GROUND-TRUTHING SATELLITE CHLOROPHYLL CONCENTRATION ALGORITHMS IN LAKE ERIE

Measurements of chlorophyll concentration obtained from water samples are used to evaluate the effectiveness of four satellite chlorophyll concentration algorithms in Lake Erie. These bio-optical algorithms estimate near-surface chlorophyll concentration using satellite measurements of reflected radiation at several wavebands within the visible part of the spectrum. The goals of this study were to determine which of the algorithms, if any, are best suited to Lake Erie, and to determine whether any of algorithms were uniformly valid over all areas (or all bio-optical regimes) of the lake. Each of the four algorithms considered (OC2, OC4, Coastal Chlorophyll, and Morel-3) was applied to data collected between 1998 and 2002 by the Sea-viewing Wide Field of view Sensor (SeaWiFS) satellite. The in-water data used for ground-truthing included chlorophyll concentration measurements from 30 hydrographic stations distributed over the Western, Central and Eastern Basins of the lake. Some of these stations were occupied multiple times over the period 1998-2002. The effectiveness of each algorithm in the Lake Erie environment was assessed by statistical comparisons of satellite-derived estimates of chlorophyll concentration with spatially and temporally co-located measurements of chlorophyll concentration from the shallowest in-water observation on each sampling cast. In the Central and Eastern Basins, correlations between the satellite and in situ estimates were acceptable, as compared with validation results for open ocean environments. In these basins, the OC2 algorithm had the best overall performance. None of the algorithms performed well in the relatively turbid and shallow Western Basin. Correlations, biases and root-mean-square error between the in-situ observations and satellite-derived estimates from the four algorithms will be discussed for Lake Erie as a whole and for individual basins within the lake.