Landslide prediction based on improved principal component analysis and mixed kernel function least squares support vector regression model

Landslide probability prediction plays an important role in understanding landslide information in advance and taking preventive measures.Many factors can influence the occurrence of landslides,which is easy to have a curse of dimensionality and thus lead to reduce prediction accuracy.Then the generalization ability of the model will also decline sharply when there are only small samples.To reduce the dimension of calculation and balance the model’s generalization and learning ability,this study proposed a landslide prediction method based on improved principal component analysis(PCA)and mixed kernel function least squares support vector regression(LSSVR)model.First,the traditional PCA was introduced with the idea of linear discrimination,and the dimensions of initial influencing factors were reduced from 8 to 3.The improved PCA can not only weight variables but also extract the original feature.Furthermore,combined with global and local kernel function,the mixed kernel function LSSVR model was framed to improve the generalization ability.Whale optimization algorithm(WOA)was used to optimize the parameters.Moreover,Root Mean Square Error(RMSE),the sum of squared errors(SSE),Mean Absolute Error(MAE),Mean Absolute Precentage Error(MAPE),and reliability were employed to verify the performance of the model.Compared with radial basis function(RBF)LSSVR model,Elman neural network model,and fuzzy decision model,the proposed method has a smaller deviation.Finally,the landslide warning level obtained from the landslide probability can also provide references for relevant decision-making departments in emergency response.

From active fault segmentation to risks of earthquake hazards and property and life losses—A case study from the Xianshuihe-Xiaojiang fault zone

Strong earthquakes generally rupture along active faults,and associated ground motion can cause earthquake disasters,property losses,and casualties from kilometers to tens of kilometers away.Therefore,one of the most effective ways to find earthquake’s dangerous parts of faults is to study the seismic hazards on fault segments.After that,we can also evaluate the probabilities of landslides hazard,property losses,and casualties.In this study,using fault slip rates and magnitude-frequency relationship as constraints,we calculated the earthquake occurrence rates for the segments along the Xianshuihe-Xiaojiang fault zone.We obtained 11 sites of single-segment or multi-segment rupturing risk.We also provided these potential events conditional probabilities in the next 30 years.For the 11 potential earthquakes,we calculated the property loss of residential buildings in the ground motion field.The most significant property loss is CNY 7.65 billion caused by the single-segment rupturing of the F19 segment on the Anninghe fault.We applied the deep learning neural network method in predicting the number of casualties for the potential earthquakes,showing that the most significant event is the multi-segment rupturing of the F29 and F30 segments on the Anninghe fault with the predicted death number of 279-317.We also evaluated the probabilities of earthquake landslides after the potential earthquakes.The results show that areas with intense compressional tectonic stress are highly unstable and prone to earthquake induced landslides,including the southern section of the Yuke fault,the southern section of the Xianshuihe fault,and the conjugated area between the southern section of the Daliangshan fault and the Lianfeng fault.These areas have a considerable number of earthquake landslides with probabilities>10%.The methodology and results will give us a new effective way of applying active fault data in earthquake hazard and risk analysis and provide a scientific path for earthquake prevention,disaster reduction,and emergency rescue preparation.