Atmospheric Correction of GOCI Using Quasi-Synchronous VIIRS Data in Highly Turbid Coastal Waters

Jie Wu,Sravanthi Nukapothula,Chuqun Chen

doi:10.3390/rs12010089

Abstract

The Geostationary Ocean Color Imager (GOCI) sensor, with high temporal and spatial resolution (eight images per day at an interval of 1 hour, 500 m), is the world’s first geostationary ocean color satellite sensor. GOCI provides good data for ocean color remote sensing in the Western Pacific, among the most turbid waters in the world. However, GOCI has no shortwave infrared (SWIR) bands making atmospheric correction (AC) challenging in highly turbid coastal regions. In this paper, we have developed a new AC algorithm for GOCI in turbid coastal waters by using quasi-synchronous Visible Infrared Imaging Radiometer Suite (VIIRS) data. This new algorithm estimates and removes the aerosol scattering reflectance according to the contributing aerosol models and the aerosol optical thickness estimated by VIIRS’s near-infrared (NIR) and SWIR bands. Comparisons with other AC algorithms showed that the new algorithm provides a simple, effective, AC approach for GOCI to obtain reasonable results in highly turbid coastal waters.

Highlights

The total radiance measured by an ocean color sensor is primarily composed of the water-leaving radiance, sea surface radiance, Rayleigh scattering radiance caused by air molecules, and aerosol scattering radiance
For the quasi-synchronous VIIRS (QSV) algorithm, the Rayleigh-corrected reflectance values were obtained by using the SeaDAS 7.4 l2gen processor, and these values were used as the inputs of the aerosol scattering correction procedure, which was performed by using Interactive Data Language (IDL)
An alternative atmospheric correction (AC) algorithm using quasi-synchronous Visible Infrared Imaging Radiometer Suite (VIIRS) data for Geostationary Ocean Color Imager (GOCI) in highly turbid waters is presented in this paper

Summary

Introduction

The total radiance measured by an ocean color sensor is primarily composed of the water-leaving radiance, sea surface radiance, Rayleigh scattering radiance caused by air molecules, and aerosol scattering radiance (which includes aerosol single-scattering radiance and interactive scattering radiance between molecules and aerosols). It is key to accurately estimate the aerosol scattering by aerosols and Rayleigh–aerosol interactions in order to determine the water-leaving radiance. The AC algorithm developed by Gordon and Wang [4] ( named the GW94 algorithm) works quite well. It estimates the aerosol optical properties based on the black pixel assumption, according to which the water-leaving radiance at near-infrared (NIR) bands is assumed to be zero because the water can strongly absorb the light in these bands. In turbid coastal waters, the AC is more complicated because this assumption is rarely valid due to the significant suspended sediments backscattering in the NIR bands [5]

Objectives

Methods

Results

Conclusion