基于降雨强度-历时的安徽省黄山市滑坡分组阈值研究

目的]对安徽省黄山市降雨型滑坡的降雨强度—历时阈值进行研究，拟为黄山市各区县精细化滑坡预警提供技术支持。方法]以黄山市2004—2019年历史滑坡点及其对应的降雨数据为研究样本，基于经验性阈值Caine模型，通过比较滑坡前有效降雨量的方式确定诱发滑坡降雨历时，建立各区县降雨强度—历时(I-D)阈值曲线。考虑到I-D曲线在灾害预警中的实用性，以歙县滑坡点为例，将该县诱发滑坡的降雨事件按照气象预报的雨强标准分为小中雨组、大雨组和暴雨及以上组。根据雨强的分组情况建立歙县复合型降雨I-D阈值。结果]黄山市降雨型滑坡多由中长历时降雨诱发，复合型I-D阈值更能体现各雨况对滑坡的诱发作用，此阈值可直接使用气象预报的降雨对滑坡进行预警。结论]I-D阈值具有强烈的区域适用性，依据气象预报对降雨强度分组的滑坡复合型降雨阈值曲线，可使用气象预报的降雨数据进行区域滑坡预警。

Landslide Classification Threshold in Huangshan City of Anhui Province Based on Rainfall Intensity-Duration

Objective] The rainfall intensity-duration threshold of rainfall-induced landslides in Huangshan City, Anhui Province was studied in order to provide technical support for detailed landslide early warning in all districts and counties of Huangshan City. Methods] Rainfall duration of induced landslides was determined by comparing the effective rainfall before landslides at historical landslide points, and by using corresponding rainfall data in Huangshan City from 2004 to 2019. The rainfall intensity-duration (I-D) threshold curve for each district and county was established based on the empirical threshold Caine model. Considering the applicability of I-D curve in disaster early warning, and taking the landslide point in Shexian County as an example, the rainfall events that induced landslides in Shexian County were divided into the small and moderate rain group, the heavy rain group, and the heavy rain or above group according to the rain intensity standard of the meteorological forecast. The I-D threshold of combined rainfall in Shexian County was established according to the grouping classification of rain intensity. Results] Most of the rainfall-induced landslides in Huangshan City were induced by long and medium duration rainfall. The complex I-D threshold could better reflect the inducing effect of various rainfall conditions on landslides, and this threshold could be directly used for early warning of landslides using precipitation forecasts. Conclusion] I-D threshold has strong regional applicability, and the landslide rainfall threshold grouped by rainfall intensity according to precipitation forecasts is suitable for the construction of a landslide early warning system in this region.