Method and system for inverting global sea level change based on satellite gravity data

Abstract

The invention discloses a method and a system for inverting global sea level change based on satellite gravity data, and belongs to the field of geophysical monitoring and remote sensing. The method comprises the following steps: acquiring satellite gravity data and multi-source auxiliary data, calculating time-varying characteristics of a gravity field after preprocessing, and constructing an inversion model coupling land mass change and ocean thermal expansion; and finally outputting high-resolution global sea level change data by combining a machine learning and data fusion technology optimization result. According to the method, the problems of insufficient precision and low resolution of a traditional method are solved, the sea level change monitoring accuracy can be effectively improved, and key data support is provided for climate change research and disaster prevention and reduction.

Description

Technical Field

[0001] This invention relates to a method and system for inverting global sea level change, and more particularly to a method and system for inverting global sea level change based on satellite gravity data, which involves multi-sphere mass coupling and multi-algorithm cascade optimization. Background Technology

[0002] Global sea-level change is a crucial indicator of climate change, and its monitoring is essential for understanding Earth system processes such as glacial melting, ocean thermal expansion, and groundwater migration. Currently, sea-level change monitoring primarily relies on two types of technologies:

[0003] Tide gauge observation: Sea level is measured directly through coastal stations, but due to the uneven distribution of stations (mostly concentrated on the edge of the continent), it is difficult to cover the global ocean, especially lacking data for polar and remote sea areas.

[0004] Satellite altimetry technology: It measures sea level height using radar altimeters and has global coverage capabilities, but it is easily affected by dynamic factors such as ocean currents, tides, and atmospheric pressure, requires complex corrections, and cannot directly reflect changes in seawater quality.

[0005] Satellite gravity measurement technologies (such as GRACE and GRACE-FO satellites) can directly invert the mass migration of the Earth's surface by monitoring the spatiotemporal changes of the Earth's gravity field, providing a novel observation method for sea level changes (especially sea level changes caused by mass changes). However, existing satellite gravity inversion methods have the following limitations:

[0006] (1) The data noise is high, especially the short-period signal is easily affected by instrument error and atmospheric interference;

[0007] (2) The indirect effects of changes in land mass such as glaciers, permafrost, and groundwater on sea level were not fully coupled;

[0008] (3) The inversion resolution is low (usually hundreds of kilometers), making it difficult to capture the fine sea level change characteristics at the regional scale (such as the tropical western Pacific warm pool and polar marginal seas).

[0009] Therefore, in the existing technology, there is no solution that can simultaneously solve the technical challenges of "multi-sphere factor coupling, multi-source data collaboration, and high precision and high resolution". This results in limited monitoring accuracy and causal analysis capabilities for global sea level change, which cannot meet the needs of refined climate change research and precise early warning for disaster prevention and mitigation. There is an urgent need for a high-precision, multi-factor coupled satellite gravity inversion method to improve the accuracy and spatial resolution of global sea level change monitoring. Summary of the Invention

[0010] The core objective of this invention is to overcome the limitations of existing technologies and provide a pioneering method and system for retrieving global sea level changes based on satellite gravity data using a multi-sphere mass coupling model and a three-level precision optimization link. This achieves a technological breakthrough in "precise decomposition of causes of sea level changes, effective suppression of errors, and progressive improvement in resolution," thereby solving problems such as insufficient accuracy, inadequate consideration of multi-sphere mass coupling, and low spatial resolution in existing technologies for retrieving global sea level changes using satellite gravity data.

[0011] To achieve the above objectives, the present invention provides a method for retrieving global sea level changes based on satellite gravity data, comprising the following steps:

[0012] S1: Acquire multi-source data and preprocess it. Satellite gravity data includes Level-2 time-varying gravity field data, precise orbital parameters, and instrument calibration data from the GRACE-FO satellite; auxiliary data covers satellite altimetry data (Jason series), glacier ablation data (ICESat-2), groundwater storage change data (GLDAS), atmospheric-ocean dynamic model data (ECMWF), and Argo buoy ocean temperature and salinity data, achieving full coverage of multi-sphere data from gravity to altimetry to land and ocean.

[0013] S2: Calculate the change in the gravity field. Based on preprocessed satellite gravity data, calculate the temporal rate of change of the Earth's gravity field anomaly. The paper innovatively proposes an equivalent water height conversion scheme based on 60th-order spherical harmonic analysis and 300km Gaussian smoothing, which accurately converts gravitational field anomalies into equivalent water height changes. This enables the quantitative characterization of surface mass migration, and improves the sensitivity of mass migration identification by 30% compared to traditional low-order spherical harmonic analysis.

[0014] S3: Construction of a multi-sphere mass-coupled sea-level change inversion model;

[0015] S4: Error correction and optimization. The random forest algorithm is introduced into the inversion results to correct errors, and the corrected sea level change value is output. The Kalman filter multi-source data fusion algorithm is used to fuse satellite gravity inversion results and satellite altimetry data to improve spatial resolution.

[0016] S5: Output Results and Visualizations. Outputs a 0.5°×0.5° gridded monthly / yearly sea-level change time series, and uses a 3D visualization system to intuitively present the spatial distribution of sea-level change and hotspots of rise rate, providing direct data support for subsequent applications.

[0017] Furthermore, in step S1, the multi-source data includes satellite gravity data and auxiliary data.

[0018] Furthermore, the satellite gravity data includes: time-varying gravity field data, orbital parameters, and instrument calibration data of the GRACE / GRACE-FO satellite;

[0019] The auxiliary data acquired include: satellite altimetry data, glacier ablation data, groundwater storage change data, and atmospheric and ocean dynamic model data.

[0020] Furthermore, in step S1, the preprocessing steps for multi-source data are as follows:

[0021] S11: Adaptive wavelet threshold denoising method is used to denoise satellite gravity data: The db4 wavelet basis function is used to perform a 5-level decomposition of the Stokes coefficients, and the adaptive threshold algorithm is used to specifically remove high-frequency instrument noise. Compared with the traditional fixed threshold denoising, the noise suppression efficiency is improved by more than 25%.

[0022] S12: Correcting time-varying baselines and subtracting geophysical effects: Simultaneously completes the correction of time-varying baselines and the subtraction of geophysical effects such as solid tides / polar tides, solving the problems of traditional correction methods being singular and having large residual errors;

[0023] S13: Perform spatiotemporal matching on auxiliary data (unify to the same time base and spatial grid): Interpolate all data to a 0.5°×0.5° spatial grid and UTC time base to achieve spatiotemporal alignment of heterogeneous data, laying the foundation for subsequent multi-source fusion.

[0024] Furthermore, in step S3, this invention proposes a three-element coupled decomposition model of "equivalent water height - land mass - ocean thermal expansion," breaking through the limitation of traditional models that only consider internal ocean processes. Specifically, it decomposes the equivalent water height variation... Decomposed into sea level change driven by changes in seawater mass Sea level change driven by non-mass factors (mainly thermal expansion of seawater) ,Right now:

[0025]

[0026] in, By accurately obtaining the data by deducting the contributions of terrestrial water storage migration such as glacial melting and changes in groundwater storage, cross-sphere coupling of land-sea quality is achieved. From satellite altimetry data and The difference was inverted and constrained by Argo buoy ocean temperature profile data to ensure the accuracy of the thermal expansion component.

[0027] Furthermore, in step S4, the machine learning model is a random forest: it innovatively uses three error sources—atmospheric disturbance, ocean tidal residuals, and satellite orbital errors—as input features to train a random forest correction model, specifically suppressing systematic errors and reducing the absolute error of sea level change inversion by 15-20%; while the multi-source data fusion algorithm is a Kalman filter: it constructs a Kalman filter model with an adaptive covariance matrix, deeply fusing satellite gravity inversion results with satellite altimetry data, achieving a leapfrog improvement in spatial resolution from 300km to 100km, and can accurately capture fine changes in sea level at the regional scale.

[0028] Another aspect of the present invention provides a system for retrieving global sea level changes based on satellite gravity data, comprising:

[0029] Data acquisition module: used to collect satellite gravity data and auxiliary data;

[0030] Adaptive preprocessing module: Integrating the above-mentioned innovative wavelet denoising, multi-physics field correction, and spatiotemporal registration algorithms, it can automatically match preprocessing parameters according to the input data type, improving preprocessing efficiency and adaptability;

[0031] High-precision gravity field calculation module: Built-in 60th order spherical harmonic analysis and Gaussian smoothing algorithm to realize rapid solution of gravity field anomalies and time-varying equivalent water height;

[0032] Multi-sphere coupled inversion model module: solidifies the above three-element coupled decomposition model, and supports dynamic access and quantitative decomposition of factors such as land water storage and ocean thermal expansion;

[0033] Optimized correction module: Integrates random forest error correction and Kalman filter fusion algorithms to achieve end-to-end accuracy improvement of inversion results;

[0034] Output visualization module: Generates grid data and visualization results.

[0035] In summary, the present invention has the following advantages over the prior art:

[0036] Accuracy Improvement: A three-level accuracy optimization link of "wavelet denoising - random forest correction - Kalman filter fusion" is constructed to achieve a dual breakthrough in error suppression and resolution improvement, reducing inversion error by 15-20%;

[0037] Multi-factor coupling: For the first time, a sea-level change decomposition model with multi-sphere mass coupling was realized, achieving the coordinated inversion of land-sea-atmosphere multi-sphere processes, filling the technical gap of traditional models that ignore the influence of land mass migration, and more comprehensively reflecting the mass balance of the Earth system;

[0038] High resolution: The resolution is improved from 300km to 100km, which can capture fine changes at regional scales (such as tropical seas and polar edges);

[0039] Highly practical: The inversion results can accurately distinguish the contribution ratio of mass change and thermal expansion, which not only improves the accuracy of sea level monitoring, but also provides a core technical tool for climate change attribution analysis. It has both monitoring and research value, and the output results can directly serve application scenarios such as climate change research and sea level rise risk assessment. Attached Figure Description

[0040] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0041] Figure 1 A flowchart of the method for retrieving global sea level changes based on satellite gravity data provided by this invention;

[0042] Figure 2 This is a system module structure diagram;

[0043] Figure 3 This is a comparison chart of the inversion results and the tide gauge data. Detailed Implementation

[0044] It should be noted that, unless otherwise specified, the embodiments and features described in the present invention can be combined with each other. The present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0045] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the present invention. As used herein, the singular form may also include the plural form unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0046] Unless otherwise specifically stated, the relative arrangement, numerical expressions, and values ​​of the components and steps set forth in these embodiments do not limit the scope of the invention. It should also be understood that, for ease of description, the dimensions of the various parts shown in the drawings are not drawn to actual scale. Techniques, methods, and devices known to those skilled in the art may not be discussed in detail, but where appropriate, such techniques, methods, and devices should be considered part of the specification. In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not as limitations. Therefore, other examples of exemplary embodiments may have different values. It should be noted that similar reference numerals and letters in the following figures denote similar items; therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0047] See Figure 1 As shown, this invention provides a method for retrieving global sea level changes based on satellite gravity data, comprising the following steps:

[0048] S1: Acquire multi-source data and preprocess it.

[0049] As a preferred option, multi-source data includes satellite gravity data and auxiliary data.

[0050] Acquire satellite gravity data: including time-varying gravity field data (such as Level-2 Stokes coefficients), orbital parameters, and instrument calibration data of GRACE / GRACE-FO satellites.

[0051] Acquire auxiliary data, including satellite altimetry data (such as Jason series), glacier ablation data (such as ICESat laser altimetry data), groundwater storage change data (such as GLDAS model data), and atmospheric and ocean dynamic model data (such as ECMWF reanalysis data).

[0052] As a preferred approach, the preprocessing steps for multi-source data are as follows:

[0053] S11: Denoise the satellite gravity data (using wavelet threshold denoising method).

[0054] S12: Correct time-varying baselines and subtract geophysical effects (such as solid tides and polar tides);

[0055] S13: Perform spatiotemporal matching on auxiliary data (unify to the same time base and spatial grid).

[0056] S2: Calculate the change in the gravity field. Based on preprocessed satellite gravity data, calculate the temporal rate of change of the Earth's gravity field anomaly. ); The gravitational field anomaly was converted into an equivalent water height change through spherical harmonic analysis ( This reflects the change in the equivalent water column height caused by the migration of surface mass.

[0057] S3: Construct a sea-level change inversion model;

[0058] As a preferred option, the coupled model construction process includes: converting the equivalent water height variation ( ) decomposed into sea level change caused by changes in seawater mass ( Sea level change caused by non-mass factors (such as thermal expansion of seawater) ),Right now:

[0059]

[0060] in: pass This is obtained after deducting the contribution of changes in terrestrial water storage (glaciers, groundwater, etc.); From satellite altimetry data and Difference inversion is performed, and constraints are applied by combining ocean temperature data (such as Argo buoy data).

[0061] S4: Error correction and optimization. A machine learning model (such as random forest) is introduced to correct the errors in the inversion results. Input features include atmospheric disturbances, ocean tidal residuals, satellite orbit errors, etc., and the output is the corrected sea level change value. A multi-source data fusion algorithm (such as Kalman filter) is used to fuse satellite gravity inversion results and satellite altimetry data to improve spatial resolution (from 300km to 100km).

[0062] S5: Output results and visualizations, outputting a global gridded (e.g., 0.5°×0.5°) time series of sea level changes (monthly / yearly scale); displaying the spatial distribution and trends of sea level changes (e.g., hotspots of rise rate) through a 3D visualization system.

[0063] See Figure 2 As shown, another aspect of the present invention provides a system for retrieving global sea-level change based on satellite gravity data. The system's module structure diagram presents an end-to-end, full-link global sea-level change retrieval technology system. Each module is sequentially connected according to the logical hierarchy of "data input - preprocessing - core calculation - model inversion - accuracy optimization - result output," forming a closed-loop process from multi-source heterogeneous data acquisition to visualization results delivery. This enables high-precision, high-resolution retrieval and causal analysis of sea-level change. The specific module functions and their connections are as follows:

[0064] 1. The data acquisition module is the data source entry point for the entire system, providing basic data support for subsequent processes. The collected satellite gravity data (GRACE / GRACE-FO) and auxiliary data (satellite altimetry, glacier melting, groundwater storage, etc.) will be directly sent to the adaptive preprocessing module for standardization processing.

[0065] 2. The adaptive preprocessing module addresses the heterogeneity of the input data by using innovative algorithms to perform data denoising, correction, and spatiotemporal registration. The standardized data is then passed to the high-precision gravity field calculation module, laying a unified data foundation for gravity field feature extraction.

[0066] 3. The high-precision gravity field calculation module completes the quantification of gravity field anomalies and time-varying equivalent water height based on preprocessed data. Its calculation results are the core input for the multi-sphere coupled inversion model module to carry out sea level change decomposition.

[0067] 4. The multi-sphere coupled inversion model module relies on the three-element coupled decomposition model to achieve accurate separation of mass factors and non-mass factors in sea level change. The preliminary inversion results will enter the optimization and correction module for accuracy upgrade.

[0068] 5. The optimization and correction module uses machine learning correction and multi-source data fusion to suppress errors and improve resolution of the inversion results, and delivers high-quality inversion data to the output visualization module.

[0069] 6. The output visualization module is the system's deliverable, transforming optimized data into gridded time-series data and visualization graphics, providing intuitive data support for scenarios such as climate change research.

[0070] Example:

[0071] I. Data Source

[0072] Satellite gravity data: Level-2 Stokes coefficient data from the GRACE-FO satellite from 2018 to 2023 (provided by NASA / DLR), with a time resolution of 30 days and a raw spatial resolution of 300 km;

[0073] Satellite altimetry data: Jason-3 satellite sea level height (SSH) data, spatial resolution 1°×1°;

[0074] Auxiliary data include: ocean thermal expansion data (such as CMIP6 ocean model data): used to quantify the non-mass factor (thermal expansion) in sea level change, complementing the mass factor results of gravity inversion; land surface hydrological data (such as hydrological station observation data): used to assist in verifying the inversion results of "land water storage change" in the multisphere inversion module; and geophysical correction data (such as solid tide model, polar motion parameters, and atmospheric load data): used for multiphysics field correction in the adaptive preprocessing module, removing interference from non-target signals on satellite gravity data.

[0075] GPS crustal deformation data: used to correct the impact of vertical crustal movement on sea level observation / inversion results, and improve the inversion accuracy of regional sea level changes.

[0076] Glacier data: ICESat-2 Antarctic / Greenland ice sheet ablation rate data;

[0077] Groundwater data: terrestrial water storage change data from the GLDAS model;

[0078] Ocean data: Argo buoy ocean temperature and salinity profile data, ECMWF atmosphere-ocean dynamic model data.

[0079] II. Preprocessing Procedure

[0080] Satellite gravity data denoising: The db4 wavelet basis function is used to decompose the Stokes coefficients into 5 levels, and high-frequency noise is removed by an adaptive thresholding method. The adaptive threshold is automatically calculated from the data noise variance.

[0081] Geophysical effects corrections cover solid tides (using the IERS 2010 model) and polar tides (using the Wahr model);

[0082] Spatiotemporal matching: Interpolate all data to a 0.5°×0.5° grid, and unify the time base to UTC time.

[0083] III. Model Parameter Settings

[0084] Spherical harmonic analysis order: 60th order spherical harmonic expansion (corresponding to spatial resolution ~300km), noise smoothed by Gaussian filtering (radius 300km);

[0085] Machine learning model: Random Forest algorithm, training set is data from 2018-2021, test set is data from 2022-2023, input feature dimension is 10 types of error source parameters;

[0086] Kalman filtering: The state equation is a sea-level change trend model, the observation equation is a fusion model of gravity inversion values ​​and altimetry observation values, and the noise covariance matrix adopts an adaptive iterative update strategy.

[0087] IV. Result Verification

[0088] Figure 3 This is the core accuracy verification chart of the global sea level change inversion method based on satellite gravity data proposed in this invention. By comparing the sea level mass change (equivalent sea level height) results obtained from satellite gravity inversion from 2003 to 2023 with the measured sea level data from independent tide gauge stations in the Pacific region (such as Honolulu) over time, the inversion accuracy and reliability of this method are intuitively verified.

[0089] From the key quantitative indicators in the figure, the coefficient of determination R² of the inversion results and the tide gauge observation results is 0.940, indicating that the two have a strong linear correlation and the inversion results can accurately reproduce the long-term trend of sea level change; the root mean square error RMSE is 0.54 mm and the mean absolute error MAE is 0.43 mm, which shows that the numerical deviation of the inversion results is at a low level.

[0090] From the perspective of practical application accuracy, the annual average sea-level rise rate retrieved by this method has an error of ≤0.3mm / year, which is more than 40% higher than the traditional satellite gravity retrieval method (error ≥0.5mm / year). It can also effectively capture the interannual fluctuation characteristics of sea-level changes. Furthermore, the retrieval results presented in the figure highlight the sea-level rise component contributed by land mass migration (glacial melting, groundwater changes, etc.). Combined with satellite altimetry data, the influence of thermal expansion can be further analyzed comprehensively, providing crucial data support for the accurate attribution of the causes of sea-level changes.

[0091] Overall, the comparison chart fully demonstrates the advanced nature of the "three-level precision optimization link" and "multi-layer coupling inversion model" constructed in this invention, providing strong verification evidence for the practical promotion and application of the method.

[0092] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for retrieving global sea level change based on satellite gravity data, characterized in that, Including the following steps: S1: Acquire multi-source data and preprocess it; S2: Calculate the change in the gravity field. Based on preprocessed satellite gravity data, calculate the temporal rate of change of the Earth's gravity field anomaly. The gravitational field anomaly was converted into an equivalent water height change through spherical harmonic analysis. This reflects the change in equivalent water column height caused by the migration of surface mass; S3: Construct a sea-level change inversion model; S4: Error correction and optimization. A machine learning model is introduced to correct the errors in the inversion results and output the corrected sea level change value. A multi-source data fusion algorithm is used to fuse satellite gravity inversion results and satellite altimetry data to improve spatial resolution. S5: Output results and visualizations.

2. The method for retrieving global sea level change based on satellite gravity data according to claim 1, characterized in that, In step S1, the multi-source data includes satellite gravity data and auxiliary data.

3. The method for retrieving global sea level change based on satellite gravity data according to claim 2, characterized in that, The satellite gravity data includes: time-varying gravity field data, orbital parameters, and instrument calibration data of the GRACE / GRACE-FO satellite; The auxiliary data acquired include: satellite altimetry data, glacier ablation data, groundwater storage change data, and atmospheric and ocean dynamic model data.

4. The method for retrieving global sea level change based on satellite gravity data according to claim 1, characterized in that, In step S1, the preprocessing steps for the multi-source data are as follows: S11: Denoising satellite gravity data using wavelet thresholding method; S12: Correction of time-varying baselines and subtraction of geophysical effects; S13: Perform spatiotemporal matching on the auxiliary data to unify it to the same time base and spatial grid.

5. The method for retrieving global sea level change based on satellite gravity data according to claim 1, characterized in that, In step S3, the sea level change inversion model process includes: converting the equivalent water height change... Decomposed into sea level changes caused by changes in seawater mass Sea level change caused by non-mass factors ,Right now: 。 6. The method for retrieving global sea level change based on satellite gravity data according to claim 1, characterized in that, In step S4, the machine learning model is a random forest; the multi-source data fusion algorithm is a Kalman filter.

7. The method for retrieving global sea level change based on satellite gravity data according to claim 6, characterized in that, Input features include atmospheric disturbances, ocean tidal residuals, and satellite orbital errors.

8. A system for retrieving global sea level changes based on satellite gravity data, characterized in that, include: Data acquisition module: used to collect satellite gravity data and auxiliary data; Preprocessing module: used for data denoising, correction, and spatiotemporal matching; Gravity field calculation module: used to solve for gravity field anomalies and time-varying equivalent water height; Inversion Model Module: Constructs a multi-factor coupled decomposition model of sea-level change; Optimized correction module: Improves inversion accuracy through machine learning and data fusion; Output visualization module: Generates grid data and visualization results.

Images