透過您的圖書館登入
IP:54.234.136.147
  • 學位論文

集水區定量降水估計誤差對於水位模擬影響之研究

Study on Impact of Watershed QPE Error to River Stage Simulation

指導教授 : 張哲豪
若您是本文的作者,可授權文章由華藝線上圖書館中協助推廣。

摘要


現階段防洪預警作業技術,可藉由即時水文氣象觀測或預測資料內容,銜接水文水理模式,模擬計算得到現況與預測河川水位高程,如水利署目前已正式上線的「洪水早期預警系統」。就系統之預測準確度而言,大多數預測水位站未來一小時的水位高程誤差可以在 20~50 公分以內;但少部分預測水位站預測準確度仍待改善。預測準確度影響主要來自水位預報模式過程中,文氣象觀測、預報資料等輸入因子,本研究擬以其因子建立水位模擬之誤差分析流程,作為未來水位預報模式影響評估之開端。 在預報上為盡可能掌握水文氣象資料高變異性,目前在世界主要進行防洪測預報之先進國家,在水文氣象資料來源上均採用多種水文氣象資料來源的結合,以增加水情演算與推斷的可靠度。由於多種資料內容在輸出格式、空間維度、與時間維度等,都存有相當差異,整合上也就面臨資料格式與維度處理課題。故本研究將誤差分析流程分為觀測降雨之敏感度,以及預報降雨之不確定性。 本研究在水文氣象資料處理過程中,分析觀測降雨數值於空間轉換與內插、數值精度等差異,對於集水區平均降雨量以及河川水位高程計算時之影響,藉以評估觀測降雨數值、空間反應於水位計算時之敏感度。在考慮預報降雨本身不確定性下,本研究針對2008卡玫基颱風及2010凡那比颱風之降雨事件,採用蒙地卡羅法模擬定量降水預報,分析模擬定量降水預報對於水位預報模式之影響。

並列摘要


The current/forecasting water level can be calculated through the connection between instant hydrology and meteorology observation or data forecasting and hy-drologic/hydraulic simulation models in the flood prevention and precaution tech-nology at the present stage, such as the online “Flood Early Warning System" pro-vided by the Water Resources Agency. As for the forecasting accuracy, the change in the water level within one hour of the reading being submitted from forecasting stations with the highest readings can fall in the range of 20 to 50 centimeters, but the accuracy of some of the forecasting stations is awaiting improvement. The fac-tors that may influence the accuracy primarily comes from the process of water level forecasting patterns and the input factors like hydrographic and meteorology obser-vation and forecasting data. This study plans to use the factors to build up a error analytic process to water level simulation in order to create a threshold for evaluat-ing the impact of water level forecasting patterns in the future. In order to control the high variability of hydrographic and meteorology data from the forecasting, within the countries which carry out flood prevention forecasts, all adopt the combination of a variety of hydrographic and meteorology data of their resources in order to increase the reliability of the calculation and to forecast the hydrology. Due to the differences in outputting format and space and time dimen-sions between multiple types of data, there will be some problems when facing inte-gration. Therefore, this study divides bias analyzing processes into the sensitivity of precipitation observation and the uncertainty of precipitation forecasting. In hydrographic and meteorology data processing, the study analyzes the vari-ants to space transforming, interpolation and number accuracy in the observation of precipitation number analysis, and tries to evaluate the the sensitivity of water level calculations of space and precipitation observation reactions. In considering the un-certainty of precipitation forecasting itself, this study adopts the Monte Carlo Algo-rithm to simulate Quantitative Precipitation Forecasts for the precipitation cases of Typhoon KALMAEGI, 2008 and FANAPI, 2010, to analyze the impact of simulat-ing Quantitative Precipitation Forecasts to water level forecasting patterns.

參考文獻


[23] 郭冠霆,高效率流量量測法之可靠度分析,碩士論文,國立台北科技大學土木與防災研究所,台北,2009。
[22] 陳信中,蘭陽溪洪水預報模式之研究,碩士論文,國立台灣大學生物環境系統工程學研究所,台北,2006。
[17] 連以婷,水文模式之參數不確定性分析,碩士論文,國立台灣大學生物環境系統工程學研究所,台北,2010。
[20] 許盈松、周湘俊、楊舒雲,「流量率定曲線分析方法評估研究-以濁水溪為例」,農業工程學報,第50卷,第4期,2004,第62-76頁。
[14] 康富智,防洪系統溢堤風險分析模式之研究,碩士論文,國立成功大學水利及海洋工程學系專班,台南,2007。

延伸閱讀