Abstract: The multi-scale encoding system for landslide and rock fall hazard and interpretation with Monte Carlo model is applied to hazard assessment in China. Counties and cities of China are encoded according to landslide and rock fall hazard information. Monte Carlo simulation is used to interpret some codes as reference, and the other codes are interpreted by matching them with reference. The whole China is zoned into five regions according to interpreted code hazard degrees that include very low, low, medium, high and very high. The correlation analysis shows that landslide and rock fall codes have high correlations with codes of geology factor and geomorphology factor. Verification of BP neural network, comparison between results interpreted by Monte Carlo model and historical disaster record and comparison between proposed approach and other approach indicate that results given by proposed approach are sufficient accuracy.

Key words: landslide and rock fall, code, Monte Carlo simulation, hazard, China

摘要： 将滑坡崩塌危险性多尺度编码及其蒙特卡罗解译方法(简称编码方法)应用于中国滑坡崩塌危险度识别的研究。在全国范围内, 以县市为基本单元对滑坡崩塌危险性信息进行编码, 并通过蒙特卡罗模型解译部分单元的编码作为参照组, 其余单元的编码则可通过与参照组单元编码的快速匹配运算确定其危险度。根据编码解译结果, 将中国按照滑坡崩塌危险度分为高、较高、中、较低、低5个类区。相关分析表明, 滑坡崩塌危险性编码与地质码及地形码相关性比较高。 神经网络的检验、蒙特卡罗解译结果与历史灾害记录的对比, 以及编码方法计算结果与其他方法计算结果的对比, 共同验证了编码方法具有较高的准确性。

关键词: 滑坡崩塌, 编码, 蒙特卡罗, 危险度, 中国