A simple numerical model of carbon/nitrogen cycling within the Bering/Chukchi Seas and early field observations of total DOC and chlorophyll biomass in the North Sea first suggested that ~50% of the CZCS color signal, sensed above the 70-m isobath of these sub-polar shelves, might be composed of CDOC (colored dissolved organic carbon). A more complex, spectral bio-optical model of the western English Channel indeed indicated that 39%of the color signal after the spring bloom and 46-76% during the fall overturn was CDOC contamination of Case-II shelf waters. A similar third spectral model, embedded within our more complex food web model of multiple phytoplankton groups, replicated some recent bio-optical observations at the JGOFS time series site in the Sargasso Sea. Our most recent analysis of the BATS data suggests that absorption of light by CDOC at both 412 nm and 442 nm is equivalent to that by pigments for two 30-day periods, after the spring bloom and during the fall overturn, even within Case-I waters around Bermuda! Finally, in the southern Caribbean Sea, away from the influence of the Orinoco River, preliminary model simulations and a 1995-1996 in situ time series in the Cariaco Basin concur in suggesting that past CZCS imagery overestimated seasonal changes of phytoplankton biomass here as well.

In the otherwise oligotrophic Caribbean Sea, distinct seasonal periods of wind-forced upwelling, deposition of Fe-rich Saharan aerosols, and river discharge provide a unique opportunity to examine the persistence of CDOC, provided by 1) physical injection from the aphotic zone, 2) biological release from surface blooms of the cyanophyte, Trichodesmium, and 3) lateral supply from land. In response to elements 1 and 2 of the NSF/NASA SMP AO, the hypothesis will be tested that after injection of new nitrogen, by coastal upwelling during the winter/spring, by nitrogen-fixation in the summer, and by river runoff in the fall, CDOC of distinct physical and biological origins will continue to contaminate the color signals of the southern Caribbean Sea, seen now by the SeaWiFS and OCTS sensors. Using concurrent AVHRR imagery of SST and aerosol optical thickness to distinguish different periods of upwelling and of iron-deposition, the time-dependent introduction of CDOC to surface waters, by both physical supply and phytoplankton loss, will be analyzed in relation to the photolysis there of CDOC and to the changing contamination of SeaWiFS/OCTS images. Validation data of the bio-optical part of the analyses will consist of in-water optical measurements made along the Venezuelan coast, from the mouth of the Orinoco River to the stenohaline Cariaco Basin, as part of a separately funded NASA-SIMBIOS project at USF (F.E. Muller-Karger).

Any analysis of CDOC must be conducted, of course, within the context of temporal and spatial variability of the total DOC stocks, which in turn are only one component of carbon cycling in the sea. Our present time-dependent, 3-d ecological models (driven by wind-forced, baroclinic flow fields with a turbulence closure scheme) consider daily C/N cycling by a minimal set of diatoms, Synechococcus, zooplankton, and bacteria, modulated by ambient temperature and the availability of labile and refractory DOC , light, nitrate, ammonium, and DIC over periods of weeks to an annual cycle of habitat change. A new phytoplankton state variable, with the physiological, buoyancy, and palatability properties of Trichodesmium, would be added to consider in the southern Caribbean Sea 1) the air-sea exchange of carbon dioxide in relation to both local/scatterometer wind fields and the source of new nitrogen, 2) the seasonal and spatial patterns of primary and new production at 4.4 km resolution, in relation to aeolian, riverine, and marine influxes of nutrients, and 3) the in situ color signatures of pigments and CDOC in relation to the total carbon pools of POC and DOC, as measured by F.E. Muller-Karger in a separately funded NSF/CONCIT-CARIACO project on the Venezuelan shelf.