Satellite And Ground-based InSAR Time Series Modeling And Solution Method For Landslide Dynamic Monitoring And Early Warning

Landslide is one of the most widely distributed natural disasters in the world.It often brings huge catastrophic damage,resulting in a large number of casualties and property losses.Accurate and effective monitoring of landslide is not only the basis of early warning of landslide,but also an important means of landslide risk assessment and disaster prevention and reduction.The development,evolution and failure of landslide are accompanied by a large number of changes in measurable variables.The surface displacement is the characteristic value that most intuitively reflects the deformation of landslide.The deformation state of landslide can be predicted by analyzing the displacement monitoring data,so as to reduce the loss caused by landslide through early warning.It is a difficult and important task.Synthetic aperture radar interferometry(InSAR)is a newly developed space-earth observation technology.It has become a powerful technical means to measure surface deformation with high spatial-temporal resolution.It has the technical advantages of high measurement accuracy,large operation range and no restriction by weather conditions.At present,it has been widely used in the research fields of volcanoes,earthquakes,glaciers,landslides and urban surface subsidence.Nowadays,with the launch of the new generation of high-resolution SAR satellite platforms,the higher spatial-temporal resolution and coverage provide new opportunities for the routine monitoring of geological disasters such as landslides.At the same time,a variety of ground-based InSAR platforms are increasingly used in landslide emergency monitoring.The data availability of these sensor platforms with different imaging geometry further enriches the extraction ability of deformation information required for comprehensive monitoring and interpretation of landslide geological disasters.However,a large number of data processing and complexity make the application of InSAR in landslide monitoring more challenging.To meet the above challenges,this paper first introduces the unique observation geometry and imaging mode of satellite and ground-based InSAR,discusses in detail the target positioning spatial framework when they are applied to landslide monitoring,and expounds the projection conversion method based on the target positioning spatial framework.Then,the surface migration geometric framework of InSAR landslide monitoring is further analyzed,and the relevant experimental analysis also verifies the logical correctness and feasibility of the geometric framework.Finally,based on the analysis of the geometric framework of surface movement of InSAR landslide monitoring,and making full use of the respective outstanding advantages of satellite and ground-based InSAR in landslide monitoring,the concept of landslide whole life cycle monitoring based on satellite and ground-based time series InSAR is proposed,and a widely applicable improved tangent angle landslide early warning criterion is combined to realize the purpose of landslide InSAR dynamic monitoring and early warning.In this process,combined with the monitoring characteristics of satellite and ground-based InSAR,the main difficulties faced in the whole life cycle monitoring of InSAR landslide are analyzed.This paper mainly focuses on two aspects: the insufficient temporal resolution of satellite InSAR landslide dynamic monitoring and the difficulty of accurate positioning and interpretation of ground-based InSAR landslide emergency monitoring deformation.Specifically,the free and open access to the earth observation SAR satellite images has enabled the implementation of landslide time-series monitoring,but the low temporal resolution under the satellite revisit period restrictions is difficult to meet the requirements of quasi-real-time dynamic monitoring of landslide.Moreover,with the increase of SAR satellite platforms,a corresponding data integration algorithm is urgently needed to obtain the complementary deformation information from every single observation orbit.In this work,an integrated algorithm for landslide multi-source deformation optimization estimation based on Kalman filter(KF)is proposed to improve the temporal resolution and realize the dynamic monitoring and prediction of landslide.Specifically,the landslide migration coordinate system is established firstly,and the InSAR deformation results in line of sight(LOS)are projected to the downslope direction.Then,with the introduction of the acceleration component into the process covariance matrix of prediction model and the observation noise variance weight determination,the downslope displacements of multiplatform InSAR observations are dynamically integrated into a unified time series by KF dynamic prediction and correction.So as to achieve the high temporal resolution monitoring of landslide.For validation purpose,the Baige landslide in Tibet,China is selected as the test area,and 55Sentinel-1A ascending images,42 Sentinel-1A descending images and 10 ALOS-2PALSAR-2 images collected over this area from May 2017 to April 2019 are used to estimate the high temporal time series.The temporal resolution of landslide monitoring is successfully improved from 12 days with a single orbit to the shortest 1 day and repeatedly 2–5 days with multiple platforms,and the prediction of subsequent displacement is also realized.The monitoring example results verify the effectiveness and reliability of the proposed method.Prospectively,with the continuous increase of the InSAR satellite platforms,this proposed algorithm can provide quasi-real-time monitoring data for landslide and better assist relevant emergency response,which is necessary for the dynamic monitoring and early warning of landslide.For landslides that have been proved to be in the accelerated deformation stage(especially the critical sliding stage)by satellite InSAR,or landslides that have occurred and are in the rescue and disaster relief stage,the main monitoring demand is the emergency monitoring of landslides,and the core requirement is the accuracy and timeliness of monitoring data(minute level to hour level).In recent 10 years,ground-based synthetic aperture radar monitoring technology(GB-InSAR)has developed rapidly.Its extremely high monitoring accuracy(up to sub-millimeter level)and extremely high spatial-temporal resolution(minute level time update,decimeter grid resolution)are more and more widely used in landslide monitoring and emergency response,playing a vital role.Although GB-InSAR has extremely high deformation monitoring accuracy and resolution,due to the special slant range projection imaging mode of radar,its fan-shaped grid coordinates are difficult to accurately match with the local topography in geospatial space,which makes it difficult to accurately interpret the deformation area and disaster targets in landslide emergency monitoring.For the purpose of identifying the deformation target and its position accurately,it is urgent to perform high-accuracy geocoding of the GB-InSAR images and transform them from 2D plane coordinate system to 3D local coordinate system.To overcome the difficulty of control points selection of GB-InSAR images in complex scattering environment by using traditional methods,high-resolution DSM obtained by unmanned aerial vehicle(UAV)aerial photogrammetry is used to establish the high-accuracy GB-InSAR coordinate transformation model for the first time,and an accurate GB-InSAR image geocoding method based on solution space search is proposed.In this method,three modules are used to achieve the purpose of geocoding: the unification of coordinate elements framework,the establishment of transformation model,and the solution space search of minimum Euclidean distance.Applying this method to Laoguanjingtai(LGJT)landslide monitoring experiment in Hailuogou Glacier,the sub-pixel geocoding precision is realized,and the effectiveness and accuracy of this method are verified by contrastive analysis.The method proposed in this paper can be applied to the accurate 3D interpretation and spatial-temporal law analysis of GB-InSAR deformation monitoring and it is worth to be popularized and applied.Finally,based on the three-stage evolution law of landslide surface displacement,the characteristics of InSAR time series curve are studied,and the time series is introduced to improve the tangent angle landslide early warning criterion,and a landslide early warning model combining time series InSAR monitoring and improved tangent angle criterion is proposed.As a verification,based on the satellite InSAR high-temproal resolution quasi real-time monitoring time series and GB-InSAR real-time emergency monitoring time series,the landslide early warning attempt and stability analysis are carried out for the Baige landslide of Jinsha River and the LGJT landslide of Hailuogou Glacier.The deformation stage of landslide is well divided,and its corresponding early warning level is determined.Finally,through the dynamic updating landslide time series InSAR monitoring in this paper,combined with the improved tangent angle early warning criterion,the landslide deformation stage division and stability analysis are effectively carried out,so as to achieve the purpose of landslide monitoring and early warning by InSAR.This method effectively makes up for the deficiency of InSAR technology in landslide dynamic monitoring and early warning,and has wide applicability and great popularization and application value.