大高滩水库溃坝洪水模拟及损失评估研究

    Research on Flood Simulation and Loss Assessment of Dagaotan Reservoir Dam Break

    • 摘要: 水库溃坝洪水具有突发性强、破坏范围广的特点,科学评估其风险对下游区域防灾减灾至关重要。现有的一维模型无法反映洪水横向动态、而传统的二维模型又存在计算效率无法满足实时预警时效性的问题。为此,本研究通过将GPU加速的二维模型与渐进式溃坝边界条件相结合,构建了一种高效精准的水库下游洪水模拟与损失评估方法。以大高滩水库为例,构建了土石坝溃坝模型、水库下游水动力洪水演进模型、损失分析模型为一体的耦合评估模式。结合该水库坝体材质,采用逐渐溃坝模型计算不同溃坝工况下溃口流量,作为二维水动力学模型的流量边界,利用高分辨率DEM数据为水动力学模型网格赋值,用土地利用数据为模型糙率参数赋值,模拟淹没水深、淹没范围和淹没历时等洪水风险因素,并使用基于GIS的空间权重叠置分析方法评估经济损失。研究结果表明,大高滩水库下游全溃洪水先蔓延至低洼区并积水,随后淹没范围与水深扩大,达最大后逐步消退;区域洪水演进呈上游受冲强烈、下游滞洪缓冲规律,上游特征点水深峰值出现早、数值高且积水可完全消退,下游特征点水深峰值延后、数值低且积水也能消退,河边低洼区域则因地形影响存在 24h 滞水;设计洪水位局部溃坝、校核洪水位局部溃坝、全溃三种工况下,淹没水深与面积呈递增趋势,且总淹没面积的显著增加主要源于高水深区域面积扩大;溃坝严重程度与经济损失呈正相关,全溃方案造成的经济损失(4.39亿元)远超局部溃坝方案,且其对民居和农业的破坏性尤为突出。本研究可为大高滩水库下游的防洪避险提供具体参考,帮助相关部门更科学地制定应急方案和减灾措施,在实际防汛工作中具有应用价值。

       

      Abstract: Reservoir dam break flood are characterized by their sudden onset, intense destructive power, and wide impact range. Scientifically assessing their risks is therefore crucial for effective disaster prevention and mitigation in downstream regions. However, while existing 1D models cannot capture lateral flood dynamics, conventional 2D models lack sufficient computational efficiency to support real-time early warning applications. To address these challenges, this study develops an efficient and accurate flood simulation and damage assessment framework for downstream reservoir regions by integrating a GPU-accelerated 2D hydrodynamic model with progressive dam-break boundary conditions. Taking the Dagaotan Reservoir as a case study, a coupled evaluation model is constructed, consisting of an earth–rock dam failure model, a downstream hydrodynamic flood evolution model, and an economic loss assessment module. Based on the reservoir’s dam material properties, a gradual dam failure model is applied to compute the outflow hydrograph under different breach scenarios, which serves as the discharge boundary for the 2D hydrodynamic model. High-resolution DEM data are used to parameterize the computational grid, while land-use data inform the spatial distribution of roughness coefficients. Key flood risk indicators—inundation depth, extent, and duration—are simulated, and economic losses are estimated using GIS-based spatial overlay analysis. The results indicate that floodwaters initially spread to low-lying regions and accumulate, followed by expansion of the inundation area and water depth, eventually reaching a peak before gradually receding. The downstream flood evolution exhibits a pattern of severe upstream flooding with detention and attenuation effects in downstream areas. Peak water depths appear earlier and are higher at upstream points, while downstream peaks are delayed and lower. Under the three dam-break scenarios—partial breach under design flood level, partial breach under check flood level, and complete breach—both inundation depth and extent increase progressively. The expansion of high-depth zones contributes most to the total increase in inundated area. Economic loss is positively correlated with the severity of dam failure. The total breach scenario results in an estimated loss of 439 million yuan, substantially exceeding that of partial breach scenarios, with particularly severe impacts on residential and agricultural areas. This study provides a valuable reference for flood control, risk management, and emergency planning in the downstream area of the Dagaotan Reservoir, offering practical implications for real-world flood mitigation efforts.

       

    /

    返回文章
    返回