A method for estimating river flow by using altimetry satellites to deduce the cross-sectional area of a river

Abstract

The present invention provides a method for estimating river flow rate by applying an altimetry satellite to deduce the cross-sectional area of a river channel, including extracting and preprocessing the river width from optical images, extracting the river width from preprocessed radar images; extracting the water level from preprocessed altimetry satellite data; inversely deriving the river channel topography from elevation boundaries; inversely deriving the river channel topography from remote sensing images; constructing the river channel topography; and inversely deriving the river flow rate. The method for estimating river flow rate by applying an altimetry satellite to deduce the cross-sectional area of a river channel proposed by the present invention extracts the river width and water level from multi-source remote sensing satellite data to inversely derive the river channel topography, and establishes an empirical formula of cross-section - flow rate to achieve river flow rate prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of remote sensing inversion of river flow, and in particular to a method for estimating river flow by applying an altimetry satellite to deduce the cross-sectional area of a river channel. Background Art

[0002] Existing studies have predicted river flow by extracting river width and water level from multi-source remote sensing data and establishing empirical formulas of river width-flow and water level-flow.

[0003] Currently, the inversion of river flow is only carried out through single factors such as river width and water level, with a small applicable range and being greatly affected by river morphology. To improve the prediction accuracy of river flow, some studies have considered multiple factors such as river width and water level for river flow inversion. Compared with single-factor flow inversion, the accuracy and applicability of multi-factor flow inversion have been improved.

[0004] These studies use empirical formulas established by river width and water level to invert and predict flow, which requires a large amount of measured flow data for support. When the terrain is complex or the river bed changes, the error of the inverted flow is relatively large.

[0005] Chinese Patent Application No. CN114037905A discloses a method and system for remote sensing inversion of water surface salinity in estuaries based on space-ground collaboration. On the one hand, the method uses relatively accurate in-situ measured data in the estuary area, constructs a model with the measured data of synchronous in-situ measurement stations as output data, and improves the inversion accuracy of the remote sensing inversion model. On the other hand, the method improves the traditional remote sensing inversion model that only inverses water surface salinity through spectral information or indirect water quality parameters, adds geographical grid information as an input parameter, and fully utilizes the longitude and latitude information of in-situ measurement stations and remote sensing images as well as the spatially continuous data characteristics of remote sensing images, especially suitable for achieving high-precision inversion of water surface salinity in complex estuary areas. This method depends on the measured data of in-situ measurement stations. If the data of ground stations are inaccurate or incomplete, it will directly affect the accuracy of the inversion results.

[0006] Chinese Patent Application No. CN110133655A discloses a method for monitoring and inverting river runoff based on multi-source radar remote sensing technology, including the following steps: Step 1, preprocessing of multi-source radar remote sensing data: preprocessing of SAR images and preprocessing of radar altimeter data; Step 2, extracting river width based on SAR images; Step 3, extracting relative water depth of the river based on the radar altimeter; Step 4, constructing a runoff calculation model based on the Manning formula; Step 5, monitoring the change of river runoff in time series. This method depends on the measured lowest water level data. If the historical lowest water level data is missing, the water level data obtained only by the radar altimeter will directly affect the accuracy of flow inversion. Summary of the Invention

[0007] To solve the above technical problems, the present invention proposes a method for estimating river flow by applying an altimetry satellite to deduce the cross-sectional area of a river channel, which extracts the river width and inverts the river channel topography based on multi-source remote sensing satellite data, and establishes an empirical formula of cross-section - flow to achieve river flow prediction.

[0008] The object of the present invention is to provide a method for estimating river flow by applying an altimetry satellite to deduce the cross-sectional area of a river channel, including preprocessing and extraction of the river width from optical images, and further comprising the following steps:

[0009] Step 1: Preprocessing of radar images and extraction of river width;

[0010] Step 2: Preprocessing of altimetry satellite images and extraction of water level;

[0011] Step 3: Inverting the river channel topography from elevation boundaries;

[0012] Step 4: Inverting the river channel topography from remote sensing images, where the remote sensing images include the radar images, the optical images, and the altimetry satellite images;

[0013] Step 5: Constructing the river channel topography;

[0014] Step 6: Inverting the river flow.

[0015] Preferably, the extraction and preprocessing of the river width from the optical images include the following sub-steps:

[0016] Step 01: Screening and preprocessing the surface reflectance product data to obtain a river optical image dataset;

[0017] Step 02: Calculating the Normalized Difference Water Index (NDWI) for the dataset, extracting the water body threshold through the Otsu method, and masking to extract the river water body range. The formula is

[0018] NDWI = (Green - NIR) / (Green + NIR);

[0019] where Green is the reflectance of the green band and NIR is the reflectance of the near-infrared band;

[0020] Step 03: Removing small holes in the water body and smoothing the water body contour through morphology, and deleting areas with large slopes in the water body range;

[0021] Step 04: Applying a skeletonization algorithm to extract the river centerline, calculating the cross-sectional angle of the river to obtain the median line orthogonal angle, and extracting the river width.

[0022] In any of the above solutions, preferably, the method for removing small holes in the water body and smoothing the water body contour through morphology includes: performing image opening operation and image closing operation on the image,

[0023] The morphological opening operation on the image smooths the contour of the object, disconnects narrow necks, and eliminates thin protrusions;

[0024] The morphological closing operation on the image bridges narrow discontinuities and slender gullies, eliminates small holes, and fills in breaks in the contour line;

[0025] Among them, the morphological opening operation on the image first erodes and then dilates; the morphological closing operation on the image first dilates and then erodes.

[0026] Preferably, in any of the above solutions, the method for deleting the area with a large slope in the water body range includes: calculating the slope of the DEM obtained on the GEE platform, and performing masking on the water body to extract the area with a small slope.

[0027] Preferably, in any of the above solutions, step 1 includes screening Sentinel-1 GRD image data with the interferometric wide swath mode of IW, and performing Refined Lee filtering for denoising, geocorrection, and radiometric normalization data processing to obtain a river radar image dataset.

[0028] Preferably, in any of the above solutions, step 2 includes downloading ICEsat-2 ATL03 data, denoising the altimetry satellite image using the adaptive DBSCAN algorithm, performing Gaussian fitting on the water level to obtain the expectation, calculating the elevation conversion coefficient for elevation conversion, and obtaining the river water level in a unified coordinate system.

[0029] Preferably, in any of the above solutions, the method for calculating the elevation conversion coefficient for elevation conversion includes converting the coordinates between different geodetic coordinate systems into corresponding coordinates through seven parameters, and the formula is

[0030] X' = X + dx - Y * rz + Z * ry + S

[0031] Y' = Y + X * rZ + Z * rx + S

[0032] Z' = Z - X * ry + Y * rx + S

[0033] Among them, (X, Y, Z) are the coordinates of the original coordinate system, (X', Y', Z') are the coordinates of the target coordinate system, dx, dy, and dz are translation parameters, rx, ry, and rz are rotation parameters, and S is the scale parameter.

[0034] Preferably, in any of the above solutions, step 3 includes obtaining multi-period river channel boundary elevations by combining water levels extracted from multi-period altimetry satellite data with river boundary information, interpolating the multi-period boundary elevations using the Kriging interpolation method to obtain river channel boundary elevation points above the lowest water level, and then applying the Kriging interpolation method to extrapolate and invert the underwater elevation of the entire river elevation boundary, and combining the inverted underwater terrain to obtain the complete river channel terrain.

[0035] Preferably, in any of the above solutions, step 4 includes calculating the water depth values at corresponding points by interpolating the river channel terrain from the elevation boundary and the multi-period water level data obtained from altimetry satellites, using the radiance of the blue and green bands of the optical remote sensing image Sentinel-2 and the water depth as prior data for the logarithmic ratio model, calculating relevant parameters, establishing a logarithmic ratio water depth inversion model, and inverting the multi-period underwater terrain.

[0036] Preferably, in any of the above solutions, step 5 includes retaining the overlapping area of the elevation boundary and the river channel terrain inverted from the remote sensing image, using the river channel terrain result inverted from the elevation boundary as the main result for the non-overlapping area, and using the multi-period river channel terrain inverted from the remote sensing image as a reference to fuse the river channel terrain to obtain the river channel terrain.

[0037] Preferably, in any of the above solutions, step 5 further includes using the radiance of the blue and green bands of the optical remote sensing image Sentinel-2 and the water depth as prior data for the logarithmic ratio model, calculating relevant parameters, and establishing the construction of a logarithmic ratio water depth inversion model:

[0038]

[0039] In the formula, n 、 m 1 、 m 0 are empirical parameters, are the radiance values of the two bands, and is the water depth value.

[0040] Preferably, in any of the above solutions, step 6 includes calculating the multi-period river channel cross-sectional areas according to the inverted river channel terrain and the multi-period water levels obtained from altimetry satellites, establishing an empirical formula of cross-section - flow rate by relating the cross-sectional area to the measured flow rate, and performing accuracy evaluation. The empirical formula of cross-section - flow rate is

[0041] Q ≈ k A+ b

[0042] Where Q is the river channel flow rate, A is the cross-sectional area of the water passage, k and b are empirical relationship coefficients.

[0043] Preferably, in any of the above solutions, the method for accuracy evaluation includes using the root mean square error (RMSE) and the correlation coefficient R to evaluate the accuracy. The calculation formula for the root mean square error (RMSE) is:

[0044]

[0045] The calculation formula for the correlation coefficient R is:

[0046]

[0047] where Q p is the measured river flow rate, q i is the predicted river flow rate, N is the total number of samples.

[0048] The present invention proposes a method for estimating river flow rate by using an altimetry satellite to deduce the cross-sectional area of a river channel, which realizes the accurate inversion of the river channel topography. The accuracy of river flow rate inversion based on multi-source remote sensing satellite data is high, and accurate prediction of river flow rate is achieved.

[0049] GEE platform: Google Earth Engine, which is a cloud platform provided by Google for online visualization calculation and analysis processing of a large amount of global-scale earth science data (especially satellite data).

[0050] DEM: Digital Elevation Model, which realizes the digital simulation of the ground topography through limited terrain elevation data.

[0051] DBSCAN: A density-based spatial clustering algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 FIG. is a flowchart of a preferred embodiment of the method for estimating river flow rate by using an altimetry satellite to deduce the cross-sectional area of a river channel according to the present invention.

[0053] Figure 2 FIG. is a flowchart of another preferred embodiment of the method for estimating river flow rate by using an altimetry satellite to deduce the cross-sectional area of a river channel according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] The present invention will be further described below with reference to the drawings and specific embodiments. Embodiment 1

[0055] As Figure 1 shown, step 100 is executed, and the optical image preprocessing and river width extraction include the following sub-steps:

[0056] Execute step 101 to screen the surface reflectance product data and perform preprocessing to obtain a river optical image dataset.

[0057] Execute step 102 to calculate the Normalized Difference Water Index (NDWI) for the dataset, extract the water body threshold through the Otsu method, and mask and extract the river water body range. The formula is

[0058] NDWI = (Green - NIR) / (Green + NIR);

[0059] where Green is the reflectance of the green band and NIR is the reflectance of the near-infrared band.

[0060] Execute step 103 to remove small holes in the water body and smooth the water body contour through morphology, and delete the areas with large slopes in the water body range; the method of removing small holes in the water body and smoothing the water body contour through morphology includes: performing image opening operation and image closing operation on the image,

[0061] The image opening operation smooths the contour of the object, disconnects narrow necks, and eliminates thin protrusions;

[0062] The image closing operation closes narrow discontinuities and slender gullies, eliminates small holes, and fills in breaks in the contour line;

[0063] Among them, the image opening operation first erodes and then dilates; the image closing operation first dilates and then erodes.

[0064] The method of deleting the areas with large slopes in the water body range includes: calculating the slope of the DEM obtained on the GEE platform and masking and extracting the areas with small slopes of the water body.

[0065] Execute step 104 to apply the skeletonization algorithm to extract the river centerline, calculate the cross-sectional angle of the river to obtain the median line orthogonal angle, and extract the river width.

[0066] Execute step 110 for preprocessing the river width extraction from radar images, including screening Sentinel-1 GRD image data with the interferometric wide swath mode of IW, performing Refined Lee filtering for denoising, geocorrection, and radiometric normalization data processing to obtain a river radar image dataset.

[0067] Execute step 120 for preprocessing the water level extraction from altimetry satellite images, including downloading ICEsat-2 ATL03 data, denoising the altimetry satellite images using the adaptive DBSCAN algorithm, performing Gaussian fitting on the water level to obtain the expectation, calculating the elevation conversion coefficient for elevation conversion, and obtaining the river water level under a unified coordinate system.

[0068] The method for performing elevation conversion by calculating the elevation conversion coefficient includes converting the coordinates between different geodetic coordinate systems into corresponding coordinates through seven parameters. The formula is

[0069] X' = X+ dx-Y*rz + Z*ry + S

[0070] Y'= Y + X*rZ +Z*rx + S

[0071] Z'=Z-X*ry + Y*rx + S

[0072] where (X,Y,Z) are the coordinates of the original coordinate system, (X',Y',Z') are the coordinates of the target coordinate system, dx, dy, and dz are translation parameters, rx, ry, and rz are rotation parameters, and S is the scale parameter.

[0073] Execute step 130 to invert the river channel topography from the elevation boundary, including obtaining the elevation of the river channel boundary for multiple periods by combining the water levels extracted from multi-period altimetry satellite data with the river boundary information, interpolating the elevation points of the river channel boundary above the lowest water level using the Kriging interpolation method for the multiple-period boundary elevations, and then extrapolating and inverting the underwater elevation for the entire river elevation boundary using the Kriging interpolation method, and combining the inverted underwater topography to obtain the complete river channel topography.

[0074] Execute step 140 to invert the river channel topography from remote sensing images, including calculating the water depth values at corresponding points by interpolating the river channel topography from the elevation boundary and using the multi-period water level data obtained from altimetry satellites, using the radiance of the blue and green bands of the optical remote sensing image Sentinel-2 and the water depth as the prior data for the logarithmic ratio model, calculating the relevant parameters, and establishing a logarithmic ratio water depth inversion model to invert the multi-period underwater topography.

[0075] Execute step 150 to construct the river channel topography, including retaining the overlapping areas of the river channel topography inverted from the elevation boundary and the remote sensing images, using the river channel topography result inverted from the elevation boundary as the main for the non-overlapping areas, and using the multi-period river channel topography inverted from the remote sensing images as a reference to perform river channel topography fusion to obtain the river channel topography. The radiance of the blue and green bands of the optical remote sensing image Sentinel-2 and the water depth are used as the prior data for the logarithmic ratio model, and the relevant parameters are calculated to establish the construction of the logarithmic ratio water depth inversion model:

[0076]

[0077] In the formula, n 、 m 1 、 m 0 are empirical parameters, are the radiance values of the two bands, and is the water depth value.

[0078] Execute step 160, river flow inversion, including calculating the cross-sectional areas of multiple river channels based on the inverted river channel topography and multiple water levels obtained by altimetry satellites, establishing an empirical cross-section-flow formula between the cross-sectional area and the measured flow, and performing accuracy evaluation. The empirical cross-section-flow formula is

[0079] Q ≈ k A + b

[0080] where Q is the river channel flow, A is the cross-sectional area of the water passage, k and b are empirical relationship coefficients.

[0081] The method for accuracy evaluation includes performing accuracy assessment through the root mean square error RMSE and the correlation coefficient R . The calculation formula for the root mean square error RMSE is:[[]]

[0082]

[0083] The calculation formula for the correlation coefficient R is

[0084]

[0085] where Q p is the measured river flow,[[]] q i is the predicted river flow,[[]] N is the total number of samples. Embodiment 2

[0086] As Figure 2 shown, a method for river flow inversion based on the collaborative application of multi-mode remote sensing satellites according to the present invention includes:[[]]

[0087] Step S1. Optical image river width extraction and preprocessing: On the Google Earth Engine platform, Sentinel-2 Level-2A surface reflectance products with less than 10% are screened, and the data is preprocessed to obtain a river optical image dataset. The NDWI (Normalized Difference Water Index) is calculated for the dataset, and the water body threshold is extracted through the OTSU threshold method (maximum inter-class variance method), and the river water body range is extracted by masking. Data processing such as removing small holes in the water body and smoothing the water body contour through morphology, and deleting data in areas with large slopes in the water body range is performed to improve the accuracy of water body extraction. The skeleton algorithm is applied to extract the river centerline, the median line orthogonal angle is obtained by calculating the river cross-section angle, and the river width is extracted. Among them, NDWI (Normalized Difference Water Index): NDWI = (Green - NIR) / (Green + NIR);

[0088] Step S2. Radar image preprocessing and river width extraction: On the Google Earth Engine platform, Sentinel-1 GRD images with the interferometric wide swath mode of IW are screened, and the data is processed by Refined Lee filtering for denoising, geocorrection, and radiometric normalization to obtain a river radar image dataset. The SDWI (Dual-Polarization Water Index) is calculated for the dataset, and the water body threshold is extracted through the OTSU threshold method (maximum inter-class variance method), and the river water body range is extracted by masking. Data processing such as removing small holes in the water body and smoothing the water body contour through morphology, and deleting data in areas with large slopes in the water body range is performed to improve the accuracy of water body extraction; the skeleton algorithm is applied to extract the river centerline, the median line orthogonal angle is obtained by calculating the river cross-section angle, and the river width is extracted. Among them, SDWI (Dual-Polarization Water Index): SDWI = EXP(VV * VH / 1000)

[0089] Step S3. Altimetry satellite preprocessing and water level extraction: Download ICEsat-2 ATL03 data from the National Aeronautics and Space Administration (NASA), denoise the image using the adaptive DBSCAN algorithm, perform Gaussian fitting on the water level to obtain the expectation, calculate the elevation conversion coefficient for elevation conversion, and obtain the river water level under a unified coordinate system.

[0090] Step S4. Inverting the river channel topography from elevation boundaries: The water levels extracted from multi-period altimetry satellite data are combined with river boundary information to obtain the elevation of the river boundary for multiple periods. The Kriging interpolation method is used to interpolate the elevation points of the river boundary above the lowest water level from the elevation of the boundary for multiple periods. Then, the Kriging interpolation method is applied to extrapolate and invert the underwater elevation outside the elevation boundary of the entire river, and the complete river channel topography is obtained by combining the inverted underwater topography.

[0091] Step S5: Inverting the river channel topography from remote sensing images: The water depth values at corresponding points are calculated by interpolating the river channel topography from the elevation boundary and the multi-period water level data obtained by altimetry satellites. The radiance of the blue and green bands of the optical remote sensing image Sentinel-2 and the water depth are used as prior data for the logarithmic ratio model, and the relevant parameters are calculated to establish a logarithmic ratio water depth inversion model to invert the multi-period underwater topography.

[0092] Step S6. Constructing the river channel topography: Retain the overlapping area of the river channel topography inverted from the elevation boundary and the remote sensing images. For the non-overlapping area, the result of the river channel topography inverted from the elevation boundary is the main one, and the multi-period river channel topography inverted from the remote sensing images is used as a reference to fuse the river channel topography to obtain the river channel topography.

[0093] Step S7. Inverting the river flow: Based on the inverted river channel topography and the multi-period water levels obtained by altimetry satellites, the cross-sectional areas of the river channel for multiple periods can be calculated. An empirical formula of cross-section - flow is established between the cross-sectional area and the measured flow, and the accuracy is evaluated.

[0094] In step S5, the radiance of the blue and green bands of the optical remote sensing image Sentinel-2 and the water depth are used as prior data for the logarithmic ratio model, and the relevant parameters are calculated to establish the construction of the logarithmic ratio water depth inversion model:

[0095]

[0096] In the formula n ,, are empirical parameters, and, are the radiance values of the two bands, and is the water depth value.

[0097] In step S7, the accuracy evaluation of river flow inversion includes the following:

[0098] (1) Root mean square error (RMSE), and the calculation formula is:

[0099]

[0100] (2) Correlation coefficient (R), and the calculation formula is:

[0101]

[0102] Q p : Measured river flow;

[0103] q i : Predict the river flow rate;

[0104] N : Total number of samples.

[0105] Through the above technical solutions disclosed in the present invention, the following beneficial results are obtained. By extracting multi - period river widths from remote sensing satellite data and accurately inverting the river channel topography from water levels, and by constructing a cross - section - flow empirical formula for flow inversion, the accurate prediction of river flow rate is achieved, providing effective technical support for river flow inversion. Embodiment III

[0106] A river flow inversion system based on the collaborative application of multi - mode remote sensing satellites, comprising:

[0107] A remote sensing data processing module, configured to obtain data products of Sentinel - 1, 2 satellites and ICEsat - 2 satellites, and perform data pre - processing. Extract the river width using Sentinel - 1 and 2 satellite images, and extract the water level using ICEsat - 2 satellite images.

[0108] A river channel topography inversion module, configured to apply the river width and water level to extract multi - temporal river boundaries, and interpolate to obtain the river channel topography; establish a water depth inversion model using optical images and water depth to obtain multi - temporal river channel topography, and fuse the two to obtain the river channel topography.

[0109] A flow prediction module, configured to calculate the cross - sectional area of the river channel in multiple periods through the inverted river channel topography and multi - period water levels obtained by altimetry satellites, and establish a cross - section - flow empirical formula between the cross - sectional area and the measured flow rate.

[0110] An accuracy detection module, configured to detect the prediction accuracy of river flow rate using the root - mean - square error formula.

[0111] To better understand the present invention, the above has been described in detail in combination with specific embodiments of the present invention, but it is not a limitation of the present invention. Any simple modification made to the above embodiments based on the technical essence of the present invention still belongs to the scope of the technical solutions of the present invention. Each embodiment in this specification focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the system embodiments, since they basically correspond to the method embodiments, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiments.

Claims

1. A method for estimating river flow by using altimetry satellite to deduce river flow area, characterized in that: The following steps are involved: Step 1: Optical image preprocessing and river width extraction; Step 2: Radar image preprocessing and river width extraction; Step 3: Preprocessing of altimetry satellite images to extract water levels; Step 4: Inverting the river terrain by elevation boundary, including obtaining the multi-period river boundary elevation by combining the water level extracted from the multi-period altimetry satellite data with the river boundary information, interpolating the multi-period boundary elevation by the Kriging interpolation method to obtain the river boundary elevation points above the lowest water level, and then applying the Kriging interpolation method to extrapolate the entire river elevation boundary to invert the underwater elevation, and combining the inverted underwater terrain to obtain the complete river terrain, wherein the river boundary information is determined by the river width extracted by the optical image and the river width extracted by the radar image; Step 5: Inverting river topography from remote sensing images, wherein the remote sensing images include the radar images, the optical images, and the altimetry satellite images; Step 6: Construct river terrain, including retaining the elevation boundary and the overlapping area of ​​the river terrain inverted by remote sensing images. The non-overlapping area is mainly based on the river terrain result inverted by the elevation boundary. The multi-period river terrain inverted by remote sensing images is used as a reference to fuse the river terrain to obtain the river terrain. Step 7: Invert the river flow. According to the inverted river channel topography and the multi-period water level obtained by the altimetry satellite, the multi-period river channel cross-sectional area is calculated. The cross-sectional area-flow empirical formula is established by the cross-sectional area and the measured flow to invert the river flow, and the accuracy is evaluated. The accuracy evaluation method includes evaluating the accuracy through the root mean square error RMSE and the correlation coefficient R. The calculation formula of the root mean square error RMSE is: ， Correlation coefficient R The calculation formula is: ， in, Q p To measure river flow, q i To predict river flow, N is the total number of samples; The step 1 includes the following sub-steps: Step 11: Filter and preprocess the surface reflectance product data to obtain the river optical image dataset; Step 12: Calculate the normalized water index NDWI for the data set, extract the water body threshold by the maximum inter-class variance method, and extract the river water body range by masking. The formula is: NDWI= (Green-NIR) / (Green+NIR); Among them, Green is the reflectance of the green band, and NIR is the reflectance of the near-infrared band; Step 13: Remove small holes and smooth water body contours in the water body through morphology, and delete the area with a large slope in the water body range; the method of removing small holes and smooth water body contours in the water body through morphology includes: Perform image opening and image closing operations on optical images. The image opening operation smoothes the contours of the object, disconnects narrow necks and eliminates thin protrusions; The image closing operation bridges narrow discontinuities and elongated gullies, eliminates small holes, and fills in breaks in contour lines; The image opening operation first erodes and then expands; the image closing operation first expands and then erodes; Step 14: Apply the skeletonization algorithm to extract the center line of the river, calculate the cross-sectional angle of the river to obtain the median orthogonal angle, and extract the river width.

2. The method for estimating river flow by using altimetry satellite to deduce river flow area as claimed in claim 1, characterized in that: The method for deleting the area with a larger slope in the water body range includes: performing slope calculation on the DEM obtained in the GEE platform, and performing masking on the water body to extract the area with a smaller slope.

3. The method for estimating river flow by using altimetry satellite to deduce river flow area as claimed in claim 2, characterized in that: The step 2 includes screening the Sentinel-1 GRD radar image data with the interference wide-band mode of IW, and performing Refined Lee filtering denoising, geographic correction, and radiometric normalization data processing to obtain a river radar image data set.

4. The method for estimating river flow by using altimetry satellite to deduce river flow area as claimed in claim 3, characterized in that: The step 3 includes downloading ICEsat-2 ATL03 data, denoising the altimetry satellite image using the adaptive DBSCAN algorithm, performing Gaussian fitting on the water level to obtain the expectation, calculating the elevation conversion coefficient to perform elevation conversion, and obtaining the river water level in a unified coordinate system.

5. The method for estimating river flow by using altimetry satellite to deduce river flow area as claimed in claim 4, characterized in that: The method for calculating the elevation conversion coefficient for elevation conversion includes converting the coordinates between different geodetic coordinate systems into corresponding coordinates through seven parameters. The formula is: X' = ​​X + dx-Y*rz + Z*ry + S Y' = Y + X*rZ +Z*rx + S Z'=ZX*ry + Y*rx + S Among them, (X, Y, Z) are the coordinates of the original coordinate system, (X', Y', Z') are the coordinates of the target coordinate system, dx, dy and dz are translation parameters, rx, ry and rz are rotation parameters, and S is the scale parameter.

6. The method for estimating river flow by using altimetry satellite to deduce river flow area as claimed in claim 5, characterized in that: The step 5 includes obtaining the river topography through interpolation of the elevation boundary and calculating the water depth value of the corresponding point using the multi-period water level data obtained by the altimetry satellite, using the blue and green band radiation brightness and water depth of the optical image Sentinel-2 as the prior data of the logarithmic ratio model, calculating the relevant parameters, establishing the logarithmic ratio water depth inversion model, and inverting the multi-period underwater topography.

Images