Reservoir storage period wide area landslide geological disaster rapid identification method

Abstract

The invention relates to the technical field of landslide geological recognition, in particular to a reservoir water storage period wide-area landslide geological disaster rapid recognition method, which comprises the following steps: acquiring SAR data and DEM data of a target area covered by a reservoir water storage period; performing baseline estimation according to the SAR data acquired in the same target area at different times, and removing long baseline data to obtain a short baseline data set; performing differential interference processing on the short baseline data set based on external DEM data to obtain a differential interference result; mutual verification is carried out on interference pair combinations corresponding to the multi-period differential interference results of the same main image or adjacent time periods by using the short baseline data set, and potential hidden danger points of the reservoir area are determined; and performing field verification on the potential hidden danger points of the reservoir area by using an unmanned aerial vehicle aerial photography technology, and determining landslide disaster points. According to the invention, sudden landslide disaster points caused by water level changes can be found quickly, efficiently and accurately.

Description

Technical Field

[0001] The invention relates to the technical field of landslide geological identification, and in particular to a method for quickly identifying wide-area landslide geological disasters during a reservoir impoundment period. Background Art

[0002] During the operation period of large-scale hydropower projects, the environmental geological conditions of the reservoir area have been greatly changed due to reasons such as reservoir water storage, which has aggravated the scale and frequency of bank slope geological disasters, bringing great safety risks and economic losses. However, the disaster source area is located in the middle and upper parts of the mountains, most of which are sparsely populated and have high vegetation coverage. It is characterized by high position and concealment. It is difficult to detect such disaster hazards in advance by traditional means. Many sudden disasters are not included in the hidden danger database, resulting in a passive situation in the prevention and control of geological disasters during the operation period of hydropower projects. Generally, large-scale bank slope geological disasters in mountainous reservoirs often have the characteristics of hidden surface deformation, extreme triggering factors, complex instability patterns, and sudden disaster risks. Therefore, it is particularly important to find a method to quickly identify geological disaster hazards during the reservoir bank impoundment period.

[0003] Differential Interferometric Synthetic Aperture Radar (DInSAR) further eliminates elevation information from the interference phase and obtains the differential phase caused by surface changes within the time interval between two observations, thereby quantitatively inverting the surface deformation.

[0004] The existing technology uses time series differential interferometry technology to invert surface deformation using time series interferometric pairs composed of multiple main images. However, when this method performs time series analysis on newly added images together with a large number of original images, it has low sensitivity to newly added images and long processing time. It has certain limitations for special situations with high suddenness and the need to quickly and efficiently discover hidden dangers. Summary of the invention

[0005] In order to solve the technical problem that the existing time series analysis of images has low sensitivity and certain limitations, the purpose of the present invention is to provide a method for rapid identification of wide-area landslide geological disasters during the reservoir impoundment period. The technical scheme adopted is as follows: Obtain SAR data and DEM data covering the target area during the reservoir impoundment period; Baseline estimation is performed based on SAR data acquired at different times in the same target area, and long baseline data are removed to obtain a short baseline data set. Perform differential interferometry processing on the short baseline dataset based on external DEM data to obtain differential interferometry results; The short baseline data set is used to verify the interference pairs corresponding to the multi-period differential interference results of the same main image or adjacent time periods to determine the potential hidden dangers in the reservoir area; Unmanned aerial photography technology is used to conduct field verification of potential risk points in the reservoir area and determine landslide hazard points.

[0006] Preferably, the SAR data includes satellite ascending and descending orbit SAR data, and the spatial resolution of the DEM data is 30m.

[0007] Preferably, the short baseline data set is subjected to differential interferometry processing based on external DEM data to obtain a differential interferometry result, specifically including: The interferometric phase of the satellite's observation of the target area includes the reference ellipsoid phase, terrain phase, deformation phase, atmospheric phase, and noise phase; The SAR images of the same target area before and after deformation at different times are combined into an interference pair, and the interference processing is performed to generate an interference map. Based on the dual-track DInSAR method, the terrain phase in the interference map is removed using DEM data to obtain the deformation phase of the surface and the differential interference result.

[0008] Preferably, the use of the short baseline data set to mutually verify the interference pair combinations corresponding to the multi-period differential interference results of the same main image or adjacent time periods to determine the potential risk points in the reservoir area specifically includes: The main image is a deformed SAR image. By comparing different interference patterns corresponding to the interference combination based on the same main image, mutual verification is performed to determine the potential hidden danger points in the reservoir area.

[0009] Preferably, the use of drone aerial photography technology to conduct field verification of potential risk points in the reservoir area to determine landslide disaster points specifically includes: Through data preprocessing, aerial triangulation, image matching, image fusion, and digital product production, DOM data of the target area is generated and a three-dimensional model is established. The micro-topography and geomorphological features observed by aerial photography are used to determine whether the deformation zone has micro-topography and geomorphological features of landslides, and the deformation zone with micro-topography and geomorphological features is identified as a landslide hazard point.

[0010] The embodiments of the present invention have at least the following beneficial effects: In order to ensure the reliability of InSAR monitoring results, the present invention avoids the combination of long baseline image pairs, selects short baseline interference pairs for combined analysis, and effectively reduces the impact of spatial incoherence. The short baseline D-InSAR method is used to quickly identify geological disasters in the reservoir area during the water storage period. Field verification is then carried out based on drone aerial photography technology, and the landslide disaster points are determined using the micro-topography and geomorphic features of the landslide, excluding non-hazardous areas caused by atmospheric disturbances, vegetation, etc. It is possible to quickly, efficiently and accurately discover sudden landslide disaster points caused by water level changes, ensure the safe operation of hydropower stations, and better serve the disaster prevention and mitigation of water conservancy and hydropower projects. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0012] Figure 1 It is a flowchart of the steps of a method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period provided by the present invention; Figure 2 It is a result schematic diagram of theoretical displacement accuracy in the result of baseline estimation provided by the present invention; Figure 3 It is a result schematic diagram of theoretical elevation accuracy in the result of baseline estimation provided by the present invention; Figure 4 is the interference pattern provided by the present invention; Figure 5 is the interference pattern after flattening provided by the present invention; Figure 6 is the filtered interference pattern provided by the present invention; Figure 7 The present invention provides an unmanned aerial photography image. DETAILED DESCRIPTION

[0013] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a method for rapid identification of wide-area landslide geological disasters during the reservoir impoundment period proposed by the present invention, its specific implementation method, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0014] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0015] The specific scheme of the method for rapid identification of wide-area landslide geological disasters during the reservoir impoundment period provided by the present invention is described in detail below with reference to the accompanying drawings.

[0016] See also Figure 1 , which shows a flowchart of the steps of a method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period provided by an embodiment of the present invention, the method comprising the following steps: Step S100, obtaining SAR data and DEM data covering the target area during the reservoir impoundment period.

[0017] Specifically, the SAR data includes satellite ascending and descending orbit SAR data, and the spatial resolution of the DEM data is 30 m.

[0018] In this embodiment, the Sentinel-1A / B data released by the Copernicus Open Access Hub and the Alaska Satellite Facility Data Search are mainly used for experiments.

[0019] Sentinel-1 is a binary system consisting of two satellites, Sentinel-1A and Sentinel-1B, operating in a near-polar sun-synchronous orbit of 693 kilometers. In the area covered by both satellites, the revisit period can be further reduced to 6 days. The satellite carries a C-band synthetic aperture radar, which can provide continuous images (day, night and all weather conditions), with all-weather and all-day imaging capabilities. It can provide continuous daytime and night images of various meteorological conditions and medium- and high-resolution land, coastal and ice measurement data for natural resource monitoring, flood monitoring, earthquake monitoring and other surface deformation measurements. Since the Sentinel (Sentinel-1A / B) satellites are repeated orbit observations, the data release time is generally periodic. The data release time during the water storage period is predicted based on the published time, so as to quickly and efficiently identify potential hidden dangers at the first time.

[0020] More specifically, in this embodiment, Sentinal-1A / B satellite ascending / descending orbit SAR data covering the target area and ALOS WORLD 3D DEM data with a spatial resolution of 30 m are collected.

[0021] Step S200 , performing baseline estimation based on SAR data acquired at different times for the same target area, removing long baseline data, and obtaining a short baseline data set.

[0022] The baseline is the relative position difference of the radar antenna when two images of the same target area are acquired at different times. The baseline is divided into vertical baseline and horizontal baseline. In SAR interferometry, the vertical baseline will affect the terrain and deformation information in the interferometric phase. Accurately estimating the baseline is crucial for subsequent terrain mapping and deformation monitoring.

[0023] The Sentinel-1A / B satellites are repeated orbit observations. When imaging the same target area twice, due to factors such as side-view imaging, the imaging coverage of the same target area deviates twice, and the vertical component of the baseline affects the coherence of the image to a certain extent. When the vertical component of the baseline is larger, the coherence is worse, and the interference effect is also worse. Therefore, a short baseline data set is selected for differential interferometry processing for baseline estimation. It should be noted that coherence is one of the important indicators for measuring the quality of interferometric images. It indicates the degree of similarity between two images at corresponding positions, reflects the stability and reliability of the interference phase, and the coherence coefficient is usually used to quantify the coherence.

[0024] In this embodiment, through baseline estimation, short baseline images with a maximum time baseline of 36 days and a spatial baseline of no more than 200m are selected to form a short baseline interference image pair combination to reduce the impact of spatial incoherence on the deformation area. Subsequent differential interference processing is performed. The baseline estimation can be used to evaluate the quality of the interference image pair, calculate the baseline, orbit offset (in range and azimuth) and other system parameters, select two images for processing, and obtain the relationship between the theoretical elevation accuracy and deformation accuracy and coherence measured by the two image pairs, that is, the situation where the elevation and deformation change with the change of coherence. Figure 2 As shown in , the accuracy of the deformation increases with the increase of coherence, such as Figure 3 As shown, the accuracy of elevation also increases with the increase of coherence.

[0025] Step S300: Perform differential interferometry processing on the short baseline data set based on external DEM data to obtain a differential interferometry result.

[0026] Specifically, the interferometric phase of the satellite observation of the target area includes the reference ellipsoid phase, terrain phase, deformation phase, atmospheric phase and noise phase, which can be expressed as:

[0027] in, represents the interferometric phase of the satellite's observation of the target area, represents the reference ellipsoid phase, represents the terrain phase, represents the deformation phase, represents the atmospheric phase, represents the noise phase.

[0028] The SAR images of the same target area before and after deformation at different times are combined into an interference pair, and the interference processing is performed to generate an interference map. Based on the dual-track DInSAR method, the terrain phase in the interference map is removed using DEM data to obtain the deformation phase of the surface and the differential interference result.

[0029] In the actual interferometric phase, the noise phase presents random characteristics and presents high-frequency signals in the phase model, which can be suppressed by low-pass phase filtering. The atmospheric phase is seriously affected by the terrain and appears as a high-frequency signal in the phase information. Therefore, the influence of the atmospheric phase is weakened by high-pass filtering. In addition, the reference ellipsoid phase, terrain phase, and deformation phase account for the majority.

[0030] When extracting terrain information based on InSAR technology, the reference ellipsoid phase refers to the interference phase obtained by observing the same target at different times with synthetic aperture radar and obtaining satellite images and then interfering when there is no terrain undulation, that is, the interference phase generated by the observation plane. This is the main component of the original interference phase of the ground target before the terrain and deformation are generated. Assuming that the deformation information is zero, the interference pair of short time baseline and short space baseline is selected, and the terrain phase can be obtained by removing the reference ellipsoid phase. When using DInSAR technology to extract deformation information, there is still a terrain phase caused by ground elevation.

[0031] In this embodiment, a dual-track differential method is used for differential interferometry, the central idea of ​​which is to form an interference pair of two radar images before and after deformation in different phases in the same area, and use a known external DEM of the same area to remove the interference phase generated by the terrain to obtain deformation information. The external DEM and the main image first need to be accurately matched, the elevation of the DEM is simulated into the form of terrain interference fringes (i.e., simulated terrain phase), and then the terrain phase simulated by the DEM is removed from the interference pattern to obtain deformation information.

[0032] That is, its core is to use DEM data to remove the interference phase caused by the terrain in the interference phase of the same target point after the interference processing of two SAR images, and then remove the flat ground effect, so as to retain only the deformation phase in the interference phase, so as to realize the monitoring of ground deformation.

[0033] It should be noted that the dual-track method mainly includes the following steps: SLC data input, SAR image registration, interference generation, removal of terrain phase and flat ground phase, phase filtering, phase unwrapping, and conversion of phase to deformation information and geocoding. The technical means adopted in the present invention does not perform phase unwrapping, but directly performs subsequent processing on the filtered interference results, thereby reducing the error caused by phase unwrapping.

[0034] More specifically, based on SAR data processing software, SAR images covering the same research area without loss of coherence are selected for differential interferometry processing. In this embodiment, two images are interferometrically processed, one is acquired before the surface deformation occurs, which is the auxiliary image, and the other is acquired after the deformation, which is the main image. The two images are interferometrically processed to generate an interferogram, such as Figure 4 The interference pattern includes terrain phase and deformation phase, as shown in Figure 5 Then, external DEM data is introduced, and the DEM data is simulated as terrain phase. The terrain phase simulated by DEM is subtracted from the interference graph to obtain the deformation phase of the surface, which is the result of differential interference processing, as shown in Figure 6 shown.

[0035] Step S400, using the short baseline data set to mutually verify the interference pair combinations corresponding to the multi-period differential interference results of the same main image or adjacent time periods, to determine the potential hidden danger points in the reservoir area.

[0036] The differential interferometry results obtained using the short baseline data set are interpreted to quickly and efficiently discover potential hidden dangers. In the radar interferometry phase, if the surface is deformed, the reference ellipsoid phase and terrain phase are removed from the overall interference phase to obtain the deformation phase, and the deformed area often shows discontinuity of phase information. From the interference diagram after removing the reference ellipsoid phase and terrain phase, it can be seen that the phase discontinuity is generally manifested as a mutation area of ​​the interference fringes. Based on this rule, the short baseline data set is used to interpret the results of multi-period differential interferometry based on the same main image or adjacent time periods, so as to achieve mutual verification between the interference pair combinations, and then quickly, accurately and efficiently discover potential hidden dangers in the reservoir area.

[0037] In the interferogram after removing the reference ellipsoid phase and terrain phase, the phase discontinuity is generally manifested as a mutation area of ​​the interference fringes. Based on this, the short baseline data set is used to interpret the results of multi-period differential interferometry based on the same main image or adjacent time periods. Using A / B binary satellites and multi-orbit joint observations, DInSAR ascending and descending orbits can be jointly identified every 6 days. The same image is processed to achieve mutual verification between the interference pair combinations, and potential hidden dangers in the reservoir area are discovered quickly, accurately and efficiently, thereby conducting large-dataset, wide-area, rapid and dynamic geological disaster monitoring.

[0038] More specifically, the A / B dual satellites are used to image the same area within a specific time interval (for example, 6 days) to obtain multiple sets of SAR images. These images need to have sufficient overlapping areas for effective interference processing. In this way, the rise and fall changes of the surface can be identified by identifying the results of interference pairs of different orbits or different times. For the same image, there are corresponding interference results of multiple different interference pairs. By comparing these results for mutual verification, areas with hidden dangers can be quickly and accurately found, that is, potential hidden danger points in the reservoir area. The mutual verification process is a well-known technology and will not be introduced in detail here.

[0039] Step S500, using drone aerial photography technology to conduct field verification of potential risk points in the reservoir area to determine landslide disaster points.

[0040] Specifically, the ground resolution of the designed image during operation is 0.02m. After calculation, the relative altitude should be maintained at about 180m, and the heading and lateral overlaps are 75%. Then the POS and images obtained by aerial photography and the field control points collected by RTK are imported into the DJI Zhitu software for aerial triangulation, and finally a fine three-dimensional model with geographic coordinates and a ground resolution of 2cm and a digital orthophoto DOM are generated. Through the micro-topography and geomorphic features observed by aerial photography, it is judged whether there are micro-topography and geomorphic features of landslides in the deformation area (such as saber trees and various cracks growing on the landslide body, sliding scratches on the landslide wall, ground uplift at the landslide shear outlet, landslide drums, etc.), and the deformation area with micro-topography and geomorphic features is determined as a landslide disaster point.

[0041] Based on the short-baseline DInSAR method, aerial survey flights are carried out to obtain aerial photography images within the area of ​​the potential danger points identified. Through data preprocessing, aerial triangulation, image matching, image fusion, digital product production and other steps, high-precision DOM data of the study area is generated and a three-dimensional stereo model is established for fine processing. The deformation area is interpreted in detail through the micro-topography and geomorphic features observed by aerial photography. It is determined whether the deformation area has micro-topography and geomorphic features of landslides, and then it is determined as a landslide hazard point, such as Figure 7 As shown in Figure 2, non-landslide disaster points caused by atmospheric disturbances, vegetation and other error terms are excluded when identifying landslide geological disasters based on the short baseline DInSAR method.

[0042] In summary, when the present invention uses the short baseline DInSAR method to identify hidden danger points, it can be seen from the interference diagram of the terrain phase after removing the reference ellipsoid phase that the places where the phase is discontinuous are generally manifested as mutation areas of interference fringes. Based on this rule, the short baseline data set is used to interpret the results of multi-period differential interference based on the same main image or adjacent time periods, realize mutual verification between interference pair combinations, and quickly, accurately and efficiently discover potential hidden danger points in the reservoir area. Then, large data sets and wide-area rapid and dynamic geological disaster monitoring are carried out.

[0043] The deformation measurement hazard identification method of short-baseline DInSAR cannot perform terrain analysis and verification based on SAR images, while drone aerial photography can interpret micro-topography very well, but it is limited by flight costs and cannot achieve wide-area full coverage. Therefore, drone aerial photography technology is used to verify the hazard points identified by the short-baseline DInSAR method in the field. Aerial images are processed to establish a three-dimensional model to determine whether there are micro-topography features of landslides in the hazard point area, and then determine the landslide disaster point. The reservoir area geological disaster hazard identification method system combining the two technologies has great advantages and high application value.

[0044] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period, characterized in that: The method comprises the following steps: Obtain SAR data and DEM data covering the target area during the reservoir impoundment period; Baseline estimation is performed based on SAR data acquired at different times in the same target area, and long baseline data are removed to obtain a short baseline data set. Perform differential interferometry processing on the short baseline dataset based on external DEM data to obtain differential interferometry results; The short baseline data set is used to verify the interference pairs corresponding to the multi-period differential interference results of the same main image or adjacent time periods to determine the potential hidden dangers in the reservoir area; Unmanned aerial photography technology is used to conduct field verification of potential risk points in the reservoir area and determine landslide hazard points.

2. The method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period according to claim 1 is characterized in that: The SAR data includes satellite ascending and descending orbit SAR data, and the spatial resolution of the DEM data is 30m.

3. The method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period according to claim 1 is characterized in that: The short baseline data set is subjected to differential interferometry processing based on external DEM data to obtain differential interferometry results, specifically including: The interferometric phase of the satellite's observation of the target area includes the reference ellipsoid phase, terrain phase, deformation phase, atmospheric phase, and noise phase; The SAR images of the same target area before and after deformation at different times are combined into an interference pair, and the interference processing is performed to generate an interference map. Based on the dual-track DInSAR method, the terrain phase in the interference map is removed using DEM data to obtain the deformation phase of the surface and the differential interference result.

4. A method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period according to claim 3, characterized in that: The short baseline data set is used to mutually verify the interference pair combinations corresponding to the multi-period differential interference results of the same main image or adjacent time periods to determine the potential hidden danger points in the reservoir area, specifically including: The main image is a deformed SAR image. By comparing different interference patterns corresponding to the interference combination based on the same main image, mutual verification is performed to determine the potential hidden danger points in the reservoir area.

5. The method for rapid identification of wide-area landslide geological disasters during reservoir impoundment period according to claim 1 is characterized in that: The use of drone aerial photography technology to conduct field verification of potential risk points in the reservoir area and determine landslide disaster points specifically includes: Through data preprocessing, aerial triangulation, image matching, image fusion, and digital product production, DOM data of the target area is generated and a three-dimensional model is established. The micro-topography and geomorphological features observed by aerial photography are used to determine whether the deformation zone has micro-topography and geomorphological features of landslides, and the deformation zone with micro-topography and geomorphological features is identified as a landslide hazard point.

Images