Validating a new algorithm for estimating aerosol optical depths over land from MODIS imagery

Abstract

Aerosols greatly affect the signals of satellite sensor imagery for remote sensing of land surfaces and play a dual role in global climate change and the hydrological cycle. However, there has not been a reliable method for estimating aerosol properties over land directly from multispectral remotely sensed imagery. In a recent study, a new algorithm to estimate aerosol optical depths (AODs) from Moderate-Resolution Imaging Spectroradiometer (MODIS) imagery suitable for all land surfaces was proposed. It is based on a sequence of imagery over a period of time with the assumption that the surface property is relatively stable and atmospheric conditions vary much more dramatically. Although this algorithm was validated over several sites, more validation was necessary. In this study, this algorithm was validated using 3-month measurements at 25 AErosol RObotic NETwork (AERONET) sites in North America. The validation results show that this algorithm can estimate AODs with close agreement with the AERONET measurements [R2=0.69, root mean square error (RMSE) 0.06].