Remote Sensing Extraction And Suspectibility Assessment Of Earthquake-triggered Debris Flows In The Wenchuan County

Wenchuan County,with complex geological conditions,intense tectonic movement and the bigger terrain up-downs. In these regions,mountain hazards,such as collapse,landslide,debris flow disasters,happens more frequently,the hazards cause serious threat to people’s life, property security, and social economic development. After the earthquake, the distribution and evolution characteristics of secondary geological disasters is an important basis for disaster prevention andmitigation.But due to the disaster area conditions and inconvenience traffic, it’sonly have to survey along the both sides of river and road.Remote sensing has advantages,such as,non-time limit,faster to access information,lower cost, it provide a convenient channel for regional dynamic monitoring of geological disasters.On the basis of field investigation, combined with remote sensing technology,this study use both decision tree and maximum likelihood classification method to establish the rule base extraction of geological hazard information,effectively extract4period post-earthquake landslides and other geological disasters from2009to2013,dynamic monitoring the distribution and evolution of the geological disasters.And on this basis in2013,we established the rule extraction about Wenchuan County debris flow information,in order to further extract the verification of the debris flow.At last,by correlation model,susceptibility evaluatingof Wenchuan County gully.The main research harvests are as follows:1) Based on the multi-source and multi temporal remote sensing data, combinedwith the basic geographic data, field survey data and GIS techniques, the secondarygeological disaster remote sensing interpretation mark were established by theimage analysis of the secondary geological disaster and spectroscopy with itssurrounding disaster pregnant environment difference.2) Select the training sample data from different years in Wenchuan County,analysis the spectral characteristics about each disasters type, information extractionis proposed based on decision tree model, realizes the extraction4year periodlandslide information from2009to2013.3) Use the maximum likelihood and the decision tree method to classify theWenchuan County landslide, after Accuracy verification, the decision tree method inthe extraction of landslides and other information is obviously better than themaximum likelihood method.4) Based on the landslide remote sensing information extraction,the39debrisflow gullies along Duwen highway were taken as training object,established7indicators extraction rules in2013,such as the relative elevation, average slope,watershed area, disaster area, the formation of area and density, circulation areadisaster area of debris flow, the debris flow remote sensing were extraction inWenchuan County. The results show that, the debris flow gully misjudgment numberwere10,accounting for Wenchuan County total4.4%, accounting for18.8%of thetotal number of the identification results,the identification accuracy to meetemergency needs.5) By the method of the grey correlation model,the debris flow gully susceptibility evaluation in Wenchuan County were complete after selecting the10 analysis factor in five aspects.Evaluation results show that moderate debris flow number occurred in Wenchuan county was137, accounting for the total areaof debris flow gully60.35%,concentrated in the Minjiang River.Along the Minjiang River on both sides,67.4%channel is very high susceptibility, especially inMinjiang water left, reached100%.Extremely high and high susceptibility debrisflow are mainly distributed in4townships,such as Yingxiu,Geng Da, Miansi,Yinxing,accounting for25.5%of the total debris flow gully,and most serious in the Yingxiu and Yinxing township. Low and very low susceptibility of debris flow are mainly distributed in Wolong,Long Xi and the caopo,27.3%of the total debris flow.