Dust Detection Over East Asia From Multispectral and Multi‐Temporal Himawari‐8/AHI Thermal Infrared Observations

The frequent outbreak of dust storms every spring is one the extreme weathers in East Asia. Geostationary meteorological satellite thermal infrared (TIR) imagery provides observations for high-frequency, all-day, large-scale monitoring sources and transport. This study proposes an integrated method to detect over Asia by using multispectral multi-temporal Himawari-8/Advanced Himawari Imager (AHI) TIR observations, based on a Feedforward Neural Network (FNN) model Robust Principal Component Analysis Anomalous Dust Detection (ADD) method. FNN trained map Brightness Temperatures (BTs) three category probabilities (i.e., cloud, dust, clear sky), ADD calculates score normalizing BT difference anomalous variation. Integrated Index (IDI) reflects quantified confidence AHI pixel, which defined as sum probability score. IDI was evaluated 2 months Sentinel-5P TROPOMI Ultraviolet Aerosol achieved total detection accuracy above 90%. A comparison with Cloud Lidar Infrared Pathfinder Satellite Observation nighttime shows that can also at night. times series massive 14–16 March 2021 storm show its capabilities trace monitor