中国河北太行山、燕山地区，地形起伏大，短时强降雨频繁，山洪灾害十分活跃，严重影响了流域内人们的生产和生活。保险作为转移灾害风险最有效的手段之一，日益成为学界研究的热点。

本研究在假定雨洪同频的条件下，以河北省石家庄市小作河流域为例，利用HEC-HMS和FLO-2D耦合模型，分析了15种年遇型降雨情景下山洪的危险性，并根据小作河流域房屋暴露特点及已有文献中的房屋脆弱性曲线，定量评估了山洪灾害农房损失风险；而后在足额办理承保手续的条件下，以保险人承担80%经济损失的标准下，定量分析了小作河流域内山洪灾害农房保险的损失风险；进而依据相关政策，利用巨灾保险费率精算模型，厘定了政策性山洪灾害农房保险的费率（即在政府补助下，居民自己承担的费率），以期为山区小流域山洪灾害农房保障及风险防范工作提供科学依据。得到以下主要结论：

（1）基于HEC-HMS水文模型和FLO-2D水动力模型，构建了可用于小流域山洪灾害危险性评估的水文水动力耦合模型，并利用典型山洪事件（即7·19山洪事件）对构建的耦合模型进行了验证。其中，HEC-HMS模型在洪峰流量及事件总流量的模拟上，精度高，相对误差不足2%；FLO-2D模型在山洪淹没范围上模拟精度高达85.12%，在淹没水深上的模拟精度较高，均方根误差（Root Mean Square Error, RMSE）和纳什效率系数（Nash and Sutcliffe Efficiency, NSE）分别为0.11和0.86。

（2）基于小作河流域居民房屋分布特点，参照相关文献中山洪灾害房屋的脆弱性曲线，分析了15种年遇型降雨情景下，小作河流域内山洪灾害房屋的损失风险。研究表明，重现期1年一遇（1a）到10000年一遇（10000a）降雨情景下，小作河流域房屋的淹没范围1.30-2.89km²，按照当地2017年房屋建造价格994元/m²计算，损失约为2579-73707万元。高风险区域（损失≥20万/900m²）主要分布在流域下游人口聚集区域；当降雨重现期为10000a时，高风险的房屋面积为1.08km²，占比37.37%；损失约为27544万元。

（3）基于灾害、风险和保险三者间的关系，以小作河流域内不同降雨情景下山洪灾害房屋损失风险为基础，在足额办理承保手续的条件下、保险人承担80%经济损失的赔付标准下和考虑保险公司成功运行条件下，利用非寿险精算中经典的聚合风险模型厘定了小作河流域山洪灾害农房保险的纯费率约为2.01‰、保险费率约为3.03‰。为更好地推广山洪灾害农房保险产品，在相关政策辅助下，厘定了小作河流域山洪灾害农房政策性保险的费率约为0.91‰。按照投保人足额办理承保手续的条件，民众在进行山洪灾害农房保险投保时，只需缴纳0.90元/平方米的保费，户均需缴纳保费47元。如将该政策性保险推广整个河北省，费率在0.52~5.94‰之间变化，人均需缴纳的保费为22~258元，平均为78元，该费用约为河北省人均可支配收入的6‰。本研究所设计山洪灾害农房政策性保险的投保费用在绝大多数群众的可承受范围之内。

针对山丘区小流域的山洪灾害风险及其农房保险展开深入研究，有助于提高小作河流域山洪灾害风险现状评估的准确性，有助于推广流域内山洪灾害农房保险产品，进一步减轻流域内“因灾致贫、因灾返贫”的现象，对山丘区城镇安全与发展具有重要的现实意义。

The Yanshan Mountain and Taihang Mountain, Hebei Province, China, have large terrain undulations, frequent short-term heavy rainfall, and active mountain torrent disasters, which have seriously affected the production and life of people in the basin. Insurance, as one of the most effective means to transfer disaster risks, has increasingly become a hotspot in academic research.

In this paper, assuming the same frequency of rain and flood, taking the Xiaozuo River Basin in Shijiazhuang City, Hebei Province as an example, using HEC-HMS and FLO-2D coupling models, the risk of mountain torrents under 15 rainfall scenarios is analyzed, and based on the Xiaozuo River Basin The housing exposure characteristics and the vulnerability curve in the existing literature have quantitatively evaluated the risk of rural housing losses caused by mountain torrents. Then, under the conditions of full underwriting procedures, the insurer will assume 80% of the economic losses. The risk of loss of rural housing insurance for mountain torrent disasters in the river basin; and in accordance with relevant policies, the actuarial model of catastrophe insurance rates was used to determine the policy-based mountain torrent disaster housing insurance rate (that is, the rate that residents would bear under government subsidies) ), in order to provide a scientific basis for the protection of farm houses and risk prevention of mountain torrent disasters in small watersheds in mountainous areas. The main conclusions are as follows:

(1) Based on the HEC-HMS hydrological model and the FLO-2D hydrodynamic model, a hydrological and hydrodynamic coupling model that can be used to assess the risk of mountain torrent disasters in small watersheds is constructed, and a typical mountain torrent event (i.e., 7.19 mountain torrent event) is used to construct the coupling model was verified. Among them, the HEC-HMS model has high accuracy in the simulation of flood peak flow and total event flow, with a relative error of less than 2%; the FLO-2D model has a simulation accuracy of 85.12% in the range of mountain flood submergence, and has a higher simulation accuracy in the submerged water depth, Root Mean Square Error (RMSE) and Nash and Sutcliffe Efficiency (NSE) are 0.11 and 0.86, respectively.

(2) Based on the distribution characteristics of residential houses in the Xiaozuo River Basin, and referring to the vulnerability curves of houses in mountain torrent disasters in related documents, the loss risks of houses in the Xiaozuo River Basin under mountain torrent disasters under 15 annual rainfall scenarios are analyzed. Studies have shown that under the rain scenario with a return period of 1 year to 10,000 years, the submerged area of houses in the Xiaozuo River Basin is 1.30-2.89km², calculated according to the local house construction price of 994 yuan/m² in 2017. The loss was approximately 25.79-737.07 million yuan. High-risk areas (loss ≥200,000 yuan/900m²) are mainly distributed in the lower reaches of the river basin with concentrated populations; when the rainfall return period is 10,000 years, the area of high-risk houses is 1.08km², accounting for 37.37%; the loss is about 275.44 million yuan.

(3) Based on the relationship between disasters, risks, and insurance, and based on the risk of housing losses caused by mountain floods under different rainfall scenarios in the Xiaozuo River Basin, the insurer shall bear 80% of the economic loss compensation under the conditions of full contract procedures Under the standard and considering the successful operation of the insurance company, using the classic aggregate risk model in non-life insurance actuarial calculations, the net premium rate of the rural housing insurance for mountain flood disasters in the Xiaozuo River Basin is determined to be about 2.01‰, and the insurance premium rate is about 3.03‰. In order to better promote the insurance products of mountain torrent disaster farm housing, with the assistance of relevant policies, the premium rate of the policy insurance for mountain torrent disaster farm housing in the Xiaozuo River Basin is determined to be about 0.91‰. In accordance with the conditions for the insured to go through the underwriting procedures in full, the people only need to pay a premium of 0.90 yuan/square meter for the rural housing insurance for mountain flood disasters, and each household needs to pay a premium of 47 yuan. If the policy insurance is promoted throughout Hebei Province, the premium rate varies from 0.52 to 5.94‰, and the premium per capita is 22 to 258 yuan, with an average of 78 yuan, which is about 6‰ of the per capita disposable income of Hebei Province. The research institute designed the insurance cost of the rural housing policy insurance for mountain torrent disasters to be within the affordable range of the vast majority of the people.

In-depth research on the flash flood risk and rural house insurance in small watersheds in hilly areas improves the accuracy of the assessment of the current situation of flash flood risk in the Xiaozuo River Basin, promotes the flash flood rural house insurance products in the watershed, and reduces poverty due to flash floods in the river basin, which is important practical significance for the safety and development of villages and towns in mountain areas.