Abstract
We tested variants of semianalytic algorithms for estimating phytoplankton chlorophyll pigment in Case 2 waters. Since 2002 we sampled 279 stations in 22 estuaries, bays, and near shore at seven National Estuarine Research Reserves between Delaware and Texas (USA). The following median values and ranges were observed: chlorophyll a = 17.4 μg/L3 (0.2-490.1); total suspended solids = 23.4 mg/L dry weight (0.7-191.1); and CDOM absorbance (440 nm) = 3.11 m-1 (0.00-21.08). Spectroradiometers measured volume reflectance at each station. Sampling was designed to capture upriver to coastal mixing gradients. Algorithms utilized features in the red and lower NIR, with interference adjustments for CDOM absorption and non-algal particle scatter using bands in either the green (550 nm) or NIR between 723-739 nm. Data from an additional 53 inland lake, reservoir, and river stations were included for comparison in algorithm testing. Our best two algorithms were re-parameterized using matchups with AISA Eagle imagery. Examples of pigment classification using these adjusted models are presented for five studies: dinoflagellate bloom in Maryland, tidal watershed in Georgia, estuarine reserve and neighboring refinery in Mississippi, shallow ponds important to Whooping Cranes in Texas, and a transect capturing transitions between the Nueces River and Corpus Christi Bay in Texas. These analyses were useful in addressing issues important to coastal management.