Publications

McKinna, Lachlan I. W.; Fearns, Peter R. C.; Weeks, Scarla J.; Werdell, P. Jeremy; Reichstetter, Martina; Franz, Bryan A.; Shea, Donald M.; Feldman, Gene C. (2015). A semianalytical ocean color inversion algorithm with explicit water column depth and substrate reflectance parameterization. JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS, 120(3), 1741-1770.

Abstract
A semianalytical ocean color inversion algorithm was developed for improving retrievals of inherent optical properties (IOPs) in optically shallow waters. In clear, geometrically shallow waters, light reflected off the seafloor can contribute to the water-leaving radiance signal. This can have a confounding effect on ocean color algorithms developed for optically deep waters, leading to an overestimation of IOPs. The algorithm described here, the Shallow Water Inversion Model (SWIM), uses pre-existing knowledge of bathymetry and benthic substrate brightness to account for optically shallow effects. SWIM was incorporated into the NASA Ocean Biology Processing Group's L2GEN code and tested in waters of the Great Barrier Reef, Australia, using the Moderate Resolution Imaging Spectroradiometer (MODIS) Aqua time series (2002-2013). SWIM-derived values of the total non-water absorption coefficient at 443 nm, a(t)(443), the particulate backscattering coefficient at 443 nm, b(bp)(443), and the diffuse attenuation coefficient at 488 nm, K-d(488), were compared with values derived using the Generalized Inherent Optical Properties algorithm (GIOP) and the Quasi-Analytical Algorithm (QAA). The results indicated that in clear, optically shallow waters SWIM-derived values of a(t)(443), b(bp)(443), and K-d(443) were realistically lower than values derived using GIOP and QAA, in agreement with radiative transfer modeling. This signified that the benthic reflectance correction was performing as expected. However, in more optically complex waters, SWIM had difficulty converging to a solution, a likely consequence of internal IOP parameterizations. Whilst a comprehensive study of the SWIM algorithm's behavior was conducted, further work is needed to validate the algorithm using in situ data.

DOI:
10.1002/2014JC010224

ISSN:
2169-9275