圖書典藏及影音數位平台

本計畫持續蒐集泥砂濃度觀測資料，歷年已累積32場颱風事件觀測成果，並透過110年0530豪雨等事件，驗證人工採樣與自動化監測趨勢之高度一致性，確保全洪程水砂歷線之連續性與準確性。人工測站透過現場取樣與比重瓶試驗，詳細記錄入、出庫泥砂濃度，提供自動化儀器校正依據並進行粒徑分析；自動化測站則利用時域反射儀（TDR）監測不同深度之泥砂濃度，針對異重流之形成、分布高程、厚度及抵達時間進行精準監測。此雙重監測模式確保了資料之全面性，有效掌握泥砂運移之整體歷程。 經對歷年累積數據進行規律分析，泥砂運移及異重流之形成受降雨強度、入流特性與水庫操作等多重因素影響。綜整97至113年重大颱風特性，確立異重流自大溪坪至壩前約有2.5小時之關鍵應變時間。研究進一步開發異重流預警水砂關係圖，證實結合入庫濃度與累積雨量或砂量之指標，預警準確率可達100%。 本計畫亦針對壩前濁度傳輸歸納出三類模式，並定義垂直同步指數(VSI)與衝擊上升速率(RTR)等即時監測參數，用以量化濁度變化之動態特性。歷年全洪程觀測所累積之長期數據，除深化對泥砂運移與異重流規律之認識外，亦對水庫防洪操作及未來防淤規劃提供科學支撐。

This project has continuously collected sediment concentration data, accumulating observations from 32 typhoon events over the years. Through events such as the May 30, 2021 heavy rain, the high consistency between manual sampling and automated monitoring trends has been validated, ensuring the continuity and accuracy of sediment hydrographs throughout flood events. Manual stations utilize field sampling and pycnometer tests to record inflow and outflow sediment concentrations, providing a calibration basis for automated instruments and conducting particle size analysis. Automated stations employ Time Domain Reflectometry (TDR) to monitor sediment concentrations at various depths, enabling precise monitoring of the formation, vertical distribution, thickness, and arrival time of turbid density currents . This dual-monitoring mode ensures data comprehensiveness and effectively captures the entire process of sediment transport. Based on the analysis of historical data, the transport of sediment and the formation of turbid density currents are influenced by multiple factors, including rainfall intensity, inflow characteristics, and reservoir operations. By synthesizing the characteristics of major typhoons from 2008 to 2024, a critical response window of approximately 2.5 hours from Daxiping to the dam front has been established. This study further developed early warning water-sediment relationship charts, demonstrating that the combination of inflow concentration and indicators such as cumulative rainfall or sediment load can achieve a warning accuracy of up to 100%. Furthermore, this project categorized three turbidity transport modes at the dam front and defined real-time monitoring parameters, such as the Vertical Synchronicity Index (VSI) and the Rise Transition Rate (RTR), to quantify the dynamic characteristics of turbidity variations. The long-term data accumulated from years of flood monitoring not only deepens the understanding of sediment transport and turbid density current patterns but also provides robust scientific support for reservoir flood control, desilting operations, and future strategic planning.