Development of GK-2A AMI Aerosol Detection Algorithm in the East-Asia Region Using Himawari-8 AHI Data

Abstract

The GEO-KOMPSAT 2A (GK-2A) satellite is a next-generation geostationary meteorological satellite with multi-wavelength channels. The GK-2A Advanced Meteorological Imager (AMI) aerosol detection production (ADP) algorithm can detect aerosol types (e.g., dust, haze, volcanic ash), using visible and infrared channels. This algorithm is not employed in existing meteorological satellites and can detect four different aerosol types in the daytime and three at night. Here, cloud and aerosol detection were performed using Advanced Himawari Imager (AHI) data with channels similar to those of simulation data. Comparing the results with the Communication, Ocean and Meteorological Satellite (COMS) aerosol index (AI) and AHI RGB aerosol data revealed generally similar dust classification and excellent daytime haze and nighttime aerosol detection performances. Regarding detection of haze, dust, and undefined aerosols generated across East Asia, a percent correct (PC) of 0.76, probability of detection (POD) of 68%, and false alarm rate (FAR) of 33% were obtained for GK-2A ADP of dust when validated with Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) Vertical Feature Mask (VFM) aerosol type data. Furthermore, the detection performance was ≥87% for haze and undefined aerosols, and the FAR was <30%. When Moderate Resolution Imaging Spectroradiometer (MODIS) L2 aerosol optical depth (AOD) data were used for validation (with and without AOD), the POD was 0.66 (ocean) and 0.87 (land), while the FAR was 0.29 (ocean) and 0.35 (land). Thus, aerosol detection performances of GK-2A ADP are similar to those of existing satellites and superior to satellites that use limited optical channels.