A two-dimensional landslide deformation decomposition method based on a local surface parallel flow model

Abstract

The present invention discloses a two-dimensional deformation decomposition method for landslides based on a local surface parallel flow model, and relates to the field of landslide deformation monitoring. First, ascending and descending SAR data are processed in time series using the StaMPS technology to obtain the one-dimensional deformation of the landslide along the LOS direction. Second, the imaging geometry information and LOS deformation information of the ascending and descending SAR data are combined to invert the optimal sliding direction and sliding inclination of the landslide movement based on the local surface parallel flow model. Finally, the LOS deformation information of the landslide, the imaging geometry information of the ascending and descending SAR data, and the inverted optimal sliding direction and sliding inclination are combined to calculate the two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface. The new two-dimensional deformation decomposition method based on the local surface parallel flow model is applicable to translational loess landslides, and helps prevent and mitigate landslide disasters.

Description

Technical Field

[0001] The present invention relates to the technical field of landslide deformation monitoring, and in particular to a two-dimensional landslide deformation decomposition method based on a local surface parallel flow model. Background Art

[0002] In recent years, due to the frequent occurrence of extreme rainfall events caused by global climate change, many catastrophic loess landslides have occurred. Therefore, it is necessary to conduct long-term stability monitoring of landslides to reduce the losses caused by landslide disasters. The rapid development of remote sensing technology has provided new ideas for landslide monitoring. Among them, spaceborne synthetic aperture radar interferometry (InSAR), as a new type of space-based earth observation technology, has the characteristics of wide coverage, high monitoring accuracy, all-day, all-weather, and high spatial resolution. It provides technical support for landslide deformation monitoring from regional to local scales.

[0003] However, InSAR technology can only capture one-dimensional surface deformation along the radar's line of sight. This not only hinders interpretation of the magnitude and direction of landslide movement, but also makes it difficult to communicate with researchers unfamiliar with one-dimensional observation geometry during multidisciplinary collaboration. Therefore, capturing multidimensional deformation of landslide surfaces is crucial for studying the dynamic evolution of both the surface and interior of landslides.

[0004] In response to the problems existing in conventional InSAR monitoring, this application provides a two-dimensional landslide deformation decomposition method based on a local surface parallel flow model. Summary of the Invention

[0005] The purpose of this paper is to provide a universal two-dimensional landslide deformation decomposition method. This method uses currently freely available SAR (Synthetic Aperture Radar) data (Sentinel-1) to recover the two-dimensional deformation of landslides. This method can be used to study the spatiotemporal evolution trends and failure modes of landslides and provide reliable advice for further landslide management.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions:

[0007] A two-dimensional landslide deformation decomposition method based on a local surface parallel flow model includes the following steps:

[0008] The time series of ascending and descending SAR data was processed to obtain the millimeter-level one-dimensional deformation rate of the landslide along the LOS (Line-of-Sight) direction;

[0009] Combining the one-dimensional deformation rate in the LOS direction with the imaging geometry information of ascending and descending SAR images, the optimal sliding direction and sliding angle of the landslide movement are inverted based on the local surface parallel flow model.

[0010] The two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface is calculated by combining the one-dimensional deformation rate in the LOS direction, the imaging geometric information of the ascending and descending orbit SAR images, and the optimal sliding direction and sliding inclination angle of the landslide obtained by inversion.

[0011] Furthermore, the millimeter-level one-dimensional deformation rate of the landslide along the LOS direction is obtained, specifically including:

[0012] SAR data are preprocessed using auxiliary DEM (Digital Elevation Model) data and precise orbit data to generate a single primary image interferogram.

[0013] The amplitude deviation threshold method is used to preliminarily select the PS (Persistent Scatterer) points on the interferogram to obtain the PS candidate point set;

[0014] In the iterative operation, the phase stability analysis of the PS candidate points is performed to screen out the final PS points;

[0015] The spatially correlated phases, including atmospheric phase, orbital error, and spatially correlated DEM error phases, are removed by low-pass filtering;

[0016] The 3D phase unwrapping algorithm is used to obtain the absolute interference phase value of the PS point;

[0017] The absolute interferometric phase at the PS point is subjected to temporal high-pass filtering and spatial low-pass filtering. The result is the remaining disturbance phase containing the atmospheric effect, which is subtracted from the unwrapped phase and the final estimated LOS deformation value is obtained through geocoding.

[0018] Furthermore, the optimal sliding direction and sliding angle of the landslide movement are inverted, specifically including:

[0019] Kriging interpolation is used to perform spatial interpolation on the LOS deformation rate results of the ascending and descending orbits;

[0020] Resample the interpolated LOS deformation rate results of the ascending and descending orbits to have the same spatial resolution;

[0021] For any PS point after resampling, all PS points within the area were extracted with a buffer of 50 m, and it was stipulated that they belonged to the same local sliding surface;

[0022] Based on the local surface parallel flow model, the least squares algorithm is used to invert the optimal sliding direction and sliding angle of the landslide movement at any PS point.

[0023] Furthermore, the two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface is calculated, specifically including:

[0024] Based on the optimal sliding direction and sliding inclination of the landslide obtained by inversion, the sliding surface coordinate system of the landslide movement is constructed;

[0025] In the sliding surface coordinate system, the two-dimensional deformation of the landslide along the sliding direction and the sliding surface normal direction is calculated by combining the one-dimensional deformation rate in the LOS direction and the imaging geometry information of the ascending and descending orbit SAR images.

[0026] Furthermore, StaMPS (Stanford Method for Persistent Scatterers) technology is used to process the time series of ascending and descending orbit SAR data.

[0027] Furthermore, the least squares algorithm is used to invert the optimal sliding direction and sliding inclination of the landslide movement.

[0028] The present invention provides a two-dimensional landslide deformation decomposition method based on a local surface parallel flow model, which has the following beneficial effects:

[0029] (1) The present invention provides a universal two-dimensional landslide deformation decomposition method, which can obtain the two-dimensional deformation of the landslide along the sliding direction and the sliding surface normal direction using only two SAR data sets with different imaging geometries without any other auxiliary data;

[0030] (2) The two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface obtained by the present invention has high accuracy, which is conducive to interpreting the amplitude and direction of the landslide movement on an overall scale;

[0031] (3) Compared with the traditional global surface parallel flow model, the local surface parallel flow model proposed in this paper can capture the motion details of different parts of the landslide body, which is conducive to the detailed study of the spatiotemporal evolution of the landslide;

[0032] (4) The present invention can be easily applied to other translational loess landslides around the world to help prevent and mitigate landslide disasters. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 A flow chart of the two-dimensional decomposition technology of landslide deformation provided by an embodiment of the present invention;

[0034] Figure 2 A schematic diagram of a two-dimensional decomposition geometry of a landslide deformation provided by an embodiment of the present invention;

[0035] Figure 3 A diagram of the LOS deformation rate of the lifting rail provided in an embodiment of the present invention;

[0036] Figure 4 A diagram illustrating the accuracy of deformation results provided by an embodiment of the present invention;

[0037] Figure 5The optimal sliding direction and sliding inclination angle distribution diagram of landslide movement provided by the embodiment of the present invention;

[0038] Figure 6 A two-dimensional deformation map of a landslide provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0039] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0040] The present invention provides a two-dimensional landslide deformation decomposition method based on a local surface parallel flow model, comprising the following steps:

[0041] Step 1: Use StaMPS technology to perform time series processing on the ascending and descending SAR data to obtain the millimeter-level one-dimensional deformation rate of the landslide along the LOS direction. The LOS deformation calculation in Step 1 specifically includes:

[0042] Step 1.1: Preprocess the SAR data using auxiliary DEM data and precise orbit data to generate a single main image interferogram;

[0043] Step 1.2: Use the amplitude deviation threshold method to preliminarily select PS points on the interference graph to obtain the PS candidate point set;

[0044] Step 1.3, perform phase stability analysis on the PS candidate points in the iterative operation to screen out the final PS point set;

[0045] Step 1.4: remove spatially correlated phases, including atmospheric phases, orbital errors, and spatially correlated DEM error phases, by low-pass filtering.

[0046] Step 1.5, use the 3D phase unwrapping algorithm to obtain the absolute interferometric phase value of the PS point;

[0047] In step 1.6, the absolute interferometric phase at the PS point is subjected to temporal high-pass filtering and spatial low-pass filtering. The result is the remaining perturbation phase containing atmospheric effects, which is subtracted from the unwrapped phase and the final estimated LOS deformation value is obtained through geocoding.

[0048] Step 2: Combining the one-dimensional deformation rate in the LOS direction with the imaging geometry information of the ascending / descending SAR image, based on the local surface parallel flow model, the optimal sliding direction and sliding angle of the landslide movement are inverted using the least squares algorithm. Step 2 includes:

[0049] Step 2.1: Use Kriging interpolation to perform spatial interpolation on the LOS deformation rate results of the ascending and descending orbits;

[0050] Step 2.2: resample the interpolated LOS deformation rate results of the ascending and descending orbits to have the same spatial resolution;

[0051] Step 2.3: For any PS point after resampling, use a 50-meter buffer zone to extract all PS points within the area and stipulate that they belong to the same local sliding surface;

[0052] In step 2.4, based on the local surface parallel flow model, the least squares algorithm is used to invert the optimal sliding direction and sliding inclination of the landslide movement at any PS point.

[0053] Step 3: Combine the one-dimensional deformation rate in the LOS direction, the imaging geometry information of the ascending / descending SAR image, and the optimal sliding direction and sliding inclination angle obtained by inversion to calculate the two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface. Step 3 includes:

[0054] Step 3.1: Based on the optimal sliding direction and sliding inclination of the landslide obtained by inversion, the sliding surface coordinate system of the landslide movement is constructed;

[0055] Step 3.2: In the sliding surface coordinate system, the one-dimensional deformation rate in the LOS direction and the imaging geometry information of the ascending and descending orbit SAR image are combined to calculate the two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface.

[0056] The above steps are analyzed in detail as follows:

[0057] Figure 1 The flowchart of the two-dimensional decomposition technology of landslide deformation in the example area solved by this technical solution includes data preprocessing, StaMPS time series processing, inversion of optimal sliding direction and sliding angle of landslide, and two-dimensional landslide deformation decomposition.

[0058] Figure 2 This is a two-dimensional decomposition diagram of the landslide in the example area solved by this technical solution; the kinematic law of the landslide shows that there is a friction surface between the moving rock and soil and the stable bedrock, which is the sliding surface of the landslide. Figure 2 The geometric relationship between the "sliding surface" coordinate system and the "north-east-height" coordinate system of the landslide can be established, and then the two-dimensional deformation of the landslide can be calculated.

[0059] Figure 3 This is the LOS deformation rate diagram of the lifting rail in the example area solved by this technical solution; according to Figure 3It can be clearly seen that the deformation is mainly concentrated in the northern part of the landslide group, with the maximum annual average deformation rate reaching -51mm / a in the ascending LOS direction and 49.5mm / a in the descending LOS direction; the deformation rate is fastest at the leading edge adjacent to the highway excavation, and slower at the trailing edge. The overall deformation is directed from the leading edge to the trailing edge towards the mountain highway; it is worth noting that a local deformation zone was found in the southern area of the landslide group, with a deformation magnitude between 15-30mm / a, which is speculated to be related to the local resurgence of the landslide.

[0060] Figure 4 This is a verification diagram of the deformation results of the example area solved by this technical solution; Figure 4 It can be seen that the InSAR and GNSS (Global Navigation Satellite System) measurements have highly similar displacement trends and are consistent in magnitude. The mean absolute deviation (MAD) and root mean square error (RMSE) of the GNSS monitoring values and the ascending and descending InSAR observations on the same observation date are calculated as follows: Figure 4 As shown in the figure, the maximum mean absolute error and root mean square error are 2.64 mm and 2.3 mm respectively. In general, the time series InSAR and GNSS monitoring results are consistent with each other, indicating that the InSAR deformation monitoring results of this paper are highly reliable.

[0061] Figure 5 The optimal sliding direction and sliding inclination of the landslide in the example area solved by this technical solution; Figure 5 As can be seen from the left figure, the sliding dip angle in the study area is between 0 and 15 degrees, and the sliding dip angle of the landslide group in the lower part of the slope is significantly greater than that in the stable area above the slope; the calculated sliding dip angle is slightly smaller than the natural surface slope of the slope. This is because the landslide group is a shallow soil landslide with a gentle dip angle, and it mainly slides along the soil interface. Figure 5 As can be seen from the right figure, the direction of landslide movement is basically along the slope direction, and is nearly perpendicular to or intersects with the valley direction at the front edge of the landslide group at a large angle, which shows that the displacement of the landslide group is mainly driven by gravity.

[0062] Figure 6 This is a two-dimensional deformation map of the landslide in the example area solved by this technical solution; Figure 6As shown in the figure, landslide bodies L3 to L6 all have significant deformation, and their deformation modes are different. Taking landslide L4 as an example, its deformation in the sliding direction and normal direction are relatively significant, and the tensile cracks at the rear edge and the shear cracks on both sides have connected and penetrated each other to form a closure, which indicates that the deformed body has separated from the surrounding rock and soil medium to form a landslide block with higher freedom of movement. It is worth noting that significant deformation in the sliding direction and normal direction appeared at the southern slope corner of landslide L1. This is due to the local revival of the old landslide caused by concentrated heavy rain in summer. The surface type in this area is wasteland and a small amount of cultivated land, which is far away from the village. Even if a landslide occurs, it is speculated that accumulation will occur in the valley without affecting the upper village because the sliding direction is towards the front valley. In contrast, although landslide bodies L2 to L6 are smaller in scale, the mountain road is located below the open area of their front edges. If they become unstable, the consequences will be more serious and should be paid special attention.

[0063] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. A two-dimensional landslide deformation decomposition method based on a local surface parallel flow model, characterized in that: The following steps are involved: The time series processing of ascending and descending orbit SAR data was performed to obtain the millimeter-level one-dimensional deformation rate of the landslide along the LOS direction; Combining the one-dimensional deformation rate in the LOS direction and the imaging geometric information of the ascending / descending SAR image, based on the local surface parallel flow model, the optimal sliding direction and sliding inclination of the landslide movement are inverted; the inversion of the optimal sliding direction and sliding inclination of the landslide movement specifically includes: using Kriging interpolation to spatially interpolate the ascending / descending LOS deformation rate results; resampling the interpolated ascending / descending LOS deformation rate results to have the same spatial resolution; for any PS point after resampling, using a 50-meter buffer zone, extracting all PS points in the area and stipulating that they belong to the same local sliding surface; based on the local surface parallel flow model, using the least squares algorithm to invert the optimal sliding direction and sliding inclination of the landslide movement at any PS point; The two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface is calculated by combining the one-dimensional deformation rate in the LOS direction, the imaging geometric information of the ascending and descending orbit SAR images, and the optimal sliding direction and sliding inclination angle of the landslide obtained by inversion.

2. A two-dimensional landslide deformation decomposition method based on a local surface parallel flow model according to claim 1, characterized in that: The obtaining of the millimeter-level one-dimensional deformation rate of the landslide along the LOS direction specifically includes: SAR data are preprocessed using auxiliary DEM data and precise orbit data to generate a single main image interferogram; The amplitude deviation threshold method is used to preliminarily select PS points on the interference graph to obtain the PS candidate point set; In the iterative operation, the phase stability analysis of the PS candidate points is performed to screen out the final PS points; The spatially correlated phases, including atmospheric phase, orbital error, and spatially correlated DEM error phases, are removed by low-pass filtering; The 3D phase unwrapping algorithm is used to obtain the absolute interference phase value of the PS point; The absolute interferometric phase at the PS point is subjected to temporal high-pass filtering and spatial low-pass filtering. The result is the remaining disturbance phase containing the atmospheric effect, which is subtracted from the unwrapped phase and the final estimated LOS deformation value is obtained through geocoding.

3. The method for decomposing two-dimensional landslide deformation based on a local surface parallel flow model according to claim 1, characterized in that: The calculation of the two-dimensional deformation of the landslide along the sliding direction and the normal direction of the sliding surface specifically includes: Based on the optimal sliding direction and sliding inclination of the landslide obtained by inversion, the sliding surface coordinate system of the landslide movement is constructed; In the sliding surface coordinate system, the two-dimensional deformation of the landslide along the sliding direction and the sliding surface normal direction is calculated by combining the one-dimensional deformation rate in the LOS direction and the imaging geometry information of the ascending and descending orbit SAR images.

4. The method for decomposing two-dimensional landslide deformation based on a local surface parallel flow model according to claim 1, characterized in that: The time series processing of the ascending and descending orbit SAR data utilizes the StaMPS technology.

5. The method for decomposing two-dimensional landslide deformation based on a local surface parallel flow model according to claim 1, characterized in that: The optimal sliding direction and sliding inclination angle of the inverse landslide movement adopt the least squares algorithm.

Images