透過您的圖書館登入
IP:3.238.79.169
  • 學位論文

六小時雨量預報資訊之即時校正與分析

Real time correction and analysis to 6-hour-precipitation forecast

指導教授 : 張哲豪

摘要


河川警戒水位主要視洪水上漲速率及沿岸地區應變時間而定,其目的為減低在防汛期間內,因外水溢淹所造成沿岸地區民眾生命財產損失。但洪水上漲速率往往因暴雨特性、河道沖淤現象等集水區地文、地貌因子影響而有所改變,進而造成沿岸地區應變時間減少。為使得沿岸地區應變時間增加,使用各種不同的水文水理模式進行水位預報已成為一種趨勢。但目前預報時間長短需視雨量預報時間而定,因此為了取得更多反應時間,延長雨量預報長度具有相當重要性。 目前水文水理模式大多以QPF(Quantitative Precipitation Forecast,以下簡稱QPF)定量降水預報,作為水文演算之輸入來源,惟QPF定量降水預報長度僅三小時,對於增加各沿岸地區之反應時間略顯不足。因此,本研究組合QPF定量降水預報,及中尺度數值雨量預報WRF(Weather Research and Forecasting Model,以下簡稱WRF)兩種雨量預報模式成果,將雨量預報時間自三小時延長至六小時,以提供後續水文水理模式進行下游河川水位預報,增加沿岸地區反應時間。 且為提升組合後雨量預報之可靠度,透過經濟部水利署所開發FEWS_Taiwan水情測預報整合平台內所提供之General Adapter模組,銜接由國家高速網路與計算中心吳祥禎博士所開發之即時校正模式。透過FEWS_Taiwan平台內一致的資料格式交換技術,將組合後雨量預報成果進行即時校正,提供各水文水理模式可靠度較高之雨量預報成果。

並列摘要


River warning levels often vary according to flood rising rates and riverside disaster response time. These warning levels are initially set to protect the lives and property of citizens living by the riverside and minimize the chances of damage due to external water overflow during flood control periods. However, flood rising rates are frequently subject to change due to physiographical and geomorphological factors in the catchment area, such as rainstorm characteristics as well as erosion and deposition along the river course. In order to increase the disaster response time of riverside regions, water level forecasts using various hydrological and hydrographical methods has become a common measure. In spite of this, currently the time scale of such forecasts depends heavily on rainfall forecasts, therefore in order to lengthen disaster response time, extending the time scale of rainfall forecasts is a critical issue. At present, most hydrological and hydrographical methods utilize Quantitative Precipitation Forecast (QPF) for quantification of rainfall forecasts; these results in turn serve as input for hydrologic calculations. The downside is this method provides a mere three-hour forecast, which does not allow sufficient time for disaster response along riverside areas. In light of this issue, this study combined two rainfall forecast methods, QPF and the Weather Research and Forecasting Model (WRF), thus extending forecast time from three hours to six hours. These results can serve as information for subsequent downstream hydrological and hydrographical methods, and moreover improve disaster response time of riverside areas. In addition, to improve reliability of this combined forecasting approach, this study applied a real-time correction method developed by Dr. Shiang-Jen Wu of the National Center for High Performance Computing. This was done through utilization of the General Adapter module of the FEWS_Taiwan hydrological prediction integrated platform developed by the Water Resources Agency, Ministry of Economic Affairs. Through consistent data exchange techniques embedded in the FEWS_Taiwan platform, real-time correction was performed on preliminary rainfall forecasts of the QPF-WRF combined measure, thus producing highly reliable rainfall forecast results.

參考文獻


[10] 蔡孟原,雷達定量降水估計應用在河川洪水預報之研究,碩士論文,臺灣大學生物環境系統工程學研究所,台北,2009。
[11] 蔡孟原、許銘熙、傅金城、林李耀、王安翔,結合雷達定量降水估計在河川洪水預報模式,農業工程學報,57卷,2期,2011。
[31] Tsai, M. Y.、Hsu, M. H.、Fu, J. C.、Lin, L. Y.、Wang, A. H., ”Radar-based quantitative precipitation estimation for flood forecast in rivers,” Journal of Taiwan Agricultural Engineering, 57, 4, 2011, 49-68.
[8] 馮威勝,建構fews_taiwan 與模式連結之標準作業方式,碩士論文,國立台北科技大學土木與防災研究所,台北,2010。
[9] 黃鵬豪,應用qpesums高解析降雨資料改良洪水預報模式之研究。,碩士論文,臺灣大學生物環境系統工程學研究所,台北,2008。。

被引用紀錄


陳柏愷(2012)。介接淹水模式於即時作業平台〔碩士論文,國立臺北科技大學〕。華藝線上圖書館。https://doi.org/10.6841%2fNTUT.2012.00344

延伸閱讀