SeaWiFS retrievals of chlorophyll in Chesapeake Bay and the mid-Atlantic bight

This paper presents three years (1998–2000) of chlorophyll a (chl a) data from the Sea-viewing Wide Field-of-view Sensor (SeaWiFS) for Case 2 waters of Chesapeake Bay and the middle Atlantic bight (MAB) to describe phytoplankton dynamics on seasonal to interannual time scales. We used extensive data on inherent and apparent optical properties in conjunction with satellite retrievals to: (1) characterize the bio-optical properties of the study area relevant to processing and interpreting SeaWiFS data; (2) test the applicability of the SeaWiFS bio-optical algorithm (OC4v.4) for the estuarine and coastal waters; (3) evaluate the accuracy of the SeaWiFS remote sensing reflectance (R_RS) and chl a products on regional and seasonal bases using in situ observations. The characteristically strong absorption by chromophoric dissolved organic matter (a_cdom) and non-pigmented particulate matter (a_d) in estuarine and coastal waters contributed to overestimates of chl a using OC4v.4 applied to in situ radiances for the Bay (mean ratio 1.42±1.20) and the MAB (2.60±1.36). Values of R_RS from SeaWiFS in the blue region of the spectrum were low compared to in situ R_RS, suggesting that uncertainties remain in atmospheric correction. Direct comparisons of SeaWiFS retrievals of chl a with in situ chl a for the Bay showed larger biases and uncertainties (mean ratio 1.97±1.85) than for chl a estimated from OC4v.4 applied to in situ R_RS. The larger biases were attributed to errors in SeaWiFS radiances and the larger uncertainties to time-space “aliasing” of satellite observations and in situ measurements. To reduce the time differences between SeaWiFS and in situ data, we compared chl a obtained from continuous underway fluorometric measurements on selected ship tracks to SeaWiFS chl a and showed that SeaWiFS captured phytoplankton dynamics in much of the Bay. The agreement of SeaWiFS chl a with in situ chl a was strongest in the mid- (regions 3, 4) to lower Bay (regions 1, 2), and deteriorated toward the upper Bay (regions 5, 6), in part due to a reduction of sensitivity and an increase of noise for SeaWiFS products in the highly absorbing, low R_RS waters of the upper Bay. A three-year time-series of SeaWiFS and in situ data showed that SeaWiFS accurately and reliably captured seasonal and interannual variability of chl a associated with variations of freshwater flow. Significant short-term variability of chl a in summer that was unresolved with shipboard data was detected in the SeaWiFS time-series and the implications are discussed. The overall performance of SeaWiFS in the mid- to lower Bay and the MAB, combined with high spatial (～1 km²) and temporal (～100 clear scenes per year) resolution, indicate current SeaWiFS products are valuable for quantifying seasonal to interannual variability of chl a in estuarine and coastal waters.