Developing Ocean Color Remote Sensing Algorithms for Retrieving Optical Properties and Biogeochemical Parameters in the Optically Complex Waters of Long Island Sound.

摘要

The optical properties of the sea determine how light penetrates to depth, interacts with water-borne constituents, and re-emerges as scattered rays. By inversion, quantifying change in the spectral light field as it reflects from the sea unlocks information about the water's optical properties, which can then be used to quantify the suspended and dissolved biogeochemical constituents in the water. Retrieving bio-optical properties is relatively straightforward for the open ocean where phytoplankton-derived materials dominate ocean color. In contrast, the presence of land-derived material contributes significantly to the optical signature of nearshore waters, making the development of ocean color algorithms considerably more challenging. A hypothesis of this research is that characterization of the spectral nature of bio-optical properties in these optically complex waters facilitates optimization of semi-analytical algorithms for retrieving these properties.;The main goal of this research is to develop an ocean color remote sensing algorithm for the highly turbid, estuarine waters of Long Island Sound (LIS) Bio-optical data collected in LIS showed it to be strongly influenced by the surrounding watershed and characterized by exceptionally high absorption associated with phytoplankton, non-algal particulate material, and chromophoric dissolved material compared to other coastal environments world-wide. Variability in the magnitudes of inherent optical properties, IOPs (e.g. absorption, scattering and attenuation coefficients), is explained by local influences such as major river outflows, as well as seasonal changes. Nevertheless, ocean color parameters describing the spectral shape of IOPs---parameters to which algorithms optimization is sensitive---are fairly constant across the region, possibly a result of the homogenizing influence of vigorous tidal and subtidal mixing or relative regional homogeneity in the biogeochemical nature of terrigenous material.;Field observations of IOPs, biogeochemical properties (e.g. chlorophyll concentration and total suspended materials, TSM), and sea-surface reflectances are used to select, optimize and validate an ocean color algorithm for LIS. Optimization revealed that a 640 nm reflectance channel and a spectrally varying f/Q factor (i.e. relating reflectances to IOPs) were critical to good performance. The algorithm is applied---as an example of its utility---to satellite imagery to quantify TSM across the region during record flooding events of March 2010. Analysis shows that the delivery of TSM during these storms was strongly dependent on localized land-use characteristics.