Multi-factors fusion method of debris flow prediction based on genetic programming

Debris flow is a frequent geological disaster, which often poses a great threat to the safety of people’s lives and property. Outbreak of debris flow is not only related to rainfall, but also related to many geological and environmental factors. In this paper, the watershed area, ratio of loose materials and the average slope of the gully bed are taken as the geological factors,the maximum hourly rainfall intensity (T) and the total rainfall (R) are taken as the rainfall index, the sample database is established by means of geological factors and rainfall index, genetic programming is used to establish a prediction model for the critical rainfall index of debris flow, which overcomes the shortcomings of the previous warning model that used rainfall as a single indicator, model verification results show that the model has high warning accuracy and strong adaptability, which can realize timely warning.