Matchup Comparisons: Processing Details |
|---|
The l2gen comparisons were done in three parts using SeaWiFS high-resolution (1km) scenes of the Northeast US from Sept 1997 - Jan 1999: 1) Four months of data were processed using the jun_21_99 test version of l2gen and two methods (the STANDARD and 68NIR, a.k.a. Siegal) to create monthly composite images (January has 28 scenes; April, 42; and July, 22 are shown here).Click here for monthly composite results.
2) Seven problem scenes were processed using the aug_31_99 version of l2gen by three methods (STANDARD, 68NIR, and STUMPF-ARNONE, with 412 iteration).Click here for problem scene results.
3) Nineteen typical scenes were processed using the aug_31_99 version of l2gen by three methods (STANDARD, 68NIR, and 78NIR, a.k.a. Siegal). 3) Twenty test scenes were processed using the new URI/NOAA Aeroplus processing scheme. The following SEADAS flags were used to eliminate (mask) pixels from the monthly mean chlorophyll and the monthly summary of the frequency of negative water-leaving radiance: LAND SUNGLINT HIGHLT STRAYLIGHT CLDICE EPSILON Monthly mean chlorophyll (geometric mean) concentration was computed by averaging all non-masked log10-transformed chlorophyll observations for each pixel.