Lake and reservoir water storage capacity change multi-source radar remote sensing monitoring method for African area without ground data

Abstract

The invention discloses a multi-source radar remote sensing monitoring method for lake and reservoir water storage change in an African area without ground data, which is used for solving the problems that the traditional monitoring technology for the African area without ground data is high in cost and difficult to monitor and analyze the lake and reservoir water storage change in extreme weather and dense time phases. The method comprises the following steps: collecting multi-source radar remote sensing data and carrying out data preprocessing and screening; obtaining a lake and reservoir area change sequence by using SAR image data, and obtaining a lake and reservoir water level change sequence by using spaceborne radar altimeter data; establishing a lake and reservoir area and water level change relation model, and reconstructing a lake and reservoir area and water level change time sequence; and constructing a lake and reservoir water storage change estimation model, monitoring the water storage change of each lake and reservoir in the to-be-detected area, and analyzing the water storage change characteristics of the lake and reservoir. The method is suitable for lake and reservoir water storage change monitoring in a region without ground data, data dependence and weather influence can be reduced, and lake and reservoir water storage change information with high time resolution is obtained.

Description

Technical Field

[0001] The invention relates to the fields of remote sensing technology and water resource monitoring technology, and in particular to a multi-source radar remote sensing monitoring method for lake and reservoir water storage capacity changes in areas without ground data in Africa. Background Art

[0002] Although lakes and reservoirs only occupy about 2% of the Earth's land surface, they are indispensable water resources for human social economy and play an important role in regulating floods, providing agricultural irrigation, promoting aquaculture, and improving the ecological environment. The traditional method of monitoring changes in lake and reservoir water storage capacity by relying on field measurement data is not only time-consuming and labor-intensive, but also difficult to meet the monitoring needs of a large range, high temporal frequency and low frequency.

[0003] Remote sensing technology has the advantages of low cost, strong timeliness, and the ability to conduct large-scale synchronous observations. In recent years, it has received increasing attention in surface water resource monitoring, especially in areas without ground observation data. At present, some scholars have used satellite remote sensing technology to carry out research on monitoring changes in lake and reservoir water storage, but most of the existing methods are based on optical remote sensing images and satellite altimeter data. This method is greatly affected by weather and it is difficult to obtain high-temporal resolution information on changes in lake and reservoir water storage.

[0004] Both synthetic aperture radar (SAR) and radar altimeters are not affected by clouds and rain, and have all-day and all-weather observation capabilities. Applying multi-source radar remote sensing technologies including SAR and radar altimeters to lake and reservoir water storage monitoring can reduce dependence on ground data and weather influences, and provide a monitoring method for obtaining lake and reservoir water storage change information with high temporal resolution in areas without ground observation data. Summary of the invention

[0005] The purpose of the present invention is to provide a multi-source radar remote sensing monitoring method for lake and reservoir water storage capacity changes in areas without ground data in Africa, which is used to accurately extract lake and reservoir water storage capacity change information in areas without ground observation data with high temporal resolution.

[0006] The present invention adopts the following technical solution: a multi-source radar remote sensing monitoring method for lake and reservoir water storage capacity changes in areas without ground data in Africa, comprising the following steps:

[0007] Step S1, collecting multi-source radar remote sensing data of each lake and reservoir in the area to be measured, including: SAR image data and satellite-borne radar altimeter data, and performing data preprocessing respectively;

[0008] Step S2, using the pre-processed SAR image data of each lake and reservoir, by introducing the SBGFRLS active contour model, to obtain the time series of the area of ​​the lake and reservoir;

[0009] Step S3, using the pre-processed satellite radar altimeter data of each lake and reservoir, calculate the lake and reservoir water level value on the day when the satellite radar passes through and perform deviation correction to obtain the lake and reservoir water level time series;

[0010] Step S4, based on the area time series obtained in step S3 and the water level time series obtained in step S4, a lake area and water level change relationship model is established, and the lake area and water level change time series are reconstructed;

[0011] Step S5: Based on the reconstructed lake area and water level change time series, a lake and reservoir water storage capacity change estimation model is constructed to monitor the water storage capacity changes of each lake and reservoir in the measured area and analyze the characteristics of lake and reservoir water storage capacity changes.

[0012] Preferably, step S1 data preprocessing is performed as follows:

[0013] Step S1.1, the SAR image data of each lake and reservoir are subjected to orbit correction, radiation correction, speckle noise suppression using the RefinedLee filtering method, terrain correction and decibel processing in turn;

[0014] Step S1.2: Obtain the sub-satellite point coordinates based on the latitude and longitude information of the satellite-borne radar altimeter data, and extract the altimetry data of the point.

[0015] Preferably, step S2 uses satellite-borne SAR images to obtain lake and reservoir area time series, the method is as follows:

[0016] Step S2.1: For the preprocessed SAR image, select the VV and VH bands of different polarization modes, calculate the water body index SDWI, and obtain the SDWI image

[0017] Step S2.2, using the K-means clustering algorithm to classify the SDWI image, generate a binary map containing two categories of water body and non-water body, and obtain the initial water body contour map;

[0018] Step S2.3, using the water body contour map obtained in the previous step as the initial contour of the SBGFRLS active contour model, and determining the water body boundary through an iterative formula;

[0019] Step S2.4, segment the water body boundary into connected components, select the connected components with larger areas as the final water body areas, count the number of water body pixels in the final water body distribution map and calculate their areas, and establish the area time series of each lake and reservoir.

[0020] Preferably, in step S3, the time series of lake and reservoir water levels are obtained using satellite-borne radar altimeter data, and the method is as follows:

[0021] Step S3.1: for the pre-processed satellite radar altimeter data, use the water body boundary determined in step S2.3 as the lake and reservoir vector boundary data, filter out the elevation points falling on the water surface, and perform geophysical correction and waveform re-tracking correction using the ICE-algorithm on the elevation points within the water surface in turn;

[0022] Step S3.2, arrange the height measurement points retained in the previous step in ascending order of latitude to form a one-dimensional array, calculate the mean error σ of the data in the array, use the 3σ criterion to eliminate outliers, and average the remaining height measurement data to obtain the water level value;

[0023] Step S3.3, executing steps S3.1 and S3.2 respectively for different radar altimeter data, and classifying the data according to different geoids, selecting a reference benchmark, and converting the water levels calculated by different radar altimeters into the same elevation benchmark;

[0024] Step S3.4, using the overlapping observation period data between satellites to perform bias correction on the high water level data in turn, combining the obtained bias-corrected data series, and taking the arithmetic mean of the single-day water level values ​​in the overlapping period;

[0025] Step S3.5: Perform Gaussian filtering on all water level values ​​to establish a water level time series for the lake or reservoir.

[0026] Preferably, step S4, a relationship model between lake area and water level change is established. Since it is difficult to obtain area and water level data on the same day, cubic spline curve interpolation is performed on the water level time series data according to the area time series to reconstruct the lake area and water level change time series.

[0027] Preferably, in step S5, a lake reservoir water storage capacity change estimation model is constructed based on the lake reservoir area and water level change time series, the lake reservoir water storage capacity change is monitored, and the lake reservoir water storage capacity change characteristics are analyzed.

[0028] Compared with the prior art, the present invention adopts the above technical solution and has the following technical effects:

[0029] 1. The present invention provides a method for monitoring changes in lake and reservoir water storage capacity that is suitable for areas without ground data. It can reduce data dependence and weather influence, and obtain lake and reservoir water storage capacity change information with high temporal resolution.

[0030] 2. The method of the present invention utilizes multi-source radar remote sensing data to achieve large-scale water body monitoring and rapid acquisition of change information, reducing the consumption of manpower and material resources and greatly saving monitoring costs. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flow chart of the multi-source radar remote sensing monitoring method for lake and reservoir water storage capacity changes of the present invention;

[0032] Figure 2 A flow chart of a method for constructing a time series of lake and reservoir areas provided in an embodiment of the present invention;

[0033] Figure 3 A flow chart of a method for constructing a lake reservoir water level time series provided by an embodiment of the present invention;

[0034] Figure 4 The rendering of the local boundary extraction of Lake Victoria in 2021 provided by the embodiment of the present invention;

[0035] Figure 5 This is a graph of the construction and reconstruction results of the water level, area and time of Lake Victoria in 2021 provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0036] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the application is further elaborated in detail below in conjunction with the accompanying drawings. The described embodiments are only a part of the embodiments involved in the present invention. All non-innovative embodiments of other researchers in the field on this embodiment belong to the protection scope of the present invention. At the same time, for the step numbering in the embodiment of the present invention, it is only set for the convenience of explanation, and the order between the steps is not limited in any way. The execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0037] The present invention provides a multi-source radar remote sensing monitoring method for lake and reservoir water storage capacity changes in areas without ground data in Africa. Figure 1 As shown, the following steps are included:

[0038] Step S1, collecting multi-source radar remote sensing data of each lake and reservoir in the area to be measured, including: SAR image data and satellite-borne radar altimeter data, and performing data preprocessing respectively;

[0039] Step S2, using the pre-processed SAR image data of each lake and reservoir, by introducing the SBGFRLS (Sparse Bregman Gradient Flow with Reinitialization and Level Set) active contour model, the area time series of the lake and reservoir is obtained;

[0040] Step S3, using the pre-processed satellite radar altimeter data of each lake and reservoir, calculate the lake and reservoir water level value on the day when the satellite radar passes through and perform deviation correction to obtain the lake and reservoir water level time series;

[0041] Step S4, based on the area time series obtained in step S3 and the water level time series obtained in step S4, a lake area and water level change relationship model is established, and the lake area and water level change time series are reconstructed;

[0042] Step S5: Based on the reconstructed lake area and water level change time series, a lake and reservoir water storage capacity change estimation model is constructed to monitor the water storage capacity changes of each lake and reservoir in the measured area and analyze the characteristics of lake and reservoir water storage capacity changes.

[0043] In one embodiment of the present invention, Lake Victoria, a tectonic lake located in east-central Africa, is selected as the target area, and Sentinel-1 remote sensing images are used as SAR images to construct area time series, and Sentinel-3 and Jason-3 data are used as satellite-borne radar altimeters to construct water level time series. The specific process is as follows:

[0044] Step 1: Obtain Sentinel-1 data for the entire year of 2021 as SAR images, such as Figure 2 As shown, orbit correction, radiation correction, speckle noise suppression using the Refined Lee filtering method, terrain correction and decibel processing are performed in sequence to obtain the pre-processed SAR image.

[0045] Get Sentinel-3 data and Jason-3 data for the whole year of 2021 as spaceborne radar altimeter data, such as Figure 3 As shown, the coordinates of the sub-satellite point are obtained according to the longitude and latitude information, and the altimetry data of the point is extracted.

[0046] Step 2: Based on the SAR image data in January 2021, Figure 4 As shown in the figure, for the preprocessed SAR image, select the VV and VH bands of different polarization modes, and calculate the water body index SDWI according to the following formula to obtain the SDWI image:

[0047] SDWI=ln(10·VV·VH)-8

[0048] Where VV and VH represent the co-polarization backscatter value and cross-polarization backscatter value corresponding to the SAR image, respectively.

[0049] Since lakes and reservoirs are closed and have boundaries, the K-means clustering algorithm is used to classify the SDWI images to generate a binary map containing two major categories: water bodies and non-water bodies, and obtain the initial water body contour map.

[0050] Furthermore, in this embodiment, the SBGFRLS active contour model is introduced to optimize the water body boundary extraction method and improve the accuracy of the water body contour.

[0051] The initial water body contour is used as the initial contour of the SBGFRLS active contour model, and the level set function Φ is initialized according to the following formulas; the inner and outer mean values ​​c of the contour curve are calculated: 1 and c 2 ; Evolve the level set function; Regularize the level set function using Gaussian filtering; Check whether the evolution of the level set function has converged. If not, return to recalculate the inner and outer means of the contour curve.

[0052] The final water body boundary is obtained when the iteration stops.

[0053]

[0054]

[0055] In the formula, I(x, y) represents the original image; Ω represents the feature space of the image; c 1 is the grayscale mean of the image within the curve; c 2 is the grayscale mean of the image outside the curve; α is the constant speed; SPF is the pressure sign function; represents the gradient of the pressure sign function; represents the gradient of the level set function φ; div represents the divergence operator; represents the evolution equation of the level set function φ.

[0056] In addition to the lakes and reservoirs that need to be extracted, there are many rivers, puddles and other water bodies around the image that will also be extracted together. In this embodiment, the final water body boundary is further segmented into connected components, and the connected components with larger areas are screened out and retained as the final water body area to form a water body distribution map.

[0057] Then, the number of water body pixels in the water body distribution map was calculated and multiplied by the ground resolution of a single pixel to estimate the actual area of ​​the water body.

[0058] S=10*10*N

[0059] In the formula, S is the final water area result, and N is the total number of pixels occupied by lake water bodies.

[0060] Furthermore, based on the SAR image data from February to December 2021, step 2 is repeated respectively to obtain the monthly area data of Lake Victoria in 2021 and establish the time series of the area of ​​Lake Victoria in 2021.

[0061] Step 3: For the Sentinel-3 satellite-borne radar altimeter data, use the vector boundary data of Lake Victoria to filter out the elevation points that fall on the water surface, and then perform geophysical correction and waveform re-tracking correction using the ICE algorithm on the filtered elevation point data.

[0062] Arrange the corrected elevation point data from small to large latitude to form a one-dimensional array, use the 3σ criterion to eliminate outliers, and calculate the average to get the water level value of the day;

[0063]

[0064] Where, σ is the mean error; H i is a single water level value in a cycle; n is the number of water level values ​​in the cycle; X is the mean water level in the cycle, and m is the number of water level values ​​remaining after removing abnormal data.

[0065] The data measured by the radar altimeter on Jason-3 are also corrected and eliminated in the above steps to obtain a set of water level values.

[0066] It should be noted that since different radar altimeters have different elevation systems, there is a certain difference between them, and they need to be classified according to the elevation system to facilitate the subsequent calculation of the height measurement data into the same elevation system.

[0067] Specifically, the water level data are classified according to different elevation systems. The first type of altimetry data is preprocessed to obtain altimetry water level values ​​based on the EGM2008 (global very high-order earth gravity field model) geoid model. The second type of altimetry data is processed to obtain altimetry water level values ​​based on the non-EGM2008 geoid model. The first type of altimetry water level values ​​based on the EGM2008 geoid model are used as the reference benchmark to convert water levels observed by different satellites into the same elevation benchmark.

[0068] Since both altimeters in this embodiment are based on the EGM2008 geoid model, no reference system conversion is required. However, it is necessary to consider the systematic errors between different altimeters. By using the overlapping observation period data between satellites, the altimetry data are corrected for deviations in turn according to the following formula, thereby eliminating the systematic errors:

[0069]

[0070] In the formula, The high water level measured by satellite A on the i-th day after eliminating the system error; To eliminate the system error before the A satellite measured the high water level on the i-th day; is the water level value monitored by satellite A on the jth day during the overlapping observation period; is the water level value monitored by satellite B on the jth day during the overlapping observation period; n is the sample size of paired data during the overlapping observation period.

[0071] Furthermore, the bias-corrected data series obtained in the previous step are combined to obtain the daily water level of the crossover period:

[0072]

[0073] In the formula, H i is the final water level value on the i-th day of the crossover period; is the water level value of satellite A on the i-th day in the crossover period after bias correction; is the water level value of satellite B on the i-th day in the crossover period after bias correction.

[0074] Then, all daily water levels were subjected to Gaussian filtering to establish the Lake Victoria water level time series.

[0075] Step 4: Since different satellites arrive above Lake Victoria to collect data at different dates, and the collection times of Sentinel-1 remote sensing image data and Sentinel-3 and Jason-3 satellite-borne radar altimeter data are also different, it is necessary to establish a model for the relationship between the area and water level changes of Lake Victoria based on the lake area and water level values ​​in different time periods.

[0076] like Figure 5 As shown, the Victoria Lake area time series obtained in step 3 is used to perform cubic spline interpolation on the Victoria Lake water level time series data obtained in step 4 to obtain the Victoria Lake water level value on the date corresponding to the area image, and to reconstruct the Victoria Lake area and water level change time series.

[0077] Step 5. Since the lake surface and the lake bottom are of different sizes, based on the area of ​​the reconstructed Lake Victoria and the time series of water level changes, Lake Victoria is divided into multiple units from the lake surface to the lake bottom according to the height. The volume of each unit corresponds to the water storage in a period of time. All the units are added together to estimate the water storage capacity of the entire Lake Victoria.

[0078] Specifically, assuming that each unit has the geometric characteristics of a prism or a prism, the water level changes in adjacent time periods, and the change in water storage is reflected in the change in water level. The volume before and after the change corresponds to a unit volume. The monthly change in lake water storage is calculated using the following volume calculation formula:

[0079]

[0080] Where ΔV i is the change in water storage between two consecutive dates, S i-1 and S i are the water area of ​​two consecutive dates, h i and h i-1 They are the water level elevations on the days corresponding to two adjacent dates.

[0081] Furthermore, based on the monthly changes in the water storage capacity of the lake reservoir, the monthly changes in the water storage capacity of Lake Victoria in 2021 can be monitored. It can be found that there are obvious seasonal changes in the water storage capacity of Lake Victoria. In May and October of each year, the water storage capacity increases, corresponding to the local rainy season; although there are fluctuations from June to September, the change in water storage always remains in a negative state, indicating that the water storage capacity continues to decrease, corresponding to the local dry season, and the extraction results are relatively reasonable.

[0082] The above is only a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A multi-source radar remote sensing monitoring method for lake and reservoir water storage capacity changes in areas without ground data in Africa, characterized in that: The following steps are involved: Step S1, collecting multi-source radar remote sensing data of each lake and reservoir in the area to be measured, including: SAR image data and satellite-borne radar altimeter data, and performing data preprocessing respectively; Step S2, using the pre-processed SAR image data of each lake and reservoir, by introducing the SBGFRLS active contour model, the area time series of the lake and reservoir is obtained; Step S3, using the pre-processed satellite radar altimeter data of each lake and reservoir, calculate the lake and reservoir water level value on the day when the satellite radar passes through and perform deviation correction to obtain the lake and reservoir water level time series; Step S4, based on the area time series obtained in step S3 and the water level time series obtained in step S4, a lake area and water level change relationship model is established, and the lake area and water level change time series are reconstructed; Step S5: Based on the reconstructed lake area and water level change time series, a lake and reservoir water storage capacity change estimation model is constructed to monitor the water storage capacity changes of each lake and reservoir in the measured area and analyze the characteristics of lake and reservoir water storage capacity changes.

2. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 1 is characterized in that: In the step S1, SAR image data is collected and preprocessed, and orbit correction, radiation correction, terrain correction, and Refined Lee filtering method are used to suppress speckle noise and decibel processing are performed in sequence; altimeter data of different satellite-borne radars are collected and preprocessed, the coordinates of the sub-satellite point are obtained according to the longitude and latitude information, and the altimetry data of the sub-satellite point are extracted.

3. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 1 is characterized in that: In step S2, the time series of lake and reservoir areas are obtained as follows: Step S2.1: For the preprocessed SAR image, select the VV and VH bands of different polarization modes, calculate the water body index SDWI, and obtain the SDWI image: SDWI=ln(10·VV·VH)-8 (1) Where VV and VH represent the co-polarization backscatter value and cross-polarization backscatter value corresponding to the SAR image, respectively; Step S2.2, using the K-means clustering algorithm to classify the SDWI image, generate a binary map containing two categories of water body and non-water body, and obtain an initial water body contour map; Step S2.3, using the initial water body contour as the initial contour of the SBGFRLS active contour model, and determining the water body boundary through iteration; Step S2.4, segment the water body boundary into connected components, select the connected components with an area greater than the preset lake area threshold as the water body area, form a water body distribution map, count the number of water body pixels in the water body distribution map and estimate the actual area of ​​the water body, and establish a time series of the lake area.

4. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 3 is characterized by: In step S2.3, the water body boundary is determined by iteration, as follows: S2.3.

1. Initialize the level set function Φ and calculate the inner and outer means of the contour curve; S2.3.2, evolving level set function, using Gaussian filtering to regularize the level set function; S2.3.

3. Check whether the evolution of the level set function has converged. If not, return to execute steps S2.3.1 to S2.3.2 to recalculate the inner and outer means of the contour curve. When the iteration stops, the water boundary of the lake is obtained. The iteration formula is as follows: Where I(x,y) is the original image; Ω is the feature space of the image; c1 is the grayscale mean of the image inside the curve; c2 is the grayscale mean of the image outside the curve; α is the constant speed; SPF is the pressure sign function, represents the gradient of the pressure sign function; represents the gradient of the level set function φ; div represents the divergence operator; represents the evolution equation of the level set function φ.

5. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 3 is characterized by: In step S2.4, the actual area of ​​the water body is estimated by multiplying the number of water body pixels in the water body distribution map by the ground resolution of a single pixel. The formula is as follows: S=10*10*N (4) In the formula, S is the estimated water area, and N is the total number of pixels occupied by lake and reservoir water bodies. The number of water body pixels is converted into the actual water body area through formula (4), and the lake and reservoir area time series is established.

6. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 3 is characterized by: In step S3, the time series of lake and reservoir water levels are obtained as follows: Step S3.1: for the pre-processed satellite radar altimeter data, the water body boundary determined in step S2.3 is used as the lake and reservoir vector boundary data, the elevation points falling within the water surface are screened out, and the elevation points within the water surface are sequentially subjected to geophysical correction and waveform re-tracking correction using the ICE-algorithm to obtain the corrected elevation point data; Step S3.2, the corrected elevation point data are arranged in ascending order of latitude to form a one-dimensional array, the mean error σ of the data in the array is calculated, the outliers are eliminated using the 3σ criterion, and the retained elevation point data are averaged to obtain the water level value of the lake reservoir on the day when the satellite-borne radar passes over; Step S3.3: for the altimeter data of different satellite-borne radars, respectively execute steps S3.1 and S3.2 to obtain the water level value of the lake reservoir on the day when each satellite-borne radar passes through, classify and pre-process the water level value data according to the different satellite-borne radar altimeter elevation systems, obtain the altimeter water level value based on the EGM2008 model and the altimeter water level value based on the non-EGM2008 model, and use the altimeter water level value based on the EGM2008 model as a benchmark to return the altimeter water level values ​​of different satellite-borne radars to the same elevation system; Step S3.4, performing deviation correction on the water level values ​​in the overlapping observation periods between the satellite-borne radars, and combining the deviation-corrected height water level data to calculate the final water level value for a single day in the overlapping observation period; Step S3.5: Perform Gaussian filtering on all daily final water level values ​​to establish a lake and reservoir water level time series.

7. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 6 is characterized by: In step S3.2, the mean error σ of the data in the array is calculated according to formula (5), and the elevation point data after removing the outliers are averaged according to formula (6) to obtain the water level value H of the lake reservoir on the day when the satellite-borne radar passes. avg : Where, σ is the mean error; H i is a single water level value in a cycle; X is the mean water level in the cycle; n is the number of water level values ​​in the cycle; and m is the number of water level values ​​remaining after removing abnormal data.

8. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 6 is characterized by: In step S3.4, the measured high water level data is subjected to deviation correction, and the formula is as follows: Where: To eliminate the elevation system error before the A satellite measured the water level on the i-th day; is the water level value monitored by satellite A on the jth day during the overlapping observation period; is the water level value monitored by satellite B on the jth day during the overlapping observation period (reference benchmark); n is the sample size of paired data during the overlapping observation period, As a reference benchmark, we get The water level value measured by satellite A on the i-th day after the elevation system error is eliminated; Calculate the final water level value of a single day during the overlapping observation period using the following formula: In the formula, H i is the final water level value on the i-th day during the overlapping observation period; is the water level value of satellite A on the i-th day during the overlapping observation period after bias correction; is the water level value of satellite B on the i-th day during the overlapping observation period after bias correction.

9. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 6 is characterized by: In step S4, since the collection time of SAR image data and satellite-borne radar altimeter data is different, the lake and reservoir water level time series obtained in step S3.5 is interpolated with cubic spline curves according to the time of the lake and reservoir area time series obtained in step S2.4, so as to obtain the lake and reservoir area and water level data corresponding to the same time, and reconstruct the lake and reservoir area and water level change time series.

10. The method for multi-source radar remote sensing monitoring of lake and reservoir water storage capacity changes according to claim 9 is characterized in that: In step S5, a lake reservoir water storage change estimation model is constructed as follows: Step S5.1: According to the lake area and water level change time series, the lake is divided into several units from the lake surface to the lake bottom according to the height. Assuming that each unit has the geometric characteristics of a prism or a prism, the volume ΔV of each unit is calculated respectively. i : In the formula, S i-1 and S i are the water area on two consecutive dates respectively; h i and h i-1 are the water level elevations on two consecutive dates; ΔV i is the change in water storage between two adjacent dates; Step S5.2: Set the volume of each unit ΔV i As the change in the water storage capacity of a lake or reservoir between two consecutive dates, the volumes of all units are added together to estimate the water storage capacity of the entire lake or reservoir.

Images