Validation of SeaWiFS ocean color satellite data using a moored databuoy

[1] The Plymouth Marine Bio-Optical Databuoy (PlyMBODy) was deployed in the western English Channel to measure L-wn over an extended period of time in order to vicariously validate ocean color data from the NASA Sea-Viewing Wide Field-of-View Sensor (SeaWiFS). Several experiments have shown that the buoy had radiometric accuracy similar to standard sea-going optical profiling equipment ( about +/- 8%). There were 15 match-ups of good quality SeaWiFS and PlyMBODy data during 10 months of deployment between May 1997 and September 1998. Differences between PlyMBODy measurements of Lwn and SeaWiFS data ( processed using SeaDAS version 4.1) were found to be spectrally variable and sometimes large. The differences were small in the blue-green and green bands ( SeaWiFS bands 3 - 5, 490, 510, and 555 nm) where the biases were < 3% and the mean absolute differences were similar to 20%. Bands 1, 2, and 6 ( 412, 443, and 670 nm) were significantly underestimated by SeaWiFS: the average differences were 55%, 19%, and 58%, respectively. The mean absolute differences were also much larger in these bands (30-77%). SeaWiFS radiometric measurements were generally sufficiently accurate to estimate chlorophyll-a concentration using OC4v4 to within the mission target of +/- 35% but with a high bias of similar to 8%, though the performance of the algorithm itself was not tested. Bias and scatter of SeaWiFS Lwn measurements relative to PlyMBODy measurements were much poorer when SeaDAS version 3.3p5, rather than version 4.1, was used. The upgrade from SeaDAS 4.0p3 to 4.1 made little difference to the agreement.