Font Size: a A A

Spatial And Temporal Risk Analysis Of Landslide Disaster In Changbai Mountain Area

Posted on:2018-02-10Degree:MasterType:Thesis
Country:ChinaCandidate:X W LanFull Text:PDF
GTID:2310330515454980Subject:Structural engineering
Abstract/Summary:PDF Full Text Request
In recent several decades,geological conditions in ChangBai Mountain area have changed continuously as well as human construction activities in certain region are also increasing accelerately,people have gradually cognitived the relevant landslide hazard risk.As a result,it is a high time that we should take effective approaches to solve the problem of landslide disaster forecast in space and time perspective.Meanwhile,the evaluation of geological hazard areas is not only the basis of analyzing damage but also the scientific evidence to formulate prevention plan of disaster.In this circumstance,firstly,the weight of the influence factor of landslide is analyzed separately.Then,the holistic analysis of the study area carried out by RES and safety entropy.Later,our paper utilizes GIS to evaluate the overall landslide hazard risk in study area,then to evaluate each landslide sub-domain.Figure out classification graph according to the dangerous degree.We can conclude that the extremely dangerous area,the dangerous area,the middle safe area,the safe area and the steadfast safe area account for the total area of the syudy area 9.9%,22.3%,30.2%,24.0%,13.6%,respectively.In deeply investigating a lot of information,we also found the vast majority of landslide disaster is closely related to precipitation.In fields relevant to water resources,many phenomena in accordance with Markov property.Meanwhile,when we take much quantitative analysis method into consideration theoretically,we have learned the merits about the prediction theory of Markov process such as high accuracy,low volume of historical data needed,less dependent on historical data and prediction effect improved greatly,respectively.From another opinion in actual situation,many rainfall induced landslides often happened abruptly and bring unpredictable harm to human beings.So this article collects a large amount of cases about rainfall induced landslide disaster in history(because of many realistic factors,this essay mainly analyzes rainfall induced landslides in YanBian Korean autonomous state),our essay also uses Markov process to analyze precipitation state in study area and recurrence period of rainfall induced landslide in different kinds of precipitation conditions.Finally,we use actual rainfall induced landslide disaster to testify the rationality of our conclusion.The verification results are satisfactory.It is worth mentioning that geological conditions are poor in ChangBai Mountain region,most part of its areas belongs to mountainous topography,which it is easy to cause landslides during rainfall.Through the practice,we have proved that the research on rainfall induced landslide prediction method in YanBian Korean antonomous state can also be applied to other parts in the study area.Likely,we can draw the right conclusions in appropriate analysis.In terms of results form analysis of Markov process,precipitation forecast is the basic work of rainfall induced landslide prediction.The accuracy of precipitation forecast affects the evaluation of precipitation state and rainfall induced landslides.So we take advantange of BP neural network prediction model for simulating annual precipitation in certain syudy area.When we compare forecast data and the real value,we found that the prediction accuracy is about 85%.So we can verify the feasibility of the method of annual rainfall forecasting.On the basis of our previous achievements,it is worth mentioning that if we are likely to predict the landslide disaster caused by the precipitation in a certain year successfully,afterwards,this result can also be combined with landslide hazard map,in this method,we can find out and defense against the high degree of landslide hazard areas as well as reduce disaster losses.Finally,we implement landslide risk assessment in the unity on space and time.By combining the spatial analysis and time period,it is easy to know that landslides will occur in a particular area at a given time.This combination of temporal and spatial risk analysis can,to some extent,overcome the limitations of single angle analysis,such as from a spatial perspective or only from a time perspective,and can improve the accuracy of landslide hazard prediction.In summary,this essay not only analyzes the possibility of landslide disaster in study area,but also discusses the characters of rainfall induced landslide disaster in this region.Our eassy also gives the rainfall prediction corresponds to the certain study region.Meanwhile,our conclusions will give some proposal suggestions to prevent and mitigate landslide disaster in certain area.
Keywords/Search Tags:landslide, GIS, landslide hazard risk, Markov process, rainfall induced landslide, recurrence period, BP neural network, prediction model
PDF Full Text Request
Related items