A blended inherent optical property algorithm for global satellite ocean color observations

Abstract

Water inherent optical properties (IOPs) can be derived from satellite-measured normalized water-leaving radiance (nLw(λ)) spectra. In this study, we evaluate the performance of the quasi-analytical algorithm (QAA) and the near-infrared (NIR)-based IOP algorithm using a Hydrolight simulation data set covering a wide range of water types that span from clear open ocean to turbid coastal/inland waters. The NIR-based algorithm produces significantly improved IOP retrievals over turbid coastal and inland waters, while the QAA algorithm performs well in the open ocean and less turbid coastal waters. Based on the advantages of the NIR-based and QAA-based algorithms, a combination of the NIR- and QAA-based algorithm has been proposed using satellite-measured nLw(745) as the threshold in order to produce accurate IOP products for both the open ocean and turbid coastal/inland waters. The new combined IOP algorithm can produce reasonably accurate IOP data for all water types, and can be easily implemented into the satellite ocean color data processing. The La Plata River Estuary region is used as an example to show the difference in performance of IOP retrievals from the Visible Infrared Imaging Radiometer Suite (VIIRS) measurements between 2012 and 2017 with the NIR-based, QAA-based, and NIR-QAA combined IOP algorithms. We also demonstrate that the NIR-QAA combined algorithm can be applied to VIIRS global ocean color observations to derive good quality IOP products in China's east coastal region, the US east coastal region, and the region of Mississippi River Estuary and tributaries.