結合水文及數值模式應用於河川水位預報—以高屏溪為例

Combining hydrological and numerical models applied to water stage forecast- A Case Study in Gaoping river

洪逸鈞(I-chun Hung)

指導教授 ： 吳瑞賢

國立中央大學/工學院/土木工程學系/碩士(2015年)

摘要

由於極端氣候及颱風事件所造成之降雨不確定性，如何建立適當的洪水預警機制是本文所探討的問題。因應2009莫拉克風災以及2010凡那比颱風於台灣本島南部帶來嚴重的豪雨，致使高屏地區飽受水患之苦，故本研究選定高屏溪流域進行水文模擬及河川水位的討論。
本研究將整塊高屏流域分為四個區塊，分別為上游地區的荖濃溪、旗山溪、隘寮溪流域，以及下游地區的高屏溪流域。上游地區以水文逕流模式進行河川流量模擬，並將模擬出的流量當作給定之下游邊界條件。而下游則以數值模式建立一維河道與二維網格，並利用水文模式模擬之出流量進一步模擬出河川水位。
模式的檢定驗證以過去颱風事件作為依據，且以各流域的雨量站降雨資料為主要參數。驗證完成後再將模式耦合多組系集預報降雨，再進行各組不同降雨預報的水位模擬。研究結果可知，藉由雨量站所提供之雨量資料模擬出的水位值，洪峰誤差值皆小於4%。而天兔事件的洪峰時間在里嶺大橋水位站只有提前2小時，下游的萬大大橋水位站的模擬時間也相當吻合。而根據定量降雨數據所模擬出的水位數值，愈接近峰值發生時間的模擬水位，峰值也會愈接近。

關鍵字

水文模式 ； 定量降雨 ； 系集預報

並列摘要

This study aims to explore how to establish appropriate flood forecast in response to rainfall uncertainty caused by extreme weather and typhoons. Considering Typhoon Morakot (2009) and Typhoon Fanapi (2010) lead to the severe rainfall and flood in southern Taiwan, this study selected Gaoping River Basin as study site.
For research purpose, this study divided Gaoping River basin into four blocks, including upstream region: Laonong river, QiShan river and Ailiao river, and downstream region: Gaoping river. We used hydrological model to calculate surface runoff on upstream area and applied numerical model for channel to downstream calculations.
Past typhoon events were examined to calibrate model parameters in this study, while rainfall stations data for each river basin was used as main parameter to validate modeling performance. Lastly, the model was joined with quantitative precipitation and ensemble forecast to simulate different water stage.
The research result indicated that peak error values are less than 4% when setups for flash flood in this study were used. The peak time of Usagi event presents a relatively minor difference from the estimated peack time: 2 hours in advance, and the simulation time for Wanda Bridge downstream water level stations are in good agreement with actual event. Additionally, as the simulated water stage result based on quantitative precipitation data gets closer to simulated water stage result based on actual peak value event, the peak value appears to be more accurate.

參考文獻

8. Venkatesh Merwade，Creating SCS Curve Number Grid using HEC-GeoHMS，Purdue University，(2012)

9. Venkatesh Merwade，Hydrologic Modeling using HEC-HMS，Purdue University，(2012)

11. Shih ,D.S. , Chen, C.H. , Yeh, G.T., “Improving our understanding of flood forecasting using earlier hydro-meteorological intelligence.”, Journal of Hydrology 512 ,pp 470-481. (2014).

12. Yeh, G.T., Shih, D.S., Cheng, J.C.,“An integrated media, integrated processes watershed model .” , Comput. Fluids 45 (1), pp 2-13.(2011).

14. Shih, D.S., Yeh, G.T.,“Identified model parameterization, calibration and validation of the physically distributed hydrological model, WASH123D in Taiwan. ”Journal of Hydrology. Eng. 16 (2), pp.126-136. (2011).