A method for evaluating the suitability of GNSS monitoring point selection for landslide disasters

Abstract

The present invention discloses a method for evaluating the suitability of GNSS monitoring point location for landslide disasters, which relates to the technical field of disaster monitoring. The present invention includes determining the location of the landslide disaster area and the floor area of the global navigation satellite system (GNSS) monitoring point; using differential interferometric synthetic aperture radar (D-InSAR) technology to process the synthetic aperture radar (SAR) data of the landslide disaster area to obtain the prior deformation information of the landslide disaster point; using a digital elevation model (DEM) to calculate the slope, aspect, and surface roughness of the landslide disaster area; using an orthoimage (DOM) to calculate the vegetation index of the landslide disaster area; preprocessing the five index factors, including unifying the spatial resolution and normalizing the numerical values of the index factors; using the analytic hierarchy process to determine the weights of the five indexes; and performing weighted superposition on the five indexes to obtain an evaluation map of the suitability of GNSS monitoring station location in the landslide disaster area.

Description

Technical Field

[0001] The present invention relates to the technical field of disaster monitoring, and in particular to a method for evaluating the suitability of GNSS monitoring point selection for landslide disasters. Background Art

[0002] Due to its characteristics such as global coverage, all-weather operation, continuity, and real-time performance, GNSS technology is widely used in the three-dimensional deformation monitoring of landslide disasters. When using GNSS to monitor the deformation of landslides, the selection of GNSS monitoring points is particularly important.

[0003] Currently, it is mainly through professional personnel going to the landslide disaster site to judge the key positions that can reflect the deformation characteristics through experience, and then determining the layout positions of GNSS monitoring points after considering the influences such as field power supply and GNSS multipath.

[0004] However, the method of determining the layout positions of GNSS monitoring points by professional personnel on-site inspection has strong subjectivity, is time-consuming and laborious, has low work efficiency, and is difficult to promote and apply. Summary of the Invention

[0005] In view of the above problems, there is an urgent need to study a method for evaluating the suitability of GNSS monitoring point selection in landslide disaster areas. The present invention proposes a method for evaluating the suitability of GNSS monitoring point selection for landslide disasters, which combines deformation prior information and topographic data of the disaster area. This method can comprehensively evaluate the suitability of GNSS monitoring point selection in the landslide area without the need for professional personnel to reach the landslide site, improving the operation efficiency of GNSS monitoring point selection for landslide disasters.

[0006] A method for evaluating the suitability of GNSS monitoring point selection for landslide disasters proposed by the present invention includes the following steps:

[0007] Determine the location of the landslide disaster area and the occupied area of the GNSS monitoring points;

[0008] Use differential interferometric synthetic aperture radar (D-InSAR) technology to process the synthetic aperture radar (SAR) data of the landslide disaster area to obtain a deformation map of the landslide disaster area;

[0009] Use a digital elevation model (DEM) to calculate the slope, aspect, and surface roughness of the landslide disaster area;

[0010] Use an orthoimage (DOM) to calculate the vegetation index of the landslide disaster area;

[0011] Preprocess the five index factors of deformation, slope, aspect, surface roughness, and vegetation index in the landslide disaster area. The preprocessing includes unifying the spatial resolution and normalizing the values of the index factors;

[0012] Use the analytic hierarchy process to determine the weights of five indicators: deformation, slope, aspect, surface roughness, and vegetation index in the landslide disaster area;

[0013] Perform weighted overlay on the five indicators of deformation, slope, aspect, surface roughness, and vegetation index in the landslide disaster area to obtain the suitability evaluation map for the GNSS monitoring station location in the landslide disaster area.

[0014] Further, the obtaining of the deformation map of the landslide disaster area specifically includes:

[0015] The method of using D-InSAR for surface deformation measurement is the two-track plus external DEM method; the processing process includes:

[0016] Download two scenes of ascending or descending orbit data of Sentinel-1 covering the landslide disaster area, and obtain the single-look complex image SLC through preprocessing;

[0017] Set the SAR data with an earlier acquisition time as the master image, and the other image as the slave image. Then register and resample the slave image to the same radar image coordinate space as the master image. After registration and resampling, determine the one-to-one correspondence of the homologous pixels;

[0018] Multiply the complex data of the corresponding pixels of the master and slave images conjugately to obtain the initial interferogram. At this time, each pixel in the initial interferogram is still complex data, and the phase of this complex number is the interference phase. The range of the interference phase is [-π, π). At this time, the interference phase includes the reference ellipsoid phase, terrain relief phase, surface deformation phase, atmospheric delay phase, and interference noise phase components;

[0019] The reference ellipsoid phase is the phase component caused by the interaction between the spatial attitudes of the two SAR satellite imagings and the reference ellipsoid. This phase can be removed by constructing a model based on the satellite orbit parameters at the acquisition times of the master and slave images;

[0020] In order to obtain the surface deformation phase, it is necessary to subtract the terrain phase from the interferogram, calculate the terrain phase contribution through external DEM data, and then subtract this contribution from the interference phase;

[0021] Since there are still atmospheric delay phase and noise phase in the interferogram, it is necessary to filter the differential interferogram to weaken its noise signal and improve the signal-to-noise ratio of the interference data. After phase filtering, the phase quality of the differential interferogram can be effectively improved, and the accuracy of phase unwrapping can be improved. However, there will still be significant phase noise in some severely decoherent areas, and these decoherent areas need to be masked;

[0022] The deformation phase recorded in the differential interferogram is wrapped in the interval [-π, π), and the true surface deformation information cannot be directly characterized. Phase unwrapping processing is required to restore the integer number of phase cycles for each pixel;

[0023] The product directly obtained by SAR interferometry processing is in the range-Doppler coordinate frame. For the convenience of later processing, it needs to be projected and converted to a commonly used geographic coordinate system.

[0024] Furthermore, the calculation of the slope, aspect, and surface roughness of the landslide disaster area using the digital elevation model DEM specifically includes:

[0025] Slope, aspect, and surface roughness are commonly used indicators in GIS terrain analysis and can be obtained by calculating the DEM with the help of software such as ArcGIS or SAGA;

[0026] The fitting surface method can be used for slope and aspect calculation. The fitting surface can be a quadratic surface, that is, a 3x3 window, with the center of each window being an elevation point. The slope and aspect calculation formulas for the center point e are:

[0027]

[0028]

[0029]

[0030] In the formula, a, b, c, d, f, g are the elevation values of the corresponding pixels, cellsize is the spatial resolution of the pixel, dx and dy are the change rates of the elevation values in the x and y directions, and slope and aspect respectively represent the slope and aspect of the central pixel e;

[0031] The surface roughness reflects the elevation difference value between the central pixel and its surrounding pixels, which characterizes the degree of unevenness of the central pixel; the surface roughness calculation formula is:

[0032]

[0033] In the formula, TRI represents the surface roughness of the central pixel e, and the other symbols represent the elevation values of each pixel.

[0034] Furthermore, the calculation of the vegetation index of the landslide disaster area using the orthoimage DOM specifically includes:

[0035] The vegetation index can reflect the degree of surface vegetation coverage to a certain extent;

[0036] The orthophoto of the landslide disaster area contains information on three visible light bands: red, green, and blue. By calculating the band information of each pixel, the vegetation index of the corresponding pixel can be obtained. When the vegetation index is high, it indicates a high vegetation content or a high probability of being a plant at that pixel.

[0037] The calculation formula for the vegetation index is:

[0038] VI = (2G - R - B) - (1.4R - G);

[0039]

[0040] In the formula, VI is the vegetation index value of the corresponding pixel, and r, g, and b are the values of the red, green, and blue bands of each pixel.

[0041] Furthermore, the preprocessing of the deformation index factor for the landslide disaster area specifically includes:

[0042] For the deformation map obtained by D-InSAR technology, its spatial resolution is generally large. Therefore, it is necessary to resample the spatial resolution of the deformation map to the construction area of GNSS monitoring points. The resampling method uses the nearest neighbor interpolation method. The purpose of interpolation is mainly to unify the spatial resolution and ensure that the deformation information does not change.

[0043] After unifying the spatial resolution, the magnitude of the deformation map may vary for different landslides and also has a large difference from the magnitudes of other index factors. Therefore, it is necessary to normalize the deformation value of the deformation map to the interval [0, 1]. The closer the normalized value is to 1, the greater the relative deformation degree in the disaster area, and it is more necessary to monitor. Therefore, the normalized deformation value also reflects the influence degree of deformation on the suitability evaluation of GNSS monitoring point siting.

[0044] The normalization formula for the deformation index factor is:

[0045]

[0046] In the formula, X norm represents the normalized value, X represents the value before normalization, abs represents the absolute value operation, X max , X min respectively represent the maximum and minimum values of the absolute values of each pixel before normalization.

[0047] Furthermore, the preprocessing of the slope index factor for the landslide disaster area specifically includes:

[0048] By calculating the DEM, the slope value of the landslide disaster area can be obtained;

[0049] Unify the spatial resolution of the DEM with the footprint area of the GNSS monitoring points. If the DEM resolution is higher than the footprint area, use the average resampling method for downsampling; otherwise, use the nearest neighbor interpolation method for upsampling.

[0050] Considering the construction difficulty and cost issues during the construction of GNSS monitoring points, in the actual construction work of monitoring points, areas with a slope less than 30 degrees are generally selected for the construction of GNSS monitoring points; therefore, locations with a slope greater than 30 degrees are considered unsuitable for the construction of GNSS monitoring points, and pixels with a slope greater than 30 degrees are masked during the preprocessing of the slope index factor and do not participate in the subsequent siting suitability evaluation.

[0051] Normalize the masked slope index factor; when the slope is smaller, the construction difficulty of the GNSS monitoring station is lower and the cost is also lower, that is, it is more suitable for the construction of GNSS monitoring points, and its value should also be closer to 1 after normalization.

[0052] The normalization formula for the slope index factor is:

[0053]

[0054] In the formula, Slope norm represents the normalized slope index factor, Slope represents the slope before normalization, Slope max is the maximum slope value of 30° before normalization, Slope min is the minimum slope value of 0° before normalization.

[0055] Furthermore, preprocess the aspect index factor of the landslide disaster area, specifically including:

[0056] Unify the spatial resolution to the construction area of the GNSS monitoring points for subsequent weighted overlay calculation.

[0057] The value range of the aspect index factor is 0 - 360°. Its main impact on the GNSS monitoring points is the sunshine duration. Since the GNSS monitoring stations in the wild are mainly powered by solar energy, the longer the illumination duration, the more sufficient the solar power supply, that is, it is more suitable for the construction of GNSS monitoring points. Therefore, normalizing the aspect index factor to [0,1] is carried out through its aspect range.

[0058] Divide the aspect into four categories: shady slope, semi-shady slope, semi-sunny slope, and sunny slope, and their corresponding normalized values are set to 0.2, 0.4, 0.6, and 0.8 respectively. The larger the normalized value, the more suitable the corresponding pixel position is for the construction of GNSS monitoring points.

[0059] Furthermore, preprocess the surface roughness index factor of the landslide disaster area, specifically including:

[0060] Unify the spatial resolution to the floor area of the GNSS monitoring points for subsequent weighted overlay calculation;

[0061] The surface roughness reflects the unevenness between the corresponding pixel position and adjacent pixels. The greater the surface roughness, the less suitable it is for the construction of GNSS monitoring points;

[0062] The normalization formula for surface roughness is:

[0063]

[0064] In the formula, TRI norm represents the normalized surface roughness index factor, TRI represents the surface roughness before normalization, TRI max , TRI min are the maximum and minimum values of the surface roughness before normalization.

[0065] Furthermore, the preprocessing of the vegetation index indicator factor for the landslide disaster area specifically includes:

[0066] Unify the spatial resolution to the floor area of the GNSS monitoring points for subsequent weighted overlay calculation;

[0067] Because the vegetation near the GNSS monitoring points will have a certain multipath effect on the GNSS monitoring station, in actual work, the monitoring points are generally built at locations without vegetation or with less vegetation coverage;

[0068] Obtain the vegetation index through orthophoto image calculation. This index reflects the vegetation coverage of the surface to a certain extent, that is, the greater this value, the more serious the vegetation coverage;

[0069] The normalization formula for the vegetation index is:

[0070]

[0071] In the formula, VI norm represents the normalized vegetation index indicator factor, VI represents the vegetation index before normalization, VI max , VI min are the maximum and minimum values of the vegetation index before normalization.

[0072] Furthermore, the obtaining of the suitability evaluation map for the GNSS monitoring station in the landslide disaster area specifically includes:

[0073] The key to the analytic hierarchy process is to construct a judgment matrix, the purpose of which is to determine the weights of each indicator factor. The Saaty scale is usually used to construct the judgment matrix, and the Saaty scale range is [1, 9];

[0074] Based on the 5 index factors used in the GNSS monitoring point site selection suitability evaluation: deformation, slope, aspect, surface roughness, and vegetation index, a 5x5 judgment matrix P is constructed. The elements in this matrix should satisfy the formula: P i,j ·P j,i = 1; where i and j in the formula represent the row and column numbers respectively;

[0075] Normalize the judgment matrix, that is, make the sum of each column of the normalized matrix equal to 1. The specific calculation formula is:

[0076] For each index factor, the range of its weight value is [0, 1]. The higher the weight value, the greater its impact on the GNSS monitoring point site selection suitability. The calculation formula for the relative weight is: In the formula, w i represents the weight of each index factor, represents the value of the element in the l-th row and j-th column of the normalized judgment matrix;

[0077] The consistency test is determined by the size of the consistency ratio CR value. When CR ≤ 0.1, it is considered that the established analytic hierarchy process model is reasonable; the calculation of the CR value is the ratio of the consistency index CI to the random consistency index RI given, where the value of RI is related to the number of index factors. The GNSS monitoring point site selection suitability evaluation involves 5 index factors. Therefore, through the random consistency index table, the value of RI can be determined to be 1.12; the calculation formula for CI is: In the formula, λ max is the largest eigenvalue of the judgment matrix.

[0078] After passing the consistency test, weighted superposition of each index factor can be used to obtain the GNSS monitoring point site selection suitability map. The weighted superposition formula is:

[0079] S i = w deformation ·X i + w slope ·Slope i +

[0080] w aspect ·Aspect i + w TRI ·TRI i + w VI ·VI i ;

[0081] In the formula, S i represents the GNSS monitoring point suitability of the i-th pixel in the landslide disaster area, w deformation , wslope , w aspect , w TRI , w VI respectively represent the weights of the deformation index, slope index, aspect index, terrain roughness index, and vegetation index determined by the analytic hierarchy process. X i , Slope i , Aspect i , TRI i , VI i respectively represent the values of the five index factors of deformation, slope, aspect, terrain roughness, and vegetation index corresponding to the i-th pixel after preprocessing.

[0082] Compared with the prior art, a method for evaluating the suitability of GNSS monitoring point selection for landslide disasters provided by the present invention has the following beneficial effects:

[0083] The present invention uses the differential interferometric synthetic aperture radar D-InSAR technology to obtain the prior deformation information of the landslide area, then calculates the slope, aspect, terrain roughness, and vegetation index of the landslide area based on the terrain data. After preprocessing the five index factors of deformation, slope, aspect, terrain roughness, and vegetation index, the weights of each index factor are determined by the analytic hierarchy process. Finally, the suitability map of GNSS monitoring point selection in the landslide area is obtained through weighted overlay; compared with the method of manual site selection by on-site inspection, the method disclosed by the present invention can more quickly and objectively evaluate the suitability of GNSS monitoring point selection for the overall landslide area, and is easy to be programmed and implemented, having good popularization and application value. Brief Description of the Drawings

[0084] Figure 1 is a flowchart of a method for evaluating the suitability of GNSS monitoring point selection for landslide disasters provided by an embodiment of the present invention;

[0085] Figure 2 is a schematic diagram of a 3x3 window of DEM raster data provided by an embodiment of the present invention;

[0086] Figure 3 is a schematic diagram of the nearest neighbor interpolation method provided by an embodiment of the present invention;

[0087] Figure 4 is an orthophoto map of an unmanned aerial vehicle in the landslide area provided by an embodiment of the present invention;

[0088] Figure 5 is a deformation map obtained by D-InSAR provided by an embodiment of the present invention;

[0089] Figure 6 is a preprocessing result map of the deformation index factor provided by an embodiment of the present invention;

[0090] Figure 7 The preprocessing result diagram of the slope index factor provided by the embodiment of the present invention;

[0091] Figure 8 The preprocessing result diagram of the slope aspect index factor provided by the embodiment of the present invention;

[0092] Figure 9 The preprocessing result diagram of the surface roughness index factor provided by the embodiment of the present invention;

[0093] Figure 10 The preprocessing result diagram of the vegetation index factor provided by the embodiment of the present invention;

[0094] Figure 11 The suitability evaluation result diagram of GNSS monitoring points provided by the embodiment of the present invention. Specific implementation manners

[0095] The following combines the attached Figures 1 to 11 , and further describes the specific implementation manners of the present invention. The following embodiments are only used to more clearly illustrate the technical solutions of the present invention, and cannot be used to limit the protection scope of the present invention.

[0096] Embodiment 1: As Figures 1 - 11 shown, a method for evaluating the suitability of GNSS monitoring point selection for landslide disasters provided by the present invention includes the following steps:

[0097] S1. Determine the location of the landslide disaster area and the occupied area of the GNSS monitoring point. Determine the location of the landslide disaster area where the suitability evaluation of GNSS monitoring point selection is required, and this location can be represented by the WGS-84 coordinate system; give the approximate occupied area of the GNSS monitoring point, which can generally be set to 1m×1m or 2m×2m, and the specific size can be given according to the actual construction size of the GNSS monitoring station.

[0098] S2. Use D-InSAR to obtain the deformation diagram of the landslide disaster area. The commonly used method for surface deformation measurement using D-InSAR is the two-track method with external DEM, and the main processing processes include: SAR data preprocessing, registration and resampling of the master and slave SAR images, differential interference of the master and slave images, removal of the reference ellipsoid phase, removal of the terrain phase, interference map filtering and masking, phase unwrapping of the interference map, and projection conversion.

[0099] S3. Calculate the slope, slope aspect, surface roughness, and vegetation index factor. The slope, slope aspect, and surface roughness can be calculated from the high-resolution DEM, and the vegetation index can be calculated from the visible light band information included in the orthoimage. Among them, the slope characterizes the steepness of the mountain, the slope aspect represents the orientation of the mountain, the surface roughness reflects the unevenness of the surface, and the vegetation index reflects the degree of ground vegetation coverage to a certain extent.

[0100] S4. Preprocessing of evaluation index factors for site suitability. There are five evaluation index factors for site suitability: deformation, slope, aspect, surface roughness, and vegetation index. The spatial resolutions and the magnitudes of the values of each index factor may be completely different. Before performing overlay analysis, it is necessary to preprocess each index factor first. The preprocessing mainly includes two steps: unifying the spatial resolution and normalizing the values of the index factors.

[0101] S5. Evaluating the site suitability of GNSS monitoring points using the analytic hierarchy process. The analytic hierarchy process consists of three levels: the goal level, the criterion level, and the scheme level. In the evaluation of the site suitability of GNSS monitoring points, the scheme level is the places where GNSS monitoring points can be arranged, the criterion level is the 5 selected index factors, and the goal level is the site suitability of GNSS monitoring points within the landslide disaster area. First, the locations with a slope less than 30 degrees within the landslide disaster area are used as the positions where GNSS monitoring points can be arranged, that is, the scheme level; then, according to the importance of each index factor, they are sorted to construct a judgment matrix; finally, the constructed judgment matrix is subjected to a consistency test. If the consistency test is passed, the judgment matrix is considered reasonable, and the obtained weights of the index factors are used for weighted overlay of each index factor to obtain the site suitability map of GNSS monitoring points.

[0102] Specifically:

[0103] S2. Use D-InSAR to obtain the deformation map of the landslide disaster area. ① Download two scenes of Sentinel-1 ascending or descending orbit data covering the landslide disaster area, and obtain the single-look complex image SLC through preprocessing; ② Set the SAR data with an earlier acquisition time as the master image, and the other as the slave image. Then register and resample the slave image to the same radar image coordinate space as the master image. After registration and resampling, the one-to-one correspondence of homologous pixels is determined; ③ Multiply the complex data of the corresponding pixels of the master and slave images conjugately to obtain the initial interferogram. At this time, each pixel in the initial interferogram is still complex data, and the phase of this complex number is the interference phase. The range of the interference phase is [-π, π). At this time, the interference phase includes components such as the reference ellipsoid phase, terrain relief phase, surface deformation phase, atmospheric delay phase, and interference noise phase; ④ The reference ellipsoid phase is the phase component caused by the interaction between the spatial attitudes of the two SAR satellite imagings and the reference ellipsoid. This phase can be removed by constructing a model based on the satellite orbit parameters at the acquisition times of the master and slave images; ⑤ In order to obtain the surface deformation phase, it is necessary to subtract the terrain phase from the interferogram. The contribution of the terrain phase can be calculated through external DEM data, and then this contribution can be subtracted from the interference phase; ⑥ Since there are still atmospheric delay phase and noise phase in the interferogram, it is necessary to filter the differential interferogram to weaken its noise signal and improve the signal-to-noise ratio of the interference data. After phase filtering, the phase quality of the differential interferogram can be effectively improved, and the accuracy of phase unwrapping can be improved. However, there will still be significant phase noise in some severely decorrelated areas, and these decorrelated areas need to be masked; ⑦ The deformation phase recorded in the differential interferogram is wrapped in the interval [-π, π), and the true surface deformation information cannot be directly characterized. It is necessary to perform phase unwrapping processing to restore the phase integer number of each pixel; ⑧ The product directly obtained by SAR interferometry is in the range-Doppler coordinate frame. For the convenience of later processing, it is necessary to project and convert it to the commonly used geographic coordinate system.

[0104] S3. Calculation of slope, aspect, surface roughness, and vegetation index. Slope, aspect, and surface roughness are commonly used indicators in GIS terrain analysis and can be calculated from the DEM using common software (such as ArcGIS, SAGA, etc.). Among them, the calculation of slope and aspect generally uses the fitting surface method, and the fitting surface generally uses a quadratic surface, that is, a 3x3 window, as Figure 2 shown. The center of each window is an elevation point. The slope and aspect calculation formulas for the center point e are:

[0105]

[0106]

[0107] Wherein, a, b, c, d, f, and g are the elevation values of corresponding pixels, cellsize is the spatial resolution of the pixel, dx and dy are the change rates of the elevation value in the x and y directions, and tan -1 represents the arctangent function, and slope and aspect respectively represent the slope and aspect of the central pixel e.

[0108] The surface roughness reflects the elevation difference value between the central pixel and its surrounding pixels, which characterizes the unevenness degree of the central pixel. The calculation formula of the surface roughness is: Wherein, TRI represents the surface roughness of the central pixel e, and the rest of the symbols represent the elevation values of each pixel.

[0109] The vegetation index can reflect the surface vegetation coverage degree to a certain extent. The orthophoto image of the landslide disaster area contains the information of three visible light bands of red, green, and blue. By calculating the band information of each pixel, the vegetation index of the corresponding pixel can be obtained. When the vegetation index is high, it indicates that the vegetation content of the pixel is high or the possibility of being a plant is high. For simplicity, the present invention uses the vegetation index to characterize the surface vegetation coverage degree. The calculation formula of the vegetation index is:

[0110] VI = (2G - R - B) - (1.4R - G)

[0111]

[0112] Wherein, VI is the vegetation index value of the corresponding pixel, and r, g, and b are the values of the three bands of red, green, and blue of each pixel.

[0113] S4. Pretreatment of the evaluation index factors for site suitability. There are a total of five evaluation index factors for site suitability: deformation, slope, aspect, surface roughness, and vegetation index. Considering that the spatial resolutions of the various index factors may be different and the magnitudes of their values are different, pretreatment is performed on each index factor.

[0114] ① Pretreatment of the deformation index factor: For the deformation map obtained by the D-InSAR technology, its spatial resolution is generally large. For example, the spatial resolution of Sentinel-1 data is about 12 m in the case of multi-look 5:1, which is much larger than the occupied area of the GNSS monitoring point (1 m × 1 m or 2 m × 2 m). Therefore, it is necessary to resample the spatial resolution of the deformation map to the construction area of the GNSS monitoring point, and the resampling method uses the nearest neighbor interpolation method. Its schematic diagram is as Figure 3As shown. The main purpose of this interpolation is to unify the spatial resolution and ensure that the deformation information remains unchanged. After unifying the spatial resolution, the magnitudes of the deformation maps may vary in different landslides and have significant differences from the magnitudes of other index factors. Therefore, it is necessary to normalize the deformation values of the deformation maps to the range of [0, 1]. The closer the normalized value is to 1, the greater the relative deformation degree in the disaster area, and it is more necessary to conduct monitoring. Therefore, the normalized deformation value also reflects the influence degree of deformation on the suitability evaluation of GNSS monitoring point selection. The normalization formula is: In the formula, X norm represents the normalized value, X represents the value before normalization, abs represents the absolute value operation, X max , X min respectively represent the maximum and minimum values of the absolute values of each pixel before normalization.

[0115] ② Preprocessing of slope index factor. By calculating the DEM, the slope values of the landslide disaster area can be obtained. First, unify the spatial resolution of the DEM with the occupied area of the GNSS monitoring point. If the DEM resolution is higher than the occupied area, the average value resampling method is used for downsampling, otherwise the nearest neighbor interpolation method is used for upsampling. Considering the construction difficulty and cost issues during the construction of GNSS monitoring points, in the actual monitoring point construction work, generally, areas with a slope less than 30 degrees are selected for the construction of GNSS monitoring points. Therefore, the present invention considers that locations with a slope greater than 30 degrees are not suitable for the construction of GNSS monitoring points. Therefore, the second step in the preprocessing of the slope index factor is to mask the pixels with a slope greater than 30 degrees and not participate in the subsequent suitability evaluation of site selection; finally, normalize the masked slope index factor; when the slope is smaller, the construction difficulty and cost of the GNSS monitoring station are lower, that is, it is more suitable for the construction of GNSS monitoring points, and its value should also be closer to 1 after normalization. Therefore, the normalization formula for the slope index factor is:

[0116]

[0117] In the formula, Slope norm represents the normalized slope index factor, Slope represents the slope before normalization, Slope max , Slope min are the maximum (30°) and minimum (0°) values of the slope before normalization.

[0118] ③ Preprocessing of slope aspect index factor. The preprocessing of the slope aspect index factor also first unifies the spatial resolution to the construction area of the GNSS monitoring points for subsequent weighted superposition calculation. The value range of the slope aspect index factor is 0 - 360°. Its main impact on the GNSS monitoring points is the sunshine duration. Since the GNSS monitoring stations in the wild are mainly powered by solar energy, the longer the illumination duration, the more sufficient the solar power supply, that is, it is more suitable for the construction of GNSS monitoring points. Therefore, the normalization of the slope aspect index factor to [0, 1] is carried out through its slope aspect range. The slope aspect is divided into four categories: shady slope, semi-shady slope, semi-sunny slope, and sunny slope, and their corresponding normalized values are set to 0.2, 0.4, 0.6, and 0.8 respectively. The larger the normalized value, the more suitable the corresponding pixel position is for the construction of GNSS monitoring points. The corresponding relationship between the slope aspect and the normalized value is shown in Table 1.

[0119] Table 1 Corresponding relationship between slope aspect and normalized value

[0120] Slope aspect Slope aspect classification Normalized value 0～45° North-facing slope 0.2 45～135° Semi-north-facing slope 0.4 135～225° South-facing slope 0.8 225～315° Semi-south-facing slope 0.6 315～360° North-facing slope 0.2

[0121] ④ Preprocessing of surface roughness index factor. The preprocessing of the surface roughness index factor also first unifies the spatial resolution to the floor area of the GNSS monitoring points for subsequent weighted superposition calculation. The surface roughness reflects the unevenness between the corresponding pixel position and adjacent pixels. The larger the surface roughness, the less suitable it is for the construction of GNSS monitoring points. Therefore, its normalization formula is:

[0122]

[0123] In the formula, TRI norm represents the surface roughness index factor after normalization, TRI represents the surface roughness before normalization, TRI max , TRI min are the maximum and minimum values of the surface roughness before normalization.

[0124] ⑤ Preprocessing of vegetation index factor. The preprocessing of the vegetation index factor also first unifies the spatial resolution to the floor area of the GNSS monitoring points for subsequent weighted superposition calculation. Since the vegetation near the GNSS monitoring points will have a certain multipath effect on the GNSS monitoring stations, in actual work, the monitoring points are generally built at locations without vegetation or with less vegetation coverage. The present invention obtains the vegetation index through orthophoto image calculation. This index reflects the vegetation coverage degree of the surface to a certain extent, that is, the greater this value, the more serious the vegetation coverage. Therefore, the normalization formula of the vegetation index is:

[0125]

[0126] In the formula, VI normIndicates the normalized vegetation index indicator factor, VI represents the vegetation index before normalization, VI max , VI min are the maximum and minimum values of the vegetation index before normalization.

[0127] S5. Evaluate the suitability of GNSS monitoring point selection using the analytic hierarchy process. There are mainly 5 steps in using the analytic hierarchy process to evaluate the suitability of GNSS monitoring point selection:

[0128] ①Construct a judgment matrix. The key to the analytic hierarchy process is to construct a judgment matrix, the purpose of which is to determine the weights of each indicator factor. The Saaty scale is usually used to construct the judgment matrix, and the Saaty scale ranges from [1, 9]. The meanings of each scale are shown in Table 2.

[0129] Table 2 Saaty scale and its meanings

[0130] Scale Meaning 1 The two indicators are equally important 3 The previous indicator is slightly more important than the latter one 5 The previous indicator is significantly more important than the latter one 7 The previous indicator is extremely more important than the latter one 9 The previous indicator is strongly more important than the latter one 2,4,6,8 The intermediate value of the above judgment The reciprocals of 1 to 9 The importance of swapping the comparison order of the two indicators

[0131] Based on the 5 indicator factors used in the evaluation of the suitability of GNSS monitoring point selection: deformation, slope, aspect, surface roughness, and vegetation index, a 5x5 judgment matrix P can be constructed. The elements in this matrix should satisfy the formula: P i,j ·P j,i = 1, where i and j in the formula represent the row and column numbers respectively.

[0132] ②Normalize the judgment matrix. Normalizing the judgment matrix means making the sum of each column of the normalized matrix equal to 1. The specific calculation formula is:

[0133] ③Calculate the relative weights. For each indicator factor, the range of its weight value is [0, 1]. The higher the weight value, the greater its impact on the suitability of GNSS monitoring point selection. The calculation formula for the relative weight is: In the formula, w i represents the weights of each indicator factor, represents the value of the element in the l-th row and j-th column of the normalized judgment matrix.

[0134] ④Consistency test. The consistency test is determined by the size of the consistency ratio (CR) value. When CR ≤ 0.1, it is considered that the established analytic hierarchy model is reasonable; the calculation of the CR value is given by the ratio of the consistency index (CI) to the random consistency index (RI) . Among them, the value of RI is related to the number of indicator factors. The evaluation of the suitability of GNSS monitoring point selection involves 5 indicator factors. Therefore, the value of RI can be determined to be 1.12 through the random consistency index table; the calculation formula of CI is: In the formula, λmax is the largest eigenvalue of the judgment matrix.

[0135] ⑤ Index factor weighted superposition. After passing the consistency test, the GNSS monitoring point siting suitability map can be obtained by weighted superposition of each index factor. The weighted superposition formula is:

[0136] S i = w deformation ·X i + w slope ·Slope i + w aspect ·Aspect i + w TRI ·TRI i + w VI ·VI i

[0137] In the formula, S i represents the GNSS monitoring point suitability of the i-th pixel in the landslide disaster area. w deformation , w slope , w aspect , w TRI , w VI respectively represent the weights of the deformation index, slope index, aspect index, surface roughness index, and vegetation index determined by the analytic hierarchy process. X i , Slope i , Aspect i , TRI i , VI i respectively represent the values of the five index factors of deformation, slope, aspect, surface roughness, and vegetation index corresponding to the i-th pixel after preprocessing.

[0138] Example 2: To verify Example 1, the experimental area of the present invention is a landslide in the Jinsha River Basin. The experimental data used includes the digital elevation model (DEM) and orthoimage (DOM) of this area obtained by an unmanned aerial vehicle, and two scenes of Sentinel-1 SAR data with acquisition times close to that of the unmanned aerial vehicle images. The spatial resolution of the DEM and DOM is 0.15 meters. The schematic diagram of the DOM in the landslide area is as Figure 4 shown.

[0139] In the embodiments of the present invention, the location of the landslide disaster area has been determined. Therefore, it is also necessary to determine the floor area of the GNSS monitoring points. Since this area is near the Jinsha River Basin and the average altitude is 3000m, considering the difficulty of subsequent on-site GNSS monitoring point construction operations, it is planned to build simple GNSS monitoring points, so the construction area is set to 1m×1m. Then, it is necessary to obtain the prior deformation information of this area. First, it is necessary to download two scenes of Sentinel-1 SAR data that are relatively close to the time of UAV image acquisition and cover the landslide area, as well as external DEM and precise orbit data. Use D-InSAR technology to process the SAR data. After operations such as SAR data preprocessing, removal of topographic phase, interferogram filtering and masking, interferogram phase unwrapping, and projection conversion, the deformation values between the imaging dates of the two scenes of data can be obtained. The focus of the present invention is on the evaluation method for the suitability of GNSS monitoring point siting. Therefore, the specific process of D-InSAR will not be described in detail. The landslide area deformation map obtained by D-InSAR is as Figure 5 shown, and its spatial resolution is approximately 12m. From Figure 5 it can be seen that the deformation of this landslide is mainly concentrated in the lower part of the slope near the Jinsha River, and the line-of-sight deformation magnitude reaches the decimeter level. If this landslide becomes unstable, it will accumulate in the Jinsha River, posing a certain risk. Therefore, it is necessary to deploy GNSS monitoring equipment on this landslide body for real-time deformation monitoring.

[0140] Next, it is necessary to calculate the slope, aspect, surface roughness, and vegetation index indicator factors using the high-resolution DEM and DOM obtained by the UAV. Currently, commonly used GIS software can achieve the calculation of the above indicator factors. The software used in the embodiments of the present invention is SAGA 8.3. The calculation of slope, aspect, and surface roughness can be completed by inputting DEM data using the "Terrain Analysis" module. Among them, for slope and aspect, the "Slope, Aspect, Curvature" tool under the "Morphometry" module is used; for surface roughness, the "Terrain Ruggedness Index (TRI)" tool under the "Morphometry" module is used. For the calculation of the vegetation index, since the DOM is a collection of 3-band raster data, it can be obtained through simple raster operations. The tool used is "Grid->Calculus->Grid Calculator". After opening this tool and the calculation formula used for data calculation, the vegetation index result can be obtained.

[0141] After obtaining the five index factors of deformation, slope, aspect, surface roughness, and vegetation index, since their spatial resolutions and magnitudes are different, it is necessary to perform preprocessing before conducting the site suitability evaluation. The preprocessing mainly includes unifying the spatial resolution and normalizing the values. For unifying the spatial resolution, the "Grid->Tools->Resampling" in SAGA 8.3 software can be used. The average sampling method is used to reduce the spatial resolution, and the nearest neighbor interpolation method is used to increase the spatial resolution. The normalization operation is performed on each index factor, and its essence is still raster calculation, which can be completed using the "Grid->Calculus" toolbox in SAGA 8.3 software.

[0142] For the deformation index factor, its spatial resolution is about 12, while the occupied area of the GNSS monitoring points is set to 1m x 1m. Therefore, it is necessary to first increase the spatial resolution of the deformation index factor to 1m through the nearest neighbor interpolation method, and then normalize the deformation value to [0, 1]. The slope, aspect, surface roughness, and vegetation index are calculated from the high-resolution products provided by the UAV, with a spatial resolution of 0.15m. Therefore, it is necessary to first use the average sampling method to reduce the spatial resolution to 1m, and then perform the corresponding normalization operation. The preprocessing results of the five index factors of deformation, slope, aspect, surface roughness, and vegetation index in the embodiment are as Figure 6 , Figure 7 , Figure 8 , Figure 9 , Figure 10 shown. As can be seen from Figure 7 , there are many places with a slope greater than 30 degrees in the landslide disaster area. The areas with larger values of the preprocessed slope index factor are mainly in the upper and middle parts of the slope body, while the lower parts are smaller. As can be seen from Figure 8 , there are only two types of aspects in the landslide disaster area: shady slopes and semi-shady slopes, and most of them belong to shady slopes. As can be seen from Figure 9 , the smaller values of the surface roughness in the landslide area are mainly in the upper and middle parts of the slope body, that is, Figure 9 the areas with larger values of the surface roughness index factor in Figure 10 . As can be seen from

[0143] , after preprocessing each index factor, it is necessary to determine the weights of each index factor. First, use the Saaty scale to construct the judgment matrix P:

[0144]

[0145] The first column (row) in the matrix corresponds to the deformation index factor, the second column (row) corresponds to the slope index factor, the third column (row) corresponds to the surface roughness index factor, the fourth column (row) corresponds to the aspect index factor, and the fifth column (row) corresponds to the index exponent factor. It can be seen from matrix P that the deformation index factor is the most important because deformation is the most important purpose of monitoring. If the monitoring points are built at locations without deformation, the significance of the deformation monitoring work is limited. The next in importance are the slope, surface roughness, vegetation index, and aspect. This is because the slope and surface roughness will affect the construction of the monitoring points. In landslide monitoring, it is first necessary to deploy the monitoring equipment within the deformation area, so its importance is second only to the deformation index factor. Finally, the vegetation cover will affect the data, resulting in poor subsequent deformation monitoring results. Therefore, it is considered that the vegetation index factor is more important than the aspect factor.

[0146] Next, a consistency test (CR ≤ 0.1) should be carried out to ensure the rationality of the judgment matrix. After calculation, the consistency index CR of this matrix is 0.017, and it passes the consistency test, indicating that the site suitability evaluation model is reasonable. After passing the consistency test, the weights of each index can be calculated. After calculation, the weights of the five index factors of deformation, slope, aspect, surface roughness, and vegetation index are: 0.4408, 0.2596, 0.0492, 0.1516, 0.0988 respectively.

[0147] Finally, using the weights obtained by the analytic hierarchy process method, weighted superposition is performed on the five index factors, and the site suitability map of the GNSS monitoring points in the landslide area can be obtained. The suitability map is classified into four grades: "unsuitable", "barely suitable", "relatively suitable", and "very suitable". The corresponding suitability value ranges for each grade are: [0, 0.25), [0.25, 0.5), [0.5, 0.75), [0.75, 1). The site suitability map of the GNSS monitoring points for this landslide is as Figure 11 shown. It can be seen from Figure 11 that most of the locations in the landslide area are barely suitable for the construction of GNSS monitoring points. There are also some areas in the bottom and upper-middle parts of the slope body that belong to the relatively suitable grade. There is no location in the whole area that is very suitable for deploying GNSS monitoring points. This also shows that the GNSS monitoring in this landslide area is difficult, and the "relatively suitable" area should be mainly considered in the subsequent construction site selection of GNSS monitoring points.

[0148] Generally speaking, the present invention provides a method for evaluating the suitability of GNSS monitoring points for landslide disasters. Compared with the prior art, its beneficial effects are as follows:

[0149] 1. Five evaluation index factors, namely deformation, slope, aspect, surface roughness and vegetation index, are selected in the present invention. At the same time, considering the deformation monitoring requirements, construction costs of GNSS monitoring points and influences such as GNSS multipath during the site selection process of GNSS monitoring points, the analytic hierarchy process can be used to assign weights to multiple indicators, and the suitability of GNSS monitoring point site selection in landslide disaster areas can be objectively, reasonably and quantitatively evaluated.

[0150] 2. The present invention obtains prior deformation information of the disaster area, as well as the digital elevation model and orthophoto image of the disaster area through D-InSAR. Before professionals arrive at the disaster site, the overall suitability of GNSS monitoring point site selection in the landslide disaster area can be mastered. The evaluation results can clearly show the positions suitable for arranging GNSS monitoring points within the entire landslide disaster area, and can provide very important reference data for the GNSS monitoring point site selection work.

[0151] The above-described embodiments are only preferred specific embodiments of the present invention, and the protection scope of the present invention is not limited thereto. Any simple changes or equivalent replacements of technical solutions that can be obviously obtained by those skilled in the art within the technical scope disclosed by the present invention all belong to the protection scope of the present invention.

Claims

1. A method for evaluating the suitability of GNSS monitoring point selection for landslide disasters, characterized in that, It includes the following steps: Determine the location of the landslide disaster area and the floor area of the GNSS monitoring points; Use the differential interferometric synthetic aperture radar (D-InSAR) technology to process the synthetic aperture radar (SAR) data of the landslide disaster area and obtain the deformation map of the landslide disaster area; Use the digital elevation model (DEM) to calculate the slope, aspect, and surface roughness of the landslide disaster area; Use the digital orthophoto map (DOM) to calculate the vegetation index of the landslide disaster area; Preprocess the five index factors of deformation, slope, aspect, surface roughness, and vegetation index in the landslide disaster area. The preprocessing includes unifying the spatial resolution and normalizing the numerical values of the index factors; Use the analytic hierarchy process to determine the weights of the five indexes of deformation, slope, aspect, surface roughness, and vegetation index in the landslide disaster area; Perform weighted superposition on the five indexes of deformation, slope, aspect, surface roughness, and vegetation index in the landslide disaster area to obtain the suitability evaluation map for the location selection of the GNSS monitoring station in the landslide disaster area.

2. The suitability evaluation method for the location selection of GNSS monitoring points for landslide disasters according to claim 1, wherein, The specific steps for obtaining the deformation map of the landslide disaster area include: The method of using D-InSAR for surface deformation measurement is the two-track plus external DEM method. The processing process includes: Download the ascending or descending track data of two scenes of Sentinel-1 covering the landslide disaster area, and obtain the single-look complex image (SLC) through preprocessing; Set the SAR data with an earlier acquisition time as the master image, and the other image as the slave image. Then register and resample the slave image to the same radar image coordinate space as the master image. After registration and resampling, determine the one-to-one correspondence of the homologous pixels; Multiply the complex data of the corresponding pixels of the master and slave images conjugately to obtain the initial interferogram. At this time, each pixel in the initial interferogram is still complex data, and the phase of this complex number is the interference phase. The range of the interference phase is [-π, π). At this time, the interference phase includes the reference ellipsoid phase, terrain relief phase, surface deformation phase, atmospheric delay phase, and interference noise phase components; The reference ellipsoid phase is the phase component caused by the interaction between the spatial attitudes of the two SAR satellite imaging and the reference ellipsoid. This phase can be removed by constructing a model based on the satellite orbit parameters at the acquisition times of the master and slave images; In order to obtain the surface deformation phase, it is necessary to subtract the terrain phase from the interferogram. Calculate the terrain phase contribution through external DEM data, and then subtract this contribution from the interference phase; Since there are still atmospheric delay phase and noise phase in the interferogram, it is necessary to filter the differential interferogram to weaken its noise signal and improve the signal-to-noise ratio of the interference data. After phase filtering, the phase quality of the differential interferogram can be effectively improved, and the accuracy of phase unwrapping can be improved. However, there will still be significant phase noise in some severely decoherent areas, and these decoherent areas need to be masked; The deformation phase recorded in the differential interferogram is wrapped in the interval [-π, π) and cannot directly represent the real surface deformation information. It is necessary to perform phase unwrapping processing to restore the phase integer number of each pixel. The product directly obtained by SAR interferometric processing is in the range-Doppler coordinate framework. For the convenience of subsequent processing, it needs to be projected and converted to the commonly used geographic coordinate system.

3. The suitability evaluation method for the location selection of GNSS monitoring points for landslide disasters according to claim 1, wherein, Calculating the slope, aspect, and surface roughness of the landslide disaster area using the digital elevation model DEM specifically includes: Slope, aspect, and surface roughness are commonly used indicators in GIS terrain analysis and can be calculated from the DEM with the help of software such as ArcGIS or SAGA; The fitting surface method can be used for slope and aspect calculations. The fitting surface can be a quadratic surface, that is, a 3x3 window. The center of each window is an elevation point. The calculation formulas for the slope and aspect of the center point e are: where a, b, c, d, f, g are the elevation values of the corresponding pixels, cellsize is the spatial resolution of the pixels, dx and dy are the change rates of the elevation values in the x and y directions, and tan -1 represents the arctangent function, and slope and aspect respectively represent the slope and aspect of the central pixel e; The surface roughness reflects the elevation difference between the central pixel and its surrounding pixels, which characterizes the degree of unevenness of the central pixel; the calculation formula for surface roughness is: In the formula, TRI represents the surface roughness of the central pixel e, and the other symbols represent the elevation values of each pixel.

4. The suitability evaluation method for the location selection of GNSS monitoring points for landslide disasters according to claim 1, characterized in that, Calculating the vegetation index of the landslide disaster area using the orthoimage DOM specifically includes: The vegetation index can reflect the degree of surface vegetation cover to a certain extent; The orthoimage of the landslide disaster area contains information on the three visible light bands of red, green, and blue. By calculating the band information of each pixel, the corresponding vegetation index of each pixel can be obtained. When the vegetation index is high, it indicates that the vegetation content of the pixel is high or the possibility of being a plant is high; The calculation formula for the vegetation index is: VI = (2G - R - B) - (1.4R - G); In the formula, VI is the vegetation index value of the corresponding pixel, and r, g, and b are the values of the three bands of red, green, and blue of each pixel.

5. The suitability evaluation method for the location selection of GNSS monitoring points for landslide disasters according to claim 1, characterized in that Preprocessing the deformation index factor of the landslide disaster area specifically includes: For the deformation map obtained by D-InSAR technology, its spatial resolution is generally large. Therefore, it is necessary to resample the spatial resolution of the deformation map to the construction area of the GNSS monitoring points. The resampling method uses the nearest neighbor interpolation method. The purpose of interpolation is mainly to unify the spatial resolution, and this interpolation method can ensure that the deformation information does not change; After unifying the spatial resolution, the magnitude of the deformation map may vary for different landslides and has a large difference from the magnitude of other index factors. Therefore, it is necessary to normalize the deformation value of the deformation map to the interval [0,1]. The closer the normalized value is to 1, the greater the relative deformation degree in the disaster area, and it is more necessary to monitor. Therefore, the normalized deformation value also reflects the influence degree of deformation on the siting suitability evaluation of GNSS monitoring points; it should be noted that the positive and negative values of each pixel value in the deformation map only represent different deformation directions, so it is necessary to take the absolute value of each pixel value in the deformation map before normalization; The normalization formula for the deformation index factor is: Wherein, X norm represents the value after normalization, X represents the value before normalization, abs represents the absolute value operation, X max , X min respectively represent the maximum and minimum values of the absolute values of each pixel before normalization.

6. The suitability evaluation method for the location selection of GNSS monitoring points for landslide disasters according to claim 1, characterized in that, Preprocessing the slope index factor of the landslide disaster area specifically includes: By calculating the DEM, the slope value of the landslide disaster area can be obtained; Unify the spatial resolution of the DEM with the floor area of the GNSS monitoring points. If the DEM resolution is higher than the floor area, the average value resampling method is used for downsampling, otherwise the nearest neighbor interpolation method is used for upsampling; Considering the construction difficulty and cost of GNSS monitoring points during the construction process, in the actual construction work of monitoring points, areas with a slope less than 30 degrees are generally selected for the construction of GNSS monitoring points; therefore, locations with a slope greater than 30 degrees are considered unsuitable for the construction of GNSS monitoring points, and pixels with a slope greater than 30 degrees are masked during the preprocessing of the slope index factor and do not participate in the subsequent siting suitability evaluation; Normalize the masked slope index factor; when the slope is smaller, the construction difficulty and cost of the GNSS monitoring station are lower, that is, it is more suitable for the construction of GNSS monitoring points, and its value should also be closer to 1 after normalization; The normalization formula of the slope index factor is: In the formula, Slope norm represents the normalized slope index factor, Slope represents the slope before normalization, Slope max is the maximum value of the slope before normalization, which is 30°, and Slope min is the minimum value of the slope before normalization, which is 0°.

7. The suitability evaluation method for the location selection of GNSS monitoring points for landslide disasters according to claim 1, characterized in that, The preprocessing of the slope aspect index factor for the landslide disaster area specifically includes: Unify the spatial resolution to the construction area of the GNSS monitoring point for subsequent weighted overlay calculation; The value range of the slope aspect index factor is 0-360°, and its main impact on the GNSS monitoring point is the sunshine duration. Since the GNSS monitoring station in the wild is mainly powered by solar energy, the longer the illumination duration, the more sufficient its solar power supply, that is, it is more suitable for the construction of GNSS monitoring points. Therefore, the slope aspect index factor is normalized to [0,1] through its slope aspect range; The slope aspect is divided into four categories: shady slope, semi-shady slope, semi-sunny slope, and sunny slope, and their corresponding normalized values are set to 0.2, 0.4, 0.6, and 0.8 respectively. The larger the normalized value, the more suitable the corresponding pixel position is for the construction of GNSS monitoring points.

8. A method for evaluating the suitability of GNSS monitoring point location for landslide disasters according to claim 1, characterized in that, The preprocessing of the surface roughness index factor for the landslide disaster area specifically includes: Unify the spatial resolution to the occupied area of the GNSS monitoring point for subsequent weighted overlay calculation; The surface roughness reflects the unevenness between the corresponding pixel position and adjacent pixels. The larger the surface roughness, the less suitable it is for the construction of GNSS monitoring points; The normalization formula of the surface roughness is: where, TRI norm represents the normalized surface roughness index factor, TRI represents the surface roughness before normalization, TRI max , TRI min are the maximum and minimum values of the surface roughness before normalization.

9. The suitability evaluation method for the GNSS monitoring point location of landslide disasters according to claim 1, characterized in that The preprocessing of the vegetation index factor for the landslide disaster area specifically includes: Unify the spatial resolution to the occupied area of the GNSS monitoring point for subsequent weighted overlay calculation; Because the vegetation near the GNSS monitoring point will have a certain multipath effect on the GNSS monitoring station, in actual work, the monitoring point is generally built at a location without vegetation or with less vegetation coverage; The vegetation index is obtained through orthophoto image calculation. This index reflects the vegetation coverage of the surface to a certain extent, that is, the greater the value, the more serious the vegetation coverage; The normalization formula of the vegetation index is: Where, VI norm represents the normalized vegetation index index factor, VI represents the vegetation index before normalization, VI max , VI min are the maximum and minimum values of the vegetation index before normalization.

10. A method for evaluating the suitability of a GNSS monitoring point location for landslide disasters according to claim 1, characterized in that, The obtaining of the siting suitability evaluation map of the GNSS monitoring station in the landslide disaster area specifically includes: The key of the analytic hierarchy process is to construct a judgment matrix, the purpose is to determine the weights of each index factor, and the Saaty scale is usually used to construct the judgment matrix. The Saaty scale range is [1,9]; Based on the five index factors used in the GNSS monitoring point site selection suitability evaluation: deformation, slope, aspect, surface roughness, and vegetation index, a 5x5 judgment matrix P is constructed, and the elements in this matrix should satisfy the formula: P i,j ·P j,i = 1; where i and j represent the row and column numbers respectively; Normalize the judgment matrix, that is, make the sum of each column of the normalized matrix equal to 1. The specific calculation formula is as follows: For each index factor, the range of its weight value is [0, 1]. The higher the weight value, the greater its impact on the suitability of GNSS monitoring point location. The calculation formula for the relative weight is: where w i represents the weight of each index factor, represents the value of the element in the l-th row and j-th column of the normalized judgment matrix; The consistency test is determined by the size of the consistency ratio CR value. When CR ≤ 0.1, the established analytic hierarchy process model is considered reasonable; the calculation of the CR value is given by the ratio of the consistency index CI to the random consistency index RI. Among them, the value of RI is related to the number of index factors. The suitability evaluation of GNSS monitoring point location involves 5 index factors. Therefore, the value of RI can be determined as 1.12 through the random consistency index table; the calculation formula of CI is: In the formula, λ max is the largest eigenvalue of the judgment matrix; After passing the consistency test, weighted overlay of each index factor can obtain the siting suitability map of the GNSS monitoring point. The weighted overlay formula is: S i = w deformation · X i + w slope · Slope i + w aspect ·Aspect i +w TRI ·TRI i +w VI ·VI i ; In the formula, S i represents the suitability of the GNSS monitoring point of the i-th pixel in the landslide disaster area, w deformation , w slope , w aspect , w TRI , w VI respectively represent the weights of the deformation index, slope index, aspect index, surface roughness index, and vegetation index determined by the analytic hierarchy process. X i , Slope i , Aspect i , TRI i , VI i respectively represent the values of the five index factors of deformation, slope, aspect, surface roughness, and vegetation index corresponding to the i-th pixel after preprocessing.

Images