VIIRS-Derived Inherent Optical Property Data over Global Coastal and Inland Waters Using the NIR-based Approach

Abstract

In open ocean and coastal/inland waters, the normalized water-leaving radiance spectra nL w (λ) are determined by water inherent optical properties (IOPs). However, for coastal/inland waters, the complex feature of the water IOPs makes it challenging for accurate retrieval of IOPs from satellite measurements. In this paper, we overview the recent progress for deriving accurate water IOPs and suspended particle size distribution (PSD) over global turbid coastal and inland waters. We show that water reflectance model in coastal and inland waters can be significantly simplified at the near-infrared (NIR) wavelengths, thus particle backscattering coefficient b bp (λ), phytoplankton absorption coefficient a ph (λ), and dissolved and detrital absorption coefficient a dg (λ) can be derived from nL w (λ) spectra from the Visible Infrared Imaging Radiometer Suite (VIIRS) observations. Specifically, the coefficient b bp (λ) and PSD in global highly turbid water can be characterized and quantified. We show that b bp (λ), a ph (λ), and a dg (λ) values derived using the NIR IOP algorithm match well with the true values in turbid coastal and inland waters. An IOP algorithm that combines the NIR-based IOP algorithm for coastal/inland waters and the Quasi-Analytical Algorithm (QAA) IOP algorithm for the open ocean is proposed to derive b bp (λ), a ph (λ), and a dg (λ) for global ocean. Using China’s east coastal region, Amazon River Estuary, and La Plata River Estuary as examples, we demonstrate that the combined IOP algorithm can produce high-quality b bp (λ), a ph (λ), and a dg (λ) data, as well as PSD products for both turbid coastal/inland waters and open ocean from VIIRS-derived nL w (λ) spectra.