|
การศึกษาความแตกต่างของปริมาณน้ำท่าและพารามิเตอร์ของแบบจำลอง SWAT จากการใช้ข้อมูลฝนตรวจวัดและข้อมูลฝนภาพถ่ายดาวเทียม |
|---|---|
| รหัสดีโอไอ | |
| Creator | เกศวรา สิทธิโชค |
| Title | การศึกษาความแตกต่างของปริมาณน้ำท่าและพารามิเตอร์ของแบบจำลอง SWAT จากการใช้ข้อมูลฝนตรวจวัดและข้อมูลฝนภาพถ่ายดาวเทียม |
| Contributor | จุติเทพ วงษ์เพ็ชร แพรววดี หงษาวง ธีรศักดิ์ ซ้ายอ่อน |
| Publisher | King Mongkut’s University of Technology Thonburi |
| Publication Year | 2565 |
| Journal Title | KMUTT Research and Development Journal |
| Journal Vol. | 45 |
| Journal No. | 1 |
| Page no. | 107-124 |
| Keyword | SWAT, Satellite Rainfall, Kaengkrachan Reservoir, Runoff, Reservoir Inflow |
| URL Website | https://journal.kmutt.ac.th/ |
| Website title | เว็บไซต์วารสารวิจัยและพัฒนา มจธ. |
| ISSN | 2697-5521 |
| Abstract | The objective of this study was to investigate the differences in runoff and SWAT model parameters that were obtained using observed rainfall data (SWAT-Station) and satellite rainfall data as prepared by JAXA Global Rainfall Watch System (SWAT-GSMaP_NRT). A bias correction method was initially employed for satellite rainfall data, which were then introduced to the SWAT model. Both SWAT-Station and SWAT-GSMaP_NRT models were calibrated and validated; R2 NSE and PBIAS were used to estimate the model performance. Low to medium performance of SWAT-Station model were noted in both calibration/validation stages, with R2 NSE and PBIAS of 0.26/0.26, 0.25/0.14 and 26.75%/-26.50%, respectively. On the other hand, better performance was noted when satellite rainfall data were introduced to the model instead of the observed data (R2 of 0.45/0.46, NSE of 0.41/0.48 and PBIAS of 25.16%/-21.19%). Calibration/validation results of SWAT-GSMaP_NRT model showed the highest performance, with R2 of 0.68/0.51, NSE of 0.68/0.45 and PBIAS of 11.93%/-13.94%. Top five sensitive parameters of both models were then investigated; most sensitive ones were noted to belong to the soil moisture parameters. The maximum value of the sensitive parameters of SWAT-GSMaP_NRT model were slightly higher than those of SWAT-Station model. Finally, annual inflows predicted by both models were examined. Difference in the annual inflows predicted by both models was 12.37%. SWAT-GSMaP_NRT estimated slightly higher inflows when compared to SWAT-Station in rainy season; lower inflows were noticeably estimated starting from the end of wet season until the end of dry season, however. Average monthly inflows in rainy/dry season were predicted to be 119/48 and 129/36 mcm by SWAT-Station and SWAT-GSMaP_NRT models, respectively. |
