The Distribution Characteristics And Hazard Assessment Of Landslide Under Various Earthquake And Rainfall Scenarios

China has a large territory,complicated natural and geographical conditions,and significant neotectonic motions,which cause earthquakes on a regular basis.Because of their great intensity and widespread dispersion,they can readily cause a variety of geological disasters such as landslides,collapses,debris flows,ground subsidence,ground cracks,and other calamities.Geological disaster sites are widely scattered,have many hidden dangers,and are difficult to prevent.At the same time,my country has the world’s most severe geological disasters and the most vulnerable population.With the advancement of remote sensing technology in recent years,there has been an increase in the number of high-quality earthquake and landslide databases.Construction of earthquake and landslide databases has made significant progress,particularly since the2008 Wenchuan earthquake.However,unlike earthquake occurrences,the compilation of landslide databases triggered by heavy rainfall events is still very sluggish,and there are currently relatively few heavy rainfall landslide databases.Furthermore,during the last 30 years,an increasing number of studies have been conducted on the prediction of rainfall-triggered landslides utilizing the infinite slope model in conjunction with the hydrological model.Given the existing abundance of regional rainfall landslide physical evaluation models,developing a quick rainfall landslide physical evaluation model suitable for broad regional scales and conducting regional rainfall landslide susceptibility assessments requires additional research.To address the aforementioned challenges,research on rainfall landslides and topography,geomorphology,and rainfall in different regions was conducted using the2013 Tianshui heavy rainfall landslide database in my country’s northwestern region and the 2019 rainfall and landslide database in the Beiling region of Longchuan,Guangdong in the southeast region.The relationship between various influencing elements,and so on,in order to assess the effect of various influencing factors on the incidence of rainfall and landslides.The susceptibility of rainfall and landslides in Tianshui region in 2013 and Longchuan region in Guangdong in 2019 was evaluated using two physical models to assess the forecast capacity of physical models in regional rainfall and landslides.We Choosed a watershed basin in the Shetang area with a high concentration of loess landslides,simulate the rainfall infiltration process and calculate the instability coefficient using the TRIGRS model under different rock,soil,and landform conditions,and conduct research on the rainfall threshold curve using the physical model.The landslide susceptibility evaluation under different rainfall conditions and the landslide susceptibility evaluation under different earthquake conditions were then performed using the TRIGRS model to produce a more objective and scientific national landslide susceptibility evaluation distribution map.Finally,using the Minxian earthquake landslide database as the research topic,the generalized additive model is applied to perform geographical prediction of landslide area distribution using mathematical statistics and GIS spatial analysis methods.This study completed a series of tasks in the six areas listed below:(1)Landslides caused by excessive rainfall in the Tianshui area in 2013: a spatial analysis This study analyzes the spatial distribution of landslides in Tianshui region,and studies different influencing factors(elevation,slope,aspect,terrain relief,lithology,river density,road density,rainfall data)and landslides based on a detailed and complete landslide database of Tianshui heavy rainfall events in 2013.Simultaneously,the 2013 Minxian earthquake landslide database was chosen to compare the development of rainfall-triggered landslides with earthquake-triggered landslides at slope locations.On the steep slopes,similar "claw" debris flows emerge,and the clustering characteristics are clearly visible.Landslides,in general,are more common in places with steeper slopes.Furthermore,landslides are more common in the Tianshui area’s S,SE,and E aspects,and these three aspects are all sunny slopes in terms of yin and yang slope qualities,indicating that landslides are more common on the sunny side.Landslides in the Tianshui area have a stronger association with the water system than those in the Minxian area.Larger landslides may reach the river in the Tianshui area,accounting for more than 5% of the total number of landslides,13%of the total landslide area,and 16% of the total landslide area.T The volume of landslides reaches the river channel,indicating that larger landslides in the Tianshui area are more likely to reach the river channel,but for the Minxian earthquake,the opposite phenomenon exists,with 7.8 percent of the number of landslides and 4.5percent of the landslides reaching the river channel.The area and 2.8 percent of the landslide volume reached the river channel,of which more than 5% of the number of landslides,13% of the landslide area,and 16% of the landslide volume reached the river channel,indicating that larger landslides in the Tianshui area are more likely to reach the river channel.It demonstrates that,during the Minxian earthquake,thethe landslides that often occur in the middle and lower part of the slope are relatively small-scale landslides,while the larger-scale landslides tend to occur near the ridge.(2)Spatial analysis of landslides caused by excessive rainfall in the Beling area in Longchuan,Guangdong in 2019: based on high-resolution pre-and post-disaster remote sensing pictures from Google Earth,as well as Sentinel 2 images in the study area as a supplement(including pre-disaster and post-disaster images)The dates are April 17 th and September 24 th,with a 10 m resolution).A database of rainfall landslides caused by significant rainfall in the Beling area of Longchuan,Guangdong in 2019 was created using the artificial visual interpretation method.Simultaneously,rainfall and landslide spatial prediction study was carried out using the Matlab version of the TRIGRS model(MAT.TRIGRSV1.0).According to the findings,this rainfall event caused 667 landslides,the majority of which occurred in Mibei and Yanhua villages in the study area’s central region.The results suggest that the majority of the landslides are located between 15° and 30°.Landslides are more likely to occur in places with relatively high terrain relief than in areas with low terrain relief.At the same time,as river distance increased,so did the density of landslides,indicating that landslides were more concentrated in places closer to the water system.The results demonstrate that the majority of the regions with a high likelihood of instability(blue areas)are on both sides of the river valley,i.e.,on quite steep slopes.The landslides themselves are essentially the same in the high-risk predicted areas.The landslide generated by the rainfall event was primarily caused by the significant rainfall event on June 10,with subsequent rains having little impact on the occurrence of the landslide.This work demonstrates that MAT.TRIGRSV1.0 has promising application possibilities in the risk prediction of regional rainfall-triggered landslides,and it provides a new application tool for the TRIGRS model’s continued promotion in rainfall and landslide prediction.(3)Based on a physical model,researchers investigated the triggering mechanism of loess landslides caused by heavy rain in the Tianshui area in 2013.The 2013 Tianshui area was chosen as the research area to assess regional rainfall landslide susceptibility using a physical model.Based on prior research,the corresponding soil data within50km2 of the study area were collected,and the FSLAM physical model was utilized to assess regional rainfall and landslide susceptibility.The change in the landslide stability coefficient due to rainfall events was compared,as were the impacts of four separate rainfall events on the capacity to initiate the landslide.When the relationship between the four heavy rainfalls and the distribution of landslides is compared,it can be seen that there is no obvious correlation between the first rainfall and the distribution of landslides,but the first and second heavy rainfalls are the largest among the four heavy rainfalls,so this rainfall can be considered.During the process,early rainfall should be prepared for the formation of later disasters;on the basis of the previous two heavy rainfalls,the scouring of heavy rainstorm and the infiltration of rainwater produced the spatial distribution state of group catastrophes in the later period once again.The catastrophe has hysteresis,and this rainfall did not generate large scale disasters;the rainfall of the third heavy rainfall is the smallest of the four rainfall processes,and its effect on the development of subsequent disasters should be reflected in the outcomes of the second heavy rainfall.The amount of rainfall in the previous heavy rainfall event was nearly identical to that in the second,and this rainfall event should provide the final push for the entire mass disaster event.(4)TRIGRS model research on rainfall and landslide threshold curves in the Tianshui region: Choose a basin in the Shetang region with a high concentration of loess landslides to conduct research on rainfall threshold curves using a physical model based on the TRIGRS model.The rainfall threshold curve was produced by simulating the duration results of 21 distinct rainfall intensity scenarios ranging from 1 mm/h to100 mm/h in the range of rainfall intensity.Overall,the scale parameter is inversely related to the shape parameter.The lower the shape parameter,the larger the scale parameter.When the scale parameter reaches a specific value,the shape parameter’s value remains essentially unaltered.Overall,as the slope increases,the shape and scale parameters drop continually,indicating that the slope has a significant influence on the fitting of the rainfall threshold curve.The shape and size parameters stay essentially unchanged as the thickness of the rock and soil increases.(5)Assessment of rainfall and earthquake-triggered landslides on a national and regional scale: We estimated the stability coefficients and phases under different rainfall conditions for the current regional rainfall in light of the challenges that exist in the application of current physical models at national and regional sizes.The associated instability likelihood,where the rainfall intensity is 100mm/day,200mm/day,and500mm/day,and the rainfall duration is 2h,4h,8h,10 h,and 20 h.In this manner,the national stability coefficient distributions under 15 rainfall scenarios based on different rainfall intensities and rainfall duration combinations are determined.We employed the Newmark model to calculate the national earthquake landslide risk prediction under different ground motion conditions in order to carry out the countrywide earthquake and landslide probability prediction and evaluation.There are ten levels of landslide prediction probability distribution map.The results suggest that locations with poor stability coefficients are primarily located in my country’s southwest and central regions,with the southwest area being primarily concentrated in the northeastern edge of the Qinghai-Tibet Plateau.Central and western Sichuan,central and northern Yunnan,and eastern Tibet are the key locations with poor anticipated slope stability coefficients.The central region is primarily centered in the Loess Plateau region,which encompasses the majority of Shaanxi,western Shanxi,northern Chongqing,and western Gansu.Taiwan’s center hilly region has a poor stability coefficient value as well.(6)Using the 2013 Minxian earthquake as an example,spatial prediction of regional landslide spatial scale information: This study uses the generalized additive model to conduct seismic research on the 2013 Minxian earthquake and landslide database,with the support of mathematical statistics and GIS spatial analysis tools.Landslide scale spatial prediction research The research region is separated into slope units based on r.slopeunits,and the slope unit is employed as the evaluation unit to carry out the spatial prediction of landslide scale.Elevation,slope,slope curvature,TWI,fault distance,epicenter distance,road distance,lithology,and other criteria that may influence the scale of the landslide are chosen,and the estimated landslide area of different slope units is determined based on these selections.The findings indicate that the total upper slope and fault distance have a substantial influence on the spatial distribution of landslide area,and that the slope gradient has a large linear effect on the occurrence of landslides.Overall,the projected landslide area is nearly identical to the actual landslide distribution.However,it should be noted that the slope unit with a relatively small landslide area and the slope unit area with a huge landslide area have a more noticeable lagging effect.Prediction findings are rather small in more developed locations.