The knowledge of aerosol optical properties is of high importance for environmental and climate studies. This presentation discusses the development of aerosol  property retrieval from the Advanced Himawari Imager (AHI), on board the next-generation geostationary meteorological satellite Himawari-8 that is a 16 channel multispectral imager to capture visible light and infrared images of the Asia-Pacific region. Observations are obtained every 10 min at a spatial resolution of 0.5-2 km, which provide great advantage of tracking aerosols and characterizing aerosol daily variability.
In proposed approach the aerosol and underlying surface are retrieved simultaneously. Namely, spectral Aerosol Optical Depth (AOD) is derived together with surface BRDF parameters using GRASP algorithm. The performance of the developed AHI/GRASP algorithm is illustrated using both synthetic data and real Himawari-8/AHI observations.
The results of synthetic tests show that both aerosol and surface parameters can be robustly derived with a high correlation for AOD, AE and surface albedo respectively. The aerosol and surface parameters retrievals from real Himawri-8/AHI data were performed over China and validated against the AERONET version 3 level 1.5 measurements as well as MODIS aerosol products. The results show good consistency of AOD retrievals with the AERONET values, with a correlation coefficient of 0.94 and an overall negative bias of -0.01. In addition, the surface albedo retrievals have high consistency with AERONET with correlations vary from 0.73 to 0.85 in various types of sites. AHI/GRASP aerosol products also show good agreement with MODIS measurements with correlations of 0.72-0.87 in different seasons.
In addition to satellite only applications, a combined retrieval using passive (Himawri-8/AHI) and active (Micro-pulse Lidar) observations was performed over Beijing-PKU site. Satellite observations and coincident lidar were inverted. Such synergetic retrieval resulted in further improvements in derived aerosol properties benefiting from additional information provided by the back-scattering observations by lidar, which significantly improved sensitivity to the vertical variability of aerosol properties, especially while the aerosol loading is low. The mean bias of AOD was further decreased from 0.04 to -0.01 when AOD is lower than 0.5.