A groundwater change monitoring method based on multi-source remote sensing

By fusing multi-source remote sensing data and using a Bayesian inversion model, the problems of sparse monitoring points and high costs in traditional groundwater monitoring have been solved. This enables large-scale, long-term monitoring of groundwater storage changes, improving resolution and accuracy, and supporting arid zone management and ecological protection.

CN123069271APending Publication Date: 2026-11-10RESEARCH INSTITUTE OF WATER CONSERVANCY & HYDROPOWER IN XINJIANG UYGUR AUTONOMOUS REGION
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611283285.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-24
Publication Date
2026-11-10

AI Technical Summary

Technical Problem

Traditional groundwater monitoring methods suffer from problems such as sparse observation well locations, uneven spatial distribution, high construction and maintenance costs, long data compilation cycles, and large parameter uncertainties, making it difficult to achieve large-scale, long-term continuous dynamic monitoring. Furthermore, using remote sensing data alone can lead to problems such as low spatial resolution, multiple signal aliasing, and delayed vegetation response.

Method used

By acquiring gravity satellite, synthetic aperture radar interferometry (InSAR), optical remote sensing, and soil moisture remote sensing data, and after unifying the spatiotemporal benchmark, a Bayesian multi-source data fusion inversion model is constructed to calculate the spatiotemporal distribution of groundwater storage changes, generate groundwater change rate maps and change sequences, and verify them with measured data.

Benefits of technology

It enables continuous monitoring of groundwater storage changes over a wide area and long time, improves spatial resolution and accuracy, reduces the influence of non-groundwater signals, provides reliable dynamic monitoring results of groundwater, and supports water resource management and ecological protection in arid areas.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123069271A_ABST
    Figure CN123069271A_ABST
Patent Text Reader

Abstract

This invention discloses a groundwater change monitoring method based on multi-source remote sensing, belonging to the field of multi-source remote sensing hydrological monitoring technology. It addresses the problems of difficulty in separating groundwater signals from single-source remote sensing data and the poor timeliness of traditional monitoring. The method acquires a unified spatiotemporal reference from multi-source remote sensing data; processes and extracts regional terrestrial water storage change signals, groundwater-related vertical deformation components, vegetation response and water area fluctuations, and root zone soil moisture change sequences; constructs a Bayesian fusion inversion model with terrestrial water storage change as a constraint to calculate the spatiotemporal distribution of groundwater storage change; generates change rate maps and sequences, and verifies the monitoring results. This invention achieves continuous monitoring and improves inversion accuracy and spatial resolution.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of multi-source remote sensing hydrological monitoring technology, and in particular relates to a method for monitoring groundwater changes based on multi-source remote sensing. Background Technology

[0002] In arid regions, rainfall is scarce and evaporation is intense, resulting in limited surface water resources. Groundwater is a crucial water source for maintaining ecological security and socio-economic development. Traditional groundwater monitoring primarily relies on groundwater observation wells, using water level gauges or manual measuring ropes to obtain groundwater level data. While this method provides accurate groundwater level information at specific locations, the limited number and uneven spatial distribution of observation wells make it difficult to reflect spatial differences in groundwater storage changes at a regional scale. In arid regions, observation wells are often concentrated near agricultural irrigation areas or residential areas, while monitoring points are sparse in large-scale deserts, mountains, and ecologically fragile areas. The construction and maintenance costs of observation wells are high, and long-term continuous monitoring requires significant investment of manpower, resources, and funding. Traditional monitoring focuses primarily on groundwater levels, and converting water level changes into groundwater storage changes requires parameters such as aquifer storage coefficients. These parameters exhibit considerable uncertainty at a regional scale, and the data collection and compilation cycle is long, making it difficult to support the needs of large-scale, long-term continuous dynamic monitoring.

[0003] In recent years, multi-source remote sensing technology has provided new means for regional groundwater monitoring. Gravity satellites can invert changes in terrestrial water storage, InSAR can acquire high-resolution surface deformation information, optical remote sensing can extract vegetation indices and water area changes, and soil moisture remote sensing products can provide surface soil moisture information. These remote sensing methods have advantages such as large coverage, stable revisit periods, and continuous data, and can effectively capture surface response information related to groundwater changes. However, using a single type of remote sensing data has limitations: the spatial resolution of land water storage changes retrieved by gravity satellites is low, typically on the order of hundreds of kilometers, and includes the comprehensive changes of multiple hydrological components such as surface water, soil water, vegetation water, snow water equivalent, and groundwater, requiring further separation; the time series of surface deformation acquired by InSAR is affected by multiple factors such as tectonic movement, groundwater extraction and recharge, soil consolidation, thermal expansion, and atmospheric delay, and its direct use for groundwater monitoring suffers from multi-signal aliasing problems; vegetation indices or water indices calculated by optical remote sensing can only provide indirect information, and vegetation response in arid areas often lags behind groundwater changes; soil moisture remote sensing products mainly reflect surface or shallow soil moisture, with limited representativeness for root zone and deep soil moisture. Therefore, there is an urgent need for a multi-source remote sensing monitoring method and system that organically integrates gravity satellite, InSAR, optical remote sensing, soil moisture remote sensing products, and a small amount of measured data to separate groundwater storage changes within a unified framework. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention proposes a groundwater change monitoring method based on multi-source remote sensing, thereby resolving the issues present in the prior art.

[0005] In a first aspect, to achieve the above objectives, the present invention provides a groundwater change monitoring method based on multi-source remote sensing, comprising the following steps:

[0006] Acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry data, optical remote sensing data, and soil moisture remote sensing products, unify the spatiotemporal benchmark, and form a multi-source remote sensing dataset;

[0007] Multi-source remote sensing datasets were processed separately to extract regional terrestrial water storage change signals, vertical deformation components related to groundwater extraction or recharge, vegetation responses and water area fluctuations associated with groundwater changes, and root zone soil moisture change sequences.

[0008] Using regional terrestrial water storage change signals as constraints, and combining vertical deformation components, root zone soil moisture change sequences, vegetation response, and water area fluctuations, a Bayesian multi-source data fusion inversion model is constructed to solve the spatiotemporal distribution of groundwater storage changes.

[0009] Generate a groundwater change rate map and change sequence for the monitoring area, and output the monitoring results after verifying them with the measured groundwater level data.

[0010] Secondly, the present invention also provides a groundwater change monitoring system based on multi-source remote sensing, for implementing a groundwater change monitoring method based on multi-source remote sensing, the system comprising:

[0011] The data acquisition module is used to acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry data, optical remote sensing data and soil moisture remote sensing products, to unify the spatiotemporal reference and form a multi-source remote sensing dataset.

[0012] The data processing module is used to process multi-source remote sensing datasets and extract regional land water storage change signals, vertical deformation components related to groundwater extraction or recharge, vegetation response and water area fluctuations associated with groundwater changes, and root zone soil moisture change sequences.

[0013] The Bayesian fusion inversion module is used to construct a Bayesian multi-source data fusion inversion model by using regional land water storage change signals as constraints, combined with vertical deformation components, root zone soil moisture change sequences, vegetation response and water area fluctuations, to solve the spatiotemporal distribution of groundwater storage changes.

[0014] The results generation and verification output module is used to generate groundwater change rate maps and change sequences in the monitoring area, and output the monitoring results after verifying them with the measured groundwater level data.

[0015] Thirdly, the present invention also provides a computer terminal device, comprising:

[0016] One or more processors;

[0017] A memory, coupled to the processor, for storing one or more programs;

[0018] When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the groundwater change monitoring method based on multi-source remote sensing in the first aspect described above.

[0019] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, it implements the steps of the groundwater change monitoring method based on multi-source remote sensing in the first aspect described above.

[0020] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the groundwater change monitoring method based on multi-source remote sensing described in the first aspect.

[0021] Compared with the prior art, the present invention has the following advantages and technical effects:

[0022] This invention provides a groundwater change monitoring method based on multi-source remote sensing. This method achieves continuous monitoring of groundwater storage changes over a large area and long time series, overcoming the problems of sparse well locations, high cost, and poor timeliness associated with traditional methods. By fusing multi-source remote sensing data, the spatial resolution and accuracy of groundwater storage change inversion are improved. Gravity satellite data provides constraints on regional water storage changes, InSAR data provides high spatial resolution deformation information, soil moisture assimilation provides corrections for surface hydrological components, and optical remote sensing provides auxiliary information on vegetation and water bodies. This invention utilizes a Bayesian fusion framework to uniformly handle the uncertainties of multi-source data, resulting in more reliable posterior estimates and uncertainties regarding groundwater storage changes. Independent component analysis separates groundwater-related components from InSAR deformation, reducing the impact of non-groundwater signals such as tectonic deformation, thermal expansion, and atmospheric delay on the inversion results. Root zone soil moisture changes are used as a correction term for surface hydrological components and subtracted from gravity satellite terrestrial water storage changes, reducing interference from shallow soil moisture changes on groundwater storage change signal extraction and improving the physical consistency of groundwater signal separation. The generated groundwater change rate maps, monthly or quarterly change sequences, and validation results can directly serve decision-making in arid areas, including water resource management, ecological protection, groundwater over-extraction control, and drought early warning. Attached Figure Description

[0023] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an undue limitation of the invention. In the drawings:

[0024] Figure 1 This is a flowchart of an embodiment of the present invention. Detailed Implementation

[0025] 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.

[0026] It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions, and although a logical order is shown in the flowchart, in some cases the steps shown or described may be executed in a different order than that shown here.

[0027] This invention is applicable to the rapid assessment and continuous tracking of large-scale, long-term groundwater storage changes in arid regions. Specifically, it can be used in operational fields such as water resource allocation, ecological protection, groundwater over-extraction control, and drought early warning in arid areas. Arid regions experience scarce rainfall and intense evaporation, resulting in limited surface water resources. Groundwater is a crucial water source for maintaining ecological security and socio-economic development. Accurately understanding the spatiotemporal changes in groundwater storage directly supports these operations. Traditional monitoring relies on groundwater observation wells, which suffer from problems such as sparse monitoring points, uneven spatial distribution, high construction and maintenance costs, long data compilation cycles, and uncertain regional-scale storage coefficients, making it difficult to meet the needs of large-scale continuous dynamic monitoring. Using gravity satellite data, InSAR data, optical remote sensing data, or soil moisture remote sensing products individually also has significant limitations: gravity satellite inversion of terrestrial water storage changes has low spatial resolution and contains mixed signals of multiple hydrological components; InSAR surface deformation time series are affected by multiple factors such as tectonic movement, groundwater extraction and recharge, soil consolidation, thermal expansion, and atmospheric delay, resulting in multiple signal aliasing; vegetation indices or water indices provided by optical remote sensing are only indirect information, and vegetation response in arid areas often lags behind groundwater changes; soil moisture remote sensing products mainly reflect surface or shallow soil moisture, with limited representativeness for root zone and deep soil moisture. This invention acquires multi-source remote sensing data and constructs a Bayesian multi-source data fusion inversion model under a unified framework to separate the spatiotemporal distribution of groundwater storage changes, providing quantifiable groundwater dynamic monitoring results for specific monitoring problems in the aforementioned fields.

[0028] Example 1

[0029] like Figure 1 As shown, this embodiment provides a groundwater change monitoring method based on multi-source remote sensing, including:

[0030] Acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry data, optical remote sensing data, and soil moisture remote sensing products, unify the spatiotemporal benchmark, and form a multi-source remote sensing dataset;

[0031] Multi-source remote sensing datasets were processed separately to extract regional terrestrial water storage change signals, vertical deformation components related to groundwater extraction or recharge, vegetation responses and water area fluctuations associated with groundwater changes, and root zone soil moisture change sequences.

[0032] Using regional terrestrial water storage change signals as constraints, and combining vertical deformation components, root zone soil moisture change sequences, vegetation response, and water area fluctuations, a Bayesian multi-source data fusion inversion model is constructed to solve the spatiotemporal distribution of groundwater storage changes.

[0033] Generate a groundwater change rate map and change sequence for the monitoring area, and output the monitoring product after verifying it with the measured groundwater level data.

[0034] In summary, the overall technical approach is as follows: First, acquire multi-source remote sensing data; second, process gravity satellite data, InSAR data, optical remote sensing data, and soil moisture remote sensing products respectively to extract regional terrestrial water storage change signals, surface vertical deformation components related to groundwater, vegetation response and water area fluctuations, and root zone soil moisture change sequences; then, using regional terrestrial water storage change as a constraint, construct a Bayesian multi-source data fusion inversion model to calculate the spatiotemporal distribution of groundwater storage change; finally, generate groundwater change rate maps and change sequences, verify them with a small amount of measured groundwater level data, and output groundwater dynamic monitoring results.

[0035] The following details the implementation of each step in sequence. To ensure coherence, the inputs and outputs of each step are clearly defined, and the data used in subsequent steps comes from the outputs of previous steps.

[0036] As one implementation method in this embodiment, the process of acquiring multi-source remote sensing data includes:

[0037] The system acquires spherical harmonic coefficient products from gravity satellite data, multi-period synthetic aperture radar interferometry images covering the monitoring area, optical remote sensing images matching the monitoring period, microwave remote sensing soil moisture inversion products, and measured soil moisture data within the monitoring area. All of these data are then uniformly converted to the same spatial reference and time standard to form a multi-source remote sensing dataset.

[0038] The specific process includes:

[0039] Step S1: Acquire multi-source remote sensing data;

[0040] Acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry (InSAR) data, optical remote sensing data, and soil moisture remote sensing products.

[0041] Gravity satellite data comes from gravity satellite missions, such as the GRACE and GRACE-FO satellites, which provide Level-2 spherical harmonic coefficient products. This data is used to invert changes in regional terrestrial water storage. When acquiring data, the corresponding spherical harmonic coefficient time series for the monitored area should be covered, with a time span potentially on a monthly scale.

[0042] InSAR data originates from synthetic aperture radar satellite imagery, such as C-band or L-band SAR data. Multiple SAR images covering the same monitoring area need to be acquired, and the images should have sufficient temporal repetition to meet the requirements of differential interferometry processing and time-series deformation extraction. When acquiring data, relevant precise orbital parameters, digital elevation model data, or existing terrain phase reference data should also be included.

[0043] Optical remote sensing data comes from optical satellite imagery, such as multispectral remote sensing data. Visible, near-infrared, and shortwave infrared data matching the monitoring period are required for calculating the enhanced vegetation index or the normalized difference water index.

[0044] Soil moisture remote sensing products are derived from microwave remote sensing soil moisture inversion products, such as soil moisture products based on active microwave, passive microwave, or a combination of active and passive methods. Additionally, a small amount of measured soil moisture data from the monitoring area should be acquired for subsequent assimilation processing.

[0045] After the data is acquired, it is uniformly converted to the same spatial reference and time base to form a multi-source remote sensing dataset. This multi-source remote sensing dataset is the common input for steps S2, S3, S4, and S5.

[0046] As one implementation method in this embodiment, the process of extracting regional terrestrial water storage change signals includes:

[0047] The spherical harmonic coefficients of gravity satellite data are truncated, and the truncated spherical harmonic coefficients are processed by decorrelation filtering and Gaussian filtering. The leakage error is corrected on the filtering results, and the regional signal amplitude is recovered by the scale factor method, forward modeling method or data assimilation method to obtain the regional land water storage change signal.

[0048] The specific process includes:

[0049] Step S2: Gravity satellite data processing to extract regional land water storage change signals;

[0050] The gravity satellite data acquired in step S1 is processed, including spherical harmonic coefficient truncation, filtering, and leakage error correction, to extract the regional terrestrial water storage change signal, denoted as... .in, Indicates time, Indicates spatial location.

[0051] First, the spherical harmonic coefficients of gravity satellite data are truncated. Since higher-order spherical harmonic coefficients have significant noise, an appropriate truncation order is usually chosen, such as truncating to the 60th or 96th order, to preserve effective regional water storage information and suppress noise.

[0052] Secondly, the truncated spherical harmonic coefficients are filtered. Filtering can be achieved using a combination of decorrelation filtering and Gaussian filtering. Decorrelation filtering reduces the correlation error between the spherical harmonic coefficients, while Gaussian filtering smooths high-frequency noise. The filtering radius can be determined based on the monitoring area and noise level; for example, a 300 km Gaussian filtering radius can be used. After filtering, a smoothed signal of terrestrial water storage change is obtained.

[0053] Then, leakage error correction is performed. In gravity satellite inversion, signals from outside the monitoring area leak into the monitoring area, and signals from inside the monitoring area leak out, leading to amplitude attenuation and spatial confusion. Leakage error correction can be performed using the scale factor method, forward modeling method, or data assimilation method. The scale factor method recovers the regional signal by establishing a statistical relationship between the water volume change signals before and after filtering; the forward modeling method uses a prior hydrological model to simulate regional water storage changes and simulate the filtering effect, thereby estimating leakage error. After correction, a more accurate regional land water storage change signal ΔTWS(t,x) is obtained.

[0054] The regional land water storage change signal ΔTWS(t,x) output in step S2 will be used as the water balance constraint for the Bayesian multi-source data fusion inversion model in step S6.

[0055] As one implementation method in this embodiment, the process of separating the vertical deformation component related to groundwater extraction or recharge includes:

[0056] Differential interferometry processing was performed on the synthetic aperture radar interferometry data, and the time series of surface deformation was obtained using the time series synthetic aperture radar interferometry method. Independent component analysis was performed on the surface deformation time series, and the vertical deformation components related to groundwater extraction or recharge were identified based on the temporal variation characteristics and spatial distribution characteristics of the independent components.

[0057] The specific process includes:

[0058] Step S3: InSAR data processing, obtaining the time series of surface deformation and separating the vertical deformation component;

[0059] Differential interferometry and independent component analysis are performed on the InSAR data obtained in step S1 to obtain the vertical deformation components related to groundwater extraction or recharge.

[0060] First, differential interferometry is performed on the InSAR data. The main image is selected, and the remaining SAR images are precisely registered with it. The interferometric phase is calculated, and terrain phase, flatland phase, and orbital errors are removed. The differential interferometric phase is then filtered and unwrapped to obtain surface deformation information along the radar line-of-sight. To obtain the time series of surface deformation, time-series InSAR methods, such as PS-InSAR or SBAS, can be used. Through time-series analysis, the deformation time series of each pixel within the monitoring area along the radar line-of-sight is obtained.

[0061] Secondly, the deformation time series is geocoded and projected onto the same spatial reference as in step S2 to obtain the surface deformation time series. In this paper, the surface deformation time series contains various deformation signals, which may include tectonic deformation, changes in surface load, thermal expansion, surface subsidence caused by pumping, and surface rebound caused by recharge.

[0062] Then, independent component analysis (ICA) was used to separate the vertical deformation components related to groundwater extraction or recharge from the surface deformation time series. ICA assumes that the observed signal is a linear mixture of several statistically independent source signals, and the independent components are obtained by demixing. After performing ICA on the deformation time series, multiple independent components can be obtained. Based on the temporal variation characteristics and spatial distribution characteristics of each component, the vertical deformation components related to groundwater extraction or recharge were identified. The identification criteria include: the component shows a significant subsidence trend during the agricultural irrigation season or the concentrated groundwater extraction period, and a rebound trend during the recharge period; the spatial distribution of the component is consistent with groundwater activity areas such as groundwater extraction areas, alluvial fans, and oases; and the correlation between the component and temporal processes such as precipitation and surface water irrigation conforms to the groundwater response characteristics.

[0063] The separated vertical deformation components related to groundwater are denoted as This component represents the time series of vertical surface deformation related to groundwater extraction or recharge, and will serve as the spatial constraint information for the Bayesian multi-source data fusion inversion model in step S6.

[0064] As one implementation method in this embodiment, the process of identifying vegetation responses and water area fluctuations associated with groundwater changes, and the process of obtaining the root zone soil moisture change sequence, include:

[0065] Radiometric calibration, atmospheric correction, and cloud masking were performed on optical remote sensing images. Enhanced vegetation index or normalized difference water index was calculated, and the changing trends, abrupt change points, and seasonal characteristics of the enhanced vegetation index time series were analyzed to identify vegetation responses associated with groundwater changes. The water body extent was extracted using the normalized difference water index, and the water body area fluctuation series was calculated. The soil moisture remote sensing products were assimilated with the measured soil moisture, and ensemble Kalman filtering, optimal interpolation, or variational assimilation methods were used to obtain the root zone soil moisture change series.

[0066] The specific process includes:

[0067] Step S4: Optical remote sensing data processing to identify vegetation response and water area fluctuations;

[0068] Using the optical remote sensing data obtained in step S1, the enhanced vegetation index (EVI) or normalized differential water index (NDWI) is calculated to identify vegetation responses and water area fluctuations associated with groundwater changes.

[0069] First, radiometric calibration, atmospheric correction, and cloud masking are performed on the optical remote sensing images to obtain surface reflectance data. Then, the Enhanced Vegetation Index (EVI) or the Normalized Difference Water Index (NDWI) is calculated as needed. The EVI can be calculated using near-infrared, red, and blue light bands and reflects vegetation growth; the NDWI can be calculated using green and near-infrared bands and highlights water body information.

[0070] After calculating the EVI or NDWI time series, vegetation responses associated with groundwater changes are identified. In arid regions, vegetation growth is closely related to groundwater depth; vegetation indices are higher when groundwater depth is shallower and may decrease when groundwater levels drop. By analyzing the trends, abrupt change points, and seasonal characteristics of the EVI time series, vegetation responses associated with groundwater changes are identified. Vegetation response information is used to indicate the potential impacts of groundwater changes on the ecosystem and serves as the basis for prior information in step S6.

[0071] Simultaneously, the extent of water bodies within the monitoring area is extracted using NDWI or enhanced water body indices, and the change in water area over time is calculated to obtain a water area fluctuation sequence. Water area fluctuations can be used to assist in determining changes in surface water storage. When lakes, reservoirs, rivers, or other water bodies exist in the monitoring area and their area fluctuations are significant, changes in surface water storage can be estimated by combining area changes with average water depth or empirical area-storage relationships. If the surface water bodies in the monitoring area are scarce or the area changes very little, the changes in surface water storage can be approximated as negligible.

[0072] The vegetation response information and water area fluctuation sequence output in step S4 will serve as auxiliary information in step S6. Specifically, the vegetation response is used to construct the Bayesian prior constraint term, and the water area fluctuation is used for optional additional surface hydrological component correction terms. .

[0073] Step S5: Assimilate soil moisture remote sensing products to obtain the root zone soil moisture change sequence.

[0074] Using the soil moisture remote sensing products and measured soil moisture data obtained in step S1, assimilation processing is performed to obtain the root zone soil moisture change sequence. , as a correction term for surface hydrological components.

[0075] Soil moisture remote sensing products typically reflect surface soil moisture at a depth of 0–5 cm or 0–10 cm. However, groundwater storage change inversion requires subtracting root zone soil moisture changes; therefore, surface soil moisture products need to be assimilated to obtain root zone soil moisture changes.

[0076] Assimilation methods can include ensemble Kalman filtering, optimal interpolation, or variational assimilation. During the assimilation process, remote sensing soil moisture products are used as observational information, measured soil moisture data are used as ground validation and bias correction information, and a soil moisture model or water balance model is used as the background field to estimate the soil moisture state in the root zone. The root zone soil moisture time series output by the assimilation system reflects the changes in root zone soil moisture content within the monitoring area.

[0077] The assimilated root zone soil moisture time series was converted into equivalent water height changes to obtain the root zone soil moisture change series. This sequence, as a correction term for the surface hydrological component, will be used to subtract it from the terrestrial water storage change signal in step S6.

[0078] The root zone soil moisture change sequence output in step S5 The output of step S2 The output of step S3 The vegetation response and water area fluctuation output from step S4 are input into step S6.

[0079] As one implementation method in this embodiment, the process of constructing a Bayesian multi-source data fusion inversion model and solving for the spatiotemporal distribution of groundwater storage changes includes:

[0080] The residual term is obtained by subtracting the root zone soil moisture change sequence and the surface water storage change corresponding to the water area fluctuation from the regional terrestrial water storage change signal; the vertical deformation component related to groundwater extraction or recharge is used as a spatial constraint; the prior distribution of groundwater storage change is adjusted by vegetation response and water area fluctuation; and the spatiotemporal distribution of groundwater storage change is solved by maximizing the posterior probability.

[0081] The specific process includes:

[0082] Step S6: Construct a Bayesian multi-source data fusion inversion model to calculate the spatiotemporal distribution of groundwater storage changes;

[0083] The regional land water storage change signal obtained in step S2 The vertical deformation component obtained in step S3 The vegetation response and water area fluctuations obtained in step S4, and the root zone soil moisture change sequence obtained in step S5. A Bayesian multi-source data fusion inversion model is used as the common input, with the land water storage change inverted by gravity satellite as the constraint, and combined with InSAR deformation information, soil moisture change and vegetation response to solve the spatiotemporal distribution of groundwater storage change.

[0084] First, establish the water balance relationship. Changes in terrestrial water storage retrieved from gravity satellites. Mainly due to changes in groundwater reserves Changes in soil moisture in the root zone Changes in surface water storage In addition, other hydrological components vary. In arid regions, the changes in other hydrological components, such as snowmelt equivalent and vegetation water content, are relatively small and can be ignored or treated as residual terms. Therefore, the water balance relationship can be written as:

[0085]

[0086] in, This includes gravity satellite observation errors and residual errors. If step S4 determines that changes in surface water storage are negligible, then... Set to zero; if it cannot be ignored, then the value estimated in step S4 is... Substitute it as a known correction term.

[0087] Since step S5 has already obtained the root zone soil moisture change sequence Step S4 can be obtained Therefore, the residual term after deducting the surface hydrological component from the change in terrestrial water storage can be constructed:

[0088]

[0089] The residual term Mainly reflects changes in groundwater reserves and error sum.

[0090] Secondly, InSAR deformation constraints are established. This includes the vertical deformation components associated with groundwater. Changes in groundwater storage There is a physical correlation. In aquifer systems, changes in groundwater storage cause changes in pore pressure and effective stress, which in turn produce vertical surface deformation. Therefore, the InSAR vertical deformation component can serve as a spatial distribution constraint for changes in groundwater storage. The following observation relationship can be established:

[0091]

[0092] in, A coefficient matrix or transfer function that reflects the magnitude of vertical surface deformation caused by changes in unit groundwater storage. These represent InSAR deformation observation errors and separation errors. It can be determined based on prior information such as the water storage characteristics of the aquifer, the thickness of the soil layer, and mechanical parameters, or it can be estimated as a parameter in a Bayesian framework.

[0093] Secondly, prior information is constructed using optical remote sensing of vegetation responses. Vegetation responses are used to indicate the impact of groundwater changes on ecosystems and can serve as prior constraints on the spatial distribution of groundwater changes. For example, in areas where the EVI (Effective Virtualization Index) decreases significantly and corresponds to an increase in groundwater depth, it can be assumed that the groundwater storage in that area is likely to decrease; conversely, in areas where the EVI is stable or increasing and related to groundwater recharge, it can be assumed that the groundwater storage is likely to increase or remain relatively stable. This prior information can be expressed as an indication of changes in groundwater storage. Adjustments to the prior mean and prior covariance.

[0094] Then, a Bayesian posterior estimation model is constructed. Let the variable to be estimated be the groundwater storage change field. After discretization, it can be represented as a vector. The observation information includes:

[0095] Land water storage residual observation vector: ;

[0096] InSAR vertical deformation observation vector: ;

[0097] Prior information comes from optical remote sensing vegetation response and spatial smoothing constraints.

[0098] According to Bayes' theorem, the variable to be estimated The posterior probability density function can be expressed as:

[0099]

[0100] in, Let be the residual likelihood function of terrestrial water storage. This is the InSAR deformation likelihood function. This is the prior distribution.

[0101] If we assume that the error follows a Gaussian distribution, then the posterior estimate can be transformed into minimizing the objective function:

[0102]

[0103]

[0104]

[0105] in, From arrive The observation matrix or operator, For the residual error covariance of terrestrial water storage, For InSAR deformation error covariance, Let be the prior mean vector. Let be the prior covariance matrix. and Optical remote sensing vegetation response information can be used to construct structures that reflect the possible distribution of groundwater changes indicated by vegetation.

[0106] The spatiotemporal distribution of groundwater storage changes can be obtained by maximizing the posterior probability or minimizing the above objective function. The results not only satisfy the overall water balance constraint of gravity satellite terrestrial water storage changes, but also integrate InSAR high spatial resolution deformation information, soil moisture assimilation correction information, and optical remote sensing vegetation response prior information, exhibiting good spatial resolution and temporal continuity.

[0107] The spatiotemporal distribution of groundwater storage changes output in step S6 This will be used as the input for step S7.

[0108] Step S7: Generate monitoring results and verify the output;

[0109] Based on the spatiotemporal distribution of groundwater storage changes obtained in step S6 It generates groundwater change rate maps and monthly or quarterly change sequences for the monitoring area, compares and verifies them with a small amount of measured groundwater level data, and outputs groundwater dynamic monitoring results suitable for management decisions in arid areas.

[0110] First, a trend analysis is performed on the time series of groundwater storage changes to calculate the rate of change of groundwater storage in each spatial unit, for example, expressed as the rate of change of equivalent water height, in mm / year or mm / month. The spatial distribution of the rate of change is then plotted as a groundwater rate of change map. This map can visually display the spatial pattern of increase and decrease in groundwater storage within the monitoring area.

[0111] Secondly, the changes in groundwater storage are statistically analyzed on a monthly or quarterly scale to generate a time series of groundwater storage changes in the monitoring area. The time series can be statistically analyzed by administrative region, watershed, aquifer unit, or ecological zone of interest to the user, facilitating management decision-making.

[0112] Then, a small amount of measured groundwater level data was used for comparative verification. The measured groundwater level data came from groundwater observation wells and could be converted into groundwater storage change information, or directly compared with the groundwater storage change series for trend consistency and correlation. Verification indicators included correlation coefficient, root mean square error, and Nash-Sutcliffe efficiency coefficient. Through verification, the uncertainty and reliability of the inversion results were assessed.

[0113] Finally, the system outputs dynamic groundwater monitoring results suitable for management decision-making in arid regions. Results can be in thematic maps, raster datasets, time-series curves, statistical reports, or data interfaces. Content includes groundwater storage change rates, monthly or quarterly change sequences, validation accuracy specifications, and uncertainty information, making it suitable for operational use in water resource management, ecological protection, groundwater over-extraction control, and drought early warning.

[0114] In addition to the above process, this embodiment uses an inland watershed in an arid region as the monitoring area. This region has scarce rainfall, strong evaporation, limited surface water, and groundwater is the main source of agricultural irrigation and ecological water use. The monitoring area includes oases, deserts, and some mountains, and the number of groundwater observation wells is limited.

[0115] In step S1, the following data is obtained:

[0116] Gravity satellite data: monthly data on the Level-2 spherical harmonic coefficients of GRACE and GRACE-FO satellites, covering the past 10 years and the monitoring area and surrounding regions.

[0117] InSAR data: Acquire C-band SAR images covering the monitoring area, with a time range consistent with gravity satellite data. The image revisit period is 12 days or 6 days, and the data volume meets the requirements for time-series InSAR processing.

[0118] Optical remote sensing data: Acquire multispectral remote sensing images, such as Sentinel-2 or Landsat images, that match the monitoring period to calculate the Enhanced Vegetation Index (EVI) and the Normalized Difference Water Index (NDWI).

[0119] Soil moisture remote sensing products: Soil moisture products based on passive microwave remote sensing inversion are acquired, and the spatial resolution can be resampled to be consistent with the InSAR results; at the same time, a small amount of measured soil moisture data in the monitoring area is acquired.

[0120] In step S2, the GRACE / GRACE-FO spherical harmonic coefficients are truncated, for example, to the 60th order. Decorrelation filtering and 300 km Gaussian filtering are applied, and leakage error correction is performed using the scaling factor method to extract the regional terrestrial water storage change signal in the monitoring area. .

[0121] In step S3, differential interferometry is performed on the SAR imagery, and the SBAS method is used to construct a deformation time series, followed by phase unwrapping and geocoding. Independent component analysis is then performed on the deformation time series. Based on the characteristics of irrigation season subsidence and recharge period rebound, as well as the consistency of spatial distribution with aquifer units, the vertical deformation components related to groundwater extraction or recharge are separated. .

[0122] In step S4, EVI and NDWI time series are calculated using multispectral remote sensing imagery. The changing trends of EVI in oasis and non-oasis areas are analyzed to identify vegetation responses associated with changes in groundwater depth. NDWI is used to extract the water body extent of lakes and reservoirs within the monitoring area, and water area fluctuations are calculated. If the water area change is not significant, then... Set to be negligible.

[0123] In step S5, the remote sensing soil moisture products and measured soil moisture data are input into the assimilation system. The ensemble Kalman filter method is used to estimate the root zone soil moisture time series and convert it into a root zone soil moisture change series. .

[0124] In step S6, a Bayesian multi-source data fusion inversion model is constructed. The output of step S2 is then used... Subtract the output of step S5 and optional This yields the water balance residual term. The output of step S3 is then... As an InSAR deformation constraint, the vegetation response output from step S4 is used to construct the prior mean and prior covariance. The spatiotemporal distribution of groundwater storage changes is then calculated by maximizing the posterior probability. .

[0125] In step S7, based on Calculate the rate of change of groundwater storage for each pixel and plot a groundwater rate of change map for the monitoring area. Compile groundwater storage change sequences for watersheds or aquifer units on a monthly or quarterly scale. Use groundwater level data from a small number of observation wells for comparative verification, and calculate correlation coefficients and error indices. Finally, output the groundwater dynamic monitoring results, including the rate of change map, monthly / quarterly change sequences, and verification accuracy description, for submission to water resources management departments.

[0126] In the above embodiments, the data transmission relationship between each step is clear. The outputs of steps S2, S3, S4, and S5 are all used as the inputs of step S6, and the output of step S6 is used as the input of step S7, which ensures the continuity and logical integrity of the monitoring process.

[0127] The terms used in this invention are explained as follows:

[0128] Regional terrestrial water storage change signal: refers to the regional-scale terrestrial water storage change obtained by gravity satellite inversion and processing, including the sum of changes in various hydrological components such as groundwater, soil water, and surface water, denoted as .

[0129] Vertical deformation components related to groundwater extraction or recharge: These refer to the surface vertical deformation signals related to groundwater storage changes, separated from InSAR surface deformation time series through independent component analysis, and denoted as... .

[0130] Vegetation response: refers to the vegetation growth status response information associated with groundwater changes, reflected by optical remote sensing enhanced vegetation index (EVI) or normalized difference water index (NDWI).

[0131] Water area fluctuation: refers to the fluctuation of water area over time as extracted by optical remote sensing water index.

[0132] Root zone soil moisture change sequence: refers to the root zone soil moisture change information obtained by assimilating soil moisture remote sensing products with measured soil moisture, denoted as... .

[0133] Surface hydrological component correction term: refers to the surface hydrological component subtracted from the terrestrial water storage change signal, mainly including root zone soil moisture change and selectable surface water storage change.

[0134] Spatiotemporal distribution of groundwater storage changes: refers to the spatiotemporal distribution of groundwater storage changes in the monitoring area obtained from the Bayesian multi-source data fusion inversion model, denoted as... .

[0135] Based on this, the groundwater change monitoring method based on multi-source remote sensing provided by this invention has the following advantages compared to the prior art:

[0136] First, it enables continuous monitoring of groundwater storage changes over a large area and long time. By acquiring regional information through multi-source remote sensing, it overcomes the problems of sparse observation well locations, high costs, and poor timeliness of traditional methods, and can quickly assess and continuously track groundwater changes in arid areas.

[0137] Second, it improves the spatial resolution and accuracy of groundwater storage change inversion. Gravity satellite data provides constraints on regional water storage changes, InSAR data provides high spatial resolution deformation information, soil moisture assimilation provides correction for surface hydrological components, optical remote sensing provides auxiliary information on vegetation and water bodies, and multi-source data fusion effectively improves the spatiotemporal resolution and accuracy of the inversion results.

[0138] Third, the uncertainty of multi-source data is handled uniformly using a Bayesian fusion framework. The Bayesian model can simultaneously consider gravity satellite errors, InSAR deformation errors, soil moisture estimation errors, and prior uncertainties of vegetation, providing a posterior estimate of groundwater storage changes and their uncertainties, resulting in more reliable results.

[0139] Fourth, by separating the groundwater-related components in InSAR deformation through independent component analysis, the influence of non-groundwater signals such as tectonic deformation, thermal expansion, and atmospheric delay on the inversion results is reduced, thereby improving the availability of InSAR information in groundwater monitoring.

[0140] Fifth, the change in root zone soil moisture is used as a correction term for the surface hydrological component and deducted from the change in land water storage in gravity satellites. This reduces the interference of shallow soil moisture changes on the extraction of groundwater storage change signals and improves the physical consistency of groundwater signal separation.

[0141] Sixth, the generated groundwater change rate maps, monthly or quarterly change sequences, and verification results can directly serve decision-making in arid areas such as water resource management, ecological protection, groundwater over-extraction control, and drought early warning, and have good application value.

[0142] Example 2

[0143] In this embodiment, a computer terminal device is provided, including:

[0144] One or more processors;

[0145] A memory, coupled to the processor, for storing one or more programs;

[0146] When the one or more programs are executed by the one or more processors, the one or more processors implement the steps of the above-described groundwater change monitoring method based on multi-source remote sensing.

[0147] In this embodiment, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the above-described groundwater change monitoring method based on multi-source remote sensing.

[0148] In this embodiment, an electronic device is also provided, including a memory and a processor. The memory stores a computer program, and the processor is configured to run the computer program to perform the steps of the above-described groundwater change monitoring method based on multi-source remote sensing.

[0149] In this embodiment, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of the above-described groundwater change monitoring method based on multi-source remote sensing.

[0150] The aforementioned program can run on a processor or be stored in memory (or a computer-readable medium). Computer-readable media includes both permanent and non-permanent, removable and non-removable media, and information storage can be achieved by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other non-transfer medium that can be used to store information accessible by a computing device.

[0151] These computer programs may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes can be implemented by different modules for different steps.

[0152] This embodiment provides such a device or system. The system, referred to as a multi-source remote sensing-based groundwater change monitoring system, includes:

[0153] The data acquisition module is used to acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry data, optical remote sensing data and soil moisture remote sensing products, to unify the spatiotemporal reference and form a multi-source remote sensing dataset.

[0154] The data processing module is used to process multi-source remote sensing datasets and extract regional land water storage change signals, vertical deformation components related to groundwater extraction or recharge, vegetation response and water area fluctuations associated with groundwater changes, and root zone soil moisture change sequences.

[0155] The Bayesian fusion inversion module is used to construct a Bayesian multi-source data fusion inversion model by using regional land water storage change signals as constraints, combined with vertical deformation components, root zone soil moisture change sequences, vegetation response and water area fluctuations, to solve the spatiotemporal distribution of groundwater storage changes.

[0156] The results generation and verification output module is used to generate groundwater change rate maps and change sequences in the monitoring area, and output the monitoring results after verifying them with the measured groundwater level data.

[0157] As one implementation method in this embodiment, the data acquisition module includes:

[0158] Gravity satellite data acquisition unit, used to acquire spherical harmonic coefficient products from gravity satellite data;

[0159] Synthetic Aperture Radar Interferometry Data Acquisition Unit is used to acquire multi-phase synthetic aperture radar interferometry images covering the monitoring area;

[0160] The optical remote sensing data acquisition unit is used to acquire optical remote sensing images that match the monitoring period.

[0161] The soil moisture remote sensing product acquisition unit is used to acquire microwave remote sensing soil moisture inversion products.

[0162] The measured soil moisture data acquisition unit is used to acquire measured soil moisture data within the monitoring area.

[0163] The unified spatiotemporal reference unit is used to uniformly transform the above data to the same spatial reference and temporal reference, forming a multi-source remote sensing dataset.

[0164] As one implementation method in this embodiment, the data processing module includes a gravity satellite data processing unit, which includes:

[0165] The spherical harmonic coefficient truncation unit is used to truncate the spherical harmonic coefficients of gravity satellite data;

[0166] The decorrelation filtering unit is used to perform decorrelation filtering on the truncated spherical harmonic coefficients.

[0167] The Gaussian filtering unit is used to perform Gaussian filtering on the decorrelation filtering result.

[0168] The leakage error correction unit is used to correct leakage errors in the filtering results. It uses the scaling factor method, forward modeling method or data assimilation method to recover the regional signal amplitude and obtain the regional land water storage change signal.

[0169] As one implementation method in this embodiment, the data processing module includes a synthetic aperture radar interferometry data processing unit, which includes:

[0170] The differential interferometry processing unit is used to perform differential interferometry processing on synthetic aperture radar interferometric measurement data;

[0171] The temporal deformation extraction unit is used to obtain the time series of surface deformation using the temporal synthetic aperture radar interferometry method.

[0172] The independent component analysis unit is used to perform independent component analysis on the time series of surface deformation, and to identify the vertical deformation components related to groundwater extraction or recharge based on the temporal variation characteristics and spatial distribution characteristics of the independent components.

[0173] As one implementation method in this embodiment, the data processing module includes an optical remote sensing data processing unit and a soil moisture assimilation unit;

[0174] The optical remote sensing data processing unit is used to perform radiometric calibration, atmospheric correction and cloud masking on optical remote sensing images, calculate the enhanced vegetation index or normalized differential water index, analyze the changing trend, abrupt change points and seasonal characteristics of the enhanced vegetation index time series, identify vegetation responses associated with groundwater changes, extract water body range using the normalized differential water index, and calculate water body area fluctuation series.

[0175] The soil moisture assimilation unit is used to assimilate remote sensing soil moisture products with measured soil moisture. It uses ensemble Kalman filtering, optimal interpolation, or variational assimilation methods to obtain the root zone soil moisture change sequence.

[0176] As one implementation method in this embodiment, the Bayesian fusion inversion module includes:

[0177] The water balance residual unit is used to obtain the residual term by subtracting the root zone soil moisture change sequence and the surface water storage change corresponding to the water body area fluctuation from the regional land water storage change signal.

[0178] Deformation constraint elements are used to constrain the vertical deformation components associated with groundwater extraction or recharge as spatial constraints.

[0179] Prior building blocks are used to adjust the prior distribution of groundwater storage changes based on vegetation response and water area fluctuations.

[0180] The posterior solution unit is used to solve the spatiotemporal distribution of groundwater storage changes by maximizing the posterior probability.

[0181] The system or apparatus is used to implement the functions of the methods in the above embodiments. Each module in the system or apparatus corresponds to each step in the method, as has been described in the method and will not be repeated here.

[0182] The above implementation method solves the problem of groundwater change monitoring based on multi-source remote sensing in related technologies, thereby ensuring that the problems existing in the prior art are resolved.

[0183] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for monitoring groundwater changes based on multi-source remote sensing, characterized in that, Includes the following steps: Acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry data, optical remote sensing data, and soil moisture remote sensing products, unify the spatiotemporal benchmark, and form a multi-source remote sensing dataset; Multi-source remote sensing datasets were processed separately to extract regional terrestrial water storage change signals, vertical deformation components related to groundwater extraction or recharge, vegetation responses and water area fluctuations associated with groundwater changes, and root zone soil moisture change sequences. Using regional terrestrial water storage change signals as constraints, and combining vertical deformation components, root zone soil moisture change sequences, vegetation response, and water area fluctuations, a Bayesian multi-source data fusion inversion model is constructed to solve the spatiotemporal distribution of groundwater storage changes. Generate a groundwater change rate map and change sequence for the monitoring area, and output the monitoring results after verifying them with the measured groundwater level data.

2. The method according to claim 1, characterized in that, The process of acquiring multi-source remote sensing data includes: The system acquires spherical harmonic coefficient products from gravity satellite data, multi-period synthetic aperture radar interferometry images covering the monitoring area, optical remote sensing images matching the monitoring period, microwave remote sensing soil moisture inversion products, and measured soil moisture data within the monitoring area. All of these data are then uniformly converted to the same spatial reference and time standard to form a multi-source remote sensing dataset.

3. The method according to claim 1, characterized in that, The process of extracting regional terrestrial water storage change signals includes: The spherical harmonic coefficients of gravity satellite data are truncated, and the truncated spherical harmonic coefficients are processed by decorrelation filtering and Gaussian filtering. The leakage error is corrected on the filtering results, and the regional signal amplitude is recovered by the scale factor method, forward modeling method or data assimilation method to obtain the regional land water storage change signal.

4. The method according to claim 1, characterized in that, The process of separating the vertical deformation components associated with groundwater extraction or recharge includes: Differential interferometry processing was performed on the synthetic aperture radar interferometry data, and the time series of surface deformation was obtained using the time series synthetic aperture radar interferometry method. Independent component analysis was performed on the surface deformation time series, and the vertical deformation components related to groundwater extraction or recharge were identified based on the temporal variation characteristics and spatial distribution characteristics of the independent components.

5. The method according to claim 1, characterized in that, The process of identifying vegetation responses and water area fluctuations associated with groundwater changes, and the process of obtaining the root zone soil moisture change sequence, include: Radiometric calibration, atmospheric correction, and cloud masking were performed on optical remote sensing images. Enhanced vegetation index or normalized difference water index was calculated, and the changing trends, abrupt change points, and seasonal characteristics of the enhanced vegetation index time series were analyzed to identify vegetation responses associated with groundwater changes. The water body extent was extracted using the normalized difference water index, and the water body area fluctuation series was calculated. The soil moisture remote sensing products were assimilated with the measured soil moisture, and ensemble Kalman filtering, optimal interpolation, or variational assimilation methods were used to obtain the root zone soil moisture change series.

6. The method according to claim 1, characterized in that, The process of constructing a Bayesian multi-source data fusion inversion model and solving for the spatiotemporal distribution of groundwater storage changes includes: The residual term is obtained by subtracting the root zone soil moisture change sequence and the surface water storage change corresponding to the water area fluctuation from the regional terrestrial water storage change signal; the vertical deformation component related to groundwater extraction or recharge is used as a spatial constraint; the prior distribution of groundwater storage change is adjusted by vegetation response and water area fluctuation; and the spatiotemporal distribution of groundwater storage change is solved by maximizing the posterior probability.

7. A groundwater change monitoring system based on multi-source remote sensing, characterized in that, The system for implementing the method of any one of claims 1-6 comprises: The data acquisition module is used to acquire multi-source remote sensing data, including gravity satellite data, synthetic aperture radar interferometry data, optical remote sensing data and soil moisture remote sensing products, to unify the spatiotemporal reference and form a multi-source remote sensing dataset. The data processing module is used to process multi-source remote sensing datasets and extract regional land water storage change signals, vertical deformation components related to groundwater extraction or recharge, vegetation response and water area fluctuations associated with groundwater changes, and root zone soil moisture change sequences. The Bayesian fusion inversion module is used to construct a Bayesian multi-source data fusion inversion model by using regional land water storage change signals as constraints, combined with vertical deformation components, root zone soil moisture change sequences, vegetation response and water area fluctuations, to solve the spatiotemporal distribution of groundwater storage changes. The results generation and verification output module is used to generate groundwater change rate maps and change sequences in the monitoring area, and output the monitoring results after verifying them with the measured groundwater level data.

8. A computer terminal device, characterized in that, include: One or more processors; A memory, coupled to the processor, for storing one or more programs; When the one or more programs are executed by the one or more processors, the one or more processors perform the steps of the method as described in any one of claims 1-6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method as described in any one of claims 1-6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 1-6.