A Study On Early Warning And Real-Time Risk Assessment Of Rainfall-Induced Landslide Based On Multi-Source Data

Landslide is a kind of frequent geological disaster,which brings great threat to people’s lives and property.As one of the control measures to prevent landslides,landslide monitoring and early warning can reduce the occurrence of landslide disaster.The casualties and property losses caused by landslide disaster can be decreased by this way to some extent.At the same time,the result of landslide risk assessment can provide bases for decision makers to make and implement decisions.In this study,a comprehensive warning model of rainfall-induced landslide based on multi-source data was established by analyzing the characteristics of rainfall-induced landslide and typical landslide warning models.Then,the quantitative evaluation model of landslide risk was established based on the probability of warning grade.Finally,based on the comprehensive landslide warning model and risk assessment model,the real-time warning and risk assessment of Slope 4 of Xinwuwei in Shenzhen city was realized.In this study,typical single index early-warning models(rainfall early warning model,physical state of soil early warning model and displacement early warning model)were used to deal with all kinds of monitoring data separately,and the corresponding early warning level could be calculated according to each type of monitoring data;Based on the variable weight theory in real-time for each type of early warning model to give different weights,a comprehensive landslide warning model was established by weighted calculation;The landslide model test data was substituted into the comprehensive early warning model to verify the comprehensive early-warning model do realize multi-index classification early warning of rainfallinduced landslide,and its advantages were analyzed: compared with the rainfall early warning model and the physical state of soil early warning model,the warning result of comprehensive early warning is more accurate;compared with the displacement warning model,the comprehensive warning model can advance the warning time of the sudden rainfall-induced landslide and provide more time to avoid the disaster.In the study of landslide risk assessment,a quantitative landslide risk assessment model was established based on early warning probability.The model used the realtime monitoring data of the slope and took the probability of the first warning grade as the slope instability probability to calculate the risk quantitatively.The real-time assessment of the slope risk was realized by this way.At the same time,through the statistical analysis of landslide and the relevant parameters of accumulation body,the landslide slip distance,the accumulation body thickness and the accumulation body width empirical formula were obtained,and the calculation method of slope body risk assessment was simplified,and facilitate the real-time assessment of slope body risk.In addition,through the statistics of real-time slope risk,the slope risk in any target period can be calculated,which is conducive to more analysis of the slope.This paper applied the comprehensive warning model and risk assessment model to Slope 4 of Xinwuwei in Shenzhen city.It can be concluded from the warning result that the slope is in a safe state at present.According to the real-time risk assessment results,when the slope instability probability is 0.5,the economic risk is 1.436 20 billion yuan,and the personnel risk is 1 844 people,indicating that the slope damage will cause great losses.According to the time period risk assessment results,the economic risk of this slope is 24.413 7 million yuan in the past year,and the personnel risk is 21 people,so the slope needs attention.