Water level monitoring method and equipment of double-satellite altimeter and medium

Abstract

The invention discloses a water level monitoring method and equipment of a double-satellite altimeter and a medium, and belongs to the technical field of hydrological remote sensing. Performing spatial resolution matching and preprocessing on the ICESat-2 and GEDI data, and extracting surface photon data; the method comprises the following steps: for a land surface region, extracting land surface cross observation points of two satellites, which are overlapped in space, by adopting a graph intersection method, screening effective paired data by combining an abnormal value elimination method, and constructing a land surface elevation correction model based on weighted least square regression; for a water body area, extracting observation photons falling in a water body range, respectively estimating average water levels measured by ICESat-2 and GEDI, and establishing a water level calibration model between the ICESat-2 and GEDI; and fusing the land elevation correction model and the water level calibration model to form a weighted correction function, correcting the GEDI observation water level, and finally obtaining a water level monitoring result of the research area. High-coverage and high-precision remote sensing monitoring of the water level of a water body in a research area is achieved.

Description

Technical Field

[0001] The present invention belongs to the technical field of hydrological remote sensing, and in particular relates to a water level monitoring method, equipment and medium using a dual-satellite altimeter. Background Art

[0002] Water level changes are key indicators in hydrological monitoring and ecological assessment. Although traditional ground hydrological stations have high measurement accuracy and temporal continuity, their construction and maintenance costs are high and they are easily restricted by terrain and transportation conditions, making it difficult to effectively deploy them over a large area. Laser altimetry satellites provide a new technical path for remote sensing of water levels, but a single satellite is constrained by factors such as orbital distribution and observation spacing, resulting in insufficient spatial coverage. To improve the spatial integrity and temporal continuity of monitoring, existing studies have attempted to fuse multi-source satellite observation data and calculate system deviations through cross-observation results to achieve error correction. However, this method relies on rich cross-observation data obtained in large water bodies, making it difficult to apply to small-scale, fragmented water body scenarios such as mining waterlogging, and can easily lead to amplified errors in water level estimation.

[0003] Prior art publication number CN112697232A describes a method and device for measuring and monitoring water levels based on multi-source satellite altimetry data. The method includes: obtaining first and second satellite altimetry data for a target object, wherein a first time series corresponding to the first satellite laser altimetry data and a second time series corresponding to the second satellite altimetry data overlap; calculating a difference modification value based on the first and second satellite altimetry data corresponding to the overlapped time series; calculating water level information for the target object within a target time series based on the first and second satellite altimetry data, and the difference modification value; and predicting the water level of the target object within a preset time period based on the water level information within the target time series. This method ignores the potential correlation between satellite observation data for different surface types and fails to effectively utilize the indirect constraints of land surface observation information on water body observation data. Summary of the Invention

[0004] In response to the problems existing in the above-mentioned prior art, the present invention provides a water level monitoring method, equipment and medium using dual satellite altimeters. This method can effectively correct low-precision satellite observation data, significantly increase the number of monitorable water bodies, and improve the overall accuracy of water level monitoring, providing more reliable basic data for the comprehensive regulation of water landscape.

[0005] To achieve the above object, the present invention provides the following technical solutions:

[0006] A water level monitoring method using dual satellite altimeters comprises the following steps:

[0007] S1: Acquire ICESat-2 and GEDI satellite data in the study area, and perform spatial resampling on the ICESat-2 data to generate a surface photon dataset that matches the spatial resolution of GEDI.

[0008] S2: The graph intersection method is used to extract the spatially overlapping land surface cross observation points in the ICESat-2 and GEDI surface photon data, and the quartile elimination method is used to screen out the effective cross observation point set for regression analysis;

[0009] S3: Based on the effective land surface cross observation point set, a weighted least squares regression model is used to construct the land surface elevation correction model of ICESat-2 and GEDI in non-water areas;

[0010] S4: Use the reflectance threshold method to extract the scope of water bodies in the study area from remote sensing images. Combined with the graph intersection method, select the ICESat-2 and GEDI observation photons that fall on the water surface area. Calculate the average elevation of ICESat-2 and GEDI observation photons in each water body as the water level observation value of each type of satellite. Build a water level calibration model between ICESat-2 and GEDI.

[0011] S5: Integrate the land surface elevation correction model and the water level calibration model to construct a weighted correction function for the GEDI observed water level;

[0012] S6: Use the weighted correction function to correct the GEDI observed water level; if there is an ICESat-2 valid laser observation point in the water body, the ICESat-2 water level observation value is used; otherwise, the corrected GEDI observed water level is used to finally generate the water level observation results of the study area.

[0013] Furthermore, in S1, the pre-processing of ICESat-2 and GEDI satellite data includes:

[0014] S101: Correlate ICESat-2 ATL03 and ATL08 data, filter out surface observation photons (Classification = 0), and calculate the average elevation value every 25 meters along the track to match the GEDI spatial resolution;

[0015] S102: Based on the topographic characteristics of the study area, remove abnormal observations in the GEDI data that do not meet the following conditions:

[0016]

[0017] Where H min Indicates the minimum value of the terrain elevation in the study area, H max Indicates the maximum value of the surface elevation in the study area; Indicates the surface observation elevation value measured by GEDI.

[0018] Furthermore, in S2, obtaining a set of valid land surface cross observation points for regression analysis includes:

[0019] S201: extracting the center-of-mass coordinate information of photons from the surface photon dataset and generating a buffer zone that matches the spot size;

[0020] S202: Using the graph intersection method, extract the land surface intersection observation points where the spatial positions of ICESat-2 and GEDI overlap, and mark the ICESat-2 and GEDI land surface observation elevation values ​​corresponding to each intersection point;

[0021] S203: At the same land surface intersection observation point, the ICESat-2 and GEDI observation elevations are paired to form observation data pairs, and the interquartile range method is used to eliminate data pairs that do not meet the following anomaly detection formula:

[0022]

[0023] Where, represents the elevation value of the i-th land surface intersection point measured by GEDI; represents the elevation value of the i-th land intersection point measured by ICESat-2; Q1 and Q3 represent the lower quartile and upper quartile of all elevation differences, respectively; IQR is the interquartile range.

[0024] Furthermore, in S3, a weighted regression analysis method is used to establish a land surface elevation correction model for ICESat-2 and GEDI in non-water areas, specifically:

[0025] S301: Based on the land surface effective intersection observation point set obtained in step S2, a weighted regression model of the following form is constructed:

[0026]

[0027] Where w i represents the weight of the i-th fitting point; α is the attenuation coefficient, which ranges from [0.1, 10] and is determined by minimizing the root mean square error of the objective function; represents the land surface elevation of the i-th fitting point observed by GEDI; represents the land surface elevation of the i-th fitting point observed by ICESat-2; a land 、b land 、c land Represents the regression coefficient of the land surface elevation correction model.

[0028] Furthermore, in S4, a water level calibration model between ICESat-2 and GEDI is constructed, specifically:

[0029] S401: Using the GEDI observed water level as the independent variable and the ICESat-2 observed water level as the dependent variable, a water level calibration model between ICESat-2 and GEDI is constructed as follows:

[0030]

[0031] Where: represents the average water level of the j-th water body observed by GEDI; represents the average water level of the jth water body observed by ICESat-2; a water 、b water 、c water represents the regression coefficient of the water level calibration model.

[0032] Furthermore, in S5, a weighted correction function of the GEDI observed water level is constructed by integrating the land surface elevation correction model and the water level calibration model. The specific process is as follows:

[0033] S501: By integrating the land surface elevation correction model and the water level calibration model, a weighted correction function for the GEDI observed water level is constructed in the following form:

[0034]

[0035] Where: represents the GEDI observed water level after weighted correction, λ represents the fusion weight, the value range is [0,1], and it is determined by minimizing the root mean square error of the objective function; a land 、b land 、c land represents the coefficient obtained from the land surface elevation correction model; a water 、b water 、c water represents the coefficients obtained from the water level calibration model; Represents the on-site measured value of the water level of the k-th water body.

[0036] A water level monitoring system with dual satellite altimeters, comprising:

[0037] The spatial resampling unit performs spatial resampling on ICESat-2 data to generate a surface photon dataset that matches the spatial resolution of GEDI.

[0038] A graphic intersection extraction unit is used to extract spatially overlapping land surface intersection observation points from ICESat-2 and GEDI surface photon data;

[0039] Quartile elimination unit, used to screen out effective cross-observation point sets for regression analysis based on the quartile elimination method;

[0040] A weighted least squares regression unit is used to construct a land surface elevation correction model for ICESat-2 and GEDI in non-water areas using information from a set of effective land surface cross-observation points.

[0041] The reflectance threshold unit is used to extract the water body range of the study area from remote sensing images;

[0042] The water level calibration unit is used to calculate the average elevation of ICESat-2 and GEDI observation photons in each water body as the water level observation value of the two types of satellites, and to build a water level calibration model between ICESat-2 and GEDI;

[0043] The weighted correction function unit is used to correct the GEDI observed water level and generate the water level observation results of the study area.

[0044] A computer device includes a processor and a memory, wherein the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute a water level monitoring method using a dual-satellite altimeter.

[0045] A computer-readable storage medium stores a computer program, which is suitable for being loaded by a processor and executed by a water level monitoring method using a dual-satellite altimeter.

[0046] Beneficial effects: The present invention deeply analyzes the response differences between ICESat-2 and GEDI laser altimeters under land and water observation conditions, and based on this difference, proposes a differentiated fitting and fusion correction mechanism for surface types to construct a more adaptable GEDI water level correction function. This method can still achieve effective correction of GEDI water level observation results when cross-water level observation data is limited, significantly improving the overall accuracy of remote sensing water level monitoring. Compared with the traditional error correction method that relies on the average deviation of cross-observations, the present invention breaks through its applicability bottleneck in small-scale and fragmented water body monitoring, and enhances the generalization ability and practicality of remote sensing technology under complex surface conditions. This method can provide more accurate and stable water level remote sensing data support for flood disaster warning, drought monitoring and refined regulation of water resources, and has good promotion and application prospects. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Schematic diagram of the flow of the water level monitoring method using dual-satellite altimeters of the present invention;

[0048] Figure 2 This is a distribution map of the main water bodies and satellite laser observation points in the mining area in an embodiment of the present invention;

[0049] Figure 3This is a regression relationship diagram of ICESat-2 and GEDI observation data in an embodiment of the present invention. DETAILED DESCRIPTION

[0050] The present invention will be further described below with reference to the accompanying drawings.

[0051] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0052] like Figure 1 As shown, the present invention provides a technical solution: a water level monitoring method using a dual-satellite altimeter, and the specific steps are as follows.

[0053] Taking the Yanzhou mining area as the experimental area, the water level monitoring method based on ICESat-2 and GEDI dual satellite altimeters proposed in the present invention is verified and explained.

[0054] 1) By querying the Earthdata official website and other public data platforms, the ICESat-2 and GEDI satellite observation strips covering the Yanzhou mining area from July to August 2021 were obtained. The spatial distribution of their laser focus in the study area is as follows: Figure 3 At the same time, in order to verify the accuracy of satellite water level monitoring results, RTK technology was used to collect water surface elevation data of 18 typical water bodies in the study area on July 19, 2021. The specific observation results are shown in Table 1.

[0055] Table 1 Water level verification data

[0056]

[0057] 2) To mitigate the impact of differences in spatial resolution between the ICESat-2 and GEDI altimeters on experimental results, this study resampled the ICESat-2 ATL03 data to align its spatial scale with the GEDI data. The specific processing flow is as follows: ATL03 data are matched with ATL08 data using their photon classification index, filtering out surface photons with a Classification attribute of 0. Subsequently, a sliding window with a step size of 25 meters is created, and the surface photon elevation values ​​within each window are averaged to obtain a surface elevation value representing the center point of each 25-meter interval, matching the spatial resolution of the GEDI satellite.

[0058] 3) For GEDI data, based on the terrain elevation range of the study area, the following conditions were set to eliminate abnormal observations and only retain data points whose elevations were within the specified threshold range:

[0059]

[0060] Where H min Indicates the minimum value of the terrain elevation in the study area, H max Indicates the maximum value of the surface elevation in the study area; Indicates the surface observation elevation value measured by GEDI.

[0061] 4) For land surface observations, based on preprocessed surface photon data, the centroid coordinates of each laser photon were first extracted, and a spatial buffer zone was generated based on the spot size. Then, the graph intersection method was used to extract overlapping ICESat-2 and GEDI observation points in the land area, and the corresponding ICESat-2 and GEDI land surface elevation values ​​were annotated for each intersection. Next, the land surface observations from the two satellites were paired within the same intersection point to construct elevation observation data pairs. To ensure data quality, the interquartile range method was used to eliminate anomalies in the paired data, retaining only data pairs that met the following criteria:

[0062]

[0063] Where, represents the elevation value of the i-th land surface intersection point measured by GEDI; represents the elevation value of the i-th land intersection point measured by ICESat-2; Q1 and Q3 represent the lower quartile and upper quartile of all elevation differences, respectively; IQR is the interquartile range.

[0064] Using this method, a total of 77 pairs of valid land surface cross-observation data points were extracted within the study area. Based on these data, a weighted least squares regression method was used to fit the ICESat-2 and GEDI observations in non-water areas, establishing a land surface elevation mapping relationship between the two and forming a land surface elevation correction model.

[0065] 5) For water body observations, the reflectance threshold method was first used to extract the distribution of water bodies within the study area from remote sensing imagery. A graph intersection method was then used to select ICESat-2 and GEDI observation photons that fell within the water bodies. The average elevation of the photons observed by both satellites was calculated for each water body, serving as the water level observation for that body. Based on these observation pairs for each water body, a water level calibration model between ICESat-2 and GEDI was constructed.

[0066] 6) The land surface elevation correction model constructed above is integrated with the water level calibration model to form a weighted correction function for the GEDI observed water level, which is mathematically expressed as follows:

[0067]

[0068] Where, represents the GEDI observed water level of the j-th water body after weighted correction, represents the average water level of the j-th water body observed by GEDI.

[0069] 7) Based on the constructed weighted correction function, the GEDI observed water level was optimized and adjusted, as shown in Table 2. The results show that the proposed method outperforms existing research and related references in terms of water level observation accuracy and adaptability, demonstrating good overall performance and potential for promotion and application.

[0070] Table 2 Comparison of the accuracy correction effects of GEDI observed water levels based on different correction methods

[0071]

[0072] 8) When generating the final water level results, if a water body has an effective ICESat-2 laser observation point, the ICESat-2 water level observation value is used first; if the water body lacks ICESat-2 observations, the corrected GEDI water level is used as a substitute. Based on this strategy, this study successfully obtained water level information for 16 typical water bodies within the Yanzhou mining area, with a coverage rate of 88.89%. Compared with using only single satellite observation data, this multi-source fusion method increases the number of monitorable water bodies to twice that of ICESat-2 and 1.6 times that of GEDI, significantly enhancing the monitoring coverage capability. In terms of accuracy, the average error of the fused observation water level results is -0.04m, the mean absolute error is 0.15m, and the root mean square error is 0.19m, fully demonstrating the comprehensive advantages of this method in spatial adaptability and monitoring accuracy.

Claims

1. A water level monitoring method using a dual-satellite altimeter, characterized in that: The following steps are involved: S1: Acquire ICESat-2 and GEDI satellite data in the study area, and perform spatial resampling on the ICESat-2 data to generate a surface photon dataset that matches the spatial resolution of GEDI. S2: The graph intersection method is used to extract the spatially overlapping land surface cross observation points in the ICESat-2 and GEDI surface photon data, and the quartile elimination method is used to screen out the effective cross observation point set for regression analysis; S3: Based on the effective land surface cross observation point set, a weighted least squares regression model is used to construct the land surface elevation correction model of ICESat-2 and GEDI in non-water areas; S4: Use the reflectance threshold method to extract the scope of water bodies in the study area from remote sensing images. Combined with the graph intersection method, select the ICESat-2 and GEDI observation photons that fall on the water surface area. Calculate the average elevation of ICESat-2 and GEDI observation photons in each water body as the water level observation value of each type of satellite. Build a water level calibration model between ICESat-2 and GEDI. S5: Integrate the land surface elevation correction model and the water level calibration model to construct a weighted correction function for the GEDI observed water level; S6: Use the weighted correction function to correct the GEDI observed water level; if there is an ICESat-2 valid laser observation point in the water body, the ICESat-2 water level observation value is used; otherwise, the corrected GEDI observed water level is used to finally generate the water level observation results of the study area.

2. The water level monitoring method using a dual-satellite altimeter according to claim 1, wherein: In S1, the preprocessing steps for ICESat-2 and GEDI satellite data are as follows: S101: Correlate ICESat-2 ATL03 and ATL08 data, filter out surface observation photons (Classification = 0), and calculate the average elevation value every 25 meters along the track to match the GEDI spatial resolution; S102: Based on the topographic characteristics of the study area, remove abnormal observations in the GEDI data that do not meet the following conditions: Where H min Indicates the minimum value of the terrain elevation in the study area, H max Indicates the maximum value of the surface elevation in the study area; Indicates the surface observation elevation value measured by GEDI.

3. The water level monitoring method using dual satellite altimeters according to claim 1, characterized in that: In S2, the steps for obtaining a set of valid land surface cross observation points for regression analysis are as follows: S201: extracting the center-of-mass coordinate information of photons from the surface photon dataset and generating a buffer zone that matches the spot size; S202: Using the graph intersection method, extract the land surface intersection observation points where the spatial positions of ICESat-2 and GEDI overlap, and mark the ICESat-2 and GEDI land surface observation elevation values ​​corresponding to each intersection point; S203: At the same land surface intersection observation point, the ICESat-2 and GEDI observation elevations are paired to form observation data pairs, and the interquartile range method is used to eliminate data pairs that do not meet the following anomaly detection formula: Where, represents the elevation value of the i-th land surface intersection point measured by GEDI; represents the elevation value of the i-th land intersection point measured by ICESat-2; Q1 and Q3 represent the lower quartile and upper quartile of all elevation differences, respectively; IQR is the interquartile range.

4. The water level monitoring method using dual satellite altimeters according to claim 1, characterized in that: In S3, a weighted regression analysis method is used to establish a land surface elevation correction model for ICESat-2 and GEDI in non-water areas. The specific steps are as follows: S301: Based on the set of effective land surface cross observation points obtained in S2, a weighted regression model of the following form is constructed: Where w i represents the weight of the i-th fitting point; α is the attenuation coefficient, which ranges from [0.1, 10] and is determined by minimizing the root mean square error of the objective function; represents the land surface elevation of the i-th fitting point observed by GEDI; represents the land surface elevation of the i-th fitting point observed by ICESat-2; a land 、b land 、c land Represents the regression coefficient of the land surface elevation correction model.

5. The water level monitoring method using dual satellite altimeters according to claim 1, characterized in that: In S4, a water level calibration model between ICESat-2 and GEDI is constructed, specifically: S401: Using the GEDI observed water level as the independent variable and the ICESat-2 observed water level as the dependent variable, a water level calibration model between ICESat-2 and GEDI is constructed as follows: Where: represents the average water level of the j-th water body observed by GEDI; represents the average water level of the jth water body observed by ICESat-2; a water 、b water 、c water represents the regression coefficient of the water level calibration model.

6. The water level monitoring method using dual satellite altimeters according to claim 1, characterized in that: In S5, a weighted correction function for the GEDI observed water level is constructed by integrating the land surface elevation correction model and the water level calibration model. The specific process is as follows: S501: By integrating the land surface elevation correction model and the water level calibration model, a weighted correction function for the GEDI observed water level is constructed in the following form: Where: represents the GEDI observed water level after weighted correction, λ represents the fusion weight, the value range is [0,1], and it is determined by minimizing the root mean square error of the objective function; a land 、b land 、c land represents the coefficient obtained from the land surface elevation correction model; a water 、b water 、c water represents the coefficients obtained from the water level calibration model; Represents the on-site measured value of the water level of the k-th water body.

7. A water level monitoring system with dual satellite altimeters, characterized in that: include: The spatial resampling unit performs spatial resampling on ICESat-2 data to generate a surface photon dataset that matches the spatial resolution of GEDI. A graphic intersection extraction unit is used to extract spatially overlapping land surface intersection observation points from ICESat-2 and GEDI surface photon data; Quartile elimination unit, used to screen out effective cross-observation point sets for regression analysis based on the quartile elimination method; A weighted least squares regression unit is used to construct a land surface elevation correction model for ICESat-2 and GEDI in non-water areas using information from a set of effective land surface cross-observation points. The reflectance threshold unit is used to extract the water body range of the study area from remote sensing images; The water level calibration unit is used to calculate the average elevation of ICESat-2 and GEDI observation photons in each water body as the water level observation value of the two types of satellites, and to build a water level calibration model between ICESat-2 and GEDI; The weighted correction function unit is used to correct the GEDI observed water level and generate the water level observation results of the study area.

8. A computer device, characterized in that: It includes a processor and a memory, the processor is electrically connected to the memory, the memory is used to store instructions and data, and the processor is used to execute the water level monitoring method of the dual-satellite altimeter according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that The computer-readable storage medium stores a computer program, which is suitable for being loaded by a processor and executing the water level monitoring method of the dual-satellite altimeter according to any one of claims 1 to 7.

Images