Improved retrieval of aerosol optical thickness from MODIS measurements through derived surface reflectance over Nanjing, China

Abstract

Determination of surface reflectance in the red and blue channels is a critical step in retrieving aerosol optical thickness (AOT) from Moderate Resolution Imaging Spectroradiometer (MODIS) measurements. The MODIS Collection 005 (C005) aerosol algorithm uses a ratio method to determine the surface reflectance in the red (0.66 μm) and blue (0.47 μm) channels from the surface reflectance in the 2.1 μm channel using global surface reflectance relationships. In this study, we attempted to improve the retrieval of AOT from MODIS measurements using a new surface parameterization derived using ground-based sunphotometer data and 6S radiative transfer code. The estimated surface reflectance in the red, blue and near-IR channel were used to derive ratio between them for use in the new retrieval from MODIS data. Our results demonstrate that the ratio of surface reflectance in the red and blue channels to the surface reflectance in the 2.1 μm channel varies seasonally in the Xianlin district of Nanjing City, China. These ratios are different from those assumed by the MODIS aerosol algorithm for the retrieval of AOT over land. The use of the appropriate ratio for the study area in a given season significantly improves the accuracy with the absolute error decreased from 0.15 to 0.08 and the relative error reduced from 31% to 17% in retrieving AOT from MODIS data.