EVALUATION OF THE QUALITY OF GRIDDED HEAVY RAINFALL OVER THE HA TINH CITY

Abstract

Gridded rainfall products which is the production generated by satellite together with weather numerical models combined with ground rain are increasingly developed based on its advantages outstanding ground rainfall measurement. Gridded rainfall products provide spatial information and its interval time is up to 30 minutes, and freely available, that makes it became very high applicability. Although the products are from the same sensors, still the algorithms used for correction, interpolation and matching are different, so there is a huge difference between the products. This study considers rain data from 04 sources including GPM, GSMaP, PERSIANN-CCS and ERA5-Land over Ha Tinh city area to assess their quality and practical applicability. Four rainfall events are used to assess the hourly rainfall compare to the data at rain gauge station in Ha Tinh city. Statistical indicators RMSE, R2, PBIAs, POD, FAR and CSI were used for the evaluation. The results show that in general, all four rain data have good quality, PERSIANN-CCS has the highest correlation with ground-observed rain, the best mean square error index, and other indicators show that GSMaP data is generally the worst quality.