MERRA-2 AOD product performance in Eastern and Southern China Sea

Abstract

Aerosol optical depth (AOD) is one of the basic parameters used to analyze physical properties of regional aerosols, but the in-situ observation or remote sensing AOD dataset could be scarce especially in ocean area. The Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) reanalysis has the longest temporal span, and its accuracy in China sea area is to be evaluated. This study provides a validation of MERRA-2 AOD products’ applicability in the eastern and southern China sea based on Aerosol Robotic Network (AERONET). The results indicated that the MERRA-2 AOD with 1-hour temporal resolution agreed with the time averaged AERONET AOD well, for its correlation coefficient is 0.887, root mean square error (RMSE) is 0.096, and mean absolute error (MAE) is 0.056. Presented analysis also revealed a systematic underestimation of AOD that MERRA-2 made, and that deviation tended to increase in higher AOD which demonstrated a slope of -0.26 when utilized linear fitting technics, but the mean bias (MB) of test dataset was only -0.001 because the AOD concentrated on lower than 0.2. These results illustrated the suitability of using MERRA-2 AOD product in aerosol researches of the China sea area.