Ecosystems at continental margins are most directly affected by human activities and are ideal candidates for the synoptic sampling available through the utilization of satellite radiometry. The introduction of the SeaWiFS ocean color instrument in the autumn of 1997 coincided with the onset of perhaps the largest El Niņo event of the century, providing us with the opportunity to monitor the impact of such events on biological and bio-optical properties in central California. The Monterey Bay Aquarium Research Institute (MBARI) has been monitoring this region using bi-weekly to monthly ship expeditions together with continuous observations from strategically placed moored and drifting platforms since 1989. Satellite observations of sea surface temperature and ocean color provide a synoptic spatial view of physical and biological properties. To monitor the impact of the 1997-98 El Niņo, these activities have been increased to include occupation of CalCoFI line 67, which transects the coast from inside Monterey Bay to approximately 400km offshore. The results from these measurements provide insights into the processes regulating biological production in a coastal upwelling ecosystem during such extreme events, and demonstrate the impact from the severe coastal runoff during peak storm events that may extend hundreds of kilometers offshore, evidenced by an apparent overestimation of chlorophyll using the standard SeaWiFS algorithms. We will present the results of our shipboard, mooring and satellite observations to detail the biological and bio-optical impacts of this extreme El Niņo event coupled with the open ocean.

Monterey Bay is located on the eastern edge of the California Current system, and is a classic example of an eastern boundary current regime. This region exhibits multiple upwelling-relaxation events from about March to September, and this signal typically dominates both the physical and biological processes. The source of this cold (ca. 9.5 C), salty (ca. 33.8 psu) upwelled water is approximately 100-150 m depth, originating in the poleward flowing California Undercurrent. During weak upwelling there is typically a source of nutrient-rich waters to the north of Monterey Bay near Davenport, California, while several days of upwelling favorable (northwesterly) winds produce a series of cold, nutrient rich plumes along the coast. A transect perpendicular to the coast such as CalCoFI line 67 typically crosses a series of these upwelling plumes before moving into the California Current, characterized by warm, low-salinity waters.

This seasonal upwelling cycle brings with it elevated macro- and micronutrient concentrations (values for nitrate and silicate in excess of 30 uM in the surface waters and dissolvable iron concentrations of >10 nM are not uncommon), which feeds a high-biomass, high-productivity diatom-dominated system from about March-October, with a low-biomass picoplankton community present during the winter. Much of this diatom-driven productivity is exported both offshore and downstream from the upwelling centers. During El Nino events, this seasonal cycle is greatly reduced, with less upwelling-favorable conditions and anomalously warm (nutrient depleted) waters upwelled and advected towards shore. Nearshore, however, phytoplankton productivity is relatively unaffected due to the continuing presence of weak upwelling. As a result, it is necessary to move further offshore to fully understand the impact of these events on coastal productivity.

In general, the OC2v2 Algorithm performs well in coastal California (Figure 3). Compared to shipboard measurements, it is apparent that SeaWiFS slightly overpredicts Chl a. This is especially true near-shore, where coastal runoff causes an extreme overestimation (Figure 4).

During February 1998 (Figure 4), this problem becomes even more apparent. Sediment can be seen all along the west coast in the true-color composite, and there is a broad region of elevated chlorophyll extending to almost 400 km off the coast. We propose that these elevated concentrations are caused by the unusually high coastal runoff associated with the El Nino. This represents a much greater influence of riverine input than is typically observed for this region.

OC2v2 results are plotted vs. concurrent shipboard chl measurements for Line 67 in January and March 1998 (above). In all months there is an apparent overestimation near shore--in February, there is also a large offshore region (above right, shaded) of apparently elevated chl values (there were no cruises during this period). The Feb. satellite data is plotted against the mean (+/- SD) for all of the 1998 cruises. There are several possibilities for the elevated chl values offshore: (1) they are real; this seems unlikely since the values are substantially higher than the mean values during all other times of the year. (2) Poor atmospheric correction; this also seems unlikely, since the February images were no worse than other time periods. (3) The severe El Nino-driven runoff caused an overestimation due to elevated CDOM or suspended sediments.

To determine whether the February SeaWiFS image represents the influence of coastal runoff, we can look at Rrs ratios indicative of CDOM and/or pigment packaging (Fig. 6A), and data from the M1 mooring for Spring, 1998 (Fig. 6B). Rrs ratios clearly indicate that Monterey Bay is likely influenced by CDOM. Pigment packaging may be occurring nearshore in the upwelling zone, but offshore where small, nutrient limited cells are prevalent, pigment packaging is less likely. Examination of the mooring data also shows the influence of coastal runoff. Starting in early January, coastal riverine input increased dramatically. Preliminary calculations (G. Friederich, EPOC 1998 poster) indicate that the outflow from SF Bay alone could produce a 400 km2 salinity anomaly per day; shipboard transects during this period showed salinity anomalies up to 300 km from shore. M1 mooring data also indicate that the OC2v2 algorithm deviates from the fluorometer at about the same time as the salinity anomalies become prevalent.

K490(chl)-OC2v2: Although we would prefer to use a component-model to determine the extent of the CDOM/sediment signal, these models are still unproven for general use. As an alternative, we can map the extent of this effect by using K490(chl)-OC2v2. In Case I waters, these algorithms perform almost equally well (tested with the SeaBAM dataset). In contaminated regions, we expect the K490(chl) relationship to exhibit overestimations in the presence of CDOM/sediment. Examination of weekly composites shows that this signal peaks in Feb. during the runoff events, and declines in April-May as runoff decreases, but with a constant overestimate near shore.

MBARI began a series of quarterly cruises in March, 1997 which occupy CalCoFI line 67 out to about 400 km. In 1997, a typical (but somewhat early) upwelling pattern was observed, with a steady gradient in nutrients and productivity from the highly productive near-shore environment out to the Upwelling Front, at approximately 123.75 W longitude. In contrast, the March 1998 transect exhibited a virtual elimination of surface nutrients, accompanied by elevated SST and very low productivity and biomass (Figure 8). Physical data demonstrated a reversal of the typical water bodies, with high salinity waters (from the California Undercurrent) offshore, and low-salinity (sub-Arctic?) waters near shore.

Figure 8. A comparison of March, 1997 and March, 1998 for CalCoFi Line 67. Nitrate, integrated chlorophyll, and integrated production are shown.

One exciting prospect from these cruises is the ability to use a bio-optical parameter, variable fluorescence (Fv/Fm, proportional to the quantum yield of PSII) as a proxy for the f-ratio, as well as an indicator of photo-physiological and nutritional status. There was a good correlation between Fv/Fm and surface f-ratios calculated with traditional 15-N measurements (Figure 9), suggesting that we can extend this relationship to the bio-optical profiles. As with the data presented in the upper left panels, the fluorescence data from 1997 show a steady decrease from onshore to offshore (Figure 10). Somewhat surprisingly, the 1998 data suggest that, aside from a dramatic decrease in health (and f-ratio) near shore, the offshore phytoplankton assemblage is well adapted and maintaining a high f-ratio. This is consistent with both a reversal in the source waters, as described above, and also with a shift in populations during 1998 to a small-cell, low diatom abundance community, which is supported by the available floristics data.

During 1997-1998, the El Nino event which originated in the equatorial Pacific strongly influenced the biological and physical patterns along the west coast of North America. Increased temperatures and thermocline depth combined with decreased upwelling favorable winds caused a suppression of the normally dominant upwelling signal in the Monterey Bay region. At the same time, the dramatic increase in riverine flow caused by increased rains resulted in a dramatic runoff signal, extending as far as 400 km offshore. We propose that a simple relationship between the OC2v2 and K490(chl) algorithms can be used to document the extent of this signal, which we attribute to a combination of CDOM and sediments. Although it is not unusual for riverine runoff to influence coastal waters, this is the first time it has been documented for this region, and the spatial extent of the signal was much larger than anticipated.

We have also demonstrated that a relatively simple measurement, Fv/Fm, can be well-correlated with a biologically important parameter, the f-ratio. During this El Nino event, the Fv/Fm signal suggests that there was a shift in both the normally upwelling-dominated physical (offshore migration of the California Undercurrent) and biological (shift from diatom to picoplankton communities) patterns in this region. These changes may help to explain the time-variant patterns in non-El Nino periods as well, and suggest that the potential influence of near-shore coastal processes may extend to hundreds of kilometers offshore.

Funding was provided by NASA grant NAG5-6563 andNAG5-7632 and by the David and Lucile Packard Foundation through MBARI. We wish to thank our colleagues who provided data from both the time-series and long-line cruises, including T. Pennington, L. Cotton, K. Hoffman, G. Friederich, K. Buck, C. Collins, and R. Dugdale. We also thank R. Michisaki and S. Lyons for help with data analysis and figure preparation.