SEARCH

SEARCH BY CITATION

Keywords:

  • ocean color;
  • SeaWiFS;
  • CDOM;
  • gelbstoff;
  • photochemistry;
  • dissolved organic carbon

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[1] Colored dissolved organic matter (CDOM), also referred to as gelbstoff, gilvin, or yellow matter, has long been known to be an important component of the optical properties of coastal and estuarine environments. However, an understanding of the processes regulating its global distribution and variability, its relationship to the total pool of dissolved organic carbon (DOC), and its influence on light availability remain largely unexplored. Satellite imagery from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) is used to characterize the global distribution of light absorption due to colored detrital and dissolved materials (CDM). The quantity CDM is considered as it is not yet possible to differentiate CDOM and detrital particulate absorption from ocean color spectra on a routine basis. Nonetheless, analysis of an extensive field data set indicates that detrital particulates make only a small contribution to CDM. A comparison of coincident field observations of CDM with SeaWiFS retrievals shows good agreement, indicating that the present procedures perform well. To first order, the basin-scale CDM distribution reflects patterns of wind-driven vertical circulation of the gyres modulated by a meridional trend of increasing CDM toward higher latitudes. The global CDM distribution appears regulated by a coupling of biological, photochemical, and physical oceanographic processes all acting on a local scale, and greater than 50% of blue light absorption is controlled by CDM. Significant differences in both CDM concentration and its contribution to blue light absorption are found spatially among the major ocean basins and temporally on variety of timescales. Significant impacts of riverine discharges can be discerned, although their effects are largely localized. Basin-scale distributions of CDM and DOC are largely unrelated, indicating that CDM is a small and highly variable fraction of the global DOC pool. This first view of the global CDM distribution opens many new doors for the quantification of global marine photoprocesses using satellite ocean color data.

1. Introduction and Motivation

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[2] Concentrations of colored dissolved organic materials (CDOM, also referred to as yellow matter, gilvin, and gelbstoff) are thought to reflect the combination of past productivity of the marine environment and terrestrial organic materials that have washed to sea [e.g., Kalle, 1966; Bricaud et al., 1981]. In a sense, CDOM represents the “wash water” left behind from the “dirty laundry” of past oceanic processes and land-ocean interactions. CDOM plays an important role in the light-induced biogeochemical cycling of many compounds as well as in determining the amount and spectral quality of light available for marine photoprocesses. Hence it is fundamentally important to quantify the CDOM distribution and understand the processes underlying its variability.

[3] Optical oceanographers working in the Baltic Sea first related CDOM to the observed freshwater fraction or salinity of a water mass [e.g., Kalle, 1938; Jerlov, 1953; Højerslev, 1982]. These observations, as well as many recent ones [Blough et al., 1993; DeGrandpre et al., 1996; Del Castillo et al., 1999; Ferrari, 2000], demonstrated that CDOM determinations decrease to nearly undetectable levels as the salinity approaches oceanic values. These results suggest a terrestrial source for CDOM. However, open ocean dissolved organic material (DOM) contains only small amounts of organic molecular markers for terrestrial materials [cf. Meyer-Schulte and Hedges, 1986; Hedges et al., 1997; Opsahl and Benner, 1997]. This implies that open ocean CDOM is derived from marine sources such as the long-term (multiyear) breakdown products of marine productivity [e.g., Kalle, 1966; Bricaud et al., 1981; Hedges et al., 1997]. Together, these results indicate that coastal CDOM distributions are regulated by land-ocean interactions; whereas for open ocean environments, distributions of CDOM will reflect local processes.

[4] Knowledge of the space-time distribution of CDOM is critical for quantifying photochemically regulated global biogeochemical cycles. For example, CDOM absorbs quanta that drive photochemical reactions regulating concentrations of several important radiative gases (compare CO, DMS, and CO2 [Mopper et al., 1991; Miller and Zepp, 1995; Kieber et al., 1996; Zepp et al., 1998]. Similar photoreactions also influence the bioavailability of organic and inorganic substrates that may be used by marine microbes [e.g., Kieber et al., 1989; Moran and Zepp, 1997]. Fluxes of ultraviolet radiation, which affect both primary and bacterial production, are controlled, at least in part, by CDOM [cf. Herndl et al., 1993; Smith and Cullen, 1995]. Substantial changes in CDOM abundance itself occur due to photobleaching [cf. Bricaud et al., 1981; Kouassi and Zika, 1992; Vodacek et al., 1997]. Observations from the surface waters of the Sargasso Sea show large seasonal variations in CDOM indicating regulation by local processes [Siegel and Michaels, 1996; Nelson et al., 1998]. Hence CDOM is a dynamic quantity with important roles in ocean photochemical and photobiological processes.

[5] The goal of this contribution is to assess the global distribution of CDOM and the factors regulating its distribution and variability. Satellite-determined ocean color spectra are decomposed into components of light absorption and backscattering using a semi-analytical ocean color algorithm (see below). The total absorption coefficient, a(λ), is often partitioned into components due to seawater, aw(λ), phytoplankton, aph(λ), CDOM, ag(λ), and detrital particulates, adet(λ), or

  • equation image

where the aw(λ) spectrum is assumed to be known [Pope and Fry, 1997]. Unfortunately, CDOM and detrital particulate absorption spectra both decrease monotonically with increasing wavelength. Hence inverse methods for quantifying ocean color spectra cannot differentiate between these two signals [e.g., Carder et al., 1991]. The combined influence of CDOM and detrital particulate absorption defines the colored detrital material absorption coefficient, acdm(λ) [≡ ag(λ) + adet(λ)]. For convenience sake, the value of acdm(440) will be denoted in the text as CDM.

[6] In this contribution, we present a global analysis of the CDM distribution diagnosed from ocean color observations made from the Sea-viewing Wide-Field-of-view Sensor (SeaWiFS). We introduce a globally optimized ocean color model for determining CDM and implement it using SeaWiFS imagery. The major spatial-temporal patterns of the global CDM fields are then presented and interpreted. The discussion section assesses the role of detrital particulates on the CDM signal, the processes regulating the global CDM distribution, the importance of land-ocean interactions and the relationship between CDM and concentrations of dissolved organic carbon (DOC). The paper concludes by suggesting new research directions possible with this global view of the “dirty water” of the ocean biosphere.

2. An Algorithm for the Global Retrieval of CDM

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[7] A semi-analytical ocean color algorithm is used to decompose SeaWiFS observations of water-leaving radiance into a distinct set of inherent optical properties (IOP) that describe the optical state of the upper ocean [e.g., Garver and Siegel, 1997; Maritorena et al., 2002]. We will refer to this globally optimized model as the GSM01 to reflect the contributions of its primary architects (Garver, Siegel, and Maritorena). The GSM01 algorithm is based upon four basic assumptions: (1) an analytical relationship between water-leaving radiance and IOP is known, (2) seawater IOP values are known, (3) a small number of dissolved and particulate constituents contribute to IOP variations, and (4) the spectral shapes of these constituents are assumed known functions. The application of these assumptions results in an analytical model for normalized water leaving radiance, LwN(λ), with only the magnitudes of the constituent IOP values to be determined. Briefly, the GSM01 ocean color model uses LwN(λ) observations to estimate CDM (≡ acdm(440)), the chlorophyll a concentration, Chl, and the particulate backscatter coefficient at 440 nm, BBP (≡ bbp(440)), or

  • equation image

where the reference wavelength, λo, is 440 nm. In equation (2), values of backscattering and absorption coefficients for pure water, bbw(λ) and aw(λ), optical closure constants, gm, downwelling irradiance at the top-of-the-atmosphere, Fo(λ), the air-sea transmission factor, t, and the real part of the seawater index of refraction, nsw, are all known constants. The parameters to be retrieved, Chl, CDM, and BBP, are underlined in equation (2).

[8] The consistent specification of the other parameters in the model, the chlorophyll specific phytoplankton absorption spectrum, aph*(λ), and the spectral slopes for the particulate backscattering coefficient (η) and the colored detrital materials (S), is more difficult to make. Garver and Siegel [1997] chose these parameters following a sensitivity analysis of field measurements from the Sargasso Sea. They found that the choice of pure water absorption spectrum, aw(λ), and the CDM spectral slope, S, were both critical for this blue water site. In a subsequent application using a global LwN(λ)-pigment data set [O'Reilly et al., 1998], the specification of the S parameter proved critical for low-Chl waters while the choice of aph*(λ) spectrum was important in green waters [Garver, 1997].

[9] These results highlighted the need for a rigorous determination of the IOP shape parameters for global applications. Toward this goal, we developed an optimal “tuning” methodology for semi-analytical ocean color models [Maritorena et al., 2002]. This approach determines, in a least squares sense, optimal values for the parameter vector by comparing model output with an extended version of the O'Reilly et al. [1998] data set. Comparisons of Chl retrievals against the global in situ data set are excellent (r2 = 91.0%; slope = 0.97; and N = 1075), at least as good as the operational SeaWiFS algorithm (r2 = 91.9%; slope = 1.04; and N = 1075). Results for CDM estimates are also excellent (r2 = 87.0%; slope = 1.01; and N = 1075). Complete details concerning the global tuning of the GSM01 algorithm are presented by Maritorena et al. [2002].

3. Implementation

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[10] We apply the globally optimized ocean color model to the global SeaWiFS satellite imagery data set [McClain et al., 1998]. Composite images of normalized water-leaving radiance (LwN(λ); 8 day composite fields sampled at 9 km for wave bands centered on 412, 443, 490, 510, 555, and 670 nm) from the first 3 years of the SeaWiFS satellite mission are used. Known issues with the atmospheric correction procedure in turbid or highly productive waters [e.g., McClain et al., 2000a, 2000b] are avoided by using only those pixels where retrievals of LwN(412) are greater than 0.2 μW cm−2 nm−1 sr−1. This cutoff is based upon the analysis of the globally distributed SIMBIOS field-satellite match-up data set [McClain et al., 2000b] and eliminates about 12% of the pixels that passed the standard (version 3) SeaWiFS data processing procedures. Truncated pixels are typically found in turbid or highly productive coastal waters and may bias the present analysis toward open ocean conditions. Global mean Chl values found using the SeaWiFS operational algorithm (OC4v4) are 0.36 mg m−3 for the entire data set while they are 0.30 mg m−3 after the truncation is applied.

[11] The potential influence of errors in the SeaWiFS atmospheric correction procedure on the present CDM retrievals can be evaluated in several ways. First, a comparison of CDM retrievals found using field and satellite determinations of LwN(λ) from the SIMBIOS match-up data set shows excellent agreement (r2 = 85.9%; slope = 1.03; and N = 134). Second, the time course of CDM retrievals made using SeaWiFS and in situ observations of LwN(λ) show good agreement in both magnitude and temporal pattern for CDM and %CDM for in situ measurements (see below). This indicates that the satellite-derived CDM signal is not adversely affected by procedures used to correct for the atmosphere in SeaWiFS imagery.

[12] The availability of spectrophotometric field data coincident with the SeaWiFS mission provides the opportunity for an end-to-end validation of the CDM estimates. A global data set of spectrophotometric CDM determinations were assembled from available ocean optics databases and observations are available from each ocean basin (although there is a significant bias toward U.S. coastal waters). All observations have been collected following established procedures [Mitchell et al., 2000]. This comparison of field and satellite CDM estimates is independent as none of the field data were used to “tune” the GSM01 model. Field determinations were selected within the 8 day time window for each SeaWIFS GAC image. This match-up procedure is similar to the SIMBIOS satellite-field data match-up procedure [McClain et al., 2000b] except that the time constraint is relaxed (which enabled many more observations to be used, N = 523 versus 32). The resulting end-to-end comparison (Figure 1) shows good consistency between the field and SeaWiFS estimates of CDM (r2 = 64.3% and power law slope = 0.935). Clearly, significant uncertainties exist for both the field and satellite determinations; however, the consistency between these estimates should instill confidence in the global CDM results.

image

Figure 1. Comparison of near-simultaneous SeaWiFS and in situ CDM observations from around the world (N = 523). The match-up procedure is described in the text. The r2 value for the log-transformed CDM estimates is 63.42%, and the fit power law relation is (SeaWiFS CDM) = 0.775(in situ CDM)0.935 (the middle dotted line). The outer dotted lines represent the 95% confidence interval for the fit, and the solid line is the 1:1 line.

Download figure to PowerPoint

4. Global Space/Time Distribution of CDM

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[13] To first order, global patterns of CDM mimic the major gyre systems and other large-scale circulation features of the world ocean (Figure 2). High values of CDM are found within regions of persistent large-scale upwelling (e.g., subarctic gyres, equatorial divergences, eastern boundary currents, etc.) while low values are observed where large-scale downwelling is expected (e.g., subtropical gyres). The basin-scale CDM distribution appears similar to chlorophyll or primary production distributions suggesting that similar processes may regulate CDM. However, a robust statistical relationship is not found between CDM and Chl (r2 = 39.7% on a per pixel basis), indicating that these factors vary independently. It is also apparent that terrestrial inputs of CDOM are not the first order controls on the global CDM distribution (Figure 2).

image

Figure 2. Global patterns of CDM (= acdm(440); upper) and %CDM (lower) for (left) the winter 1998 and (right) the summer 1998. Determinations of %CDM are defined as 100acdm(440)/(acdm(440) + aph(440)), where aph(440) is calculated using the parameterization of Bricaud et al. [1998] and Chl retrievals from the OC4v4 algorithm [O'Reilly et al., 2000]. The winter 1998 distribution is made up of 8 day composite images from December 1997 through February 1998, and the summer 1998 distribution is made up 8 day composite images from June through August 1998.

Download figure to PowerPoint

[14] The contribution made by CDM to light availability can be assessed using the fraction of the non–water absorption coefficient at 440 nm due to CDM, or %CDM. On a global basis, the mean value of %CDM is 51.1% (standard deviation 11.8%), which points to a dominant role for CDM in the absorption of blue light, even at the chlorophyll a absorption peak. The spatial distribution of %CDM shows a broad minimum in the tropical ocean (Figure 2). Values of %CDM increase toward the poles and high-latitude estimates of %CDM are greater for the Northern Hemisphere than the Southern Ocean (Figure 2). Considerably lower %CDM estimates (∼40%) are found for the western Pacific Ocean, especially during the summer of 1998 (Figure 2d). Finally, high values of %CDM are found where and when equatorial upwelling occurs.

[15] Estimates of CDM and %CDM also differ among the major ocean basins (Figure 3). For example, zonal averages of %CDM are considerably higher for the subtropical Atlantic Ocean compared to the subtropical Pacific Ocean. A possible explanation is the large contribution of major rivers (such as the Amazon or Orinoco Rivers) to the North Atlantic Ocean compared to the Pacific Ocean. Values of CDM and %CDM are also higher in the northern Indian Ocean and Arabian Sea, due to high rates of upwelling associated with Monsoon circulations [e.g., Coble et al., 1998]. The clearest waters, in terms of CDM and %CDM, occur in the subtropical Pacific Ocean (Figure 3).

image

Figure 3. Zonal averages of (top) the log10 transformed CDM absorption at 440 nm (m−1; acdm(440)) and (bottom) %CDM (%) for the Pacific (solid line), Atlantic (dashed line), and Indian (dotted line) Ocean Basins.

Download figure to PowerPoint

[16] Significant temporal changes in global CDM absorption and %CDM contribution are observed in zonally averaged latitude-time distributions (Figure 4). At a latitude of 30°N, seasonal changes in the zonally averaged CDM vary by more than a factor of 5, from ∼0.005 m−1 to more than 0.03 m−1. Seasonal cycles are apparent throughout the subtropical gyres and CDM retrievals are reduced in the local summer and larger in the winter (Figure 4a). Similar seasonal changes, though not as striking, are found throughout the record. High-latitude CDM values are considerably greater in the Northern Hemisphere compared to the Southern Hemisphere (Figure 4).

image

Figure 4. Time-latitude distributions of the zonal mean (a) CDM absorption coefficient (m−1) and (b) %CDM (%). Black regions are due to the extreme zenith angles found in each hemisphere's winter.

Download figure to PowerPoint

[17] Similarities and differences in seasonal changes of CDM can be easily observed between two sites from the North Atlantic and North Pacific subtropical gyres (Figures 5a–5d). Satellite observed mean values of CDM and %CDM near Bermuda are significantly higher (by ∼15%) than those observed off Hawaii. Similarly, retrieved CDM values off Bermuda increase by more than a factor of two from summer to spring (Figure 5a), whereas observations from near the Hawaiian Islands show weaker seasonal patterns (Figure 5c). The lower amplitude of the seasonal cycle of this site mirrors lower seasonal changes in primary production and chlorophyll concentration observed for this site compared with those for the Sargasso Sea [e.g., Michaels et al., 2000]. The fair correspondence between the satellite and field retrievals (Figures 5a–5d) implies that potential inaccuracies in atmospheric correction procedures are not unduly influencing the global patterns presented here.

image

Figure 5. Time series of (left) CDM and (right) %CDM for sites in (a) and (b) the Sargasso Sea off Bermuda (BATS), (c) and (d) the subtropical Pacific Ocean near Hawaii (HOT), and (e) and (f) the eastern equatorial Pacific ocean (0°, 155°W). The time series of sea surface temperature (SST, dotted) from the TOGA/TAO mooring at 0°, 155°W is shown in Figure 5e. In Figures 5a–5d, the satellite observations are compared with in situ radiometer estimates of CDM and %CDM (symbols).

Download figure to PowerPoint

[18] Large intraseasonal changes in CDM distribution are also observed. For example, a large CDM pulse is observed during the summer of 1998 in the equatorial Pacific Ocean (0°, 155°W), in response to the transition from El Niño to La Niña conditions (Figure 5e [e.g., Chavez et al., 1999]). After this transition, CDM values increased by ∼50% compared to the previous period. The onset of equatorial upwelling brings CDM-rich subsurface waters to the surface and sustained upwelling produces the elevated %CDM levels.

5. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[19] The present depictions of the global CDM distribution raise several issues concerning the interpretation and implications of the large-scale patterns. These include an understanding of the relative contribution of particulate materials to CDM, the nature of the processes regulating the observed patterns, and the relationship between CDM and the DOM pool. The following sections provide partial resolution of these issues; however, many more questions have arisen along the way that must be addressed in future work.

5.1. Is CDM Dissolved or Particulate?

[20] The interpretation of the global CDM distribution requires knowledge of the relative importance of detrital particulates and dissolved matter to CDM absorption. This is typically achieved in the field where dissolved and particulate absorption are partitioned via a mechanical means (i.e., filtering). Multiyear, upper layer (|z| ≤ 20 m) observations of component absorption spectra from the Sargasso Sea show detrital particulate absorption at 440 nm, adet(440), contributes only 12.2% (standard deviation 10.7% and N = 147) of the CDM signal (update of Nelson et al. [1998]). Blue light absorption is dominated by CDM as the mean value of %CDM is 69.9% (16.1% standard deviation and N = 147). Thus CDOM governs blue light absorption for this site.

[21] This comparison with field data can be extended using a global data set of upper layer, spectrophotometric observations of component absorption spectra (Table 1 (after D. A. Siegel et al., Dominance of colored dissolved organic materials in determining light availability in the sea, manuscript in preparation, 2002, hereinafter referred to as Siegel et al., manuscript in preparation, 2002)).

Table 1. Regional Observations of Relative Component Absorption Coefficientsa
RegionChl, mg m−3%CDM, %adet(440)/CDM, %N
  • a

    Values in parentheses indicate standard deviation estimates.

Sargasso Sea0.09 (0.07)69.9 (16.1)12.2 (10.7)147
Indian Ocean0.21 (0.27)61.4 (16.1)16.1 (14.8)72
Florida and Bahamas shelf region0.33 (0.48)60.9 (12.8)12.4 (9.0)272
Southern Ocean-Polar Front0.88 (0.71)56.9 (14.9)21.8 (14.4)60
California Current1.45 (2.01)52.2 (15)22.7 (14.6)638
Gulf of Maine and Georges Bank1.92 (1.46)57.3 (17.5)17.2 (12.9)403
Global1.15 (1.72)57.2 (16.3)18.3 (13.7)1593

[22] This data set is large (N = 1593) and observations are available from several distinct oceanic regions (Table 1). The data should be consistent as all of the spectrophotometric absorption observations were collected and processed following established protocols [Mitchell et al., 2000]. However, a bias for coastal waters is apparent as the data set mean chlorophyll concentration is 1.15 mg m−3, far greater than typical ocean mean values (∼0.4 mg m−3).

[23] The global component absorption data set supports the satellite results showing a dominance of CDM in blue light absorption as the data set mean %CDM is 57.2% (standard deviation 16.3% and N = 1593; Table 1). Only 18.3% (13.7% standard deviation and N = 1593) of the total CDM signal is due to detrital particles in the global data set (Table 1). There is also a rough trend of decreasing %CDM and increasing detrital particulate contribution to CDM as Chl increases. As the global spectrophotometric data set is biased for regions with higher Chl, the contribution made by adet(440) to CDM is expected to be a substantial overestimate of the actual global mean value (see Siegel et al. (manuscript in preparation, 2002) for details). The present global CDM distribution reflects to first order concentrations of CDOM and the field observations support the present satellite-based results.

5.2. Controls on the Global CDM Distribution

[24] Based upon the observed global patterns of CDM, we hypothesize that a coupling of CDOM photobleaching, biological production, and surface water renewal processes are the factors regulating the global CDM distribution and its change over time. The first line of evidence is that CDM values are low in the tropics and increase dramatically toward the poles (Figures 2, 3, and 4) following the general meridional patterns for incident solar irradiance and mixed layer depth. The combination of low light and deep mixed layers results in low average light doses for CDOM photobleaching for the high-latitude oceans compared with the tropics and subtropics. Hence CDOM should increase toward the poles as is clearly observed in the global distributions (Figures 2, 3, and 4).

[25] The consistency of the global CDM distribution (Figure 2) with characteristics of the large-scale circulation suggests vertical mixing and advective processes play an important role in the distribution of CDM. As shown previously, regions characterized by large-scale upwelling or seasonal convective mixing (compare high-latitude oceans, equatorial divergences, and coastal zones) show elevated CDM; whereas, those regions where large-scale downwelling is known to occur (i.e., subtropical gyres), CDM retrievals are much lower (Figure 2). As photobleaching is the primary sink for near surface CDM, open ocean sites will be in general depleted in CDM within the mixed layer and enriched below [e.g., Nelson and Siegel, 2002]. Hence, regions characterized by a net upward flux of CDOM-rich subsurface waters should have higher mixed layer CDM values consistent with the spatial patterns presented (Figure 2). Evidence supporting the role of vertical transport can be found in observations of elevated values of CDOM absorption below the mixed layer for a variety of stratified, open ocean sites [Nelson and Siegel, 2002]. These include CDOM observations from the Sargasso Sea [Nelson et al., 1998] (see below), the Arabian Sea [Coble et al., 1998], and the Equatorial Pacific [Pegau, 1997]. It is likely that this pattern is found throughout the global ocean although systematic studies are required [Nelson and Siegel, 2002]. Hence, the net vertical transport of subsurface CDM to the mixed layer regulates, in part, the surface layer distribution of CDM.

[26] Observations of a significant seasonal cycle in the global CDM distribution (Figures 4 and 5) provide further evidence supporting the roles of light and mixing on CDM dynamics. Low CDM values are found for midlatitude sites during the summer when the mixed layer is at its shallowest and the incident solar fluxes and stratification are most intense [e.g., Nelson et al., 1998]. Elevated CDM values are found in the fall and winter when deep mixing brings elevated subsurface CDOM concentrations to the surface layer.

[27] All told, interactions among light, mixing, and a yet unspecified production process regulate the global CDM distribution. A cartoon illustrating these processes on near-surface abundances of CDOM is shown in Figure 6. A host of biological processes may be implicated in the production of CDOM, as little is known concerning its dynamics [e.g., Nelson and Siegel, 2002]. These include phytoplankton exudation, zooplankton grazing, bacterial production, and cell lysis by viruses. Nelson et al. [1998] suggested that heterotrophic bacterial productivity is the dominant source of seasonal CDOM production in the Sargasso Sea. Production of CDOM is likely to occur throughout the water column; however, the photobleaching of near-surface CDOM makes it difficult to view this net production from a time course of CDOM stocks [Nelson et al., 1998] (thus the dashed arrow in Figure 6). Little is known about the open ocean production of “long-lived” CDOM [Nelson and Siegel, 2002]. It is likely that CDOM production is not simply related to primary production as suggested by the lack of correlation between observed CDM and chlorophyll concentration. Further, time series observations of CDOM and primary production from the Sargasso Sea (see below) show little correspondence. Hence rates of surface water photobleaching, subsurface water mass renewal, and CDOM production regulate the global CDM distribution and its temporal changes.

image

Figure 6. Hypothesized interactions regulating surface layer concentrations of CDOM in the open ocean. A set of poorly understood biological production processes, the vertical transport of “deep” CDOM into surface waters, and the loss of CDOM due to photobleaching all regulate mixed layer CDOM concentrations. Rates of photobleaching in turn are regulated by the solar flux, mixed layer depth, and mixed layer concentrations of CDOM.

Download figure to PowerPoint

5.3. Role of Land-Ocean Interactions on the Global CDM Distribution

[28] The global CDM distribution appears to be regulated by a local balance among photobleaching, production, and vertical mixing processes. However, upon closer inspection, extensive regions of possible terrestrial influence can be found. For example, the CDM distribution for the tropical Atlantic Ocean and Caribbean Sea for September 1998 shows the obvious influence of the Amazon and Orinoco River outflows (Figure 7). The shape and extent of the Amazon River outflow as mapped out by CDM into the North Equatorial Countercurrent is consistent with previous depictions [e.g., Muller-Karger et al., 1988]. Elevated CDM values of likely marine origin are found in the North African upwelling zone and along the equatorial divergence. Hence coastal CDM distributions will reflect terrestrial input; however open ocean CDM will be regulated by local processes. The determination of the boundary between these two limiting behaviors is a difficult, yet important, problem.

image

Figure 7. Monthly composite image of the CDM distribution for the equatorial Atlantic Ocean and Caribbean Sea for September 1998. Units for the color bar are m−1. The level 3 GAC image S19982441998273.L3.v3 is used.

Download figure to PowerPoint

[29] Some insights into the importance of terrestrial discharges can be made by scaling the contribution made by river inputs to the global CDM budget. Global river inputs of organic carbon to the oceans are ∼0.4 Pg C yr−1 [e.g., Schlesinger and Melack, 1981]. The annual river input of CDM can be estimated by assuming an organic carbon specific CDM absorption coefficient for river discharges (0.33 m−1 (gC m−3)−1) and that ∼70% of terrestrial CDOM is lost, presumably to photobleaching, in estuarine and near-coastal environments before it reaches the open ocean [Vodacek et al., 1997]. The parameter values used here are from the Mid-Atlantic Bight but are roughly consistent with other locations [e.g., Del Castillo et al., 1999; Ferrari, 2000; Blough and Del Vecchio, 2002]. If river CDM inputs are mixed uniformly throughout the oceans, the annual increase in global CDM is 3 × 10−5 m−1. This input is clearly too small to account for observed changes in the CDM distribution (Figures 2, 3, and 4a) and corresponds to a turnover time of nearly 500 years when compared with the global mean CDM (mean = 0.0134 m−1 and standard deviation = 0.0146 m−1). Even if all riverine CDM were contained in the upper 20 m of the water column, the annual increase would be ∼0.005 m−1, which is still too small to account for the observed seasonal variations. However, the global turnover timescale is ∼3 years, which is still too long to account for the observed seasonal and episodic changes in CDM.

[30] The above scaling analysis suggests that terrestrial discharges will have a small, yet nonnegligible, role in the distribution and dynamics of global CDM. Further, it is clear that within some coastal waters terrestrial processes can have dominant control on CDM (Figure 7). An exact partitioning of terrestrial versus marine control on CDM distribution is clearly beyond the scope of the present contribution. However, this does suggest a cause for the differences in mean CDM and %CDM values among the major ocean basins (Figure 3). Riverine inputs into the North Atlantic Ocean are more than 3 times greater than they are for the North Pacific Ocean [e.g., Budyko, 1958] .The North Atlantic Ocean should also be more responsive to terrestrial inputs than the North Pacific Ocean as its surface area is considerably smaller. Hence it seems reasonable that the elevated %CDM contributions observed for the North Atlantic Subtropical Gyre region may be due to higher relative river inputs [e.g., Opsahl and Benner, 1997]. The resolution of this hypothesis is again beyond the scope of the present manuscript although it does provide a path for quantifying coastal ocean-open ocean interactions.

5.4. Relationship Between CDM and DOC

[31] It seems reasonable that a relationship between CDM and concentrations of dissolved organic carbon (DOC) should exist as CDM is composed of a variety of colored organic compounds. Hence it has been proposed that DOC concentrations can be remotely monitored using satellite retrievals of CDM [cf. Hoge et al., 1995; Ferrari, 2000]. However, photobleaching can destroy colored compounds without significantly altering organic carbon content [e.g., Kieber et al., 1989; Mopper et al., 1991]. Thus a simple correspondence between CDM and DOC may not exist for the open ocean.

[32] Time series observations from the western Sargasso Sea show seemingly independent seasonal cycles for CDOM and DOC [e.g., Siegel and Michaels, 1996; Nelson et al., 1998; Hansell and Carlson, 2001; Siegel et al., 2001]. The seasonal cycle of CDOM (here represented as ag(325)) and DOC at the Bermuda Atlantic Time series Study site (BATS; 31.8°N, 64.5°W) can be examined following Figure 8. Deep winter mixing homogenizes the upper 150–300 m of the water column bringing up new nutrients and CDOM from depth and reducing near-surface DOC concentrations (Figure 8). The new nutrients support the spring phytoplankton bloom, increasing upper layer DOC concentrations while upper layer CDOM levels remain roughly constant (Figure 8). As described previously, there is little correspondence between increased rates of primary production and values of CDOM. The onset of summer stratification results in a reduction in mixed layer CDOM absorption (due to photobleaching) while upper layer DOC from late spring remains elevated. Throughout the summer, DOC concentrations at depth decrease due to microbial remineralization (Figure 8). Somewhat surprisingly, CDOM concentrations at depth increase due to a combination of biological processes [e.g., Nelson et al., 1998; Nelson and Siegel, 2002]. The onset of deep mixing in the fall and winter homogenizes the upper layers and the annual cycles of DOC and CDOM start again. In essence, DOC concentrations are elevated in the surface layer during the summer when CDOM is depressed while they are reduced at depth where CDOM absorption increases. All told, these uncoupled annual cycles result in a lack of correspondence between DOC concentrations and CDOM absorption at this site (Figure 8).

image

Figure 8. Time-depth distribution of (a) temperature (°C), (b) concentrations of dissolved organic carbon (DOC; μmol kg−1), and (c) CDOM (ag(325); in units of m−1) measured at the BATS site (32.1°N, 64.5°W). The red lines in Figures 8b and 8c show the mixed layer depth calculated from potential density. Methods used for the determinations are documented in the literature [Michaels and Knap, 1996; Nelson et al., 1998; Hansell and Carlson, 2001].

Download figure to PowerPoint

[33] A similar uncoupling is observed spatially when CDM distributions (Figure 2) are compared with a global climatology of upper layer (0–50 m) DOC concentrations (Figure 9)[after Hansell, 2002]. The DOC climatology is constructed using a global set of DOC observations collected from large-scale hydrographic transect cruises (Table 2) and climatological winter sea surface temperature (SST) distributions [National Oceanic and Atmospheric Administration, 1998]. Linear regression relationships are constructed for each ocean basin (Table 2) and used to construct the map shown in Figure 9. DOC estimates for the Indian and Atlantic sectors of the Southern Ocean, regions for which little data have been collected, were taken from the Pacific basin regression. The regression analysis gives DOC estimates for the Southern Ocean that are consistent with data used in this analysis as well as others [cf. Wiebinga and deBaar, 1998; Doval and Hansell, 2000]. For the Atlantic basin, the regression slope is significantly greater than for the other basins and estimated values of DOC for the colder waters of the North Atlantic Ocean are unrealistically low. Hence DOC values are not evaluated north of the North Atlantic Current. At some locations in the equatorial and South Atlantic Ocean, estimated DOC concentrations exceed 85 μmol L−1. Any value of more than 85 μmol L−1 is set to that value (Table 2), although Thomas et al. [1995] report DOC values up to 97 μmol L−1. Root mean square (rms) errors in the DOC regression relationships are less than 5 μmol L−1 (Table 2). These regression model uncertainties are much less than the 10 μmol L−1 global scale DOC changes that are the present focus.

image

Figure 9. Climatological DOC distribution from a regression analysis based upon wintertime SST values. Concentrations of DOC are in units of μmol L−1. The regression models used are presented in Table 2, and further details may be found in the text.

Download figure to PowerPoint

Table 2. Regression Models Used to Determine Surface DOC From SST
BasinCruise Data Used in Regression ModelLinear Regression Model (DOC, μmol L−1; SST, °C)r2, %RMS Deviation, μmol L−1N
AtlanticA05 (along 24°N) BATS (31.8°N, 64.5°W)if (DOC< 85), then DOC = 3.493(SST) – 9.79 else DOC = 8582.62.6185
IndianI8 (10°–45°S along 80°E)DOC = 0.795(SST) + 48.5880.32.7285
Pacific and Southern OceanP15S (0°–67°S along 170°W) APFZ-I (50°–65°S along 170°W) HOT (22.75°N, 158°W)DOC = 0.993(SST) + 52.0581.95.06319

[34] The highest values of DOC are found in the tropics and DOC concentrations decrease dramatically toward the poles (Figure 9). For high latitudes, DOC concentrations are low as deep waters with lower DOC concentrations are mixed easily to the surface. The midlatitude gyres are statically stable relative to higher latitudes and DOC resistant to immediate microbial mineralization accumulates [e.g., Hansell, 2002]. Differences in community structure between high and low latitudes (e.g., diatom dominance at high latitudes and a mixed assemblage of nanophytoplankton/picophytoplankton at lower and midlatitudes) appear to regulate meridional variations in DOC.

[35] The global DOC distribution (Figure 9) looks nothing like the global CDM distribution. The trend of elevated values of DOC for the tropics and subtropics and low DOC values for higher latitudes is exactly opposite the meridional trend found for CDM (Figure 4a). For example, the lowest CDM retrievals are found within the western Pacific warm pool (Figure 2) while the DOC values are among the highest (Figure 9). However, the DOC and CDM distributions are not related in a simple inverse fashion. Major patterns in the CDM distribution are created by large-scale vertical advective processes (Figure 2), which are not manifest in the estimated DOC distribution (Figure 9). Hence a lack of a correspondence is again found between the global DOC and CDM distributions. Where surface ocean water residence times are long, values of CDM will be low because of photobleaching while DOC concentrations are elevated due to vertical stability and a dominance of picoplankton/nanoplankton. Where the surface ocean is easily and regularly mixed with deeper water, or where upwelling takes place, high-CDM/low-DOC waters are brought to the surface. This drives the imperfect anticorrelation reported here.

[36] This lack of a correspondence between global DOC and CDM distributions is at odds with investigations made in coastal environments [e.g., Vodacek et al., 1997; Del Castillo et al., 1999; Ferrari, 2000]. Simultaneous reductions in both DOC and CDOM are found in typical surface water transects made from estuarine environments into the open ocean. Again, at issue is the extent to which CDOM is controlled by terrestrial processes. Over a limited spatial domain, land-ocean interactions increase both DOC and CDOM concentrations above typical marine levels and a good correspondence between CDOM and DOC should be observed. However, outside of this region, the correspondence will break down.

6. Summary and Conclusions

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[37] Several important conclusions can be made from the present analyses of the global time/space distribution of CDM. First, on a global basis, CDM contributes roughly the same as phytoplankton to blue light absorption (%CDM mean = 51.1% and standard deviation = 11.8%). This is surprising considering that 440 nm is the peak of the chlorophyll a absorption spectrum. For ultraviolet wavelengths, CDM will clearly dominate light absorption and must be considered in the determination of light availability for marine photoprocesses. Second, field observations of component absorption spectra demonstrate that particulate detritus is, on the average, only a small portion of the CDM signal. Hence, remote determinations of CDM will represent, to first order, concentrations of colored dissolved organic materials. Third, patterns of surface layer CDM absorption and DOC appear completely unrelated. This indicates that the processes regulating CDM and DOC variability are decoupled from each other and that only a small part of the total DOC pool is colored. Thus, CDM determinations will be of little direct use in remotely assessing DOC concentrations on a global scale. Last, we show that the global CDM distribution is regulated by local oceanic processes, not by riverine discharges of CDOM-rich waters. The effects of riverine discharges can be seen in many coastal regions; however, control of the global CDM distribution by riverine inputs appears to be the exception rather than the rule. Open ocean CDM distributions appear regulated by photobleaching losses balanced by the vertical transport of “deep” CDOM to the surface ocean. Thus the residence time of surface waters and their integrated exposure to solar energy are the important factors regulating the global CDM distribution.

[38] These global observations provide many new avenues for remotely assessing ocean biological and biogeochemical processes. For example, the light-driven cycling of dissolved organic matter can be quantified using the remotely sensed CDM distributions. This includes the photoproduction and photolysis of radiatively important trace gases such as carbon dioxide, carbon monoxide, carbonyl sulfide, and dimethyl sulfide. Changes in CDOM absorption influence the undersea ultraviolet radiation climate, which may dramatically influence rates of phytoplankton and bacterial production. All told, these satellite observations afford many exciting opportunities for assessing ocean biological and biogeochemical processes.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

[39] Discussions and encouragement from Chuck McClain, André Morel, Jim Yoder, Howard Gordon, Marlon Lewis, Dennis McGillicuddy, and Scott Doney are gratefully acknowledged. We thank Margaret O'Brien for processing data from the Bermuda BioOptics Project, Brian Langston for the satellite-in situ matchup and global IOP analyses, Christine Pequignet for the global surface DOC distribution and Karen Baith for assistance in utilizing SeaDAS. The authors gratefully acknowledge the support of NASA as part of the SeaWiFS and SIMBIOS science teams and the NSF as part of the U.S. JGOFS Bermuda Atlantic Time series Study and the U.S. JGOFS Synthesis and Modeling Program. The SeaWiFS satellite mission is a cooperative venture of the Orbital Sciences Corporation and NASA.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information
  • Blough, N. V., and R. Del Vecchio, Chromophoric DOM in the coastal environment, in Biogeochemistry of Marine Dissolved Organic Matter, edited by D. A. Hansell, and C. A. Carlson, pp. 509546, Academic, San Diego, Calif., 2002.
  • Blough, N. V., O. C. Zafiriou, and J. Bonilla, Optical absorption spectra of waters from the Orinoco River outflow—Terrestrial input of colored organic matter to the Caribbean, J. Geophys. Res., 98, 22712278, 1993.
  • 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, 4353, 1981.
  • Bricaud, A., A. Morel, M. Babin, K. Allali, and H. Claustre, Variations of light absorption by suspended particles with chlorophyll a concentration in oceanic (case 1) waters: Analysis and implications for bio-optical models, J. Geophys. Res., 103, 31,03331,044, 1998.
  • Budyko, M. I., The Heat Balance of Earth's Surface, Trans. N. A. Stepanova, Off. of Tech. Serv., Dep. of Comm., Washington, D. C., 1958.
  • 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, 20,59920,611, 1991.
  • Chavez, F. P., P. G. Strutton, C. E. Friederich, R. A. Feely, G. C. Feldman, D. C. Foley, and M. J. McPhaden, Biological and chemical response of the equatorial Pacific Ocean to the 1997–98 El Nino, Science, 286, 21262131, 1999.
  • Coble, P. G., C. E. Del Castillo, and B. Avril, Distribution and optical properties of CDOM in the Arabian Sea during the 1995 Southwest Monsoon, Deep Sea Res., Part II, 45, 21952223, 1998.
  • Del Castillo, C. E., P. G. Coble, J. M. Morell, J. M. Lopez, and J. E. Corredor, Analysis of the optical properties of the Orinoco River plume by absorption and fluorescence spectroscopy, Mar. Chem., 66, 3551, 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, 22,72722,736, 1996.
  • Doval, M., and D. A. Hansell, Organic carbon and apparent oxygen utilization in the western South Pacific and central Indian Oceans, Mar. Chem., 68, 249264, 2000.
  • Ferrari, G. M., The relationship between chromophoric dissolved organic matter and dissolved organic carbon in the European Atlantic coastal area and in the west Mediterranean Sea (Gulf of Lions), Mar. Chem., 70, 339357, 2000.
  • Garver, S. A., Variability in ocean color observations and their use in the study of upper ocean ecosystem dynamics, Ph.D. diss., Geography, Univ. of Cal., Santa Barbara, 1997.
  • Garver, S. A., and D. A. Siegel, Inherent optical property inversion of ocean color spectra and its biogeochemical interpretation, 1, Time series from the Sargasso Sea, J. Geophys. Res., 102, 18,60718,625, 1997.
  • Hansell, D. A., DOC in the global ocean carbon cycle, in Biogeochemistry of Marine Dissolved Organic Matter, edited by D. A. Hansell, and C. A. Carlson, Academic, pp. 685715, San Diego, Calif., 2002.
  • Hansell, D. A., and C. A. Carlson, Biogeochemistry of total organic carbon and nitrogen in the Sargasso Sea: Control by convective overturn, Deep Sea Res., Part II, 48, 16491667, 2001.
  • Hedges, J. I., R. G. Keil, and R. Benner, What happens to terrestrial organic matter in the ocean? Org. Geochem., 27, 195212, 1997.
  • 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, 717719, 1993.
  • Hoge, F. E., M. E. Williams, R. N. Swift, J. K. Yungel, and A. Vodacek, Satellite retrieval of the absorption coefficient of chromophoric dissolved organic matter in continental margins, J. Geophys. Res., 102, 24,84724,854, 1995.
  • Højerslev, N., Yellow substance in the sea, in The Role of Solar Ultraviolet Radiation in Marine Ecosystems, edited by J. Calkins, pp. 263281, Plenum, New York, 1982.
  • Jerlov, N. G., Influence of suspended and dissolved matter on the transparency of sea water, Tellus, 5, 5965, 1953.
  • Kalle, K., Zum problem der meerwasserfarbe, Ann. Hydrol. Ma. Mitt., 66, 113, 1938.
  • Kalle, K., The problem of gelbstoff in the sea, Oceanogr. Mar. Biol. Annu. Rev., 4, 91104, 1966.
  • Kieber, D. J., J. McDaniel, and K. Mopper, Photochemical source of biological substrates in sea water—Implications for carbon cycling, Nature, 341, 637639, 1989.
  • Kieber, D. J., J. F. Jiao, R. P. Kiene, and T. S. Bates, Impact of dimethylsulfide photochemistry on methyl sulfur cycling in the equatorial Pacific Ocean, J. Geophys. Res., 101, 37153722, 1996.
  • Kouassi, A. M., and R. G. Zika, Light-induced destruction of the absorbance property of dissolved organic matter in seawater, Toxicol. Environ. Chem., 35, 195211, 1992.
  • Maritorena, S., D. A. Siegel, and A. R. Peterson, Optimal tuning of a semianalytical ocean color model for global-scale applications, Appl. Opt., 41, 27052714, 2002.
  • McClain, C. R., M. L. Cleave, G. C. Feldman, W. W. Gregg, S. B. Hooker, and N. Kuring, Science quality SeaWiFS data for global biosphere research, Sea Technol., 39, 1016, 1998.
  • McClain, C. R., E. J. Ainsworth, R. A. Barnes, R. E. Eplee Jr., F. S. Patt, W. D. Robinson, M. Wang, and S. W. Bailey, SeaWiFS postlaunch calibration and validation analyses, part 1, vol. 9, edited by S. B. Hooker, and E. R. Firestone, NASA Tech. Memo., 2000-206892(9), 82 pp., 2000a.
  • McClain, C. R., et al., SeaWiFS postlaunch calibration and validation analyses, part 2, edited by S. B. Hooker, and E. R. Firestone, NASA Tech. Memo., 2000-206892(10), 57 pp., 2000b.
  • Meyer-Schulte, K. J., and J. I. Hedges, Molecular evidence for a terrestrial component of organic matter dissolved in ocean water, Nature, 321, 6163, 1986.
  • Michaels, A. F., and A. H. Knap, Overview of the U.S. JGOFS Bermuda Atlantic Time-series Study and the Hydrostation S program, Deep Sea Res., Part II, 32, 157198, 1996.
  • Michaels, A. F., D. M. Karl, and A. H. Knap, Temporal studies of biogeochemical dynamics in oligotrophic oceans, in The Changing Ocean Carbon Cycle in the Sea, edited by R. Hanson, H. Ducklow, and J. Field, pp. 392413, Cambridge Univ. Press, New York, 2000.
  • Miller, W. L., and R. G. Zepp, Photochemical production of dissolved inorganic carbon from terrestrial organic matter: Significance to the oceanic organic carbon cycle, Geophys. Res. Lett., 22, 417420, 1995.
  • Mitchell, B. G., et al., Determination of spectral absorption coefficients of particles, dissolved material, and phytoplankton for discrete water samples, in Ocean Optics Protocols for Satellite Ocean Color Sensor Validation, Revision 2, NASA Tech. Man. 2000-209966, edited by G. Fargion, and J. Mueller, pp. 125153, SIMBIOS Proj., Goddard Space Flight Center, Natl. Aeronautics and Space Admin., Greenbelt, Md., 2000.
  • 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, 6062, 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, 13071316, 1997.
  • Muller-Karger, F. E., C. R. McClain, and P. L. Richardson, Dispersal of the Amazon's water, Nature, 333, 5659, 1988.
  • Nelson, N. B., and D. A. Siegel, Chromophoric DOM in the open ocean, in Biogeochemistry of Marine Dissolved Organic Matter, edited by D. A. Hansell, and C. A. Carlson, pp. 547578, Academic, San Diego, Calif., 2002.
  • Nelson, N. B., D. A. Siegel, and A. F. Michaels, Seasonal dynamics of colored dissolved material in the Sargasso Sea, Deep Sea Res., Part I, 45, 931957, 1998.
  • National Oceanic and Atmospheric Administration (NOAA), NOAA World Ocean Atlas [CD-ROM], Silver Spring, Md., 1998.
  • Opsahl, S., and R. Benner, Distribution and cycling of terrigenous dissolved organic matter in the ocean, Nature, 386, 480482, 1997.
  • O'Reilly, J., S. Maritorena, B. G. Mitchell, D. A. Siegel, K. L. Carder, M. Kahru, S. A. Garver, and C. R. McClain, Ocean color algorithms for SeaWiFS, J. Geophys. Res., 103, 24,93724,953, 1998.
  • O'Reilly, J. J., et al., SeaWiFS postlaunch calibration and validation analyses, part 3, edited by S. B. Hooker, and E. R. Firestone, NASA Tech. Memo. 2000-206892(11), 49 pp., 2000. (Available as seawifs.gsfc.nasa.gov/SEAWIFS/TECH_REPORTS/PLVol11.tex_typeset.pdf).
  • Pegau, W. S., Distribution of colored dissolved organic matter (CDOM) in the equatorial Pacific, Proc. SPIE Int. Soc. Opt. Eng., 296, 508513, 1997.
  • Pope, R. M., and E. S. Fry, Absorption spectrum (380–700 nm) of pure water, 2, Integrating cavity measurements, Appl. Opt., 36, 87108723, 1997.
  • Schlesinger, W. H., and J. M. Melack, Transport of ocean carbon in the world's rivers, Tellus, 33, 172187, 1981.
  • Siegel, D. A., and A. F. Michaels, Quantification of non-algal light attenuation in the Sargasso Sea: Implications for biogeochemistry and remote sensing, Deep Sea Res., Part II, 43, 321345, 1996.
  • Siegel, D. A., et al., Bio-optical modeling of primary production on regional scales: The Bermuda BioOptics Project, Deep Sea Res., Part II, 48, 18651896, 2001.
  • Smith, R. C., and J. J. Cullen, Effect of UV radiation on phytoplankton, Rev. Geophys., 33, 12111223, 1995.
  • Thomas, C., G. Cauwet, and J.-F. Minster, Dissolved organic carbon in the equatorial Atlantic Ocean, Mar. Chem., 49, 155169, 1995.
  • Vodacek, A., N. V. Blough, M. D. DeGrandpre, E. T. Peltzer, and R. K. Nelson, Seasonal variation of CDOM and DOC in the Middle Atlantic Bight: Terrestrial inputs and photooxidation, Limnol. Oceanogr., 42, 674686, 1997.
  • Wiebinga, C. J., and H. J. W. de Baar, Determination of the distribution of dissolved organic carbon in the Indian sector of the Southern Ocean, Mar. Chem., 61, 185201, 1998.
  • Zepp, R. G., T. V. Callaghan, and D. J. Erickson, Effects of enhanced ultraviolet radiation on biogeochemical cycles, J. Photochem. Photobiol. B, 46, 6982, 1998.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction and Motivation
  4. 2. An Algorithm for the Global Retrieval of CDM
  5. 3. Implementation
  6. 4. Global Space/Time Distribution of CDM
  7. 5. Discussion
  8. 6. Summary and Conclusions
  9. Acknowledgments
  10. References
  11. Supporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.