Rainfall intensity analysis for radar rainfall evaluation in the composite area of Takhli and Sattahip radar

This study collected data on a total of 510 rain events between February 2018 and November 2020. It includes hourly rainfall data (R) from 238 ground-based automated telemetry stations and radar reflectivity data (Z) under a measurement radius of 240 km from Takhli and Sattahip radars. The Z-R relationships of Takhli radar and Sattahip radar were determined and applied to evaluate radar rainfall intensity in the composite area of these radars. The radar rainfall intensity in the composite area was analyzed using five methods: the rainfall intensity from (1) Z = 138R1.6 (only Takhli radar), (2) Z = 170R1.6 (only Sattahip radar), and the composite rainfall intensity from (3) Z = 200R1.6 (Marshall and Plamer), (4) Z = 300R1.6 (Woodley and Herndon), and (5) Z = 138R1.6 and Z = 170R1.6 (Takhli and Sattahip radar). These results were compared with the ground-based station data to determine the best method for evaluation based on the least statistical values, i.e., Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and BIAS. The results show that the composite results from Takhli and Sattahip radar were the most accurate, and this case can increase the accuracy compared to the 1st to 4th method by 22.26%, 10.25%, 3.89%, and 18.02% of the RMSE; 29.75%, 23.14%, 14.88%, and 14.88% of the MAE; and 361.54%, 42.31%, 100.00%, and 369.23% of the BIAS value, respectively.