遵义市不同地质灾害易发区滑坡临界雨量研究

目的]确定遵义市不同地质灾害易发区的滑坡临界雨量,建立阈值模型并对其进行检验评估,为该地区滑坡预测提供科学依据。方法]利用遵义市2010—2016年59次滑坡事件对应的逐小时降雨资料,采用统计方法对不同地质灾害易发区的滑坡临界雨量的历时、降雨类型对比分析,建立阈值模型并对其进行检验评估。结果]中锋型是主要的降雨类型;大于10和20mm/h强降雨基本来自滑坡发生当天,但中易区降雨历时均长于高易区的降雨历时。对于高易区来说,滑坡当日1h最大雨量与前期3d的有效雨量组合的阈值模型预报准确率最大,因此将滑坡当天1h最大雨量作为高易区滑坡发生的激发雨量。而对于中易区来说,滑坡当天3h最大雨量是其滑坡发生的激发雨量。结论]不同等级地质灾害易发区其滑坡临界雨量不同。对已经确定的滑坡阈值模型需根据新增滑坡信息进行检验评估,如果预测模型对新增滑坡预测不准确,需调整其滑坡判别线,从而对阈值模型进行订正。

Critical Rainfall of Landslides at Zunyi City in Different Geological Hazard Prone Regions

Objective] The objectives of this study were to determine the critical rainfall of landslide and establish the threshold model of landslide in different geological hazard prone areas of Zunyi City in order to provide scientific basis for landslide prediction in this area.Method] Hourly precipitation data from 2010 to 2016 were collected from automatic meteorological stations according to 59 landslides of Zunyi City in the Northern Guizhou Province. According to the duration and type of rainfall, the landslide were classified by using statistical analysis in different geological regions and the prediction model of threshold was established and assessed.Results] Heavy rainfall of more than 10 mm/h and 20 mm/h generally occurred on the day of landsliding, but the duration of rainfall in the middle susceptible area was longer than in the high susceptible area. The uniform pattern of rainfall is the main pattern. For the the high susceptible area, the forecasting accuracy of the threshold model was the best using the parameters the maximum rainfall of 1 hour during the landsliding day and the cumulative rainfall of two days before the landsliding day. Therefore, the maximum rainfall of 1 hour during the landslidubg day was the triggering rainfall. For the middle susceptible area, the forecasting accuracy of the threshold model was the best using the parameters of the maximum rainfall of 3 hours during the landsliding day and the cumulative rainfall of the landsliding day. Therefore, the maximum rainfall of 3 hours was the triggering rainfall.Conclusion] The critical rainfall of landslide varied among different geological regions. According to the new landslide information, the identified critical rainfall forecasting model is tested and evaluated. If the forecasting model is inaccurate for the new landslide, the landslide discriminant line must be adjusted so that the adjusted prediction model accurately predicts the landslide.