A New Dynamically Updated Geostationary Satellite Precipitation Estimation Algorithm for Near Real-Time Condition

Near real-time precipitation estimation from geostationary satellites plays an important role in flood forecasting, water resource management, and disaster prevention and reduction. Currently, near real-time precipitation products based on geostationary satellites still face great challenges in accurately detecting precipitation and monitoring small-scale precipitation. In this study, a novel Geostationary Satellite Precipitation Estimation (GSPE) algorithm for near real-time conditions was developed to retrieve precipitation at a spatial resolution of $0.05^{\circ }\times 0.05^{\circ }$ every 10 min both day and night. The major highlight of the GSPE is that a new precipitation detection scheme was created by introducing a newly proposed precipitation detection index (PDI) and 24-h continuous cloud microphysical parameters for the first time. Another highlight is that a dynamic updating scheme was proposed in building conversion models between the brightness temperature of geostationary satellite and precipitation to keep the accuracy and stability of the estimated precipitation. Furthermore, the 10-min temporal resolution of the estimated precipitation could accurately capture the evolution of a short precipitation process and improve the calculation of total precipitation amount. According to the validation using rain gauge observations from the Chinese mainland, the Heidke Skill Score (HSS) of the GSPE on the hourly scale could reach up to 0.39, which was improved by 11.43% compared to the GSMaP_NOW. The root mean square error (RMSE) of precipitation from the GSPE on hourly, daily, and monthly scales are 1.66, 13.65, and 97.76 mm, respectively, and was improved by 15.74%, 13.17%, and 21.01%, respectively, compared to that of the GSMaP_NOW.