Method and device for cooperatively monitoring sea level based on GNSS and InSAR, and storage medium

Abstract

The invention provides a method and device for cooperatively monitoring the sea level based on GNSS and InSAR, and a storage medium, and relates to the technical field of sea level monitoring. The method comprises the following steps: acquiring point data representing land vertical motion from a GNSS base station which is co-located with a tide station along a coastline, acquiring InSAR surface data representing earth surface deformation from a satellite, acquiring relative sea level change parameters, converting GNSS and InSAR into a same coordinate system, fusing the GNSS and InSAR to obtain land vertical motion parameters, and separating absolute sea level change parameters; and constructing and optimizing an earth crust motion-sea level coupling numerical model based on prior knowledge and the like, and predicting multi-source observation data to obtain a sea level change result. Multi-source data are fused, the influence of land motion is separated, the monitoring precision and coverage are improved, and the influence of natural and human activities on the sea level can be quantified.

Description

Technical Field

[0001] The present invention mainly relates to the technical field of sea level monitoring, and in particular to a method, device and storage medium for collaborative sea level monitoring based on GNSS and InSAR. Background Art

[0002] In traditional sea level monitoring, tide gauges, due to their sparse distribution, can only capture relative sea level change data and are unable to isolate the impact of vertical land motion. GNSS can accurately capture discrete point data on vertical land motion, but its spatial coverage is insufficient. InSAR can capture planar surface deformation fields, but is affected by factors such as nearshore tides and the atmosphere, resulting in low data reliability. Satellite altimeters fail to capture data in nearshore areas and have insufficient spatial resolution. Furthermore, existing technologies lack the ability to coordinate and integrate multi-source data, resulting in low monitoring accuracy and insufficient spatial coverage. This makes it difficult to quantify the specific contributions of natural factors and human activities to sea level change, and presents a contradiction between the high cost of high-precision technology and the unreliability of low-cost solutions. Summary of the Invention

[0003] The technical problem to be solved by the present invention is to address the deficiencies of the existing technology and provide a method, device and storage medium for collaborative sea level monitoring based on GNSS and InSAR.

[0004] The present invention solves the above-mentioned technical problem with the following technical solution: A method for collaboratively monitoring sea level based on GNSS and InSAR, comprising the following steps: Acquire GNSS point data representing vertical motion of land from GNSS reference stations pre-deployed along the coastline and co-located with tide gauges, acquire InSAR surface data representing surface deformation from Sentinel-1 or GF-3 satellites, and acquire relative sea level change parameters relative to land from the HY-2 satellite. Converting the GNSS point data and the InSAR surface data to the same coordinate system, and performing data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system to obtain a land vertical motion parameter, and separating an absolute sea level change parameter from the relative sea level change parameter and the land vertical motion parameter; constructing a crustal movement-sea level coupling numerical model based on absolute sea level change parameters, GNSS point data, and InSAR surface data, and optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model; Based on the optimized crustal movement-sea level coupling numerical model, the multi-source observation data of the sea level to be monitored are predicted to obtain the sea level change prediction results.

[0005] Another technical solution of the present invention to solve the above technical problems is as follows: a device for collaborative sea level monitoring based on GNSS and InSAR, comprising: An initial data acquisition module is used to obtain GNSS point data representing vertical motion of land from GNSS reference stations pre-deployed along the coastline and co-located with tide gauge stations, and to obtain InSAR surface data representing surface deformation from Sentinel-1 or GF-3 satellites, as well as relative sea level change parameters relative to land from the HY-2 satellite. a data preprocessing module, configured to convert the GNSS point data and the InSAR surface data into the same coordinate system, perform data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system, obtain land vertical motion parameters, and separate absolute sea level change parameters from the relative sea level change parameters and the land vertical motion parameters; a model building module for building a crustal movement-sea level coupling numerical model based on absolute sea level change parameters, GNSS point data, and InSAR surface data, and optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model; The prediction module is used to predict the multi-source observation data of the sea level to be monitored based on the optimized crustal movement-sea level coupling numerical model to obtain the sea level change prediction results.

[0006] Another technical solution of the present invention to solve the above-mentioned technical problem is as follows: A device for collaborative sea level monitoring based on GNSS and InSAR, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for collaborative sea level monitoring based on GNSS and InSAR as described above is implemented.

[0007] Another technical solution of the present invention to solve the above technical problems is as follows: a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the method for collaborative sea level monitoring based on GNSS and InSAR as described above.

[0008] The beneficial effects of the present invention are as follows: by acquiring GNSS point data, InSAR surface data and relative sea level change parameters, the land vertical motion parameters are obtained through coordinate unification and data fusion, the absolute sea level change parameters are separated, and a coupled numerical model is constructed and optimized to realize sea level change prediction. The invention can fuse multi-source data to solve problems such as sparse traditional tide gauges, insufficient GNSS coverage, and low InSAR reliability; separate the impact of land vertical motion and obtain absolute sea level change parameters; improve monitoring accuracy and spatial coverage through model prediction, and quantify the impact of natural and human activities on sea level. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 A flow chart of a method for collaborative sea level monitoring based on GNSS and InSAR provided in an embodiment of the present invention; Figure 2 This is a module block diagram of a method for collaborative sea level monitoring based on GNSS and InSAR provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0010] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only used to explain the present invention and are not used to limit the scope of the present invention.

[0011] Example 1: Figure 1 As shown, an embodiment of the present invention provides a method for collaboratively monitoring sea level based on GNSS and InSAR, comprising the following steps: S1. Acquire GNSS point data representing vertical motion of land from a GNSS reference station pre-deployed along the coastline and co-located with a tide gauge station, acquire InSAR surface data representing surface deformation from Sentinel-1 or GF-3 satellites, and acquire relative sea level change parameters relative to land from the HY-2 satellite. S2. Converting the GNSS point data and the InSAR surface data to the same coordinate system, and performing data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system to obtain a land vertical motion parameter, and separating an absolute sea level change parameter from the relative sea level change parameter and the land vertical motion parameter; S3. constructing a crustal movement-sea level coupling numerical model based on absolute sea level change parameters, GNSS point data, and InSAR surface data, and optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model; S4. Based on the optimized crustal movement-sea level coupling numerical model, the multi-source observation data of the sea level to be monitored are predicted to obtain the sea level change prediction results.

[0012] It should be understood that the Sentinel-1 satellite is an earth observation satellite, which is composed of two satellites. Each satellite carries an advanced radar instrument, which carries a C-band synthetic aperture radar (SAR). It can provide images of the earth's surface during the day, night and various weather conditions. By radar interferometry technology, by comparing the radar signals sent to the earth and reflected from the earth, the changes of the planet surface can be tracked, and the InSAR surface data characterizing the earth surface deformation can be obtained, which is widely used in the fields of geological disaster monitoring and earth surface deformation analysis, and is used to obtain InSAR surface data in the sea level monitoring in the embodiment, and assist in analyzing the land vertical motion.

[0013] The GF-3 satellite, i.e. the high-resolution three satellite, is a C-band multi-polarization synthetic aperture radar (SAR) satellite with a resolution of 1 meter. It has 12 imaging modes, covering strip, scanning, and beam, etc. Various modes, and the range of the width is large, which can meet the needs of different application scenarios. In the field of marine monitoring, it can effectively monitor the marine dynamic environment, marine disasters (such as storm surge, sea wave, etc.), marine pollution, etc.; in land application, it can be used for geological structure research, urban construction monitoring, etc. In sea level monitoring, similar to the Sentinel-1 satellite, it is also an important source of InSAR surface data, which provides the basis for analyzing the land vertical motion and the sea level change.

[0014] The HY-2 satellite is mainly used for monitoring marine dynamic environment elements such as sea surface height, significant wave height, sea surface wind field, etc. In the field of sea level monitoring, it can obtain relative sea level change parameters of the sea level relative to the land. These parameters are crucial for studying the dynamic change of the sea level and analyzing the influence of ocean circulation and climate change on the sea level, and provide key data support for subsequent separation of absolute sea level change parameters, which is an important part of the sea level monitoring data acquisition in the embodiment.

[0015] In the embodiment, by obtaining GNSS point data, InSAR surface data and relative sea level change parameters, land vertical motion parameters are obtained through coordinate unification and data fusion, absolute sea level change parameters are separated, coupled numerical models are constructed and optimized, sea level change prediction is realized, multi-source data (GNSS, InSAR, satellite altimetry) can be fused, the problems of traditional tide gauge sparsity, GNSS coverage deficiency and InSAR low reliability can be solved; the influence of land vertical motion is separated to obtain absolute sea level change parameters; the monitoring accuracy and spatial coverage are improved through model prediction, and the influence of natural and human activities on the sea level can be quantified.

[0016] Preferably, in step S2, the GNSS point data and the InSAR surface data are converted to the same coordinate system, comprising: Performing coordinate analysis on the GNSS point data and the InSAR surface data using the SNAP tool to obtain GNSS geographic coordinates and InSAR geographic coordinates; Based on the selected WGS84 as the unified benchmark, the resolved GNSS geographic coordinates and InSAR geographic coordinates are unified into geographic coordinates, and the unified GNSS geographic coordinates and InSAR geographic coordinates are transformed into planar projections using GDAL and OGR tools. The converted GNSS geographic coordinates and InSAR geographic coordinates are corrected for elevation consistency using the GMT tool, thereby completing the conversion to the same coordinate system.

[0017] It should be understood that the SNAP tool, short for Sentinel-1 Toolbox, is a professional software tool developed by the European Space Agency (ESA) for processing Sentinel-1 satellite data. The SNAP tool is used to process and analyze synthetic aperture radar (SAR) data acquired by the Sentinel-1 satellites, enabling operations such as radar data reading, calibration, interferometry, and geocoding. In this embodiment, the tool performs coordinate analysis on GNSS point data and InSAR surface data to obtain corresponding geographic coordinates, laying the foundation for subsequent data processing.

[0018] The GDAL tool, short for Geospatial Data Abstraction Library, is an open source library for reading, writing, converting, and processing raster and vector geographic data. The GDAL tool supports reading, writing, and converting a variety of geographic data formats, and has powerful spatial data processing capabilities. It can perform operations such as projection conversion, data resampling, and format conversion. In this embodiment, the parsed GNSS geographic coordinates and InSAR geographic coordinates are unified based on the selected WGS84 as a unified benchmark. The unified geographic coordinates are then subjected to a planar projection conversion using this tool, making the data easier to analyze and apply in a planar coordinate system.

[0019] The OGR tool, short for OGR Simple Features Library, is the component of the GDAL library used to process vector geographic data. OGR primarily handles vector geographic data, supporting the reading, writing, editing, and analysis of vector data, and can handle a variety of vector features, including points, lines, and surfaces. In this example, it is used in conjunction with GDAL to perform related processing tasks, such as planar projection conversion of GNSS and InSAR data, ensuring data consistency and accuracy within the same coordinate system.

[0020] GMT, short for Generic Mapping Tools, is primarily used for processing and mapping geospatial data and is widely used in fields such as earth science, oceanography, and geology. It offers a variety of data processing, mapping, and visualization capabilities, enabling it to process a wide range of geographic data, including coordinate conversion, data interpolation, and grid generation. It can also produce a variety of high-quality maps, profiles, and statistical charts, supporting a wide range of drawing styles and custom options, allowing users to flexibly set colors, legends, annotations, and other elements as needed.

[0021] In this embodiment, in the method for collaborative sea level monitoring based on GNSS and InSAR, the GMT tool is used to perform elevation consistency correction on the converted GNSS geographic coordinates and InSAR geographic coordinates to complete the conversion to the same coordinate system, ensure the uniformity of multi-source data in the elevation benchmark, and provide an accurate basis for subsequent data fusion and analysis.

[0022] In this embodiment, the SNAP tool is used to parse coordinates, the geographic coordinates are unified based on WGS84, the plane projection conversion is performed through GDAL and OGR, and the elevation consistency correction is completed with the help of the GMT tool; this ensures that multi-source data are unified under the same coordinate system, eliminates errors caused by coordinate differences, provides an accurate benchmark for subsequent data fusion, and improves data compatibility and processing efficiency.

[0023] Preferably, in step S2, data fusion processing is performed on the GNSS geographic coordinates and the InSAR geographic coordinates of the same coordinate system to obtain the land vertical motion parameter, and the absolute sea level change parameter is separated from the relative sea level change parameter and the land vertical motion parameter, including: The GNSS geographic coordinates and InSAR geographic coordinates of the same coordinate system are expressed as experimental semivariograms, which are: Among them, h is the spatial lag distance, N(h) is the number of data pairs in the distance interval, Z(x i ) is the GNSS geographic coordinate or InSAR geographic coordinate corresponding to the position x i Observed values ​​of The theoretical semivariogram is fitted based on the experimental semivariogram, and the theoretical semivariogram is: Where C0 is the nugget value (noise), C is the base value, and a is the range; The theoretical semivariogram corresponding to the GNSS geographic coordinates is defined as the primary variable and the theoretical semivariogram corresponding to the InSAR geographic coordinates is defined as the secondary variable, and a co-kriging equation system is constructed based on the primary variable and the secondary variable. The co-kriging equation system is: Among them, λ i , μ k is the weight coefficient, γ 11 is the theoretical semivariogram corresponding to the GNSS geographic coordinates, γ 12 is the cross semivariogram corresponding to the GNSS geographic coordinates and the InSAR geographic coordinates, γ 22 is the theoretical semivariogram corresponding to the InSAR geographic coordinates; the weight coefficients are solved by the matrix, and the co-kriging equations are solved based on the weight coefficients to obtain the interpolation results, which are: A land vertical motion parameter is obtained based on the interpolation result, and an absolute sea level change parameter is obtained by performing a difference operation between the relative sea level change parameter and the land vertical motion parameter.

[0024] It should be understood that the cross-semivariogram is an important tool in geostatistics for characterizing the spatial correlation between two different spatial variables, and plays a key role in multi-source data fusion methods such as Collocated Cokriging.

[0025] For example, for two spatial variables Z(x1) and Z(x2) (such as GNSS vertical motion and InSAR deformation field), their cross semivariogram γ 12 (h) describes half of the expected variance of the difference between the observations of two variables at the spatial lag distance h, that is: Among them, N(h) is the number of data pairs within the distance interval h, x i is the spatial position coordinate.

[0026] If the cross-semivariogram values ​​of GNSS and InSAR data are low (i.e., the correlation is strong), it indicates that the spatial variation trends of the two are consistent, and InSAR data can be used as effective auxiliary information to improve the interpolation accuracy of GNSS point data.

[0027] The systematic errors of InSAR data (such as phase unwrapping error within 10 km near the coast) are quantified by the cross-semivariogram function, and the high precision of GNSS is used to perform spatial correction on InSAR data so that the RMS co-interpolation error is ≤ 0.2 mm / year.

[0028] In this embodiment, GNSS and InSAR data are expressed as experimental semivariograms. After fitting the theoretical semivariograms, a set of co-kriging equations is constructed with GNSS as the primary variable and InSAR as the secondary variable. The land vertical motion parameters are solved and the absolute sea level change parameters are separated. Combining the high precision of GNSS point data with the spatial coverage advantage of InSAR surface data, the data interpolation accuracy is improved through co-kriging, which effectively integrates discrete point and surface data and improves the accuracy of land vertical motion parameters.

[0029] Preferably, in step S3, a crustal movement-sea level coupling numerical model is constructed based on absolute sea level change parameters, GNSS point data, and InSAR surface data, including: Constructing a crustal movement-sea level coupled numerical model, wherein the crustal movement-sea level coupled numerical model includes a global layer, a regional layer and an observation layer, The global layer is used to define hyperparameters, including the mean global crustal subsidence rate μ obtained from the HY-2 satellite. 沉降 and variance The regional layer is used to define the hyperparameters corresponding to the multiple geographical blocks divided, which are expressed as: The observation layer is used to define the likelihood function by combining the absolute sea level change parameter, GNSS point data and InSAR surface data, which is expressed as:

[0030] In this example, a three-layer model was constructed, consisting of a global layer (defining hyperparameters such as crustal subsidence rate), a regional layer (dividing hyperparameters into geographic regions), and an observational layer (defining likelihood functions by combining multi-source data). This layered design integrates global and regional crustal movement and sea level change data, achieving multi-scale information fusion, improving the model's adaptability to complex geographical environments, and providing a structured framework for subsequent optimization.

[0031] Preferably, in step S3, the crustal movement-sea level coupled numerical model is optimized to obtain an optimized crustal movement-sea level coupled numerical model, including: S3.1. Define prior probability distributions using geological history data, sedimentation experiment results, and human activity statistics as prior knowledge; S3.2. Based on the sequential Bayesian dynamic parameter updating framework and the defined prior probability distribution corresponding to the prior knowledge, the prior knowledge is integrated with the data of the observation layer of the crustal motion-sea level coupling numerical model; S3.3. Perform spatial optimization and spatial optimization on the crustal movement-sea level coupling numerical model that incorporates prior knowledge to obtain an optimized crustal movement-sea level coupling numerical model.

[0032] In this embodiment, geological historical data, sedimentary experimental results, and human activity statistics are used as prior knowledge to define a priori probability distribution. The prior knowledge and observation layer data are integrated through a sequential Bayesian framework to perform spatial and temporal optimization.

[0033] Prior knowledge is introduced to make up for data deficiencies, and the Bayesian framework is combined to dynamically update parameters, thereby improving the model's ability to represent complex factors such as geological evolution and human activities, and enhancing the model's robustness and predictive reliability.

[0034] Preferably, in step S3.1, geological history data, sedimentation experiment results and human activity statistics are used as prior knowledge to define a priori probability distribution, including: Importing prior knowledge, wherein the prior knowledge includes geological history data, sedimentation experiment results, and human activity statistics; The truncated normal distribution is obtained from geological history data, and the truncated normal distribution is defined as the tectonic sedimentation coefficient, which is α~N [0，+∞) (1.2,0.5 2 ); The gamma distribution is obtained from the sedimentation experiment results and is defined as the sediment compaction rate, wherein the sediment compaction rate is β~Gamma(2,0.8); A lognormal distribution is obtained from statistical data of human activities, and the lognormal distribution is defined as a human activity factor, where the human activity factor is γ~Lognormal(0.5, 0.3).

[0035] In this embodiment, the truncated normal distribution is extracted from geological history data to define the tectonic sedimentation coefficient, the gamma distribution is extracted from sedimentation experiment results to define the sediment compaction rate, and the lognormal distribution is extracted from human activity statistical data to define the human activity factor.

[0036] The probability distribution of different types of prior knowledge is quantified so that factors such as geological structure, sedimentary process, and human activities can be quantitatively described by the model, thereby improving the model's accuracy in depicting multi-source influencing factors.

[0037] Preferably, in step S3.2, the prior knowledge is fused with the data of the observation layer of the crustal movement-sea level coupling numerical model based on the sequential Bayesian dynamic parameter update framework and the prior probability distribution corresponding to the defined prior knowledge, including: the real-time collected GNSS point data and InSAR surface data are integrated into the sequential Bayesian dynamic parameter update framework in a data stream manner, and the posterior distribution is calculated through a prediction step and an update step, wherein the prediction step generates a prior distribution based on the posterior at the previous moment, and the update step calculates the posterior distribution at the next moment based on the prior distribution generated in the prediction step. 所述预测步表示为: p(θ t |D t-1 )=∫p(θ t |θ t-1 )p(θ t-1 |D t-1 )dθ t-1 , 其中,p(θ t-1 |D t-1 )为前一时刻t-1的后验分布;p(θ t |θ t-1 ) is the state transition probability, describing the change law of the parameters from time t-1 to time t; θ t is the predicted value of the observation layer corresponding to the current time t, θ t-1 is the predicted value of the observation layer corresponding to the previous moment t-1; 所述更新步表示为: p(θ t |D t )∝p(d t |θ t )p(θ t |D t-1 ), Among them, d t is the measured value of the observation layer corresponding to the current time t; p(θ t |D t ) is the posterior distribution after combining all observed data at the current time t; p(d t |θ t )为似然函数(即 ),表征在参数θ t 下观测到数据d t 的概率;p(θ t |D t-1 ) is the prior distribution obtained in the prediction step. Using the measured data d at the current moment t The updated posterior distribution is obtained by calculating the prior distribution of the prediction step through the Bayesian formula to achieve dynamic optimization of parameters.

[0038] Through the iterative process of the above prediction step and update step, the prior probability distribution corresponding to prior knowledge such as geological historical data, sedimentary experimental results, and human activity statistics is continuously integrated with the real-time data of the observation layer, so that the model parameters can be dynamically adjusted according to the latest information, improving the model's ability to represent crustal movement and sea level changes, and enhancing the model's robustness and predictive reliability.

[0039] In this embodiment, real-time GNSS and InSAR data are input into a sequential Bayesian framework in a data stream manner, and the data are fused through a prediction step (generating a prior based on the previous moment's posterior) and an update step (calculating the posterior based on new data).

[0040] Realize dynamic updating of model parameters to adapt to the time-varying characteristics of sea level changes, use real-time data to continuously optimize the model, improve the response capability to sudden or gradual changes, and ensure the timeliness of prediction results.

[0041] Preferably, in step S3.3, the crustal movement-sea level coupling numerical model incorporating prior knowledge is optimized to obtain an optimized crustal movement-sea level coupling numerical model, including: Spatial optimization process: Spatial optimization is performed on the data in the crustal movement-sea level coupling numerical model that incorporates prior knowledge based on the spatial Gaussian process expression. The spatial Gaussian process expression is: in, is the signal variance, which is used to measure the variation of the spatial signal; l is the length scale, which is used to control the range of spatial correlation; is the noise variance, which is used to characterize the random noise in the data; δ ij is the Kronecker function (1 when i = j, otherwise 0); This process quantifies the correlation between spatial locations and handles the spatial heterogeneity of the data, enabling the model to better capture the spatial distribution patterns of surface deformation and crustal movement, and improving the continuity and accuracy of the data in the spatial dimension.

[0042] Time optimization process: Assuming that the data of the crustal movement-sea level coupling numerical model drifts slowly over time, the data in the crustal movement-sea level coupling numerical model with prior knowledge is time optimized based on the time autoregressive expression. The time autoregressive expression is: Among them, φ is the autoregressive coefficient, which is used to reflect the degree of dependence of data on time. is the disturbance variance, which is used to represent the random fluctuations in the time series.

[0043] This process captures the time dependence of the data, taking into account the temporal evolution of sea level changes and crustal movement, enabling the model to adapt to the time-varying characteristics of the data and improving the ability to predict long-term trends and short-term fluctuations.

[0044] In this embodiment, the spatial correlation of data is optimized based on a spatial Gaussian process expression, and a temporal autoregressive expression is used to assume that data drifts slowly over time and optimize time series characteristics.

[0045] The spatial heterogeneity of data is processed through spatial Gaussian process, and the temporal dependence of data is captured by temporal autoregression, so that the model can simultaneously consider the changing laws of spatial and temporal dimensions, thereby improving the spatial and temporal continuity and accuracy of the prediction results.

[0046] Finally, after obtaining the optimized crustal movement-sea level coupled numerical model, a posteriori prediction test was performed.

[0047] In step S4, the multi-source observation data of the sea level to be monitored are predicted based on the optimized crustal movement-sea level coupling numerical model to obtain a sea level change prediction result.

[0048] First, multi-source data fusion is performed on the collected data using step S2.

[0049] GNSS point data: directly used as high-precision observation input into the likelihood function.

[0050] InSAR surface data: Convert to grid data through spatial interpolation and define the regional average likelihood: Satellite altimetry data: serves as an independent validation set to constrain absolute sea level parameters.

[0051] Then, the fused multi-source data are input into the optimized crustal movement-sea level coupling numerical model for prediction to obtain the sea level change prediction results.

[0052] It should be understood that by using the optimized model to predict multi-source observational data, specific quantitative results are derived, such as the specific millimeters of sea level rise or fall, and sea level heights at different time points. These values ​​do not exist in isolation but can be further interpreted as meaningful information. For example, these values ​​can be used to determine whether the trend of sea level change is rising or falling, the rate of rise or fall, and whether there are differences in sea level changes between different regions, thereby providing a basis for relevant decision-making and research.

[0053] Preferably, this embodiment improves the reliability and predictive capability of the optimized crustal movement-sea level coupling numerical model by verifying the model through a probabilistic framework, and performs the following steps: 1. Input: the posterior distribution of the model parameters optimized in step S3, the absolute sea level change parameters measured in step S2, and the GNSS or InSAR data.

[0054] 2、输出: 后验预测检验: 1) Generate prediction samples: extract parameters θ from the posterior distribution (s) ,计算预测值

[0055] 2) Uncertainty visualization: plotting 95% confidence intervals and comparing them to measured data (e.g. tide gauge records) Forecast accuracy evaluation report and uncertainty visualization results.

[0056] 3. Feedback: If the test fails, the model will be re-optimized (such as adjusting the frequency of sequential Bayesian updates), forming a closed loop of "modeling-verification-iteration" to ensure the reproducibility and reliability of the technical solution.

[0057] Preferably, this embodiment optimizes the calculation speed process of the crustal movement-sea level coupled numerical model. The specific strategy is: 1)变分推断加速 Mean field variation: Assuming parameter independence, approximate posterior distribution q(θ) = ∏ i q i (θ i ).

[0058] Stochastic Gradient Descent: Optimize variational parameters using ADVI (Automatic Differentiation Variational Inference).

[0059] 2)GPU并行加速 Matrix operations: Using CUDA to accelerate covariance matrix inversion (complexity from O(N 3 )降至O(N 2 )).

[0060] Multi-chain parallelism: Run multiple MCMC chains simultaneously on the GPU to improve sampling efficiency.

[0061] MCMC链采用自适应MCMC采样方式: Algorithm selection: Hamiltonian Monte Carlo (HMC) combined with No-U-Turn Sampler (NUTS) to avoid local optimality.

[0062] Dynamically adjust the step size: Automatically adjust the HMC step size based on the acceptance rate, with a target acceptance rate of 65%.

[0063] Parallel Chains: Launches four independent MCMC chains and verifies convergence via the Gelman-Rubin diagnostic.

[0064] During the sampling process, the Hamiltonian Monte Carlo algorithm was used to optimize the parameters to minimize the deviation between the absolute sea level change parameters predicted by the model and the measured absolute sea level change parameters. The target acceptance rate was set at 65%, and the Gelman-Rubin diagnostic was used to verify the convergence of the four MCMC chains for the prediction of absolute sea level change parameters.

[0065] Example 2: Figure 2 As shown, an embodiment of the present invention provides a device for collaboratively monitoring sea level based on GNSS and InSAR, comprising: An initial data acquisition module is used to obtain GNSS point data representing vertical motion of land from GNSS reference stations pre-deployed along the coastline and co-located with tide gauge stations, and to obtain InSAR surface data representing surface deformation from Sentinel-1 or GF-3 satellites, as well as relative sea level change parameters relative to land from the HY-2 satellite. a data preprocessing module, configured to convert the GNSS point data and the InSAR surface data into the same coordinate system, perform data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system, obtain land vertical motion parameters, and separate absolute sea level change parameters from the relative sea level change parameters and the land vertical motion parameters; a model building module for building a crustal movement-sea level coupling numerical model based on absolute sea level change parameters, GNSS point data, and InSAR surface data, and optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model; The prediction module is used to predict the multi-source observation data of the sea level to be monitored based on the optimized crustal movement-sea level coupling numerical model to obtain the sea level change prediction results.

[0066] Example 3: An embodiment of the present invention provides a device for collaboratively monitoring sea level based on GNSS and InSAR, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method for collaboratively monitoring sea level based on GNSS and InSAR as described above is implemented.

[0067] Embodiment 4: An embodiment of the present invention provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the method for collaboratively monitoring sea level based on GNSS and InSAR as described above is implemented.

[0068] The above-mentioned device and storage medium for collaborative sea level monitoring based on GNSS and InSAR can refer to the implementation content and beneficial effects of the method for collaborative sea level monitoring based on GNSS and InSAR described above, which will not be repeated here.

[0069] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0070] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and units can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0071] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of units is merely a logical functional division. In actual implementation, other division methods may be used, such as combining or integrating multiple units or components into another system, or ignoring or not implementing certain features.

[0072] Units described as separate components may or may not be physically separate, and components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of these units may be selected based on actual needs to achieve the objectives of the embodiments of the present invention.

[0073] In addition, the functional units in the various embodiments of the present invention may be integrated into a single processing unit, each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0074] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, and other media that can store program code.

[0075] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for collaborative sea level monitoring based on GNSS and InSAR, characterized in that: The steps include: Acquire GNSS point data representing vertical motion of land from GNSS reference stations pre-deployed along the coastline and co-located with tide gauges, acquire InSAR surface data representing surface deformation from Sentinel-1 or GF-3 satellites, and acquire relative sea level change parameters relative to land from the HY-2 satellite. Converting the GNSS point data and the InSAR surface data to the same coordinate system, and performing data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system to obtain a land vertical motion parameter, and separating an absolute sea level change parameter from the relative sea level change parameter and the land vertical motion parameter; constructing a crustal movement-sea level coupling numerical model based on prior knowledge, absolute sea level change parameters, GNSS point data, and InSAR surface data, and optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model; Based on the optimized crustal movement-sea level coupling numerical model, the multi-source observation data of the sea level to be monitored are predicted to obtain the sea level change prediction results.

2. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 1, characterized in that: Converting the GNSS point data and the InSAR surface data to the same coordinate system includes: Performing coordinate analysis on the GNSS point data and the InSAR surface data using the SNAP tool to obtain GNSS geographic coordinates and InSAR geographic coordinates; Based on the selected WGS84 as the unified benchmark, the resolved GNSS geographic coordinates and InSAR geographic coordinates are unified into geographic coordinates, and the unified GNSS geographic coordinates and InSAR geographic coordinates are transformed into planar projections using GDAL and OGR tools. The converted GNSS geographic coordinates and InSAR geographic coordinates are corrected for elevation consistency using the GMT tool, thereby completing the conversion to the same coordinate system.

3. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 1, characterized in that: The method comprises the following steps: performing data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system to obtain a land vertical motion parameter, and separating an absolute sea level change parameter from the relative sea level change parameter and the land vertical motion parameter, including: expressing the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system as an experimental semivariogram, wherein the experimental semivariogram is: Among them, h is the spatial lag distance, N(h) is the number of data pairs in the distance interval, Z(x i ) is the GNSS geographic coordinate or InSAR geographic coordinate corresponding to the position x i Observed values ​​of The theoretical semivariogram is fitted based on the experimental semivariogram, and the theoretical semivariogram is: Among them, C0 is the nugget value, C is the base value, and a is the range; The theoretical semivariogram corresponding to the GNSS geographic coordinates is defined as the primary variable and the theoretical semivariogram corresponding to the InSAR geographic coordinates is defined as the secondary variable, and a co-kriging equation system is constructed based on the primary variable and the secondary variable. The co-kriging equation system is: Among them, λ i , μ k is the weight coefficient, γ 11 is the theoretical semivariogram corresponding to the GNSS geographic coordinates, γ 12 is the cross semivariogram corresponding to the GNSS geographic coordinates and the InSAR geographic coordinates, γ 22 is the theoretical semivariogram corresponding to the InSAR geographic coordinates; the weight coefficients are solved by the matrix, and the co-kriging equations are solved based on the weight coefficients to obtain the interpolation results, which are: A land vertical motion parameter is obtained based on the interpolation result, and an absolute sea level change parameter is obtained by performing a difference operation between the relative sea level change parameter and the land vertical motion parameter.

4. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 1, characterized in that: A coupled crustal movement-sea level model is constructed based on absolute sea level change parameters, GNSS point data, and InSAR surface data, including: Constructing a crustal movement-sea level coupled numerical model, wherein the crustal movement-sea level coupled numerical model includes a global layer, a regional layer and an observation layer, The global layer is used to define hyperparameters, including the mean global crustal subsidence rate μ obtained from the HY-2 satellite. 沉降 and variance The regional layer is used to define the hyperparameters corresponding to the multiple geographical blocks divided, which are expressed as: The observation layer is used to define the likelihood function by combining the absolute sea level change parameter, GNSS point data and InSAR surface data, which is expressed as:

5. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 4, characterized in that: Optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model, including: The prior probability distribution is defined using geological historical data, sedimentary experimental results, and human activity statistics as prior knowledge. Based on the sequential Bayesian dynamic parameter update framework and the prior probability distribution corresponding to the prior knowledge, the prior knowledge is integrated with the observation layer data of the crustal movement-sea level coupling numerical model. The crustal movement-sea level coupling numerical model incorporating prior knowledge is spatially optimized and spatially optimized to obtain the optimized crustal movement-sea level coupling numerical model.

6. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 5, characterized in that: The prior probability distribution is defined by using geological history data, sedimentation experiment results and human activity statistical data as prior knowledge, including: importing the prior knowledge, wherein the prior knowledge includes the geological history data, sedimentation experiment results and human activity statistical data; The truncated normal distribution is obtained from geological history data, and the truncated normal distribution is defined as the tectonic sedimentation coefficient, which is α~N [0，+∞] (1.2, 0.5 2 ); The gamma distribution is obtained from the sedimentation experiment results and is defined as the sediment compaction rate, wherein the sediment compaction rate is β~Gamma(2,0.8); A lognormal distribution is obtained from statistical data of human activities, and the lognormal distribution is defined as a human activity factor, where the human activity factor is γ~Lognormal(0.5, 0.3).

7. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 5, characterized in that: Based on the sequential Bayesian dynamic parameter update framework and the defined prior probability distribution corresponding to the prior knowledge, the prior knowledge is integrated with the data of the crustal movement-sea level coupling numerical model observation layer, including: The real-time collected GNSS point data and InSAR surface data are put into the sequential Bayesian dynamic parameter update framework in a data stream manner, and the posterior distribution is calculated through the prediction step and the update step. The prediction step generates the prior distribution based on the posterior of the previous moment, and the update step calculates the posterior distribution of the next moment based on the prior distribution generated in the prediction step. The prediction step is expressed as: p(θ t |D t-1 )=∫p(θ t |θ t-1 )p(θ t-1 |D t-1 )dθ t-1 , Among them, p(θ t-1 |D t-1 ) is the posterior distribution of the previous moment t-1, θ t is the predicted value of the observation layer corresponding to the current time t, θ t-1 is the predicted value of the observation layer corresponding to the previous moment t-1; The update step is expressed as: p(θ t |D t )∝p(d t |θ t )p(θ t |D t-1 ), Among them, d t is the measured value of the observation layer corresponding to the current time t.

8. The method for collaborative sea level monitoring based on GNSS and InSAR according to claim 5, characterized in that: The crustal movement-sea level coupling numerical model incorporating prior knowledge is optimized to obtain an optimized crustal movement-sea level coupling numerical model, including: The data in the crustal movement-sea level coupling numerical model incorporating prior knowledge is spatially optimized based on the spatial Gaussian process expression, which is: in, is the signal variance, l is the length scale, is the noise variance; Assuming that the data of the crustal movement-sea level coupling numerical model drifts slowly over time, the data in the crustal movement-sea level coupling numerical model with prior knowledge is time-optimized based on a time autoregressive expression. The time autoregressive expression is: Where φ is the autoregressive coefficient, is the disturbance variance.

9. A device for collaborative sea level monitoring based on GNSS and InSAR, characterized in that: include: An initial data acquisition module is used to obtain GNSS point data representing vertical motion of land from GNSS reference stations pre-deployed along the coastline and co-located with tide gauge stations, and to obtain InSAR surface data representing surface deformation from Sentinel-1 or GF-3 satellites, as well as relative sea level change parameters relative to land from the HY-2 satellite. a data preprocessing module, configured to convert the GNSS point data and the InSAR surface data into the same coordinate system, perform data fusion processing on the GNSS geographic coordinates and the InSAR geographic coordinates in the same coordinate system, obtain land vertical motion parameters, and separate absolute sea level change parameters from the relative sea level change parameters and the land vertical motion parameters; a model building module for building a crustal movement-sea level coupling numerical model based on absolute sea level change parameters, GNSS point data, and InSAR surface data, and optimizing the crustal movement-sea level coupling numerical model to obtain an optimized crustal movement-sea level coupling numerical model; The prediction module is used to predict the multi-source observation data of the sea level to be monitored based on the optimized crustal movement-sea level coupling numerical model to obtain the sea level change prediction results.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the method for collaboratively monitoring sea level based on GNSS and InSAR as claimed in any one of claims 1 to 6 is implemented.

Images