论文部分内容阅读
针对密云水库上游密云县境内山洪泥石流灾害多发的特点,首次将Bayes判别分析法原理运用于自然灾害的预测预报。运用系统科学和山洪泥石流预报原理,对密云水库上游密云县境内影响山洪泥石流发生的系统内部要素和系统外部环境条件进行分析,确定用影响山洪泥石流发生的前15 d的实效雨量和当日雨量作为预报模型因子,并建立起一组山洪泥石流预报模型,经自身验证法和交互验证法对模型进行检验,其判断正确率均为82.4%,达到了Bayes判别分析要求的判别模型自身验证正确率>80%的要求,两种方法检验结果一致,模型表现稳定,表明所建立的模型可靠,可以投入使用。研究结果可为该地区的山洪泥石流预测预报提供一种较为简单易行的方法,为密云县山洪泥石流灾害防治提供一定的支撑,以减少山洪泥石流发生所导致的灾害损失。
In view of the frequent occurrence of flash floods and mudslides in Miyun Reservoir in the upper reaches of Miyun Reservoir, the principle of Bayes Discriminant Analysis is applied to the prediction of natural disasters for the first time. Using system science and flash flood debris flow forecasting principle, the system internal factors and system external environmental conditions that affect the flash flood in Miyun County upstream of Miyun Reservoir are analyzed, and the actual rainfall and daily rainfall of the first 15 days that affect flash flood flow are determined Model, and set up a set of flood debris flow forecasting model, the self-verification method and cross-validation method to test the model, the correct rate of judgment is 82.4%, and the discriminant model that meets Bayes discriminant analysis is verified by the correct rate of> 80 % Of the requirements of the two methods test results are consistent, the model showed stable, indicating that the established model is reliable and can be put into use. The results of this study can provide a relatively simple and feasible method for the forecasted flood and debris flow in this area, and provide certain support for the prevention and control of flash floods and mudslides in Miyun County so as to reduce the disaster losses caused by flash floods and debris flows.