Analysis of prediction and prevention countermeasures of shallow loess landslides in the construction of residential areas

Abstract

In loess plateau areas affected by perennial heavy flooding, landslide disasters occur frequently, posing serious challenges to residential construction. In this paper, the prediction and prevention countermeasures of landslides are proposed by analyzing and predicting the monitoring data of loess landslides in a residential area under construction. The study used an intelligent variable frequency landslide displacement meter for displacement monitoring, combined with the moving average method and slope model for early warning of landslide occurrence. The results show that using this method, four sudden landslides were warned with high accuracy and the maximum advance warning time was 33 hours. The slope model performs well in the prediction of landslide destabilization time, and the confusion matrix analysis of the model shows that the prediction accuracy of 3m resolution data is the highest, with a true rate of 0.68, a false-positive rate of 0.40, and a precision of 0.65. The analysis of the success curve shows that the accuracy of the model prediction is improved significantly with the improvement of the resolution of data in the digital elevation model, which is especially suitable for the prediction of point landslides. In this paper, the reliability of the slope model is verified, and a method of landslide deformation stage division based on the tangent angle model is proposed, which provides a scientific basis for the prediction and prevention of landslide disasters in residential areas under construction.