洪水灾害是世界范围内发生最为频繁的自然灾害。中国是世界上洪水风险增长最显著的地区之一。针对甘肃省武威市山洪灾害多发的实情,采用结合主客观权重的博弈论组合赋权的洪水评价方法,应用GIS技术,从致灾因子危害性、孕灾环境稳定性和承灾体脆弱性三方面出发,共选取7个指标构建洪水风险评价指标体系,得到了各评价单元的洪水风险度空间分布。评价结果表明：武威市整体洪水风险等级为较低风险,与历史洪水事件数据吻合较好。其中,洪水高风险区只占总面积的4.94%,集中分布于凉州区与古浪县交界处和天祝藏族自治县西南部地区;较低风险区与低风险区之和占总面积的43.84%,集中于民勤县的东部沙漠区。通过将单一赋权方法与博弈论组合赋权法进行了比较,表明博弈论组合赋权法可以减少单一赋权方法的主观性,提高洪水风险图的精度,为洪水风险管理提供依据。

Floods are the most frequent natural disasters in the world, and the Chinese region has shown the most significant increase in flood risk worldwide. Based on the frequency of flash floods in Wuwei City, Gansu Province, a flood risk assessment method combined with subjective and objective weights and based on game theory was adopted and GIS technology was applied. Seven indicators were selected to construct a flood risk assessment index system from three aspects: disaster dangers, disaster environment sensitivity, and suffering flexibility. Then, the risk zoning chart of flood disaster in Wuwei was drawn, which showed that the overall flood risk level of Wuwei City is low and flood risk shows a decreasing distribution characteristic from southwest to northeast. The results are in good agreement with historic flood event data in which the highest risk area accounts only for 4.94% of the total area, which is concentrated at the junction of Liangzhou District and Gulang County and southwest of Tianzhu County. The sum of the lowest risk area and the lower risk area accounts for 43.84% of the total area, which is concentrated in the eastern desert area of Minqin County. The main sources of flood risk vary among counties in Wuwei City. The risk of disaster-causing factors presents a belt-like distribution characteristic that decreases from southwest to northeast. The highest risk areas are distributed mainly in the Qilian Mountains of Tianzhu County, and the lowest risk areas are concentrated in lower reaches of the Shiyang River in the northeast. The distribution of the disaster environment sensitivity is relatively uniform except for the Qilian Mountains with higher stability. There are obvious differences in the vulnerability of disaster-bearing bodies, and the vulnerability at the junction of Liangzhou County and Gulang County is much higher than that of other regions. The entropy method and analytic hierarchy process are used for comparison with the game theory combination weighting method, which demonstrates that the combination weighting method of game theory can reduce the subjectivity of the single weighting method, improve the accuracy of the flood risk map, and provide valuable information for flood management.