Improved Hourly Estimates of Aerosol Optical Thickness Using Spatiotemporal Variability Derived From Himawari-8 Geostationary Satellite

Abstract

We developed a scheme to improve hourly estimates of aerosol optical thickness (AOT) from the Advanced Himawari Imager (AHI) onboard Himawari-8, the Japan Meteorological Agency's latest geostationary satellite. Taking advantage of the sampling characteristics of the AHI (10-min temporal and subkilometer spatial resolution), we quantify temporal and spatial variability of AOT from the observations (AOToriginal) and utilize this information to develop an hourly algorithm that produces two sets of derived AOTs: AOTpure, derived by the application of strict cloud screening to AOToriginal, and AOTmerged, derived from spatial and temporal interpolations of AOTpure. The AOTs thus obtained from the hourly algorithm were validated against measurements from the aerosol robotic network. The root-mean-square errors (RMSEs) of the AOTpure and AOTmerged products were 0.14 and 0.11, respectively, providing improvement over an RMSE of 0.20 for AOToriginal.