Spatio-temporal variability of aerosols over East China inferred by merged visibility-GEOS-Chem aerosol optical depth

Long-term visibility measurements offer useful information for aerosol and climate change studies. Recently, a new technique to converting visibility measurements to aerosol optical depth (AOD) has been developed on a station-to-station basis (Lin et al., 2014). However, factors such as human observation differences and local meteorological conditions often impair the spatial consistency of the visibility converted AOD dataset. Here we further adopt AOD spatial information from a chemical transport model GEOS-Chem, and merge visibility inferred and modeled early-afternoon AOD over East China on a 0.667 long, x 0.5 degrees lat. grid for 2005-2012. Comparisons with MODIS/Aqua retrieved AOD and subsequent spectral decomposition analyses show that the merged dataset successfully corrects the low bias in the model while preserving its spatial pattern, resulting in very good agreement with MODIS in both magnitude and spatio-temporal variability. The low bias is reduced from 0.10 in GEOS-Chem AOD to 0.04 in the merged data averaged over East China, and the correlation in the seasonal and interannual variability between MODIS and merged AOD is well above 0.75 for most regions. Comparisons between the merged and AERONET data also show an overall small bias and high correlation. The merged dataset reveals four major pollution hot spots in China, including the North China Plain, the Yangtze River Delta, the Pearl River Delta and the Sichuan Basin, consistent with previous works. AOD peaks in spring summer over the North China Plain and Yangtze River Delta and in spring over the Pearl River Delta, with no distinct seasonal cycle over the Sichuan Basin. The merged AOD has the largest difference from MODIS over the Sichuan Basin. We also discuss possible benefits of visibility based AOD data that correct the sampling bias in MODIS retrievals related to cloud-free sampling and misclassified heavy haze conditions. (C) 2016 Elsevier Ltd: All rights reserved.National Natural Science Foundation of China [41175127, 41422502, 41575018]; 973 program [2014CB441303]SCI(E)EIARTICLElinjt@pku.edu.cn111-12213