Ocean. In:
Biogeochemistry of Marine Dissolved Organic Matter, D.A. Hansell and C.A. Carlson,
eds. p. 547-578, Academic Press, San Diego, CA.
Nelson, N.B., D.A. Siegel, and A.F. Michaels, 1998: Seasonal dynamics of colored
dissolved material in the Sargasso Sea. Deep Sea Research, I, 45, 931-957.
Nelson, N.B., C.A. Carlson, and D.K. Steinberg, Production of chromophoric dissolved
organic matter by Sargasso Sea microbes. Mar. Chem. 89: 273-287, 2004.
Palter, J.B., M. S. Lozier and R.T. Barber, The effect of advection on the nutrient
reservoir in the North Atlantic subtropical gyre. Nature 437, 687-692, 2005.
Pegau, W., D. Gray, and J. Zaneveld, Absorption and attenuation of visible and nearinfrared light in water: dependence on temperature and salinity, Appl. Opt. 36, 60356046, 1997.
Schlitzer, R., 2004. Ocean Data View. http://www.awi-bremerhaven.de/GEO/ODV/.
Siegel, D.A., and A.F. Michaels, 1996: Quantification of non-algal light attenuation in
the Sargasso Sea: Implications for biogeochemistry and remote sensing. Deep-Sea
Research II, 43, 321-345.
21
Nelson et al. North Atlantic CDOM
Siegel, D.A., S. Maritorena, N. B. Nelson, D.A. Hansell and M. Lorenzi-Kayser, 2002:
Global ocean distribution and dynamics of colored dissolved and detrital organic
materials. Journal of Geophysical Research, 107, 3228, DOI: 10.1029/2001JC000965.
Siegel, D.A., S. Maritorena, N.B. Nelson and M.J. Behrenfeld, 2005a: Independence and
interdependencies of global ocean color properties; Reassessing the bio-optical
assumption. Journal of Geophysical Research, 110, C07011,
doi:10.1029/2004JC002527.
Siegel, D.A., S. Maritorena, N.B. Nelson, M.J. Behrenfeld and C.R. McClain, 2005b:
Colored dissolved organic matter and the satellite-based characterization of the ocean
biosphere. Geophysical Research Letters, 32, L20605, doi:10.1029/2005GL024310,
2005.
Smethie, W. M., and R. A. Fine, Rates of North Atlantic deep water formation calculated
from chlorofluorocarbon inventories, Deep-Sea Res., 48, 189– 215, 2001.
Smethie, W. M., R. A. Fine, A. Putzka, and E. P. Jones, Tracing the flow of North
Atlantic deep water using chlorofluorocarbons, J. Geophys. Res., 105, 14,297– 14,323,
2000.
Steinberg, D.K., N.B. Nelson, and C.A. Carlson, Production of chromophoric dissolved
organic matter (CDOM) in the open ocean by zooplankton and the colonial
cyanobacterium Trichodesmium spp. Mar. Ecol. Prog. Ser. 267: 45-56, 2004.
Talley, L.D. and M.E. Raymer, 1982. Eighteen Degree Water variability. J. Mar. Res., 40
(Suppl.), 757-777.
Toole, D.A., and D.A. Siegel, 2004: Light-driven cycling of dimethylsulfide (DMS) in
the Sargasso Sea: closing the loop. Geophysical Research Letters, 31, L09308, DOI:
10.1029/2004GL019581.
Vodacek, A., Blough, N.V., DeGrandpre, M.D., Peltzer, E.T. & Nelson, R.K. Seasonal
variation of CDOM and DOC in the Middle Atlantic Bight: Terrestrial inputs and
photooxidation. Limnol. Oceanogr. 42, 674-686, 1997.
Worthington, L. V. The 18° water in the Sargasso Sea. Deep-sea Res., 5, 297-305, 1959.
Worthington, L. V. On the North Atlantic circulation. The Johns Hopkins Oceanogr.
Stud., 6, 110 pp, 1976.
Zepp, R.G., T.V. Callaghan and D.J. Erickson, Effects of enhanced ultraviolet radiation
on biogeochemical cycles. J. Photochem. Photobiol. B: Biol., 46, 69-82, 1998.
Figure Captions
Figure 1: Distribution of data collected as part of the present study. Symbols indicate
stations where CDOM data was collected at the surface. The symbols used for the A16N
line (▲), the A20 line (o) and the A22 line (*) are used in the figures 3, 4, 6, and 8 to
denote the section on which individual samples were taken.
Figure 2: Salinity dependent correction factors at selected wavelengths used for the
UltraPath liquid waveguide spectrophotometer system in the present study. The
correction factors were determined by using artificial seawater solutions of varying
22
Nelson et al. North Atlantic CDOM
salinity. Correction factors were prepared from apparent optical density spectra (vs.
Milli-Q water) measured at a 1 nm resolution. In the field, correction factors were
interpolated for the exact salinity (based on salinometer analysis of bottle samples from
which the CDOM samples were drawn, and were subtracted from the raw absorbance
spectra to correct for the refractive index effect of seawater vs. fresh water blanks (see
Methods).
Figure 3: Distribution of CDOM (as absorption coefficient at 325 nm, m-1) in surface
water samples (depths 0-5m) collected in the present study, plotted against latitude.
Symbols denote samples from the different sections as shown in Fig. 1.
Figure 4: Selected profiles of CDOM (absorption coefficient at 325 nm, m-1) in the top
1000 m collected from different oceanographic provinces in the North Atlantic in the
summer of 2003. The impact of solar bleaching in stratified conditions can be seen in C),
contrasted with the unbleached surface waters in the subpolar gyre (A) and on the
continental shelf (B) and the very high surface lens of CDOM from the Orinoco plume
(D). Symbols denote the different sections as shown in Fig. 1.
Figure 5: Distribution of CDOM (as 325 nm absorption coefficient, m-1) along 52W
(A20 section, A,B) and 66W (A22 section, C,D) overlaid with the contours of neutral
density corresponding to the water mass definitions shown in Table 1. Note that the
contours are not regular intervals. Panels A and C cover the depth range 0-1000m, and
panels B and D cover the depth range 1000-6000 m for the A20 and A22 sections,
respectively.
Figure 6: Distribution of CDOM (as 325 nm absorption coefficient, m-1) within the
subtropical mode water (defined as neutral density interval between 26.4 and 26.6 kg m-3)
along the A20 line, 52W (A) and the A22 line, 66W (B).
Figure 7: Distribution of pCFC-12 derived age estimate (yr) along the A20 and A22
sections overlaid with neutral density contours, as Fig. 5.
Figure 8: Scatter plots of CDOM (absorption coefficient at 325 nm m-1) vs. pCFC-12 age
(yr) for water masses discussed in the present study (Table 1). A) Subtropical Mode
Water (STMW). B) Upper Antarctic Intermediate Water (uAAIW). C) Lower Antarctic
Intermediate Water (lAAIW). D) Labrador Sea Water (LSW). E) Overflow Water (OW).
F) Antarctic Bottom Water (AABW). Symbols indicate samples taken from the A20 (o)
and A22 (*) lines. Overlaid lines are least-squares linear regression lines (Table 2).
Significant relationships at the 95% confidence level (P < 0.025, two-tailed) were found
in the two AAIW and in the LSW layers. In the OW and AABW no significant linear
distribution was observed with a slope different from zero (Table 2). In the surface and
STMW water masses no relationship between CDOM and age was observed.
Figure 9: Pseudo-first-order decay constants (yr-1), estimated as the regression slopes of
CDOM vs. age, m-1yr-1 (Fig. 8) divided by the mean CDOM absorption coefficient (m-1)
23
Nelson et al. North Atlantic CDOM
for each layer (Table 2), plotted against mean absolute temperature within the layer in
question.
Figure 10: Distribution of a*cdom (m2mg-1), the DOC-specific absorption coefficient of
CDOM at 325 nm along the A20 (A) and A22 (B) sections overlaid with neutral density
contours, as in Fig. 5 and Fig. 7.
Figure 11: A) Mean depth profile for a*cdom (m2mg-1) averaged within the various layers
(Table 1). B) Mean values of a*cdom (m2mg-1) plotted vs. mean pCFC-12 age (yr) within
each layer.
24
Nelson et al. North Atlantic CDOM
Table 1: Water mass layer definitions in terms of neutral density intervals (based on
Joyce et al. 2001)
Layer
Surface
Subtropical Mode Water
Upper Antarctic
Intermediate Water
Lower Antarctic
Intermediate Water
Labrador Sea Water
Overflow Water
Antarctic Bottom Water
Abbreviation
SURF
STMW
uAAIW
Upper n(kg m-3)
18
26.4
27
Lower n(kg m-3)
25
26.6
27.5
lAAIW
27.5
27.8
LSW
OW
AABW
27.8
27.975
28.1
27.975
28.1
29
25
Nelson et al. North Atlantic CDOM
Table 2: Regression statistics for CDOM concentration vs. pCFC-12 Age (Fig. 8), by
layer, as in Table 1, for subtropical mode water and deeper layers.
Layer
STMW
uAAIW
lAAIW
LSW
OW
AABW
Mean acdom
(325 nm, m-1)
0.1117
0.1249
0.1295
0.1272
0.1254
0.1333
Mean acdom
A20 52W
0.1097
0.1261
0.1254
0.1244
0.1204
0.1326
Mean acdom
A22 66W
0.1127
0.1242
0.1320
0.1287
0.1281
0.1336
Mean T
(C)
17.59
10.33
6.31
4.24
3.05
2.25
26
CDOM Decay
(m-1yr-1)
+0.0025
-0.0010
-0.0009
-0.0007
-0.0006
-0.0002
n
R2
t-test
86
124
84
145
103
141
0.05
0.091
0.228
0.142
0.079
0.006
N.S.
P < 0.025
P < 0.025
P < 0.025
N.S.
N.S.
Nelson et al. North Atlantic CDOM
Fig. 1
27
Nelson et al. North Atlantic CDOM
Fig. 2
28
Nelson et al. North Atlantic CDOM
0.7
0.6
acdom (325nm, m-1)
S. American Continental Shelf
A16N, 20W, Jun 2003
A20, 52W, Sep 2003
A22, 66W, Oct 2003
Orinoco Plume, Caribbean
0.5
N. American Continental Shelf
0.4
0.3
Subpolar Gyre
0.2
Sargasso Sea
0.1
0
60
50
40
30
N Latitude
Fig. 3
29
20
10
Nelson et al. North Atlantic CDOM
Subpolar (6220'N)
Continental Slope (3942'N)
-200
-200
-400
-400
z (db)
0
z (db)
0
-600
-600
-800
-800
-1000
-1000
0
0.1
0.2
0.3
0.4
0.5
0
Subtropical (2934'N)
0.1
0.2
0.3
0.4
0.5
Orinoco Plume (1422'N)
0
0
-200
-200
-400
-400
-600
-600
-800
-800
-1000
-1000
0
0.1
0.2
0.3
0.4
0.5
0
acdom (325nm, m-1)
0.1
0.2
0.3
acdom (325nm, m-1)
Fig. 4
30
0.4
0.5
Nelson et al. North Atlantic CDOM
Fig. 5
31
Nelson et al. North Atlantic CDOM
0.15
A) A20 52 W
acdom (325 nm, m-1)
0.1
0.05
0.15
B) A22 66 W
0.1
0.05
40
35
30
N Latitude
Fig. 6
32
25
20
Nelson et al. North Atlantic CDOM
Fig. 7
33
Nelson et al. North Atlantic CDOM
0.3
(325 nm, m-1)
A20
A22
0.2
0.2
0.1
0.1
0
0.3
5
10
15
20
25
B) uAAIW
10
0.3
C) lAAIW
0.2
0.2
0.1
0.1
20
30
40
50
60
30
40
50
60
30
40
50
60
D) LSW
a
cdom
0.3
A) STMW
10
0.3
20
30
40
50
60
10
0.3
E) OW
0.2
0.2
0.1
0.1
10
20
30
40
50
60
F) AABW
10
pCFC-12 Age (yr)
Fig. 8
34
20
20
Nelson et al. North Atlantic CDOM
-3
x 10
8
uAAIW
lAAIW
CDOM Specific Decay k (yr-1)
7
6
LSW
5
OW
4
3
2
AABW
1
2
4
6
8
Mean Layer Temperature (C)
Fig. 9
35
10
12
Nelson et al. North Atlantic CDOM
Fig. 10
36
Nelson et al. North Atlantic CDOM
A)
37
Nelson et al. North Atlantic CDOM
Fig. 11
38