Two semianalytical remote-sensing reflectance models were evaluated and validated by use of bio-optical data collected in the Beaufort and Chukchi seas. Both models were efficient at retrieving chlorophyll concentration, phytoplankton absorption coefficients, and particulate backscattering coefficients. In contrast, they were not accurate in predicting an absorption coefficient for colored dissolved organic matter plus nonpigmented particulates. The poor model performance is attributed to the high variability in the concentrations of these colored materials. A chlorophyll-dependent reflectance model was also assessed, and it proved to be highly successful in reproducing measured reflectance spectra. A four-component, case 2 model with mean absorption spectra for phytoplankton, soluble materials, and nonpigmented particulates was employed in Hydrolight radiative-transfer model simulations. The remote-sensing reflectance spectra simulated in the radiative-transfer model were in excellent agreement with field data. The similarity between the model and the measurements confirms the accuracy of the underlying bio-optical relationships and underscores the utility of modeling for better understanding of the variability of ocean color observations. The latest SeaWiFS algorithm (OC4V4) overestimated chlorophyll by ~1.5 fold across most of the observed range of biomass (0.07–9 mg chlorophyll m−3). Regionally tuned algorithms explained >93% of the variability in the surface chlorophyll concentration.
© 2003 Optical Society of America
[Optical Society of America ]
(010.4450) Atmospheric and oceanic optics : Oceanic optics
(120.5700) Instrumentation, measurement, and metrology : Reflection
(280.0280) Remote sensing and sensors : Remote sensing and sensors
Jian Wang and Glenn F. Cota, "Remote-Sensing Reflectance in the Beaufort and Chukchi Seas: Observations and Models," Appl. Opt. 42, 2754-2765 (2003)