Star-satellite borne GNSS-R high space-time water body detection method and system

Abstract

The present application belongs to the technical field of satellite remote sensing, and discloses a kind of star-machine borne GNSS-R high space-time water body detection method and system.The present application synchronously obtains the reflectivity dataset of CYGNSS, TM-1, FY-3 satellite-borne GNSS-R and unmanned aerial vehicle airborne GNSS-R under the same space-time condition of target area;With GPS as a reference, the data of different navigation systems under the same constellation is corrected for histogram peak deviation;With CYGNSS as a reference, the histogram peak deviation correction is carried out on the data of different constellations after the same system;Water body detection is carried out based on T and corresponding seasonal Th.The present application realizes water body detection using satellite-borne and airborne GNSS-R signals, improves the inversion accuracy of water body detection, so as to promote the wide application of star-machine borne GNSS-R signal detection of water body in the fields of water resource detection and management, etc.

Description

Technical Field

[0001] This invention belongs to the field of satellite remote sensing technology, and in particular relates to a satellite-airborne GNSS-R method and system for high-temporal-spatial water body detection. Background Technology

[0002] Global Navigation Satellite System Reflectance (GNSS-R) technology utilizes the echo power, waveform, and reflectivity characteristics of L-band microwave signals from navigation satellites after reflection from the Earth's surface to retrieve surface parameters. Because the dielectric constant of water surfaces is significantly higher than that of land, their specular reflection characteristics result in a significantly higher GNSS-R reflectivity. Therefore, by using reflectivity threshold discrimination or machine learning classification, the identification and detection of water targets such as lakes, rivers, reservoirs, wetlands, and floodplains can be achieved. Compared with traditional active remote sensing, GNSS-R has advantages such as not requiring a dedicated transmitter, low cost, wide coverage, and continuous observation, and has become an important supplementary means for surface water monitoring.

[0003] Water body detection methods based on spaceborne GNSS-R mainly fall into two categories: one is the reflectance threshold discrimination method based on single-source or simple multi-source data fusion, which uses single-constellation data such as CYGNSS, FY-3, or Tianmu-1 to distinguish the land-water boundary by setting a fixed reflectance threshold; the other is the intelligent classification method based on deep learning, which uses convolutional neural networks (CNNs) or fully connected networks to extract water body characteristics end-to-end. However, the above methods all have obvious limitations in practical applications and are difficult to meet the requirements of high precision, high spatiotemporal resolution, and stable monitoring across seasons.

[0004] Based on the above analysis, the main problems and shortcomings of existing deep learning-based spaceborne GNSS-R water detection methods are as follows: (1) Inconsistent benchmarks of multi-source data: Existing methods simply splice together data from multiple constellations such as CYGNSS, Tianmu-1, and FY-3, as well as multiple systems such as GPS / BDS / Galileo / GLONASS, without considering the differences in orbital altitude, receiver gain, signal frequency, and transmission power of different constellations and different navigation systems. This results in systematic deviations in reflectivity products and chaotic internal benchmarks in the fused dataset. (2) Lack of effective calibration of spaceborne data: Existing methods rely only on internal cross-validation of spaceborne data or indirect comparison with low-resolution microwave radiometers, which cannot eliminate the absolute deviation between spaceborne observations and the real Earth surface; (3) Lack of seasonal adaptation capability: Existing methods use fixed thresholds or uniform models throughout the year, without considering the shift in water and land reflectivity characteristics caused by seasonal factors such as vegetation phenology, sediment content, and icing. This leads to an increase in false alarms and missed alarms in cross-seasonal scenarios, and a significant decrease in detection stability. Summary of the Invention

[0005] To overcome the problems existing in related technologies, the present invention discloses a satellite-airborne GNSS-R high-spatial-temporal water body detection method and system, the technical solution of which is as follows: This invention is implemented as follows: a satellite-airborne GNSS-R method for high-altitude spatiotemporal water body detection, comprising the following steps: S1. Simultaneously acquire multi-source global navigation satellite system reflectivity datasets under the same spatiotemporal conditions in the target area and grid them to obtain the CYGNSS reflectivity dataset. TM-1 reflectivity dataset and FY-3 reflectivity dataset The GNSS-R reflectivity dataset of preset sampling points in the area was acquired simultaneously using drones. ; S2. Draw probability distribution histograms for data from different systems within the same constellation. Based on the difference in peak points, and using GPS system data as a benchmark, unify the data benchmark across different systems within the same constellation to obtain the unified benchmark TM-1 GPS signal dataset. and FY-3 GPS signal dataset ; S3. Draw probability distribution histograms for different constellation data after the same system is established. Based on the difference in peak points, use CYGNSS data as a benchmark to unify the data benchmark for different constellations, and obtain the TM-1 to CYGNSS correction dataset with a unified benchmark. and FY-3 to CYGNSS calibration dataset With the CYGNSS reflectivity dataset By merging, a unified constellation of target area spaceborne GNSS-R reflectivity datasets is obtained. ; S4. Airborne GNSS-R reflectivity dataset based on preset sampling points in the target area For the target area spaceborne GNSS-R reflectivity dataset Calibration was performed to obtain a high spatiotemporal resolution reflectance dataset for the target region. ; S5. Select a pure land area and a pure water area within the target region, and extract reflectance datasets for the pure land area and the pure water area under different seasons; plot a probability distribution histogram in the same coordinate system, and take the x-coordinate of the intersection of the two curves as the reflectance threshold for the water and land areas. Threshold Including: Spring water and land reflectance thresholds Summer water and land reflectance thresholds Autumn water and land reflectance thresholds and winter water and land reflectivity thresholds ; S6, Based on reflectivity dataset and threshold Conduct water body detection in the target area.

[0006] In step 1, the preset sampling points are determined by stratified sampling based on the topographic information and land cover type of the target area.

[0007] In step 2, the standardization of data benchmarks across different systems within the same constellation includes: Based on the differences in the navigation systems to which the signals belong, the TM-1 reflectivity dataset is... Divided into four subsets: The FY-3 reflectivity dataset Divided into three sub-data sets ;in, for The subset of reflectivity data belonging to GPS signals extracted from it. for The subset of reflectance data belonging to the BDS signal extracted from it. for The subset of reflectance data extracted from the Galileo signal. for The subset of reflectance data belonging to the GLONASS signal extracted from it. for The subset of reflectivity data belonging to GPS signals extracted from it. for The subset of reflectance data belonging to the BDS signal extracted from it. for The subset of reflectance data belonging to the Galileo signal extracted from it; Draw separately The probability distribution histogram in the image is plotted, and the x-coordinate of the peak point is taken.

[0008] Calculate the systematic deviation of the Global Positioning System relative to the navigation system, including: ; ; ; ; ; In the formula, for The systematic bias in reflectivity between GPS and BDS. for The x-coordinate of the peak point of the histogram of GPS reflectivity probability distribution. for The x-axis of the peak point of the histogram of BDS reflectance probability distribution. for The systematic deviation of reflectivity between GPS and Galileo for The x-axis of the peak point in the histogram of Galileo reflectance probability distribution. for The systematic deviation of reflectivity between GPS and GLONASS for The x-axis of the peak point of the histogram of the probability distribution of reflectance in GLONASS. for The systematic bias in reflectivity between GPS and BDS. for The x-coordinate of the peak point of the histogram of GPS reflectivity probability distribution. for The x-axis of the peak point of the histogram of BDS reflectance probability distribution. for The systematic deviation of reflectivity between GPS and Galileo for The x-axis of the peak point in the histogram of Galileo reflectance probability distribution. Will Each data point in the dataset is superimposed with its corresponding systematic deviation value.

[0009] After bias correction, a new dataset is generated: ; ; ; ; ; In the formula, After systematic bias correction , After systematic bias correction , After systematic bias correction , After systematic bias correction , After systematic bias correction ; Each and By merging the datasets, a unified system dataset is obtained: ; ; In the formula, for The unified reflectance dataset after systematic bias correction and multi-system merging, for A unified reflectance dataset after systematic bias correction and merging of multiple systems.

[0010] In step 3, the standardization of data benchmarks across different constellations within the same system includes: Draw separately The probability distribution histogram in the image is plotted, and the x-coordinate of the peak point is taken. ,calculate Systematic biases relative to the other two zodiac signs: ; ; In the formula, for Compared to Systematic deviation in reflectivity for The x-axis of the peak point in the histogram of reflectance probability distribution. for The x-axis of the peak point in the histogram of reflectance probability distribution. for Compared to Systematic deviation in reflectivity for The x-axis of the peak point in the histogram of reflectance probability distribution; Will Each data point in the dataset is superimposed with its corresponding systematic deviation value. A new dataset is generated after bias correction. : ; ; In the formula, for The reflectance dataset after correction for systematic inter-constellation bias. for Reflectance dataset corrected for systematic inter-constellation bias; Will By merging, we obtain the spaceborne GNSS-R reflectivity dataset for the target region. : .

[0011] In step 4, the satellite-borne GNSS-R reflectivity dataset for the target area is processed. Calibration includes: The grid containing all sampling points is denoted as the sampling grid set, and a single grid is denoted as the sampling grid. Calculate the value within each sampling grid. Data and The arithmetic mean of the data is used as a representative value for reflectance. Using the representative reflectance values ​​of each sampling grid and its adjacent sampling grids, the computer-to-spaceborne reflectance systematic bias set is combined and a probability distribution histogram is plotted to determine the systematic bias correction value. ; Systematic biases based on satellite-airborne GNSS-R data The data is linearly adjusted: ; In the formula, This is a high spatiotemporal resolution reflectance dataset for the target region.

[0012] Furthermore, the calculation is performed within each sampling grid. Data and The reflectance values ​​of the data include: Each individual grid in the sampling grid set is taken as the target grid in turn. For the target grid: List all within this grid data Calculate the grid Reflectivity Representative Value ; List all within this grid data Calculate the grid Reflectivity Representative Value ; In the formula, For the first Airborne GNSS-R reflectivity data, The total number of airborne GNSS-R reflectivity data within the target grid. For the first Spaceborne GNSS-R reflectivity data, The total number of spaceborne GNSS-R reflectivity data within the target grid. For the first Airborne GNSS-R reflectivity dataset within each target grid For the first Spaceborne GNSS-R reflectivity dataset within a target grid.

[0013] Furthermore, the computer-to-spaceborne reflectivity system bias set and system bias correction values ​​include: Each individual grid in the sampling grid set is taken as the target grid. For each target grid, the 10 nearest neighboring sampling grids are identified. ,in, For the first The sampling grid of the first sampling grid For each of the 10 adjacent sampled grids, calculate the average airborne reflectivity difference between the target grid and these 10 adjacent grids: ; Mean of spaceborne reflectivity difference: ; In the formula, For the first Mean of airborne reflectivity neighborhood difference between each sampling grid and its 10 nearest neighbors For the first The mean of the neighborhood difference in starborne reflectivity between each sampling grid and its 10 nearest neighboring grids; calculate As the first Airborne-spaceborne reflectivity system bias for each sampling grid; Based on the bias of airborne-spaceborne reflectivity system Obtain the airborne-spaceborne reflectivity system bias set of the sampling grid. Plot the bias set of the airborne-spaceborne reflectivity system. The probability distribution histogram is used, and the x-axis corresponding to the peak point is taken as the airborne-spaceborne reflectivity system bias correction value. .

[0014] Another object of the present invention is to provide a satellite-airborne GNSS-R high-temporal-space water body detection system, which is used to implement the aforementioned satellite-airborne GNSS-R high-temporal-space water body detection method. The system includes: The data acquisition and gridding module is used to simultaneously acquire multi-source global navigation satellite system reflectivity datasets under the same spatiotemporal conditions in the target area and grid them to obtain the CYGNSS reflectivity dataset. TM-1 reflectivity dataset and FY-3 reflectivity dataset The GNSS-R reflectivity dataset of preset sampling points in the area was acquired simultaneously using drones. ; The constellation-wide data unification module plots probability distribution histograms of data from different systems within the same constellation. Based on the difference in peak points, and using GPS system data as a benchmark, it unifies the data benchmark across different systems within the same constellation, resulting in a unified TM-1 GPS signal dataset. and FY-3 GPS signal dataset ; The constellation data unification module is used to draw probability distribution histograms of data from different constellations within the same system. Based on the difference in peak points, it unifies the data benchmark across different constellations using CYGNSS data as a reference, resulting in a unified TM-1 to CYGNSS calibration dataset. and FY-3 to CYGNSS calibration dataset With the CYGNSS reflectivity dataset By merging, a unified constellation of target area spaceborne GNSS-R reflectivity datasets is obtained. ; The satellite-airborne data calibration module is based on the airborne GNSS-R reflectivity dataset of preset sampling points in the target area. For the target area spaceborne GNSS-R reflectivity dataset Calibration was performed to obtain a high spatiotemporal resolution reflectance dataset for the target region. ; The seasonal threshold determination module is used to select a pure land area and a pure water area within the target region, and extract reflectance datasets for pure land samples and pure water samples under different seasons. A probability distribution histogram is then plotted in the same coordinate system, and the x-coordinate of the intersection of the two curves is taken as the reflectance threshold for the water and land bodies. Threshold Including: Spring water and land reflectance thresholds Summer water and land reflectance thresholds Autumn water and land reflectance thresholds and winter water and land reflectivity thresholds ; Water body detection module, used for reflectivity datasets and threshold Conduct water body detection in the target area.

[0015] Another object of the present invention is to provide a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the satellite-airborne GNSS-R high-space-time water body detection method.

[0016] Another object of the present invention is to provide a computer device, including a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the satellite-airborne GNSS-R high-space-time water body detection method.

[0017] Combining all the above technical solutions, the beneficial effects of this invention are as follows: First, the satellite-airborne GNSS-R high-spatial-temporal water body detection method provided by this invention utilizes satellite and airborne GNSS-R signals to achieve water body detection, improves the inversion accuracy of water body detection, and thus promotes the widespread application of satellite-airborne GNSS-R signal detection of water bodies in water resource monitoring and management and other fields.

[0018] Secondly, this approach comprehensively utilizes multi-source data, multi-system and multi-constellation observation data, making full use of the complementary advantages of different data sources, different navigation systems and different satellite constellations in terms of spatiotemporal coverage, observation geometry and signal characteristics. This enhances the integrity and reliability of surface water detection information, improves the stability and accuracy of water body identification and inversion in complex scenarios, and thus provides technical support for GNSS-R-based water body detection research and its application in water resource monitoring, flood identification and other fields.

[0019] By using airborne GNSS-R observation data to correct spaceborne GNSS-R data, and taking advantage of the high spatial resolution, strong observation flexibility, and good ability to characterize local details of the airborne platform, errors and deviations in spaceborne observations are corrected, thereby improving the quality and consistency of spaceborne GNSS-R data, enhancing the accuracy and reliability of inversion results, and thus improving the ability to detect surface water bodies based on space-aircraft collaborative observation.

[0020] In response to the variation of reflectance characteristics of water and land areas with seasons and time phases, a seasonal and multi-temporal approach is adopted to determine the reflectance discrimination threshold between water and land. This avoids the problem of insufficient adaptability of a uniform fixed threshold under different seasonal conditions, improves the accuracy and stability of water and land differentiation, and thus effectively enhances the accuracy of surface water detection and identification under cross-seasonal and cross-temporal conditions.

[0021] Third, this invention can improve the accuracy of GNSS-R water body detection to the sub-kilometer level and daily update through multi-constellation and multi-system data fusion and airborne calibration. It is suitable for scenarios such as flood emergency monitoring, water resource scheduling, and wetland ecological assessment, which greatly reduces the dependence on SAR and optical remote sensing, reduces special launch costs, and has direct application value for hydrological monitoring service providers, insurance companies, and agricultural enterprises.

[0022] Existing GNSS-R water body detection methods, both domestically and internationally, have not solved the problem of benchmark unification between multiple constellations (CYGNSS / Tianmu-1 / FY-3) and multiple navigation systems (GPS / BDS / Galileo / GLONASS), nor have they established an airborne-satellite collaborative calibration mechanism. This invention is the first to achieve hierarchical benchmark correction and satellite-airborne reflectivity consistency fusion of data from four systems and three constellations, filling the technological gap in multi-source GNSS-R collaborative water body detection.

[0023] Systematic biases caused by differences in constellation configuration, signal system, and receiver characteristics in multi-source GNSS-R data have long hindered their fusion and application. The accuracy of spaceborne data is limited by orbital altitude and revisit period, lacking independent truth calibration methods. Fixed thresholds cannot adapt to false alarms and missed alarms caused by seasonal changes. This invention systematically solves these problems through a four-level processing chain: intra-system unification, inter-constellation unification, airborne calibration, and seasonally adaptive thresholds.

[0024] Existing technologies generally rely on a single constellation (such as CYGNSS) or a single system (GPS), resulting in uncontrollable quality of multi-source data and a technical bias where data fusion may actually reduce accuracy. Furthermore, GNSS-R is often considered only suitable for large-scale monitoring, with airborne platforms used only for algorithm verification rather than data calibration. This invention overcomes these biases by demonstrating through probability distribution peak deviation correction that multi-source data can be accurately normalized, and by using high-resolution airborne data to inversely constrain spaceborne products, thus expanding the application scenarios and accuracy boundaries of GNSS-R technology. Attached Figure Description

[0025] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the disclosure of this invention and, together with the description, serve to explain the principles of the disclosure of this invention. Figure 1 This is a flowchart of the space-airborne GNSS-R high-spatial-space water body detection method provided in the embodiments of the present invention; Figure 2 This is a diagram illustrating the specific implementation process of the space-airborne GNSS-R high-temporal-space water body detection method provided in this embodiment of the invention. Detailed Implementation

[0026] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Many specific details are set forth in the following description to provide a thorough understanding of the present invention. However, the present invention can be practiced in many other ways different from those described herein, and those skilled in the art can make similar modifications without departing from the spirit of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0027] The innovation of this invention lies in: (1) Multi-constellation and multi-navigation system hierarchical benchmark unification mechanism: For the first time, a two-level correction architecture is established, which unifies multiple systems within the same constellation and unifies across constellations. The BDS / Galileo / GLONASS data of Tianmu-1 / FY-3 are corrected with GPS as the benchmark, and then the reflectivity benchmark of the three constellations is unified with CYGNSS as the benchmark. This solves the systematic deviation problem caused by the differences in signal system and receiver characteristics of multi-source GNSS-R data, and realizes the physical consistency fusion of data from four systems and three constellations. (2) Airborne-Spaceborne Collaborative Calibration Method: Using high-resolution observations from UAV-borne GNSS-R as a true reference, systematic bias compensation is performed on the spaceborne fused dataset through statistical analysis of sampling grid difference deviations and peak correction of probability distribution. This method overcomes the limitations of existing technologies that rely solely on internal cross-validation within the spaceborne system or indirect comparison using low-resolution microwave radiometers, thereby improving the absolute accuracy of spaceborne reflectivity. (3) Seasonal adaptive water and land threshold determination strategy: In view of the shift in water and land reflectivity characteristics caused by vegetation phenology, sediment content and freezing status in different seasons, probability distribution histograms of pure land and pure water samples are extracted respectively, and the thresholds for spring / summer / autumn / winter are dynamically determined by the intersection of the two curves. This overcomes the problem of false alarms and missed alarms in cross-seasonal scenarios with fixed thresholds throughout the year, and improves the temporal stability of water body detection.

[0028] Example 1, such as Figure 1 , Figure 2 As shown, the satellite-airborne GNSS-R high-spatial-temporal water body detection method provided in this embodiment of the invention includes the following steps: S1. Simultaneously acquire multi-source global navigation satellite system reflectivity datasets under the same spatiotemporal conditions in the target area and grid them to obtain the CYGNSS reflectivity dataset. TM-1 reflectivity dataset and FY-3 reflectivity dataset The GNSS-R reflectivity dataset of preset sampling points in the area was acquired simultaneously using drones. ; In this embodiment, the tolerance range for the same spatiotemporal conditions is: time difference ≤ 6 hours, and the spatial grid resolution is uniformly 0.1°×0.1°; the preset sampling points are determined by stratified sampling based on the topographic information and surface cover type of the target area.

[0029] S2. Draw probability distribution histograms for data from different systems within the same constellation. Based on the difference in peak points, and using GPS system data as a benchmark, unify the data benchmark across different systems within the same constellation to obtain the unified benchmark TM-1 GPS signal dataset. and FY-3 GPS signal dataset ; In this embodiment, "same constellation" refers to the same satellite constellation (such as CYGNSS constellation, TM-1 constellation, FY-3 constellation); "different systems" refers to different GNSS navigation systems (GPS, BDS, Galileo, GLONASS).

[0030] S2.1 Based on the differences in the navigation systems to which the signals belong, the TM-1 reflectivity dataset is... Divided into four subsets: The FY-3 reflectivity dataset Divided into three sub-data sets ;in, for The subset of reflectivity data belonging to GPS signals extracted from it. for The subset of reflectance data belonging to the BDS signal extracted from it. for The subset of reflectance data extracted from the Galileo signal. for The subset of reflectance data belonging to the GLONASS signal extracted from it. for The subset of reflectivity data belonging to GPS signals extracted from it. for The subset of reflectance data belonging to the BDS signal extracted from it. for The subset of reflectance data belonging to the Galileo signal extracted from it; S2.2, draw them separately The probability distribution histogram in the image is plotted, and the x-coordinate of the peak point is taken. Calculate the systematic deviation of the Global Positioning System relative to the navigation system, including: ; ; ; ; ; In the formula, for The systematic bias in reflectivity between GPS and BDS. for The x-coordinate of the peak point of the histogram of GPS reflectivity probability distribution. for The x-axis of the peak point of the histogram of BDS reflectance probability distribution. for The systematic deviation of reflectivity between GPS and Galileo for The x-axis of the peak point in the histogram of Galileo reflectance probability distribution. for The systematic deviation of reflectivity between GPS and GLONASS for The x-axis of the peak point of the histogram of the probability distribution of reflectance in GLONASS. for The systematic bias in reflectivity between GPS and BDS. for The x-coordinate of the peak point of the histogram of GPS reflectivity probability distribution. for The x-axis of the peak point of the histogram of BDS reflectance probability distribution. for The systematic deviation of reflectivity between GPS and Galileo for The x-axis of the peak point in the histogram of Galileo reflectance probability distribution. In this embodiment, to address the problem that the peak points of the histogram cannot be directly extracted to calculate the systematic bias in data with multiple peaks or noise, the Freedman-Diaconis rule is first used to automatically determine the optimal bin width and construct a histogram of reflectance differences. Then, Gaussian kernel density estimation (KDE) is used for smoothing, and the bandwidth is selected by the Silverman empirical rule. For unimodal distributions, the global maximum value of the kernel density is directly taken as the peak value. For multimodal distributions, a Gaussian mixture model (GMM) is used for fitting, and the optimal number of components is automatically determined by the Bayesian information criterion (BIC). The mean of the component with the largest weight is taken as the representative value of the systematic bias.

[0031] S2.3, will Each data point in the dataset is superimposed with its corresponding systematic deviation value. After bias correction, a new dataset is generated: ; ; ; ; ; In the formula, After systematic bias correction , After systematic bias correction , After systematic bias correction , After systematic bias correction , After systematic bias correction ; S2.4, respectively and By merging the datasets, a unified system dataset is obtained: ; ; In the formula, for The unified reflectance dataset after systematic bias correction and multi-system merging, for A unified reflectance dataset after systematic bias correction and merging of multiple systems.

[0032] S3. Draw probability distribution histograms for different constellation data after the same system is established. Based on the difference in peak points, use CYGNSS data as a benchmark to unify the data benchmark for different constellations, and obtain the TM-1 to CYGNSS correction dataset with a unified benchmark. and FY-3 to CYGNSS calibration dataset With the CYGNSS reflectivity dataset By merging, a unified constellation of target area spaceborne GNSS-R reflectivity datasets is obtained. ; In this embodiment, "same system" refers to the same GNSS navigation system (such as GPS); "different constellations" refers to different satellite constellations (such as CYGNSS and TM-1); CYGNSS is used as the benchmark because it has the best data continuity, the highest time sampling density, and the highest international recognition. When the study area is located in high latitudes or the number of effective samples in the CYGNSS grid is less than 10, the system will automatically switch to TM-1 as the benchmark.

[0033] S3.1, draw them separately The probability distribution histogram in the image is plotted, and the x-coordinate of the peak point is taken. ,calculate Systematic biases relative to the other two zodiac signs: ; ; In the formula, for Compared to Systematic deviation in reflectivity for The x-axis of the peak point in the histogram of reflectance probability distribution. for The x-axis of the peak point in the histogram of reflectance probability distribution. for Compared to Systematic deviation in reflectivity for The x-axis of the peak point in the histogram of reflectance probability distribution; S3.2, will Each data point in the dataset is superimposed with its corresponding systematic deviation value. A new dataset is generated after bias correction. : ; ; In the formula, for The reflectance dataset after correction for systematic inter-constellation bias. for Reflectance dataset corrected for systematic inter-constellation bias; S3.3, will By merging, we obtain the spaceborne GNSS-R reflectivity dataset for the target region. : .

[0034] S4. Airborne GNSS-R reflectivity dataset based on preset sampling points in the target area For the target area spaceborne GNSS-R reflectivity dataset Calibration was performed to obtain a high spatiotemporal resolution reflectance dataset for the target region. ; In this embodiment, at least 30 sampling points are selected for each land surface type in the stratified sampling, with a sampling density of no less than 1 point / km². 2 The minimum spacing between sampling points is no less than 100m, and three repeated observations are performed within each point and the average is taken to ensure spatial independence and statistical significance.

[0035] S4.1. Denote the grid containing all sampling points as the sampling grid set, and denote each individual grid in the set as a sampling grid. Then, sequentially denote each individual grid in the sampling grid set as the target grid. For the target grid: List all within this grid data Calculate the grid Reflectivity Representative Value ; List all within this grid data Calculate the grid Reflectivity Representative Value ; In the formula, For the first Airborne GNSS-R reflectivity data, The total number of airborne GNSS-R reflectivity data within the target grid. For the first Spaceborne GNSS-R reflectivity data, The total number of spaceborne GNSS-R reflectivity data within the target grid. For the first Airborne GNSS-R reflectivity dataset within each target grid For the first The satellite-borne GNSS-R reflectivity dataset within the target grid. In this embodiment, if there is no [database] within the target grid... If there is no data in the target grid, skip that grid; if there is no data in the target grid, skip that grid. The data is supplemented using nearest neighbor interpolation.

[0036] S4.2. Sequentially take each individual grid in the sampling grid set as the target grid. For the target grid: Find the 10 nearest neighboring sampled grids to the target grid. ,in, For the first The sampling grid of the first sampling grid For each of the 10 adjacent sampled grids, calculate the average airborne reflectivity difference between the target grid and these 10 adjacent grids: ; Mean of spaceborne reflectivity difference: ; In the formula, For the first Mean of airborne reflectivity neighborhood difference between each sampling grid and its 10 nearest neighbors For the first The mean of the neighborhood difference in starborne reflectivity between each sampling grid and its 10 nearest neighboring grids; calculate As the first Airborne-spaceborne reflectivity system bias for each sampling grid; Based on the bias of airborne-spaceborne reflectivity system Obtain the airborne-spaceborne reflectivity system bias set of the sampling grid. Plot the bias set of the airborne-spaceborne reflectivity system. The probability distribution histogram is used, and the x-axis corresponding to the peak point is taken as the airborne-spaceborne reflectivity system bias correction value. .

[0037] In this embodiment, if there are fewer than 10 adjacent grids, all available adjacent grids are used; if the area of ​​the study region exceeds 10,000 square kilometers, the DEV is calculated by partition, where DEV is the dimensionless reflectance difference, which is usually between -3dB and +3dB. S4.3 Draw a probability distribution histogram to determine the system bias correction value. ; Systematic biases based on satellite-airborne GNSS-R data The data is linearly adjusted: ; In the formula, This is a high spatiotemporal resolution reflectance dataset for the target region.

[0038] S5. Select a pure land area and a pure water area within the target region, and extract reflectance datasets for the pure land area and the pure water area under different seasons; plot a probability distribution histogram in the same coordinate system, and take the x-coordinate of the intersection of the two curves as the reflectance threshold for the water and land areas. Threshold Including: Spring water and land reflectance thresholds Summer water and land reflectance thresholds Autumn water and land reflectance thresholds and winter water and land reflectivity thresholds ; S6, Based on reflectivity dataset and threshold Conduct water body detection in the target area.

[0039] Example 2: The space-to-airborne GNSS-R high-temporal-space water body detection system provided in this embodiment of the invention includes: The data acquisition and gridding module is used to simultaneously acquire multi-source global navigation satellite system reflectivity datasets under the same spatiotemporal conditions in the target area and grid them to obtain the CYGNSS reflectivity dataset. TM-1 reflectivity dataset and FY-3 reflectivity dataset The GNSS-R reflectivity dataset of preset sampling points in the area was acquired simultaneously using drones. ; The constellation-wide data unification module plots probability distribution histograms of data from different systems within the same constellation. Based on the difference in peak points, and using GPS system data as a benchmark, it unifies the data benchmark across different systems within the same constellation, resulting in a unified TM-1 GPS signal dataset. and FY-3 GPS signal dataset ; The constellation data unification module is used to draw probability distribution histograms of data from different constellations within the same system. Based on the difference in peak points, it unifies the data benchmark across different constellations using CYGNSS data as a reference, resulting in a unified TM-1 to CYGNSS calibration dataset. and FY-3 to CYGNSS calibration dataset With the CYGNSS reflectivity dataset By merging, a unified constellation of target area spaceborne GNSS-R reflectivity datasets is obtained. ; The satellite-airborne data calibration module is based on the airborne GNSS-R reflectivity dataset of preset sampling points in the target area. For the target area spaceborne GNSS-R reflectivity dataset Calibration was performed to obtain a high spatiotemporal resolution reflectance dataset for the target region. ; The seasonal threshold determination module is used to select a pure land area and a pure water area within the target region, and extract reflectance datasets for pure land samples and pure water samples under different seasons. A probability distribution histogram is then plotted in the same coordinate system, and the x-coordinate of the intersection of the two curves is taken as the reflectance threshold for the water and land bodies. Threshold Including: Spring water and land reflectance thresholds Summer water and land reflectance thresholds Autumn water and land reflectance thresholds and winter water and land reflectivity thresholds ; Water body detection module, used for reflectivity datasets and threshold Conduct water body detection in the target area.

[0040] In the above embodiments, the descriptions of each embodiment have different focuses. For parts that are not described in detail or recorded in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0041] Compared with existing technologies, this invention achieves physical consistency fusion of data from multiple constellations and navigation systems, effectively eliminating systematic biases; utilizes airborne high-resolution observations to achieve precise calibration of satellite-borne data, significantly improving reflectivity inversion accuracy; adopts a seasonal adaptive threshold strategy, effectively improving the stability and accuracy of cross-seasonal water body detection; and constructs a satellite-airborne collaborative high spatiotemporal resolution water body detection system, significantly enhancing its application capabilities in complex scenarios.

[0042] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by those skilled in the art within the scope of the technology disclosed in the present invention and within the spirit and principles of the present invention should be covered within the scope of protection of the present invention.

Claims

1. A satellite-airborne GNSS-R method for high-altitude spatiotemporal water body detection, characterized in that, The method includes the following steps: S1. Simultaneously acquire multi-source global navigation satellite system reflectivity datasets under the same spatiotemporal conditions in the target area and grid them to obtain the CYGNSS reflectivity dataset. TM-1 reflectivity dataset and FY-3 reflectivity dataset The GNSS-R reflectivity dataset of preset sampling points in the area was acquired simultaneously using drones. ; S2. Draw probability distribution histograms for data from different systems within the same constellation. Based on the difference in peak points, and using GPS system data as a benchmark, unify the data benchmark across different systems within the same constellation to obtain the unified benchmark TM-1 GPS signal dataset. and FY-3 GPS signal dataset ; S3. Draw probability distribution histograms for different constellation data after the same system is established. Based on the difference in peak points, use CYGNSS data as a benchmark to unify the data benchmark for different constellations, and obtain the TM-1 to CYGNSS correction dataset after unification. and FY-3 to CYGNSS calibration dataset With the CYGNSS reflectivity dataset By merging, a unified constellation of target area spaceborne GNSS-R reflectivity datasets is obtained. ; S4. Airborne GNSS-R reflectivity dataset based on preset sampling points in the target area For the target area spaceborne GNSS-R reflectivity dataset Calibration was performed to obtain a high spatiotemporal resolution reflectance dataset for the target region. ; S5. Select a pure land area and a pure water area within the target region, and extract reflectance datasets for the pure land area and the pure water area under different seasons; plot a probability distribution histogram in the same coordinate system, and take the x-coordinate of the intersection of the two curves as the reflectance threshold for the water and land areas. Threshold Including: Spring water and land reflectance thresholds Summer water and land reflectance thresholds Autumn water and land reflectance thresholds and winter water and land reflectivity thresholds ; S6, Based on reflectivity dataset and threshold Conduct water body detection in the target area.

2. The satellite-airborne GNSS-R high-temporal-spatial water body detection method according to claim 1, characterized in that, In step 1, the preset sampling points are determined by stratified sampling based on the topographic information and land cover type of the target area.

3. The satellite-airborne GNSS-R high-temporal-spatial water body detection method according to claim 1, characterized in that, In step 2, the standardization of data benchmarks across different systems within the same constellation includes: Based on the differences in the navigation systems to which the signals belong, the TM-1 reflectivity dataset is... Divided into four subsets: The FY-3 reflectivity dataset Divided into three sub-data sets ;in, for The subset of reflectivity data belonging to GPS signals extracted from it. for The subset of reflectance data belonging to the BDS signal extracted from it. for The subset of reflectance data extracted from the Galileo signal. for The subset of reflectance data belonging to the GLONASS signal extracted from it. for The subset of reflectivity data belonging to GPS signals extracted from it. for The subset of reflectance data belonging to the BDS signal extracted from it. for The subset of reflectance data belonging to the Galileo signal extracted from it; Draw separately The probability distribution histogram in the image is plotted, and the x-coordinate of the peak point is taken. Calculate the systematic deviation of the Global Positioning System relative to the navigation system, including: ； ； ； ； ； In the formula, for The systematic bias in reflectivity between GPS and BDS. for The x-coordinate of the peak point of the histogram of GPS reflectivity probability distribution. for The x-axis of the peak point of the histogram of BDS reflectance probability distribution. for The systematic deviation of reflectivity between GPS and Galileo for The x-axis of the peak point in the histogram of Galileo reflectance probability distribution. for The systematic deviation of reflectivity between GPS and GLONASS for The x-axis of the peak point of the histogram of the probability distribution of reflectance in GLONASS. for The systematic bias in reflectivity between GPS and BDS. for The x-coordinate of the peak point of the histogram of GPS reflectivity probability distribution. for The x-axis of the peak point of the histogram of BDS reflectance probability distribution. for The systematic deviation of reflectivity between GPS and Galileo for The x-axis of the peak point in the histogram of Galileo reflectance probability distribution. Will Each data point in the dataset is superimposed with its corresponding systematic deviation value. After bias correction, a new dataset is generated: ； ； ； ； ； In the formula, After systematic bias correction , After systematic bias correction , After systematic bias correction , After systematic bias correction , After systematic bias correction ; Each and By merging the datasets, a unified system dataset is obtained: ； ； In the formula, for The unified reflectance dataset after systematic bias correction and multi-system merging, for A unified reflectance dataset after systematic bias correction and merging of multiple systems.

4. The satellite-airborne GNSS-R high-temporal-spatial water body detection method according to claim 1, characterized in that, Step 3, unifying the data benchmark across different constellations within the same system, includes: Draw separately The probability distribution histogram in the image is plotted, and the x-coordinate of the peak point is taken. ,calculate Systematic biases relative to the other two zodiac signs: ； ； In the formula, for Compared to Systematic deviation in reflectivity for The x-axis of the peak point in the histogram of reflectance probability distribution. for The x-axis of the peak point in the histogram of reflectance probability distribution. for Compared to Systematic deviation in reflectivity for The x-axis of the peak point in the histogram of reflectance probability distribution; Will Each data point in the dataset is superimposed with its corresponding systematic deviation value. A new dataset is generated after bias correction. : ； ； In the formula, for The reflectance dataset after correction for systematic inter-constellation bias. for Reflectance dataset corrected for systematic inter-constellation bias; Will By merging, we obtain the spaceborne GNSS-R reflectivity dataset for the target region. : 。 5. The satellite-airborne GNSS-R high-temporal-spatial water body detection method according to claim 4, characterized in that, In step 4, the satellite-borne GNSS-R reflectivity dataset for the target area is processed. Calibration includes: The grid containing all sampling points is denoted as the sampling grid set, and a single grid is denoted as the sampling grid. Calculate the value within each sampling grid. Data and The arithmetic mean of the data is used as a representative value for reflectance. Using the representative reflectance values ​​of each sampling grid and its adjacent sampling grids, the computer-to-spaceborne reflectance systematic bias set is combined and a probability distribution histogram is plotted to determine the systematic bias correction value. ; Systematic biases based on satellite-airborne GNSS-R data The data is linearly adjusted: ； In the formula, This is a high spatiotemporal resolution reflectance dataset for the target region.

6. The satellite-airborne GNSS-R high-temporal-spatial water body detection method according to claim 5, characterized in that, Calculate within each sampling grid Data and The reflectance values ​​of the data include: Each individual grid in the sampling grid set is taken as the target grid in turn. For the target grid: List all within this grid data Calculate the grid Reflectivity Representative Value ; List all within this grid data Calculate the grid Reflectivity Representative Value ; In the formula, For the first Airborne GNSS-R reflectivity data, The total number of airborne GNSS-R reflectivity data within the target grid. For the first Spaceborne GNSS-R reflectivity data, The total number of spaceborne GNSS-R reflectivity data within the target grid. For the first Airborne GNSS-R reflectivity dataset within each target grid For the first Spaceborne GNSS-R reflectivity dataset within a target grid.

7. The satellite-airborne GNSS-R high-temporal-spatial water body detection method according to claim 5, characterized in that, The computer-to-spaceborne reflectivity system bias set and system bias correction values ​​include: Each individual grid in the sampling grid set is taken as the target grid. For each target grid, the 10 nearest neighboring sampling grids are identified. ,in, For the first The sampling grid of the first sampling grid For each of the 10 adjacent sampled grids, calculate the average airborne reflectivity difference between the target grid and these 10 adjacent grids: ； Mean of spaceborne reflectivity difference: ； In the formula, For the first Mean of airborne reflectivity neighborhood difference between each sampling grid and its 10 nearest neighbors For the first The mean of the neighborhood difference in starborne reflectivity between each sampling grid and its 10 nearest neighboring grids; calculate As the first Airborne-spaceborne reflectivity system bias for each sampling grid; Based on the bias of airborne-spaceborne reflectivity system Obtain the airborne-spaceborne reflectivity system bias set of the sampling grid. Plot the bias set of the airborne-spaceborne reflectivity system. The probability distribution histogram is used, and the x-axis corresponding to the peak point is taken as the airborne-spaceborne reflectivity system bias correction value. .

8. A satellite-airborne GNSS-R high-temporal-spatial water body detection system, characterized in that, This system is used to implement the satellite-airborne GNSS-R high-temporal-space water body detection method as described in any one of claims 1 to 7, and the system comprises: The data acquisition and gridding module is used to simultaneously acquire multi-source global navigation satellite system reflectivity datasets under the same spatiotemporal conditions in the target area and grid them to obtain the CYGNSS reflectivity dataset. TM-1 reflectivity dataset and FY-3 reflectivity dataset The GNSS-R reflectivity dataset of preset sampling points in the area was acquired simultaneously using drones. ; The constellation-wide data unification module plots probability distribution histograms of data from different systems within the same constellation. Based on the difference in peak points, and using GPS system data as a benchmark, it unifies the data benchmark across different systems within the same constellation, resulting in a unified TM-1 GPS signal dataset. and FY-3 GPS signal dataset ; The constellation data unification module is used to draw probability distribution histograms of data from different constellations within the same system. Based on the difference in peak points, it unifies the data benchmark across different constellations using CYGNSS data as a reference, resulting in a unified TM-1 to CYGNSS calibration dataset. and FY-3 to CYGNSS calibration dataset With the CYGNSS reflectivity dataset By merging, a unified constellation of target area spaceborne GNSS-R reflectivity datasets is obtained. ; The satellite-airborne data calibration module is based on the airborne GNSS-R reflectivity dataset of preset sampling points in the target area. For the target area spaceborne GNSS-R reflectivity dataset Calibration was performed to obtain a high spatiotemporal resolution reflectance dataset for the target region. ; The seasonal threshold determination module is used to select a pure land area and a pure water area within the target region, and extract reflectance datasets for pure land samples and pure water samples under different seasons. A probability distribution histogram is then plotted in the same coordinate system, and the x-coordinate of the intersection of the two curves is taken as the reflectance threshold for the water and land bodies. Threshold Including: Spring water and land reflectance thresholds Summer water and land reflectance thresholds Autumn water and land reflectance thresholds and winter water and land reflectivity thresholds ; Water body detection module, used for reflectivity datasets and threshold Conduct water body detection in the target area.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the space-to-airborne GNSS-R high-space-time water body detection method as described in any one of claims 1 to 7.

10. A computer device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the space-to-airborne GNSS-R high-temporal-space water body detection method as described in any one of claims 1 to 7.

Images