Accuracy assessments of MODIS aerosol retrievals

Abstract

Moderate Resolution Imaging Spectroradiometer (MODIS) aerosol products have been extensively used to address ecological environment and climate change issues in many countries and regions. However, the retrieval quality of the MODIS aerosol products in different regions is the international research hotpot. The paper presents aerosol optical depth (AOD) data retrieved from MODIS in Ningbo, Lin'an and Qiandaohu cities in the Yangtze River Delta. The quality of the MODIS aerosol products in these regions has been evaluated using the Aerosol Robotic Network (AERONET) data derived from our ground sunphotometer measurements in these three regions in year 2008. The accuracy of the AOD retrieved from MODIS is the best in Lin'an and Ningbo (correlation coefficient R2 is greater than 0.92), the lowest in Qiandaohu (R2 is 0.72). Results indicated that the retrieval errors are primarily due to the different underlying surface reflectance and aerosol models. The retrieval errors are not obvious when input the true aerosol models, which mainly because the underlying surface conditions have less seasonally changes in these regions. Therefore, the aerosol models and models are the dominant factors which influence the accuracy of MODIS retrieval performance. In addition, the correction of humidity is important to improve the retrieval accuracy.