[A-CG36] Satellite Earth Environment Observation

[ACG36-09] Improvement in the GSMaP histogram matching algorithm for middle and lower level clouds

*Hitoshi Hirose¹, Tashima Tomoko¹, Takuji Kubota¹, Tomoaki Mega², Tomoo Ushio² (1.Japan Aerospace Exploration Agency, 2.Department of Engineering, Osaka University)

Keywords:Satellite precipitation measurement, algorithm, Rain type classification

In Southeast Asia, where the use of ground-based rain observation is limited, there have been many attempts to utilize global satellite precipitation products for predicting the flow rate of water-related disasters. The Global Satellite Mapping of Precipitation (GSMaP; Kubota et al. 2020) combines data from microwave instruments (PMW) aboard polar-orbiting satellites and infrared (IR) radiometers aboard geostationary meteorological satellites to observe global precipitation hourly. However, the IR product are less accurate than the PMW product because they provide information only from cloud tops, and spatial heterogeneity has been reported in the precipitation estimation accuracy of GSMaP (Utsumi and Kim 2018). Therefore, we introduced a histogram matching method to correct the precipitation rate histogram of the IR product to match that of the PMW product (Hirose et al. 2021). Implementation of the histogram matching over the tropical oceans was successful in improving rain the overestimation of the IR product for upper level clouds, but not enough to improve the underestimation for middle and lower level clouds. Therefore, we used classification data of convective or stratiform rain derived from the Ku-band precipitation radar of the Global Precipitation Measurement (GPM) core satellite to classify the tropical ocean into three zones: warm-pool, subtropical subsidence zone and transition zone in between. Histogram matching was applied to each region. In addition, since the IR product tended to estimate a slightly larger precipitation area than the PMW product, the particularly weak rain sample in the IR product was excluded from the correction to eliminate the difference in precipitation area between IR and PMW products. The effect of the histogram matching will be explained in detail for each region and season.