Landslide Identification And Monitoring In The Middle Reaches Of Bailong River Using Sentinel-1A

In China's vast mountainous area,landslide disasters occur frequently,meanwhile,due to the influence of the weather,terrain,traffic and so on,there are many difficulties in the identification and monitoring of regional landslides.In order to solve the problem,this paper discussed the identification and monitoring of landslides in the middle reaches of Bailong River based on SBAS-InSAR technology by taking the images of Sentinel-1A ascending and desending tracks as the source data.And this paper identified the potential landslides in the study area by making use of the remote sensing interpretation signs in combined with the distribution of deformation points.And the author analyzed of the temporal and spatial characteristics of landslides via the GIS technology and statistical methods.Finally,the research verified the results of landslides identified and the effectiveness of deformation points monitoring.The obtained results and conclusions drawn were as follows:1.It showed that there were no large difference by analyzing visualized partition maps obtained by the images of the Sentinel-1A ascending and descending tracks respectively in study area.Superimposing historical landslides and visibility partition maps,it was found that most of the historical landslides were located in the visible region,a few of them located in the low sensitive region,yet no historical landslide located in the invisible region,which indicated that images of Sentinel-1A ascending and descending tracks were suitable for the study on the identification and monitoring of landslides in the mountainous area in China.2.The 8 potential landslides and 1 dangerous rock were identified in the study area by making use of remote sensing interpretation signs in combined with the distribution of deformation points.Furthermore,historical landslides and field investigations have verified the validity of the results,and it is concluded that the images of descending track are more suitable for potential landslide identification in the study area.3.In the condition of the no inferior morphable model,landslide monitoring is effective by using the deformation points obtained by SBAS technology in the study area.(1)By means of analyzing the dominant factors of slope,aspect,fault and stratigraphic lithology on the slope and the distribution of landslides,historical landslides were most widely distributed in slopes of 20-30 degrees,and they are yet rarely distributed in slopes of more than 40 degrees.The landslides are mainly distributed in the southwest,because of the characteristics of light intensity and humidity on the sparse vegetation.The faults are the main factor to control the landslides.The influence of formation lithology on landslides is mainly the soft-hard interbeddings and the juncture of structural line,which is the high incidence area of landslides.(2)Using the method of Geostatistical Analyst,it is found that the numbers of the deformation points and the high velocity are mainly distributed in the front,and it shows that there are faults passing through the landslides with the distribution of high velocity.From the projection trend,the landslides controlled by the faults show the inverted "U"type in the north-south direction,in another word,the velocity is high in the middle of landslides,and it is low in the two sides of landslides.Due to the de-coherence effect,the distribution and velocity of deformation points can not be judged in the east-west direction.(3)The temporal characteristics of landslides could be detected by the velocity of deformation points,the accumulative deformation and monthly deformation.The amount of the sliding velocity and the deformation are the largest from May to September,which is related to the heavy precipitation of summer,meanwhile,it is observed of abnormal monthly deformation from September to December of both 2014 and 2015,The possible reason for this phenomenon is related to snowfall and groundwater recharge.It is concluded that the images of ascending track are more suitable for the landslide monitoring.