一种基于贝叶斯理论的区域斜坡稳定性评价模型

本文结合滑坡物理模型和统计模型的优点,针对小流域滑坡稳定性分析,建立了一种基于贝叶斯理论的区域斜坡稳定性评价模型。该模型主要采用灾害自身信息来修正原始模型中的参数,解决了区域稳定性评价中参数难以确定的问题。该方法首先设定模型的初始参数分布,然后利用采样点雨前和雨后稳定性不同的信息建立验证方程,再根据马尔科夫链蒙特卡罗模拟和贝叶斯方法确定最终的参数分布,进而得出区域稳定性分布。应用该模型对福建省蔡源小流域的6 13滑坡群发性事件进行分析。结果显示,蔡源小流域地区的无量纲黏聚系数C为0.028,有效摩擦角为16.7,土壤的导水系数T和降雨量q的比值为529.026m,可能不稳定地区和不稳定地区达到76.0%。该模型利用历史灾害数据自动模拟出合适的参数,对区域滑坡稳定性评价具有重要意义。

BAYES THEORY BASED MODEL FOR REGIONAL SLOPE STABILITY ANALYSIS

A Bayesian theory based model is developed for the assessment of regional terrain stability. This model combines the statistical model with the infinite slope stability model. The uncertainty of parameters in the assessment of regional terrain stability is eliminated because the parameters can be adjusted according to the real historical events. The adjustment of parameters are based on the assumption that the failed slope is stable before it rains and unstable after it rains. First, the initial distribution of parameters is set up. Then, the final distribution of parameters is calculated through MCMC(Markov Chain Monte Carlo) methods. This model is applied to the landslide events in a watershed, in Caiyuan Village, Fujian Province, China. The parameters of the model are adjusted using the proposed approach in this study case, where the adjusted dimensionless cohesion C of the area is 0.028,the adjusted effective friction angle is 16.7, and the adjusted ratio of hydraulic conductivity to rainfall is 529.026m. The unstable areas take up to 76.0% of the whole area. This model, which automatically adjusts parameters according to historical landslide events, would be a useful tool for landslide hazard assessment.