Inversion of Remote Sensing Reflectance by Modeling the Spectral Beam Attenuation in Northeast Coast Area

Reflectance signature of ocean contains vast and useful information regarding concentrations and compositions of dissolved and suspended constituents in the sea. Since the reflectance spectrum is related to inherent optical properties (IOP) such as backscattering and absorption coefficients, retrieval of water constituents from the reflectance spectrum usually involves a parameterization of these IOPs. Both recent field observations with higher spectral resolution in coastal area and Mie calculations of absorbing particles reveal that the spectral backscattering can be modulated by the phytoplankton absorption while the beam attenuation spectrum appears to be a monotonically decreasing function. Based on these observations we chose to parameterize the spectral beam attenuation by assuming a constant backscattering ratio instead of the usual hyperbolic constraint for backscattering spectrum. Using this bio-optical model the inversion is solved with a bounded non-linear least square minimization to yield estimates of the parameters. This inversion model was applied to the reflectance data collected through several field campaigns along Northeast Coast area including Chesapeake Bay and Long Island Sound, which covers a large variety of eutrophic water environments from highly turbid water to algal blooming conditions. In addition to reflectance data the IOP data including absorption, attenuation and backscattering were recorded by an in-situ instrument package for WetLabs Inc and the water constituents such as chlorophyll, color dissolved organic matter (CDOM), total suspended solid (TSS) were obtained through the water sample analysis. With all these data available the performance of the inversion method will be tested and we will further investigate the necessary constraints to put on the model to yield accurate and reasonable estimate of its parameters for coastal environments.