CN115236667B - A method for estimating potential landslide volume by fusing multi-source SAR data - Google Patents

Abstract

The invention provides a potential landslide volume estimation method fusing multisource SAR data, which relates to the technical field of geological disaster prevention and control safety, and comprises the following steps: inSAR deformation monitoring, sliding surface determination and landslide volume estimation. The InSAR deformation monitoring comprises SAR data acquisition, deformation monitoring and landslide boundary determination; the sliding surface is determined by an Okada elastic dislocation model, and the determined content comprises the position and the form of the sliding surface; the landslide volume estimation is calculated from the integral. According to the invention, the position, the shape and the size of the sliding surface are obtained, the sliding depth is combined, the volume of the potential sliding is calculated by utilizing the plane, the square quantity of the potential sliding body is accurately judged, and the data support is provided for the subsequent sliding disaster prediction methods such as numerical simulation and the like, so that the method has a good application value.

Description

A method for estimating potential landslide volume by fusing multi-source SAR data

Technical Field

The invention relates to the field of geological disaster prevention and control safety technology, and in particular to a potential landslide volume estimation method that fuses multi-source SAR data.

Background technique

In recent years, in the field of landslide monitoring, the "air-space-ground" monitoring methods have been continuously developed, and the monitoring and prediction of disasters have been greatly improved. Synthetic Aperture Radar Interferometry (InSAR) technology, as an active remote sensing technology, has been widely used in the monitoring of landslide disasters because of its wide monitoring range, no weather restrictions, and the ability to obtain images all day and all weather.

In terms of general disaster monitoring, most of the current research is still based on monitoring of surface deformation, which cannot specifically reveal the degree of danger of landslides. In addition, for the study of landslide hazards and landslide volumes, the commonly used single-point contact detection is costly, time-consuming and limited by the amount of data. The use of a small number of points for large-scale interpolation will inevitably affect the estimation accuracy of the position and shape of the landslide surface, which is not conducive to the accurate estimation of landslide volume and deformation mode, and affects the hazard assessment of landslides.

By using InSAR technology to obtain surface deformation monitoring results and combining it with the elastic dislocation model to find the best sliding surface, the landslide volume can be estimated, providing data support for disaster risk prediction. However, existing research methods for estimating the volume of potential landslides mostly focus on contact detection such as drilling and ground penetrating radar, which is not only costly and time-consuming, but also not conducive to the sliding surface and depth detection of large and deep landslides.

Summary of the invention

The present invention provides a potential landslide volume estimation method by fusing multi-source SAR data, aiming at solving the shortcomings in the prior art.

In order to achieve the above object, the present invention provides the following technical solution: a method for estimating the volume of a potential landslide by fusing multi-source SAR data, comprising the following steps:

Obtain SAR data of potential landslide areas;

The Stacking InSAR technology is used to perform time-series InSAR processing on SAR data to obtain the line-of-sight deformation rate maps of ascending and descending orbits;

Determine the deformation boundary of the landslide surface based on the ascending and descending track line-of-sight deformation rate map, digital terrain elevation model (DEM) and optical images;

Determine the geometric and kinematic parameters required for inversion of the landslide sliding surface using the Okada elastic dislocation model;

The ascending and descending orbit line-of-sight deformation rate diagrams are converted into discrete points, and the corresponding longitude and latitude coordinates are added as surface constraints for inversion;

Input the surface constraints to the Okada elastic dislocation model to invert the sliding surface, select different smoothing factors, and perform multiple inversion calculations;

By inversion, multiple simulation results are obtained, and the simulation results are compared with the observation results. The simulation result with the best fit is selected to obtain the depth of the sliding body from the surface and the deformation of the sliding surface, and determine the deformation range and deformation direction of the sliding surface.

The volume of the landslide is calculated by integrating the deformation area of the sliding surface and the depth of the sliding body, and the deformation type of the landslide is determined according to the deformation direction of the sliding surface on the sliding surface.

Preferably, deformation monitoring is performed on the SAR data, and the deformation monitoring process includes:

Data acquisition: Download the ascending and descending orbit data of Sentinel-1, the 30-meter SRTM DEM data and the Sentinel-1 precise orbit data, and obtain the single-view complex image SLC by preprocessing the above data;

Deformation monitoring: Stacking InSAR technology is used to obtain deformation rate maps of the landslide area;

Determination of landslide deformation area: Based on the deformation monitoring results obtained, compare the ascending and descending orbit image monitoring results, combine the DEM topographic map and optical image, determine the landslide deformation boundary, and calculate the area of the boundary range.

Preferably, the specific process of obtaining the deformation rate map includes:

By setting certain spatial and temporal baseline thresholds, interferometric processing is performed between SAR images that meet the conditions to generate multiple interferometric atlases.

The phase of each interferogram mainly includes the phase caused by DEM error Phase of deformation Phase caused by atmospheric delay/> Phase caused by noise/>

By eliminating the phase Obtaining phase information generated by deformation;

The Stacking InSAR technology is used to obtain the annual average deformation rate map in the radar line of sight.

Preferably, when estimating the deformation rate, the Stacking InSAR technology obtains a deformation rate map by performing phase superposition on a plurality of interference unwrapping maps;

The weight is calculated based on the time interval between the master and slave images. The mathematical model of Stacking InSAR is as follows:

Where ph_rate is the phase deformation rate, Δt _i is the time interval between the two images of the i-th unwrapped image, is the unwrapped phase value of the i-th interference pattern.

Preferably, the geometric parameters and kinematic parameters required for inverting the sliding surface include width, depth, strike angle, dip angle, maximum sliding amount, and sliding angle, wherein the specific method for determining the parameters is as follows:

Initial depth setting: the thickness of the landslide body, iteratively optimized according to the average thickness;

Initial width setting: the width of the deformation area along the main sliding direction;

Setting of strike angle: first determine the main sliding direction of the landslide, then make a vertical line to the main sliding direction. The angle between the vertical line and the north direction is the initial strike angle.

Initial inclination angle setting: set the interval close to the slope to search, and determine the best inclination angle through continuous adjustment;

Maximum sliding amount: Set the maximum sliding limit value of the sliding surface;

Initial sliding angle: Set a certain search range, continuously adjust the maximum and minimum sliding angle values in the interval, and determine the size of the sliding angle through multiple inversion calculations.

Preferably, the process of inverting the sliding surface using the Okada elastic dislocation model comprises the following steps:

Combined with the deformation monitoring results obtained by InSAR technology, the deformation boundary and area of the landslide are determined;

Use quadtree sampling to downsample the data;

The scale and direction of the sliding surface area to be inverted are set, a dislocation inversion model is constructed, and the sliding surface is divided into multiple sub-dislocation sources along the strike and dip, and the length, width, depth of the sliding surface and the block structure of the sub-dislocation are given.

Preferably, the deformation of each sub-block dislocation is calculated by the Okada elastic dislocation model, and the specific calculation formula is:

S＝[G ^T C ^-1 G] ^-1 GC ^-1 X

Among them, S is the slip amount of each sub-dislocation source, G is the Green's function established by the connection between ground deformation and sliding surface, C is the covariance of observation data, and X is the deformation of the surface.

Preferably, the specific calculation formula of the landslide volume is as follows:

Among them, a, b are represented by a1, a2, b1, b2, and h respectively, a1, b1 are the major and minor semi-axes of the circumscribed ellipse of the sliding surface, a2, b2 are the major and minor semi-axes of the circumscribed ellipse of the landslide surface, and h represents the depth of the landslide body. The specific deformation formula of the formula is:

a＝a1+z(a2-a1)/h

b＝b2+z(b2-b1)/h

Preferably, the parameters of the landslide surface deformation include deformation range and deformation magnitude.

Compared with the prior art, the present invention has the following beneficial effects:

1. Use time-series InSAR technology to perform multi-source data deformation detection and obtain deformation monitoring results of different data. By comparison, determine the area and perimeter of the potential landslide on the surface, and combine with the elastic dislocation model to find the most likely dislocation plane for the sliding of the landslide base, and obtain the position, shape and size of the sliding surface. Combined with the depth of the landslide, use this plane to calculate the volume of the potential landslide, accurately determine the volume of the potential landslide body, and provide data support for subsequent numerical simulation and other landslide disaster prediction methods, which has good application value.

2. The present invention is different from the method of inverting the thickness of the landslide body by mass conservation. It does not need to use external parameters such as rheological parameters. It not only obtains the thickness of the potential landslide body, but also inverts the deformation information of the sliding surface. On this basis, it calculates the volume of the landslide body in combination with the deformation information of InSAR.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG1 is a specific flow chart of a method for estimating the volume of a potential landslide by fusing multi-source SAR data provided by the present invention;

FIG2 is a schematic diagram of a landslide dislocation model and parameters provided by the present invention;

FIG3 is a schematic diagram of landslide volume estimation provided by the present invention;

FIG4 is a diagram of the deformation rate of the ascending and descending rails provided by the present invention;

FIG5 is a roughness and fitting residual curve diagram provided by the present invention;

FIG6 is a diagram of the ascending and descending orbit inversion results provided by the present invention;

FIG7 is an optimal sliding distribution diagram provided by the present invention;

FIG8 is a deformation direction diagram of each landslide unit of the landslide body 1 provided by the present invention;

FIG9 is a slope diagram of the landslide body 1 provided by the present invention;

FIG10 is a deformation direction diagram of each landslide unit of the landslide body 2 provided by the present invention;

FIG. 11 is a slope diagram of the landslide body 2 provided by the present invention.

Detailed ways

The specific implementation of the present invention is further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present invention.

A potential landslide volume estimation method based on multi-source SAR data fusion combines the ground surface deformation obtained by InSAR technology and the sliding surface determined by the elastic dislocation model to estimate the landslide volume. As shown in Figure 1, the technical solution of a potential landslide volume estimation method based on multi-source SAR data fusion is as follows:

Step 1: Use the Stacking InSAR technology to process the Sentinel-1A data and obtain the line-of-sight deformation rate diagrams for ascending and descending orbits.

Step 2: Determine the parameters of landslide surface deformation, including deformation range, deformation magnitude, etc., based on the deformation rate map results, digital terrain elevation model (DEM) and optical images.

Step 3: Determine the geometric parameters and kinematic parameters required for inverting the sliding surface of the landslide using the Okada elastic dislocation model: width, depth, strike angle, dip angle, maximum sliding amount, sliding angle and other parameters; the specific method for determining the parameters is as follows:

(1) Initial depth setting: the thickness of the landslide body, which is iteratively optimized based on the average thickness.

(2) Initial width setting: the width of the deformation area along the main sliding direction.

(3) Setting of strike angle: First determine the main sliding direction of the landslide, then draw a perpendicular line to the main sliding direction. The angle between the perpendicular line and the north direction is the initial strike angle.

(4) Initial inclination angle setting: Set a range close to the slope of the slope for searching, and determine the optimal inclination angle through continuous adjustment.

(5) Maximum sliding amount: Set the maximum sliding limit value of the sliding surface.

(6) Initial sliding angle: Set a certain search range, continuously adjust the maximum and minimum sliding angle values in the interval, and determine the size of the sliding angle through multiple inversion calculations.

Step 4: Convert the deformation rate diagram of the ascending and descending track into discrete points and add the corresponding longitude and latitude coordinates as surface constraints for inversion.

Step 5: Downsample the ascending and descending orbit line-of-sight deformation rate graphs. Usually, the quadtree and uniform downsampling method is used to reduce the amount of calculation and improve the calculation speed.

Step 6: Input the surface constraints into the Okada elastic dislocation model to perform the inversion of the sliding surface. Select different smoothing factors, perform multiple inversion calculations, and select the optimal smoothing factor based on the L curve of the fitting result to reduce uncertainty and improve the stability of the inversion result. By comparing the multiple simulation results obtained by inversion with the observation results, the simulation result with the best fit is selected to obtain the depth of the sliding body and the deformation of the sliding surface, and determine the deformation range and deformation direction of the sliding surface.

Step 7: Calculate the landslide volume by integration and determine the deformation type of the landslide based on the deformation direction of the sliding surface on the sliding surface.

The present invention is mainly divided into three key steps: SAR deformation monitoring, sliding surface determination and landslide volume estimation.

1. SAR deformation monitoring includes:

1. SAR data acquisition: Download the ascending and descending orbit data of Sentinel-1, and obtain the single-view complex image SLC through preprocessing.

2. Deformation monitoring: Stacking InSAR technology is used to obtain the deformation rate map of the landslide area. First, by setting certain spatial and temporal thresholds, interference is performed between SAR images that meet the conditions to generate multiple interference map sets. For each interference map, the phase Mainly includes the phase caused by DEM error/> Phase produced by deformation/> Phase caused by atmospheric delay/> Phase caused by noise/>

Then, by removing the phase Obtain the phase information generated by deformation. During the data processing, the terrain phase in the SAR coordinate system can be obtained through external DEM data simulation, and the terrain phase obtained by differential elimination simulation can be used to obtain the differential interference map. For terrain errors, the relationship between elevation error and orbital baseline is constructed to remove them; for atmospheric phase errors, the quadratic polynomial between DEM and atmospheric phase is constructed to simulate and remove them to obtain the deformation phase map. Finally, the Stacking InSAR technology is used to obtain the annual average deformation rate map in the radar line of sight. When estimating the deformation rate, the Stacking InSAR technology is calculated based on the time interval between the master and slave images. The mathematical model of Stacking InSAR is as follows:

Where ph_rate is the phase deformation rate, Δt _i is the time interval between the two images of the i-th unwrapped image, is the unwrapped phase value of the i-th interferogram. In the process of Stacking InSAR processing, in order to reduce the error caused by atmospheric disturbance, it is necessary to use the unwrapped interferogram with atmospheric error removed, and eliminate the interferogram with large error. Under the condition of meeting the coherence condition, the interferogram with longer time baseline is selected for processing, so as to obtain a more accurate deformation rate.

3. Determination of landslide deformation area: Based on the deformation monitoring results obtained, compare the ascending and descending orbit image monitoring results, combine the DEM topographic map and optical images, determine the landslide deformation boundary, and calculate the area of the boundary range.

2. As shown in FIG. 2 , the position and shape of the sliding surface are determined based on the Okada elastic dislocation model, which specifically includes:

Remote sensing and GNSS monitoring methods can only obtain surface information, and it is difficult to determine changes in the interior of the earth through these technologies. However, for landslides, the sliding surface plays an important role in landslide activity. According to previous studies, the sliding surfaces of some large landslides can be assumed to be tectonic faults. Therefore, in order to obtain the volume of the landslide, we can use the Okada dislocation model, assuming that the landslide is an isotropic medium, and the displacement field observed by InSAR is simulated by a plane dislocation in an isotropic elastic half-space. By finding the best fit between the observed value and the simulated value, the position, shape and size of the sliding surface are determined, thereby providing data for the estimation of the landslide volume.

First, the deformation monitoring results obtained by InSAR technology are combined to determine the deformation boundary and area of the landslide, and the quadtree sampling method is used to downsample the data to improve the calculation efficiency. Then, the scale and direction of the sliding surface are set, the dislocation inversion model is constructed, and the sliding surface is divided into multiple sub-dislocation sources along the strike and dip. The length, width, depth of the sliding surface and the block situation of the sub-dislocation are given. The specific parameters are shown in Figure 2. Finally, the deformation of each sub-block dislocation is calculated using the Okada elastic dislocation model. The specific calculation formula is:

S＝[G ^T C ^-1 G] ^-1 GC ^-1 X (3)

In the formula, S is the slip amount of each sub-dislocation source, G is the Green's function established by the connection between ground deformation and sliding surface, C is the covariance of observation data, and X is the deformation of the surface. In the inversion process, since the slip amount of the sub-dislocation source sometimes jumps, a certain degree of smoothing is required to ensure the stability of the inversion result.

3. As shown in Figure 3, the specific process of estimating the landslide volume is as follows:

Combined with the surface deformation monitoring results obtained by InSAR, the upper surface area of the landslide can be determined. The lower surface area of the landslide can be determined based on the sliding surface inverted by the elastic dislocation model, and the average depth h of the landslide is obtained. In general, it is assumed that the upper surface of the landslide is nearly parallel to the sliding surface. At this time, the upper and lower surfaces of the landslide are selected to be the closest ellipse, which can be represented as Figure 3.

Assume that the major and minor axes of the circumscribed ellipse of the sliding surface are a1 and b1 respectively; the major and minor axes of the circumscribed ellipse of the landslide surface are a2 and b2 respectively, and the depth of the landslide body is h. Take the integration unit dz. The height from the top surface is z. The volume of the landslide body can be calculated by integration. The specific calculation formula is as follows:

Where a and b are represented by a1, a2, b1, b2, and h respectively:

a＝a1+z(a2-a1)/h (5)

b＝b2+z(b2-b1)/h (6)

Substituting a and b into the formula, we can find the volume V of the landslide:

Embodiment 1:

The present invention takes the potential landslide in Wadui Village as an example to conduct experiments, calculate the volume of the potential landslide body, and analyze the landslide sliding type according to the sliding surface inversion result.

1. Deformation rate map obtained by InSAR: According to the deformation rate results, the geometric size and position of the inverted sliding surface are determined. For landslide bodies with different deformation directions, multiple sliding surfaces should be selected for inversion. Its deformation rate map is shown in Figure 4, where Figure 4a is the result of the ascending orbit monitoring, Figure 4b is the result of the descending orbit monitoring, and the white box is the inversion area of the sliding surface.

2. Roughness and fitting residual curve of Okada model: As shown in Figure 5, during the inversion process of sliding distribution, the sliding distribution result is evaluated by weighing the curve between the model roughness and the fitting residual. The experiment simulates inversion multiple times and selects a smoothing factor of 0.03, when the inversion result of the model is optimal.

3. The sliding surface inversion results of the Okada model are shown in Table 1:

Table 1 Inversion results of elastic dislocation model

Sliding surface length width Depth (km) Strike Angle inclination Average sliding amount 1 1.6 0.7 0.069 20 35 0.24 2 0.9 1 0.060 25 twenty three 0.21

The comparison between the landslide deformation observation results and the sliding distribution is shown in Figure 6. In Figure 6, a, b, and c are the observation values, simulation values, and residual values of the ascending orbit, respectively, and e, f, and g are the observation values, simulation values, and residual values of the descending orbit, respectively.

The optimal sliding distribution results of the sliding surface are shown in Figure 7.

4. Estimation of landslide volume

According to the deformation rate diagram, the maximum circumscribed ellipse of the landslide surface is delineated, and the circumscribed ellipse of the sliding surface is determined in combination with the inverted sliding surface, so as to obtain the lengths of the major and minor semi-axes of the upper and lower surfaces of the landslide body, and combined with the depth value obtained by inversion, the volumes of the two landslide bodies in the area are calculated as follows:

(1) Landslide body 1

Figures 8 and 9 are the deformation direction diagrams of each landslide unit of landslide body 1 and the slope diagram of landslide body 1, respectively. According to the contents of Figures 4, 8 and 9, it is calculated that a1 = 0.35km, b1 = 0.19km, a2 = 0.6km, b2 = 0.26km, and h = 0.069km. According to the volume calculation formula, the volume of landslide body 1 is V = 1.69*10 ⁷ m ^3. Combining the deformation of the strike and dip, the deformation direction of the landslide can be roughly determined. For landslide body 1 at the foot of the slope, combined with the slope joint analysis, the landslide mainly moves along the slope, and there is deformation in both the strike and dip. The movement of landslides can generally be divided into translation, rotation or complex movement. Rotational landslides generally move downward and outward, while the movement direction of translational sliding bodies is parallel to the hillside, and the sliding surface is relatively flat in structure. According to the results of the inversion sliding surface, landslide body 1 is a translational landslide along the slope.

(2) Landslide body 2

Figures 10 and 11 are the deformation direction diagrams of each landslide unit of landslide body 2 and the slope diagram of landslide body 2, respectively. According to the contents of Figures 4, 10 and 11, it is calculated that a1 = 0.40 km, b1 = 0.24 km, a2 = 0.5 km, b2 = 0.31 km, h = 0.06 km, and it can be obtained by substituting it into the volume calculation formula that the volume of landslide body 2 is V = 2.03*10 ⁶ m ³ . According to the inversion results, landslide body 2 is a translational landslide.

The embodiments described above are only preferred specific implementation modes of the present invention, and the protection scope of the present invention is not limited thereto. Any simple changes or equivalent replacements of the technical solutions that can be obviously obtained by any technician familiar with the field within the technical scope disclosed in the present invention belong to the protection scope of the present invention.

Claims (9)

1. A method for estimating the volume of a potential landslide by fusing multi-source SAR data, characterized in that it comprises the following steps: Obtain SAR data of potential landslide areas; The Stacking InSAR technology is used to perform time-series InSAR processing on SAR data to obtain the line-of-sight deformation rate maps of ascending and descending orbits; Determine the deformation boundary of the landslide surface based on the ascending and descending track line-of-sight deformation rate map, digital terrain elevation model (DEM) and optical images; Determine the geometric and kinematic parameters required for inversion of the landslide sliding surface using the Okada elastic dislocation model; The ascending and descending orbit line-of-sight deformation rate diagrams are converted into discrete points, and the corresponding longitude and latitude coordinates are added as surface constraints for inversion; Input the surface constraints to the Okada elastic dislocation model to invert the sliding surface, select different smoothing factors, and perform multiple inversion calculations; By inversion, multiple simulation results are obtained, and the simulation results are compared with the observation results. The simulation result with the best fit is selected to obtain the depth of the sliding body and the deformation of the sliding surface, and determine the deformation range and deformation direction of the sliding surface. The volume of the landslide is calculated by integrating the deformation area of the sliding surface and the depth of the sliding body, and the deformation type of the landslide is determined according to the deformation direction of the sliding surface on the sliding surface. 2. The method for estimating the volume of a potential landslide by fusing multi-source SAR data according to claim 1, characterized in that deformation monitoring is performed on the SAR data, and the deformation monitoring process comprises: Data acquisition: Download the ascending and descending orbit data of Sentinel-1, the 30-meter SRTM DEM data and the Sentinel-1 precise orbit data, and obtain the single-view complex image SLC by preprocessing the above data; Deformation monitoring: Stacking InSAR technology is used to obtain deformation rate maps of the landslide area; Determination of landslide deformation area: Based on the deformation rate results obtained, the ascending and descending orbit image monitoring results are compared, combined with the DEM topographic map and optical images, the landslide deformation boundary is determined, and the area of the boundary range is calculated. 3. The method for estimating the potential landslide volume by fusing multi-source SAR data according to claim 2, wherein the specific process of obtaining the deformation rate map comprises: By setting certain spatial and temporal baseline thresholds, interferometric processing is performed between SAR images that meet the conditions to generate multiple interferometric atlases. The phase of each interferogram mainly includes the phase caused by DEM error Phase produced by deformation/> Phase caused by atmospheric delay/> Phase caused by noise/> By eliminating the phase Obtaining phase information generated by deformation; The Stacking InSAR technology is used to obtain the annual average deformation rate map in the radar line of sight. 4. A method for estimating the volume of a potential landslide by fusing multi-source SAR data according to claim 3, characterized in that, when estimating the deformation rate, the Stacking InSAR technology obtains a deformation rate map by performing phase superposition on a plurality of interference unwrapping maps; Among them, the weight is calculated according to the time interval between the master and slave images. The mathematical model of Stacking InSAR is as follows: Where ph_rate is the phase deformation rate, Δt _i is the time interval between the two images of the i-th unwrapped image, is the unwrapped phase value of the i-th interference pattern. 5. A method for estimating the volume of a potential landslide by fusing multi-source SAR data as claimed in claim 1, characterized in that the geometric parameters and kinematic parameters required for inverting the sliding surface include width, depth, strike angle, inclination angle, maximum sliding amount, and sliding angle, wherein the specific method for determining the parameters is as follows: Initial depth setting: the thickness of the landslide body, iteratively optimized according to the average thickness; Initial width setting: the width of the deformation area along the main sliding direction; Setting of strike angle: first determine the main sliding direction of the landslide, then make a vertical line to the main sliding direction. The angle between the vertical line and the north direction is the initial strike angle. Initial inclination angle setting: set the interval close to the slope to search, and determine the best inclination angle through continuous adjustment; Maximum sliding amount: Set the maximum sliding limit value of the sliding surface; Initial sliding angle: Set a certain search range, continuously adjust the maximum and minimum sliding angle values in the interval, and determine the size of the sliding angle through multiple inversion calculations. 6. A method for estimating the volume of a potential landslide by fusing multi-source SAR data as claimed in claim 1, characterized in that the process of inverting the sliding surface using the Okada elastic dislocation model comprises the following steps: Combined with the deformation monitoring results obtained by InSAR technology, the deformation boundary and area of the landslide are determined; Use quadtree sampling to downsample the data; The scale and direction of the sliding surface area to be inverted are set, a dislocation inversion model is constructed, and the sliding surface is divided into multiple sub-dislocation sources along the strike and dip, and the length, width, depth of the sliding surface and the block structure of the sub-dislocation are given. 7. A method for estimating the volume of a potential landslide by fusing multi-source SAR data as claimed in claim 6, characterized in that the deformation of each sub-block dislocation is calculated by the Okada elastic dislocation model, and the specific calculation formula is: S＝[G ^T C ^-1 G] ^-1 GC ^-1 X Among them, S is the slip amount of each sub-dislocation source, G is the Green's function established by the connection between ground deformation and sliding surface, C is the covariance of observation data, and X is the deformation of the surface. 8. The method for estimating the potential landslide volume by fusing multi-source SAR data according to claim 1, wherein the specific calculation formula of the landslide volume is as follows: Among them, a, b are represented by a1, a2, b1, b2, and h respectively, a1, b1 are the major and minor semi-axes of the circumscribed ellipse of the sliding surface, a2, b2 are the major and minor semi-axes of the circumscribed ellipse of the landslide surface, and h represents the depth of the landslide body. The specific deformation formula of the formula is: a＝a1+z(a2-a1)/h b＝b2+z(b2-b1)/h 9. The method for estimating the potential landslide volume by fusing multi-source SAR data according to claim 1, wherein the parameters of the landslide surface deformation include deformation range and deformation magnitude.