A landslide risk monitoring method and system based on single-track InSAR data in high mountain and canyon area

Abstract

The application discloses a kind of high mountain and gorge area landslide risk monitoring method and system based on single-track InSAR data, it is related to high mountain and gorge area landslide monitoring technical field, including steps: obtaining the single-track visual deformation point array data and digital elevation model data of target area in each phase;Obtain regional slope;Get vertical deformation data;Extract to obtain reliable coherent data points;The data filling of vertical deformation is obtained to obtain incoherent initial filling data points, and incoherent reliable filling data points are obtained;Obtain the reliable deformation time series point cloud graph corresponding to observation period;Based on reliable deformation time series point cloud graph, obtain the period deformation data of target area in observation period.The method provided by the application does not carry out filling repair to the InSAR missing data points caused by shooting shielding, carries out filling repair to the InSAR missing data points caused by landslide related physical reasons, enhances landslide risk identification data amount, and is favorable to improve landslide risk identification precision.

Description

Technical Field

[0001] This invention relates to the field of landslide monitoring technology in high mountain and canyon areas, and in particular to a method and system for landslide risk monitoring in high mountain and canyon areas based on single-track InSAR data. Background Technology

[0002] In mountainous and canyon regions with dramatic topographic relief, conventional dual-track InSAR observations struggle to simultaneously acquire effective data covering the same area. Existing technologies often suffer from insufficient data volume and inaccurate landslide risk identification due to issues such as missing data and incoherence.

[0003] Therefore, there is an urgent need for a landslide risk monitoring method and system based on single-track InSAR data in high mountain and canyon areas to solve or at least alleviate some of the above-mentioned shortcomings. Summary of the Invention

[0004] The main objective of this invention is to provide a method and system for monitoring landslide risks in high mountain and canyon areas based on single-track InSAR data, aiming to solve the technical problem of insufficient landslide risk identification data in the prior art.

[0005] To achieve the above objectives, this invention provides a method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data, comprising the following steps:

[0006] S10, acquire the single-track line-of-sight deformation lattice data and digital elevation model data of the target area in each period. The single-track line-of-sight deformation lattice data includes the initial coherent data points corresponding to the coherent area and the initial decoherent data points to be filled corresponding to the decoherent area.

[0007] S20, calculate the regional slope based on the digital elevation model data. The vertical deformation data of each initial coherent data point in the target area is calculated based on the slope of the area and satellite information.

[0008] S30, the initial coherent data points are denoised to extract reliable coherent data points;

[0009] Obtain associable fillable decoherent data points from the initial decoherent data points to be filled; perform vertical deformation data filling on the associable fillable decoherent data points based on the reliable coherent data points to obtain the initial decoherent fillable data points; and extract the reliable decoherent fillable data points from the initial decoherent fillable data points.

[0010] S40, combine the reliable coherent data points with the decoherent reliable filling data points to obtain the reliable deformation point cloud map corresponding to each period;

[0011] S50, repeat the above steps until a reliable deformation time series point cloud map corresponding to the observation period is obtained;

[0012] S60, Based on the reliable deformation time-series point cloud map, obtain the periodic deformation data of the target area within the observation period.

[0013] Furthermore, in step S30, the formula is used. The confidence level of the current data point i is calculated, where, For radar incident angle, The slope of the area is given. The angle between the slope of the current data point i and the radar azimuth angle;

[0014] The initial coherent data points corresponding to a current confidence level greater than the confidence threshold are identified as the trusted coherent data points;

[0015] The decoherent initial fill data point corresponding to the current confidence level being greater than the confidence threshold is determined as the decoherent reliable fill data point.

[0016] Furthermore, it also includes the following steps:

[0017] S70, based on the periodic deformation data, obtain the maximum vertical deformation value of the target area;

[0018] S80, determine the current risk level based on the maximum vertical deformation value of the area, the slope of the area, and a preset risk level table.

[0019] Furthermore, in step S30, the data filling for vertical deformation is performed using the linear interpolation method.

[0020] Furthermore, it also includes the step of dividing the correlated incoherent data points into current interpolation points in the core area A of the landslide body, current interpolation points in the transition area B of the landslide body, and current interpolation points in the edge area C of the landslide body based on the geological survey data, regional slope, and aspect data of the target area.

[0021] In step S30, the data filling of the associable and fillable incoherent data points by vertical deformation based on the reliable coherent data points specifically includes:

[0022] Interpolation is performed on all current interpolation points in the core area A of the landslide body based on a first number of adjacent valid reliable coherent data points; interpolation is performed on all current interpolation points in the transition area B of the landslide body based on a second number of adjacent valid reliable coherent data points; and interpolation is performed on all current interpolation points in the edge area C of the landslide body based on a third number of adjacent valid reliable coherent data points.

[0023] Furthermore, the current interpolation point corresponding to the core area A of the landslide body... Centered on the recognizable non-recognizable data points, the reliable coherent data points within the first adjacent effective number of filtering ranges are defined as the first reference data points. ;

[0024] Using formula Perform calculations to obtain the current interpolation point. Vertical deformation final interpolation ,in, The total number of the first reference data points within the filtering range. The first reference data point within the filtering range The vertical deformation value, The first reference data point With the current interpolation point Distance points, For the gradient correction coefficient of the core area;

[0025] Using formula The gradient correction coefficients for the core region are calculated.

[0026] Using formula The average deformation gradient value within the selection range was calculated. , The first reference data point within the filtering range The vertical deformation gradient value.

[0027] Furthermore, the current interpolation point corresponding to the transition zone B of the landslide body is used. Centered on the recognizable non-recognizable data points, the reliable coherent data points within the second adjacent effective number of filtering ranges are defined as the second reference data points. ,

[0028] Using formula Perform calculations to obtain the current interpolation point. The final interpolation of the vertical deformation, where, This represents the total number of second reference data points within the selected range. The second reference data point within the filtering range The vertical deformation value, The second reference data point With the current interpolation point Distance points, This is a correction system for adjusting the aspect deviation angle;

[0029] Using formula The slope aspect deviation angle adjustment correction system is obtained through calculation, where, This is a correction system for adjusting the aspect deviation angle. , The current interpolation point slope direction, The second reference data point within the filtering range slope direction, The value range is [0°, 180°].

[0030] Furthermore, the current interpolation point corresponding to the landslide edge region C is taken as... Centered on the recognizable non-recognizable data points, the reliable coherent data points within the third adjacent effective number of filtering ranges are defined as the third reference data points. ,

[0031] Using formula Perform calculations to obtain the current interpolation point. Initial interpolation of vertical deformation ,in, This represents the total number of third reference data points within the filtering range. The third reference data point within the filtering range The vertical deformation value, The third reference data point With the current interpolation point The distance points;

[0032] like Then determine ;like Then determine ; Indicates the current interpolation point Vertical deformation transition interpolation;

[0033] like Then determine ;like Then determine ; Indicates the current interpolation point The final interpolation of the vertical deformation, The current interpolation point in the edge zone C of the landslide body. Adjacent points that have been filled.

[0034] Furthermore, the value of the first number of adjacent valid numbers is less than the value of the second number of adjacent valid numbers, and the value of the second number of adjacent valid numbers is less than the value of the third number of adjacent valid numbers.

[0035] Further, determine the target area The area whose slope aspect data is consistent with the dominant direction of the landslide is the core area A of the landslide body;

[0036] Determine the target area Part of the area is the transition zone B of the landslide body;

[0037] Determine the target area The portion of the area is the edge zone C of the landslide body.

[0038] The present invention also provides a landslide risk monitoring system for high mountain and canyon areas based on single-track InSAR data, including a processing device, which is used to implement the steps of the above-mentioned landslide risk monitoring method for high mountain and canyon areas based on single-track InSAR data.

[0039] The present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data.

[0040] The landslide risk monitoring method for high mountain and canyon areas based on single-track InSAR data provided by this invention effectively converts single-track LOS deformation into vertical deformation through slope and aspect correction. Monitoring based on vertical data helps improve the accuracy of deformation monitoring in high mountain and canyon areas. By employing a method based on the analysis of initial decoherent data points to be filled, and after determining the type of factors in the decoherent area, the method does not fill or repair InSAR missing data points caused by image obstruction, but fills and repairs InSAR missing data points caused by landslide-related physical causes, thereby enhancing data reliability. Applicability analysis is used to select highly reliable data, reducing errors caused by terrain and observation geometry. This increases the amount of landslide risk identification data, which is beneficial to improving the accuracy of landslide risk identification. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0042] Figure 1 This is a flowchart illustrating a landslide risk monitoring method for high mountain and canyon areas based on single-track InSAR data in one embodiment of the present invention.

[0043] Figure 2 This is an example of an InSAR deformation point cloud map of the target region before filling, as shown in one embodiment of the present invention.

[0044] Figure 3 This is a filled InSAR deformation point cloud map (credible deformation point cloud map) of the target area in one embodiment of the present invention. Detailed Implementation

[0045] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0047] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0048] Furthermore, the use of terms such as "first" and "second" in this invention is for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of that feature. Additionally, the technical solutions of the various embodiments can be combined with each other, but only on the basis of being achievable by those skilled in the art. When the combination of technical solutions is contradictory or impossible to implement, such a combination of technical solutions should be considered non-existent and not within the scope of protection claimed by this invention.

[0049] Please refer to the appendix. Figure 1 , Figure 2 and Figure 3 This invention provides a method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data, comprising the following steps:

[0050] S10, acquire the single-track line-of-sight deformation lattice data and digital elevation model data of the target area in each period. The single-track line-of-sight deformation lattice data includes the initial coherent data points corresponding to the coherent area and the initial decoherent data points to be filled corresponding to the decoherent area.

[0051] S20, the regional slope is calculated based on the digital elevation model data;

[0052] The vertical deformation data of each initial coherent data point in the target area is calculated based on the slope of the area and satellite information.

[0053] S30, the initial coherent data points are denoised to extract reliable coherent data points;

[0054] Obtain associable fillable decoherent data points from the initial decoherent data points to be filled; perform vertical deformation data filling on the associable fillable decoherent data points based on the reliable coherent data points to obtain the initial decoherent fillable data points; and extract the reliable decoherent fillable data points from the initial decoherent fillable data points.

[0055] S40, combine the reliable coherent data points with the decoherent reliable filling data points to obtain the reliable deformation point cloud map corresponding to each period;

[0056] S50, repeat steps S10 to S40 until a reliable deformation time series point cloud map is obtained;

[0057] S60, Based on the reliable deformation time-series point cloud map, obtain the periodic deformation data of the target area within the observation period.

[0058] The landslide risk monitoring method for high mountain and canyon areas based on single-track InSAR data provided by this invention effectively converts single-track LOS deformation into vertical deformation through slope and aspect correction. Monitoring based on vertical data helps improve the accuracy of deformation monitoring in high mountain and canyon areas. By employing a method based on the analysis of initial decoherent data points to be filled, and after determining the type of factors in the decoherent area, the method does not fill or repair InSAR missing data points caused by image obstruction, but fills and repairs InSAR missing data points caused by landslide-related physical causes, thereby enhancing data reliability. Applicability analysis is used to select highly reliable data, reducing errors caused by terrain and observation geometry. This increases the amount of landslide risk identification data, which is beneficial to improving the accuracy of landslide risk identification.

[0059] Please refer to this again. Figure 2 Currently, when using InSAR data to monitor the deformation trend of high mountains and canyons, the large amount of incoherent data caused by geological conditions makes it difficult to effectively predict the deformation trend of high mountains and canyons.

[0060] Understandably, in step S10, the monitoring period for each phase can be 12 days. Single-track InSAR data is acquired through a single-track InSAR satellite, and digital elevation model (DEM) data of the target area is acquired through the DEM data processing module.

[0061] Research has revealed that the terrain in high mountain and canyon areas is highly undulating, and some areas are prone to occlusion during InSAR satellite observations. The decoherence caused by occlusion is unrelated to landslide deformation, and filling it would introduce errors. In the solution of this invention, the causes of decoherence are first distinguished, and after retaining the decoherence data points that can be correlated and filled due to landslide physical causes, interpolation filling is performed. Specifically, the single-track line-of-sight deformation matrix data includes two types of data points. One type is the initial coherent data points corresponding to the coherent region, that is, the data points with stable signals and no obvious interference in InSAR observations, which can be directly used for deformation analysis after processing based on confidence thresholds. The other type is the initial decoherence data points to be filled corresponding to the decoherent region, that is, the data points whose signals are decoherent due to landslide physical causes, terrain occlusion, etc. Among them, the decoherence data points caused by terrain occlusion are not filled, and only the decoherence data points caused by landslide physical causes are filled to ensure the rationality of data distribution.

[0062] Understandably, InSAR satellites acquire line-of-sight periodic deformation data, while the core of landslide risk monitoring is vertical deformation, that is, deformation perpendicular to the horizontal plane. Vertical deformation directly reflects the uplift and settlement characteristics of the landslide body. In the scheme of this invention, the vertical deformation data is obtained by combining the regional slope β, which can eliminate the influence of terrain undulation on deformation monitoring and improve monitoring accuracy.

[0063] In one specific embodiment of the present invention, single-track LOS deformation data, a high-precision DEM, and initial decoherent data points to be filled corresponding to the decoherent region are prepared. The terrain slope and aspect are calculated using the DEM. Combined with information such as radar incident angle and satellite orbit direction, the LOS deformation is converted into vertical deformation using a formula conversion. Specifically, the terrain slope is first calculated using the DEM data, employing a 3x3 moving window. Calculate the slope of the center pixel ;

[0064] in;

[0065]

[0066]

[0067] X represents the east-west slope of the pixel center, y represents the east-west slope of the pixel center; cellsize represents the pixel size, with units consistent with geographic elevation. to These are the elevation parameters for the moving window. as well as These represent the change step size in the corresponding direction.

[0068] Based on the calculated slope and satellite information, the vertical deformation measured by InSAR can be expressed as:

[0069] in, For radar incident angle, The angle between the satellite's orbital direction and true north is indicated by "-" for ascending and "+" for descending. It is a slope.

[0070] In another embodiment of the present invention, the regional slope β corresponding to each data point in the target area is calculated by the DEM data processing module using the maximum slope method, which facilitates the provision of a basis for subsequent landslide zoning, vertical deformation transformation and interpolation correction based on the regional slope β.

[0071] Furthermore, in step S30, the formula is used. The confidence level of the current data point i is calculated, where, For radar incident angle, The slope of the area is given. Let the angle between the slope of the current data point i and the radar azimuth be defined. The initial coherent data point corresponding to a current confidence level greater than a confidence threshold is determined as the reliable coherent data point. The decoherent initial fill data point corresponding to a current confidence level greater than a confidence threshold is determined as the decoherent reliable fill data point. It can be understood that in the scheme of this invention, a uniform confidence level can be used to filter all current data points i, or different confidence levels can be set for different regions to filter current data points i. The higher the confidence level, the more reliable the data.

[0072] Understandably, in the solution of the present invention, a confidence threshold is used to filter the initial coherent data points, and a confidence threshold is used to filter the reliable coherent data points. Preferably, the confidence threshold is 0.5 to 0.8.

[0073] More preferably, in the solution of the present invention, based on the geological survey data of the target area and the regional slope and aspect data calculated by the digital elevation model, the area where the correlated incoherent data points are located is divided into a landslide core area A, a landslide transition area B, and a landslide edge area C; specifically, based on the geological survey data, regional slope, and aspect data of the target area, the correlated incoherent data points are divided into current interpolation points in the landslide core area A, current interpolation points in the landslide transition area B, and current interpolation points in the landslide edge area C; and for the current data points i in different partitions, the following is adopted: The adaptive confidence level is specifically set as follows: the confidence level T for the core area A of the landslide body is 0.55~0.6; the confidence threshold (confidence level) T for the transition area B of the landslide body is 0.6~0.65; and the confidence level T for the edge area C of the landslide body is 0.65~0.7. In the scheme of this invention, the confidence level threshold is adjusted by region to fit the deformation characteristics of different areas. The core area deforms violently, so appropriately lowering the threshold can avoid missing high-period deformation data points. The edge area deforms weakly, so appropriately raising the threshold can ensure data reliability. At the same time, it adapts to the characteristics of the difference in data reliability between the bottom and top of the slope in the landslide scenario.

[0074] Understandably, in step S60, the periodic deformation data includes the maximum deformation value, average deformation value, deformation rate (unit mm / period), and cumulative deformation of each region; among them, the maximum deformation value and deformation rate of the landslide core area A can be used as the core indicators for judging landslide risk, which are consistent with the characteristics of severe deformation in the landslide core area to ensure the accuracy of the assessment.

[0075] Furthermore, the method includes the following steps: S70, obtaining the maximum vertical deformation value of the region based on the deformation trend of the observation period; S80, determining the current risk level based on the maximum vertical deformation value of the region, the slope of the region, and a preset level table.

[0076] Understandably, in the present invention, a preset level table with differentiated deformation thresholds is set by combining slope and geological condition models; based on the vertical deformation data in the periodic deformation data, different stability levels of 1-5 are divided for different geological conditions, with levels 1-5 representing progressively increasing risks. Specifically, optimization and adjustment are made by comprehensively considering information such as optical images, slope, and geological structure: for example, the risk level is appropriately reduced for farmland, densely vegetated areas, and gentle slopes; the risk level is correspondingly increased for steep slopes, artificially excavated and filled areas, and water catchment areas, ultimately dividing potential landslide areas and high-risk landslide areas, forming a risk assessment result with clear zoning and levels; the setting of the risk level table ensures the scientific and reasonable nature of the risk level determination.

[0077] Furthermore, in step S30, a linear interpolation method (distance) or data filling with vertical deformation is used.

[0078] Furthermore, the method includes the step of dividing the associable incoherent data points into current interpolation points in the landslide core area A, the landslide transition area B, and the landslide edge area C based on the geological survey data, regional slope, and aspect data of the target area. In step S30, the data filling of the associable incoherent data points based on the reliable coherent data points specifically includes: interpolating and filling all current interpolation points in the landslide core area A based on a first number of adjacent effective reliable coherent data points; interpolating and filling all current interpolation points in the landslide transition area B based on a second number of adjacent effective reliable coherent data points; and interpolating and filling all current interpolation points in the landslide edge area C based on a third number of adjacent effective reliable coherent data points.

[0079] Understandably, the target area is divided into the landslide core area A, the landslide transition area B, and the landslide edge area C. Step S30 mainly includes the following steps: initial coherent data point denoising and screening; screening of associative fillable incoherent data points; reasonable partitioning and interpolation filling of associative fillable incoherent data points; and screening and extraction of incoherent reliable fillable data points from the initial incoherent fillable data points.

[0080] The initial coherent data point denoising and screening process includes: employing a Gaussian filtering algorithm to denoise the initial coherent data points, eliminating the influence of random noise such as satellite observation errors and environmental interference on the periodic deformation data; and selecting the reliable coherent data points by comparing the current confidence level with a confidence threshold. The current confidence level formula incorporates topographical factors from InSAR observations (radar incident angle, regional slope, and angle), where the radar incident angle is a fixed value, the regional slope is the regional slope corresponding to the current data point i, and μ... i The angle between the slope direction of the current data point i and the radar azimuth can effectively filter out coherent data points that are less affected by terrain interference and have high data reliability.

[0081] The screening of incoherent data points that can be associated and filled includes combining geological survey data (such as stratigraphy, landslide remains, fault distribution, etc.) and DEM data of the target area to screen out incoherent data points that can be associated and filled from the initial incoherent data points to be filled. Specifically, geological survey data of the target area is obtained through geological survey equipment to clarify the reasons for the incoherence of the initial incoherent data points to be filled. Incoherent data points caused by landslide physical reasons (such as changes in the surface medium caused by landslide body creep, loosening of rock and soil, etc.) are determined to be incoherent data points that can be associated and filled. Incoherent data points caused by non-landslide reasons such as terrain shading and vegetation cover are determined to be incoherent data points that cannot be associated and filled, and are directly removed without subsequent filling. Understandably, geological survey data can clearly define the distribution range and geological characteristics of landslide bodies. Combined with the slope and aspect information of DEM data, it can accurately distinguish the causes of incoherence, avoid errors introduced by filling in incoherent data points caused by non-landslide factors, meet the core requirement of monitoring only landslide deformation, and provide geological basis for subsequent landslide body zoning.

[0082] Reasonable zoning of associative, fillable, incoherent data points includes: based on geological survey data of the target area (such as the distribution of the landslide core sliding zone and landslide accumulation area) and regional slope β and aspect data, dividing the area where the associative, fillable, incoherent data points are located into three types of landslide body sub-regions. Specifically, this is explained in accordance with the principles of slope creep and slope landslides. After determining the dominant landslide direction based on geological survey data, the target area is then defined. The area where the slope aspect data aligns with the dominant landslide direction is defined as the core area A of the landslide body. Core area A is the core sliding region of the landslide body. According to the principle of slope creep, it exhibits the most severe deformation and a significant deformation gradient, with the largest difference in landslide volume between the slope bottom and top. Therefore, it is the core area for landslide risk monitoring. The target area is determined as follows: A portion of the area is the landslide edge zone C, located at the outermost edge of the landslide (i.e., the edge of the landslide accumulation area). According to the principle of slope creep, this area exhibits the weakest deformation, the gentlest deformation gradient, the smallest landslide volume, and the highest deformation stability; the difference in landslide volume between the slope bottom and top is negligible. The target area is thus defined. Part of the area is the transition zone B of the landslide body. The transition zone B is located between the core zone A and the edge zone C of the landslide body. The deformation intensity of the transition zone B is between that of the core zone and the edge zone, serving as a transitional area for the transfer of deformation from the core zone to the edge zone. According to the principle of slope landslides, the deformation exhibits a continuous and gradual characteristic, and the slope aspect has a significant impact on the deformation. In the scheme of this invention, the target area is divided into three types of regions, strictly following the non-uniform deformation characteristics of the landslide body. The core zone experiences severe deformation, the edge zone experiences weak deformation, and the transition zone connects the two. This not only conforms to the pattern of large changes at the center and small changes at the edge in InSAR creep monitoring, but also conforms to the physical principle of inconsistent landslide volume at the bottom and top of the slope and continuous and gradual deformation in slope landslides, providing a clear basis for subsequent differential interpolation.

[0083] Interpolation filling is performed on the associative filling of incoherent data points:

[0084] Including the current interpolation point corresponding to the core area A of the landslide body. Centered on the recognizable non-recognizable data points, the reliable coherent data points within the first adjacent effective number of filtering ranges are defined as the first reference data points. ;

[0085] Using formula Perform calculations to obtain the current interpolation point. Vertical deformation final interpolation ,in, The total number of the first reference data points within the filtering range. The first reference data point within the filtering range The vertical deformation value, The first reference data point With the current interpolation point Distance points, For the gradient correction coefficient of the core area;

[0086] Using formula The gradient correction coefficients for the core region are calculated.

[0087] Using formula The average deformation gradient value within the selection range was calculated. , The first reference data point within the filtering range The vertical deformation gradient value.

[0088] The core area A of the landslide body experienced severe deformation, and the current interpolation point was obtained. Initial interpolation of vertical deformation Then, combined with the gradient correction coefficient of the core area Adjusting the fill value can avoid the problem of the existing linear interpolation fill value being too low, and conforms to the principle of deformation concentration in the InSAR creep core area; at the same time, dynamically adjusting the correction coefficient can adapt to the gradient difference between the bottom and top of the slope in the core area, which is consistent with the law that the landslide volume at the bottom and top of the slope is inconsistent.

[0089] Including the current interpolation point corresponding to the transition zone B of the landslide body. Centered on the recognizable non-recognizable data points, the reliable coherent data points within the second adjacent effective number of filtering ranges are defined as the second reference data points. ;

[0090] Using formula Perform calculations to obtain the current interpolation point. The final interpolation of the vertical deformation, where, This represents the total number of second reference data points within the selected range. The second reference data point within the filtering range The vertical deformation value, The second reference data point With the current interpolation point Distance points, This is a correction system for adjusting the aspect deviation angle;

[0091] Using formula The slope aspect deviation angle adjustment correction system is obtained through calculation, where, , The current interpolation point slope direction, The second reference data point within the filtering range slope direction, The value range is [0°, 180°].

[0092] The transition zone B of the landslide body is the connecting area between the core area and the edge area. The deformation needs to remain continuous. The data source with consistent slope aspect has a deformation pattern that is closer to the deformation characteristics of the point to be interpolated. Therefore, by adjusting the correction coefficient by the slope aspect deviation angle, abrupt deformation can be avoided, which is in line with the principle of continuous and gradual deformation of the landslide slope. At the same time, the slope aspect has a significant impact on the deformation of the transition zone. The smaller the slope aspect deviation, the stronger the deformation consistency. The graded setting of the correction coefficient is in line with the actual deformation pattern.

[0093] Including the current interpolation point corresponding to the edge region C of the landslide body. Centered on the recognizable non-recognizable data points, the reliable coherent data points within the third adjacent effective number of filtering ranges are defined as the third reference data points. ,

[0094] Using formula Perform calculations to obtain the current interpolation point. Initial interpolation of vertical deformation ,in, This represents the total number of third reference data points within the filtering range. The third reference data point within the filtering range The vertical deformation value, The third reference data point With the current interpolation point The distance points;

[0095] like Then determine ;like Then determine ; Indicates the current interpolation point Vertical deformation transition interpolation;

[0096] like Then determine ;like Then determine ; Indicates the current interpolation point The final interpolation of the vertical deformation, The current interpolation point in the edge zone C of the landslide body. Adjacent points that have been filled.

[0097] Based on the principle of slope creep (InSAR data directly reflects creep phenomena), the deformation of the edge area is weak, and its maximum deformation value should not exceed 0.3 times the average deformation value of the core area (verified by a large amount of landslide monitoring data in high mountain and canyon areas). Therefore, by using amplitude constraints, the filling value can be avoided from being too high, which fits the deformation characteristics of the edge area. At the same time, the deformation gradient of the edge area is gentle, and the deformation difference between adjacent data points should be controlled within a reasonable range, which is suitable for gradient constraints. The dual constraints not only ensure the interpolation accuracy, but also fit the stable and continuous deformation law of the edge area of ​​the landslide.

[0098] In one specific implementation, the first number of adjacent valid numbers is any integer value between 5 and 7, the second number of adjacent valid numbers is any integer value between 6 and 10, and the third number of adjacent valid numbers is any integer value between 8 and 20. Differential interpolation filling is performed using a dynamic parameter simplification method. The filling of the landslide core area A, the landslide transition area B, and the landslide edge area C is based on distance linear interpolation. Combined with the deformation characteristics of each area, key parameters such as gradient, slope aspect, and amplitude are dynamically adjusted to make the interpolation results conform to the deformation law of the area, thus solving the defects of existing single linear interpolation that ignores the non-uniformity of deformation and has low accuracy. Specifically, since the data points in the core area A of the landslide body are characterized by severe deformation and significant gradients, the interpolation filling for the core area A prioritizes selecting reliable coherent data points with close proximity and high gradients. Through gradient correction, the filling values ​​are tilted towards the high deformation area, which is consistent with the characteristics of severe deformation in the core area and the large difference in landslide volume between the slope bottom and the slope top. This aligns with the principle of concentrated deformation in the InSAR creep core area and avoids the problem of low filling values ​​in existing linear interpolation. At the same time, the correction coefficient is dynamically adjusted to adapt to the gradient difference between the slope bottom and the slope top in the core area, which conforms to the rule that the landslide volume is inconsistent between the slope bottom and the slope top. For the interpolation filling of the transition zone B of the landslide body, since the transition zone B of the landslide body has the characteristics of central deformation and continuous gradual change, in order to take into account the continuity of deformation transition, the correction coefficient is adjusted by the slope aspect deviation angle to avoid abrupt deformation and conform to the principle of continuous gradual change of landslide deformation. At the same time, the slope aspect has a significant impact on the deformation of the transition zone. The smaller the slope aspect deviation, the stronger the deformation consistency. The graded setting of the correction coefficient simplifies the calculation and conforms to the actual deformation law. For the interpolation filling of the landslide edge zone C, due to its characteristics of weak deformation and gentle gradient, amplitude constraints are used to avoid excessive filling values ​​while ensuring the continuity of deformation with adjacent filling points, thus conforming to the characteristics of weak deformation and high stability in the edge zone. Specifically, according to the principle of slope creep, the deformation of the edge zone is weak, and its maximum deformation value should not exceed 0.3 times the average deformation value of the core zone (verified by a large amount of monitoring data of landslides in high mountain and canyon areas). Therefore, amplitude constraints can avoid excessive filling values, conforming to the deformation characteristics of the edge zone. At the same time, the deformation gradient of the edge zone is gentle, and the deformation difference between adjacent data points should be controlled within a reasonable range. The dual constraints not only ensure the interpolation accuracy but also conform to the stable and continuous deformation law of the landslide edge zone.

[0099] The process of filtering and extracting reliable decoherent imputation data points from the initial decoherent imputation data points includes calculating the current threshold and comparing it with the confidence threshold, selecting decoherent initial imputation data points with a current confidence level greater than the confidence threshold T as reliable decoherent imputation data points, and removing imputation data points with insufficient confidence or low reliability to ensure the validity of the imputation data.

[0100] Understandably, a reliable deformation point cloud map can intuitively reflect the deformation distribution of the target area in each period and clearly distinguish the deformation differences between the three types of areas, A, B, and C.

[0101] Furthermore, multiple periods of reliable deformation point cloud maps are continuously acquired. These maps are then stitched and aligned chronologically to obtain a reliable deformation time-series point cloud map of the target area. This map reflects the temporal variation of landslide deformation in the target area, capturing the creep process (e.g., continuous increase in deformation in the core area and stable deformation in the edge area), providing a foundation for subsequent deformation analysis and risk assessment. Understandably, landslide creep is a long-term, continuous process. The time-series point cloud map allows for the complete tracking of landslide deformation trends, aligning with the advantages of long-term InSAR monitoring. It also conforms to the physical principle that landslides are caused by the accumulation of creep on slopes, providing time-series data support for risk level assessment.

[0102] The method of this invention adapts to complex terrain, effectively converting single-track LOS deformation into vertical deformation through slope and aspect correction, thereby improving the deformation monitoring accuracy of complex terrains in high mountain and canyon areas. It filters highly reliable data through applicability analysis to reduce errors caused by terrain and observation geometry, and enhances data reliability based on reasonable analysis. It integrates multi-source information to perform interpolation compensation for some incoherent areas, improving the completeness of landslide identification and the reliability of risk assessment. It overcomes the technical bias of using single linear interpolation in existing technologies, combining the principles of slope creep and landslide to propose a landslide... The concept of volume partitioning combined with differentiated interpolation, targeting the deformation characteristics of three types of regions, simplifies the interpolation logic with dynamic parameters, solving the shortcomings of existing methods such as low accuracy and lack of physical interpretability. At the same time, the combination of partitioning and fine-tuning the confidence threshold further improves the reliability of the data, reflecting the innovation of the technology. Furthermore, the partitioning is consistent with the non-uniform deformation characteristics of landslide bodies, the interpolation logic is consistent with the InSAR creep pattern of "intense core and weak edge" and the slope landslide pattern of "continuous and gradual deformation with inconsistent amounts at the bottom and top of the slope", and the confidence calculation is consistent with the topographic influence factors observed in InSAR.

[0103] The present invention also provides a landslide risk monitoring system for high mountain and canyon areas based on single-track InSAR data, including a processing device, which is used to implement the steps of the above-mentioned landslide risk monitoring method for high mountain and canyon areas based on single-track InSAR data.

[0104] The above embodiments are merely illustrative of several implementation methods of this application, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data, characterized in that, Includes the following steps: S10, acquire the single-track line-of-sight deformation lattice data and digital elevation model data of the target area in each period. The single-track line-of-sight deformation lattice data includes the initial coherent data points corresponding to the coherent area and the initial decoherent data points to be filled corresponding to the decoherent area. S20, the regional slope is calculated based on the digital elevation model data; The vertical deformation data of each initial coherent data point in the target area is calculated based on the slope of the area and satellite information. S30, the initial coherent data points are denoised to extract reliable coherent data points; Obtain associable fillable decoherent data points from the initial decoherent data points to be filled; perform vertical deformation data filling on the associable fillable decoherent data points based on the reliable coherent data points to obtain the initial decoherent fillable data points; and extract the reliable decoherent fillable data points from the initial decoherent fillable data points. Using formula The confidence level of the current data point i is calculated, where, For radar incident angle, The slope of the area is given. Let i be the angle between the slope direction of the current data point i and the radar azimuth angle; determine the initial coherent data point corresponding to the current confidence level being greater than the confidence threshold as the credible coherent data point; determine the decoherent initial filling data point corresponding to the current confidence level being greater than the confidence threshold as the decoherent credible filling data point; S40, combine the reliable coherent data points with the decoherent reliable filling data points to obtain the reliable deformation point cloud map corresponding to each period; S50, repeat the above steps until a reliable deformation time series point cloud map corresponding to the observation period is obtained; S60, Based on the reliable deformation time-series point cloud map, obtain the periodic deformation data of the target area within the observation period; S70, based on the periodic deformation data, obtain the maximum vertical deformation value of the target area; S80, determine the current risk level based on the maximum vertical deformation value of the area, the slope of the area, and a preset risk level table; The method also includes the steps of dividing the associable incoherent data points into current interpolation points in the core area A of the landslide body, current interpolation points in the transition area B of the landslide body, and current interpolation points in the edge area C of the landslide body based on the geological survey data, regional slope, and aspect data of the target area; in step S30, the data filling of the associable incoherent data points based on the reliable coherent data points specifically includes: interpolating and filling all current interpolation points in the core area A of the landslide body based on a first number of adjacent effective reliable coherent data points; interpolating and filling all current interpolation points in the transition area B of the landslide body based on a second number of adjacent effective reliable coherent data points; and interpolating and filling all current interpolation points in the edge area C of the landslide body based on a third number of adjacent effective reliable coherent data points.

2. The method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data according to claim 1, characterized in that, In step S30, the data filling for vertical deformation is performed using the linear interpolation method.

3. The method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data according to claim 1, characterized in that, The current interpolation point corresponding to the core area A of the landslide body Centered on the recognizable non-recognizable data points, the reliable coherent data points within the first adjacent effective number of filtering ranges are defined as the first reference data points. ; Using formula Perform calculations to obtain the current interpolation point. Vertical deformation final interpolation ,in, The total number of the first reference data points within the filtering range. The first reference data point within the filtering range The vertical deformation value, The first reference data point With the current interpolation point Distance points, For the gradient correction coefficient of the core area; Using formula The gradient correction coefficients for the core region are calculated. Using formula The average deformation gradient value within the selection range was calculated. , The first reference data point within the filtering range The vertical deformation gradient value.

4. The method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data according to claim 1, characterized in that, The current interpolation point corresponding to the transition zone B of the landslide body. Centered on the recognizable non-recognizable data points, the reliable coherent data points within the second adjacent effective number of filtering ranges are defined as the second reference data points. , Using formula Perform calculations to obtain the current interpolation point. The final interpolation of the vertical deformation, where, This represents the total number of second reference data points within the selected range. The second reference data point within the filtering range The vertical deformation value, The second reference data point With the current interpolation point Distance points, This is a correction system for adjusting the aspect deviation angle; Using formula The slope aspect deviation angle adjustment correction system is obtained through calculation, where, This is a correction system for adjusting the aspect deviation angle. , The current interpolation point slope direction, The second reference data point within the filtering range slope direction, The value range is [0°, 180°].

5. The method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data according to claim 1, characterized in that, The current interpolation point corresponding to the edge region C of the landslide body. Centered on the recognizable non-recognizable data points, the reliable coherent data points within the third adjacent effective number of filtering ranges are defined as the third reference data points. , Using formula Perform calculations to obtain the current interpolation point. Initial interpolation of vertical deformation ,in, This represents the total number of third reference data points within the filtering range. The third reference data point within the filtering range The vertical deformation value, The third reference data point With the current interpolation point The distance points; like Then determine ;like Then determine ; Indicates the current interpolation point Vertical deformation transition interpolation; like Then determine ;like Then determine ; Indicates the current interpolation point The final interpolation of the vertical deformation, The current interpolation point in the edge zone C of the landslide body. Adjacent points that have been filled.

6. The method for monitoring landslide risk in high mountain and canyon areas based on single-track InSAR data according to claim 1, characterized in that, Determine the target area The area whose slope aspect data is consistent with the dominant direction of the landslide is the core area A of the landslide body; Determine the target area Part of the area is the transition zone B of the landslide body; Determine the target area The portion of the area is the edge zone C of the landslide body.

7. A landslide risk monitoring system for high mountain and canyon areas based on single-track InSAR data, characterized in that, The system includes a processing device for implementing the steps of the landslide risk monitoring method for high mountain and canyon areas based on single-track InSAR data as described in any one of claims 1 to 6.

Images