TY - JOUR
T1 - Application of an Optimization Algorithm to Satellite Ocean Color Imagery: a Case Study in Southwest Florida Coastal Waters
AU - Hu, C.
AU - Lee, Z.
AU - Muller-Karger, Frank E
AU - Carder, Kendall L
PY - 2003/1/1
Y1 - 2003/1/1
N2 - A spectra-matching optimization algorithm, designed for hyperspectral sensors, has been implemented to process SeaWiFS-derived multi-spectral water-leaving radiance data. The algorithm has been tested over Southwest Florida coastal waters. The total spectral absorption and backscattering coefficients can be well partitioned with the inversion algorithm, resulting in RMS errors generally less than 5% in the modeled spectra. For extremely turbid waters that come from either river runoff or sediment resuspension, the RMS error is in the range of 5-15%. The bio-optical parameters derived in this optically complex environment agree well with those obtained in situ. Further, the ability to separate backscattering (a proxy for turbidity) from the satellite signal makes it possible to trace water movement patterns, as indicated by the total absorption imagery. The derived patterns agree with those from concurrent surface drifters. For waters where CDOM overwhelmingly dominates the optical signal, however, the procedure tends to regard CDOM as the sole source of absorption, implying the need for better atmospheric correction and for adjustment of some model coefficients for this particular region.
AB - A spectra-matching optimization algorithm, designed for hyperspectral sensors, has been implemented to process SeaWiFS-derived multi-spectral water-leaving radiance data. The algorithm has been tested over Southwest Florida coastal waters. The total spectral absorption and backscattering coefficients can be well partitioned with the inversion algorithm, resulting in RMS errors generally less than 5% in the modeled spectra. For extremely turbid waters that come from either river runoff or sediment resuspension, the RMS error is in the range of 5-15%. The bio-optical parameters derived in this optically complex environment agree well with those obtained in situ. Further, the ability to separate backscattering (a proxy for turbidity) from the satellite signal makes it possible to trace water movement patterns, as indicated by the total absorption imagery. The derived patterns agree with those from concurrent surface drifters. For waters where CDOM overwhelmingly dominates the optical signal, however, the procedure tends to regard CDOM as the sole source of absorption, implying the need for better atmospheric correction and for adjustment of some model coefficients for this particular region.
KW - Bio-optical inversion algorithm
KW - Case-II water
KW - Ocean optics
KW - Optimization
KW - Remote sensing
UR - https://digitalcommons.usf.edu/msc_facpub/1139
UR - https://doi.org/10.1117/12.466826
U2 - 10.1117/12.466826
DO - 10.1117/12.466826
M3 - Article
VL - 4892
JO - Proceedings of SPIE - The International Society for Optical Engineering
JF - Proceedings of SPIE - The International Society for Optical Engineering
ER -