|
Improving radar rainfall estimation accuracy in the composite area of Takhli and Sattahip radars using spatial and hourly time-varying bias adjustment |
|---|---|
| รหัสดีโอไอ | |
| Creator | Ratchawatch Hanchoowong |
| Title | Improving radar rainfall estimation accuracy in the composite area of Takhli and Sattahip radars using spatial and hourly time-varying bias adjustment |
| Contributor | Siwa Kaewplang |
| Publisher | Faculty of Engineering, Khon Kaen University |
| Publication Year | 2569 |
| Journal Title | Agricultural and Biological Engineering |
| Journal Vol. | 3 |
| Journal No. | 4 |
| Page no. | 93-102 |
| Keyword | Rainfall intensity, Bias adjustment, Radar reflectivity, Inverse Distance Weighting (IDW), Automatic ground-based telemetry station |
| URL Website | https://ph04.tci-thaijo.org/index.php/abe |
| Website title | ThaiJo |
| ISSN | 3056-932X (Online) |
| Abstract | In estimating rainfall over the composite coverage area of the Takhli and Sattahip radars, the use of a composite Z–R relationship may still result in residual errors due to differences in the physical characteristics of rainfall. These differences can be attributed to the effects of the Earth’s curvature, variations in terrain characteristics, and event-to-event differences in rainfall, including variations in the spatial distribution of raindrops over time. In this study, a total of 267 rainfall events occurring between August 2018 and August 2020 were collected and analyzed. The dataset consisted of hourly rainfall measurements from 47 automatic ground-based telemetry stations and radar reflectivity data obtained within a 240 km detection range of the Takhli and Sattahip radars. These data were used to determine pixel-specific bias adjustment factors that varied on an hourly basis using the Inverse Distance Weighting (IDW) method. The adjustment factors were estimated from neighboring pixels with known values derived from both radar observations and automatic ground-based telemetry stations located within a 20 km radius of each target pixel. The results showed that the Composite radar rainfall intensity estimated using the Z = 138R1.6 relationship for the Takhli radar and the Z = 170R1.6 relationship for the Sattahip radar, combined with the proposed hourly time-varying bias adjustment factors, provided the greatest improvement in radar rainfall estimation accuracy over the composite coverage area of the Takhli and Sattahip radars. This approach yielded the lowest RMSE (Root Mean Squared Error), MAE (Mean Absolute Error), and BIAS values compared with the other bias adjustment methods. The proposed approach improved rainfall estimation accuracy by 50.24%, 16.83%, and 8.33% in terms of RMSE, MAE, and BIAS, respectively, compared with the unadjusted composite rainfall intensity estimated using the Z = 138R1.6 relationship for the Takhli radar and the Z = 170R1.6 relationship for the Sattahip radar. |
