Groundwater storage and change monitoring and early warning system

Through a groundwater storage variable monitoring and early warning system combining satellite radar and gravity satellite data, the problem of low spatial resolution and uncertainty of groundwater storage change monitoring in the prior art is solved, high-resolution monitoring and early warning are achieved, and groundwater is reduced. Economic and ecological losses caused by over-exploitation and depletion.

CN114858238BInactive Publication Date: 2024-11-15SOUTHERN UNIVERSITY OF SCIENCE AND TECHNOLOGY +2
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202210464171.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-04-29
Publication Date
2024-11-15
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The existing technology is difficult to accurately monitor changes in groundwater reserves, especially the low spatial resolution and high uncertainty, and it is impossible to effectively warn of excessive groundwater exploitation and depletion, resulting in ecological environment and economic losses.

Method used

Adopt groundwater storage variable monitoring and early warning system, combined with satellite radar data and gravity satellite data, and synthesize aperture radar interferometry technology and gravity satellite data inversion, realize high-resolution monitoring of changes in groundwater water storage volume, and observe data through groundwater well water level time series Make abnormal judgments and issue warning messages.

Benefits of technology

High-resolution monitoring and early warning of changes in groundwater water storage volume has been achieved, early warning of groundwater over-exploitation and depletion in advance, improved monitoring accuracy and early warning efficiency, and reduced economic losses.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114858238B_ABST
    Figure CN114858238B_ABST
Patent Text Reader

Abstract

The present invention discloses a groundwater storage change monitoring and early warning system, including: a groundwater storage change monitoring module and a groundwater level early warning module. The groundwater storage change monitoring module can be used to perform high-resolution monitoring of groundwater storage changes in a target area. The groundwater level early warning module is used to warn the groundwater well water level in the target area and send early warning information, thereby playing the role of early warning of excessive exploitation and depletion of groundwater, so as to take timely measures to deal with it.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of groundwater monitoring, and in particular to a groundwater storage and change monitoring and early warning system. Background Art

[0002] Groundwater is an important component of fresh water, accounting for up to 33% of the world's total fresh water. Excessive exploitation and depletion of groundwater will cause sea level rise, affect natural runoff, and lead to ecological and environmental problems such as land salinization and deterioration of groundwater quality. At the same time, excessive exploitation of groundwater resources will directly lead to problems such as surface subsidence and cause ground deformation, which is very likely to cause buildings to tilt and walls to crack, resulting in huge economic losses. Therefore, effective monitoring of groundwater reserves changes (i.e., storage changes) is an important basis for achieving water resources management.

[0003] At present, monitoring of groundwater reserves changes mainly includes the use of traditional observation methods, remote sensing monitoring and the establishment of hydrological models. Traditional observation methods include groundwater well observation and leveling observation. Groundwater well observation mainly uses monitoring wells for measurement. This method has high requirements on the number of monitoring points and their representativeness. At the same time, it takes a lot of manpower to complete the preliminary work such as point selection and point arrangement. Due to objective conditions, it is difficult to accurately reflect the changes in groundwater reserves. Leveling observation is costly, labor-intensive, and has low accuracy. In addition, the spatial distribution is relatively sparse, making it difficult to capture the spatial details of aquifer and groundwater changes, and it is impossible to achieve uninterrupted real-time monitoring. Remote sensing monitoring mainly uses optical and thermal infrared data to study groundwater distribution, which has problems such as large uncertainty. Hydrological models are also widely used in monitoring groundwater reserves changes, but they tend to ignore the complexity of hydrological processes, resulting in inaccurate monitoring results. In addition, gravity satellites are an emerging tool for monitoring groundwater reserves changes. Commonly used monitoring methods mainly include estimating groundwater reserves changes based on the principle of water balance and calibrating hydrological models using gravity satellite data. However, groundwater storage monitoring based on gravity satellite data still faces challenges such as low spatial resolution and large uncertainty.

[0004] Therefore, there is an urgent need to provide a groundwater storage and change monitoring and early warning system.

[0005] The information disclosed in this background technology section is only intended to enhance the understanding of the overall background of the invention and should not be regarded as an acknowledgment or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a groundwater storage change monitoring and early warning system, which can perform high-resolution monitoring of groundwater storage changes in a target area and provide early warning of groundwater over-exploitation and depletion, so that timely measures can be taken to deal with it.

[0007] To achieve the above-mentioned purpose, the present invention provides a groundwater storage change monitoring and early warning system, including: a groundwater storage change monitoring module and a groundwater level early warning module; wherein,

[0008] The groundwater storage change monitoring module is used to obtain satellite radar data and gravity satellite data of the target area, and constrain the gravity satellite data based on the satellite radar data to monitor the change of groundwater storage in the target area;

[0009] The groundwater level warning module is used to obtain the groundwater well water level time series observation data of the target area, and based on the groundwater well water level time series observation data and the groundwater well water level observation model, make an abnormal judgment on the groundwater well water level change in the target area, and if the result of the judgment is that the groundwater well water level change is abnormal, then issue the groundwater well water level warning information of the target area;

[0010] The groundwater well water level observation model is trained based on historical groundwater well water level time series observation data carrying abnormal time labels of groundwater well water level changes.

[0011] In one embodiment of the present invention, the groundwater storage and change monitoring module includes:

[0012] A ground subsidence data determination unit, configured to determine ground subsidence data based on the satellite radar data by using a synthetic aperture radar interferometry technique; the ground subsidence data includes a ground subsidence area and a ground subsidence amount;

[0013] An initial change amount determination unit, used for inverting and obtaining an initial change amount of groundwater storage in the land subsidence area based on the gravity satellite data;

[0014] A monitoring unit is used to monitor the change of the groundwater storage capacity based on the correlation between the initial change amount and the ground subsidence amount.

[0015] In one embodiment of the present invention, the land subsidence data determining unit is used to:

[0016] Focusing the satellite radar data to obtain single-view complex data;

[0017] Based on the single-view complex data, a flat ground interference map is generated by using an auxiliary digital elevation model or geometric removal;

[0018] Based on the flat-earth interference pattern, the ground subsidence data is determined.

[0019] In one embodiment of the present invention, the initial change amount determination unit is used to:

[0020] Based on the gravity satellite data, the change in total land water storage is obtained by inversion;

[0021] The initial change is determined based on the total terrestrial water storage change and the GLDAS assimilation system data.

[0022] In one embodiment of the present invention, the gravity satellite data includes Level 2 spherical harmonic coefficient gravity field model data;

[0023] Accordingly, the groundwater storage and change monitoring module also includes a data processing module, which is used to:

[0024] Performing data preprocessing and data postprocessing on the Level 2 spherical harmonic coefficient gravity field model data;

[0025] The data preprocessing includes static field subtraction processing and low-order term replacement;

[0026] The data post-processing includes Gaussian smoothing and decorrelation filtering.

[0027] In one embodiment of the present invention, the groundwater level early warning module includes:

[0028] A feature extraction unit, used for inputting the groundwater well water level time series observation data into the feature extraction layer of the groundwater well water level observation model, and the feature extraction layer determines the spatiotemporal semantic features of the groundwater well water level change based on the attention mechanism;

[0029] A feature recovery unit, used for inputting the spatiotemporal semantic features into the feature recovery layer of the groundwater well water level observation model to obtain the groundwater well water level change information output by the feature recovery layer;

[0030] A decision-making unit is used to input the groundwater level change information into the groundwater threshold decision layer of the groundwater well water level observation model to obtain the judgment result output by the groundwater threshold decision layer.

[0031] In one embodiment of the present invention, the attention mechanism includes a spatial attention mechanism and / or a channel attention mechanism.

[0032] In one embodiment of the present invention, the groundwater level early warning module further includes a sending unit, which is communicatively connected to a terminal device in the target area, and is used to:

[0033] The groundwater well water level warning information is sent to the terminal device.

[0034] In one embodiment of the present invention, the sending unit is connected to the terminal device in the target area through a 5G network communication.

[0035] In one embodiment of the present invention, the decision unit is specifically used for:

[0036] Based on the groundwater threshold decision layer, comparing the groundwater level change information with the groundwater level change threshold;

[0037] If the comparison result is that the groundwater level change information is greater than or equal to the groundwater level change threshold, then determining that the result of the judgment is that the groundwater well water level change is abnormal;

[0038] Otherwise, it is determined that the result of the judgment is that the change in the groundwater well water level is normal.

[0039] Compared with the prior art, the groundwater storage change monitoring and early warning system according to the present invention includes: a groundwater storage change monitoring module and a groundwater level early warning module. The groundwater storage change monitoring module can be used to perform high-resolution monitoring of groundwater storage changes in the target area. The groundwater level early warning module is used to warn the groundwater well water level in the target area and send early warning information, which plays a role in early warning of excessive exploitation and depletion of groundwater, so that timely measures can be taken to deal with it. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 It is a structural schematic diagram of a groundwater storage and change monitoring and early warning system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The specific implementation modes of the present invention are described in detail below in conjunction with the accompanying drawings, but it should be understood that the protection scope of the present invention is not limited by the specific implementation modes.

[0042] Unless explicitly stated otherwise, throughout the specification and claims, the term “comprise” or variations such as “include” or “comprising”, etc., will be understood to include the stated elements or components but not to exclude other elements or components.

[0043] Figure 1 FIG. 1 is a schematic diagram of a groundwater storage and change monitoring and early warning system provided in an embodiment of the present invention. Figure 1 As shown, the system includes: a groundwater storage and change monitoring module 1 and a groundwater level early warning module 2. Among them,

[0044] The groundwater storage change monitoring module 1 is used to obtain satellite radar data and gravity satellite data of the target area, and constrain the gravity satellite data based on the satellite radar data to monitor the change of groundwater storage in the target area;

[0045] The groundwater level warning module 2 is used to obtain the groundwater well water level time series observation data of the target area, and based on the groundwater well water level time series observation data and the groundwater well water level observation model, make an abnormal judgment on the groundwater well water level change in the target area, and if the result of the judgment is that the groundwater well water level change is abnormal, then issue the groundwater well water level warning information of the target area;

[0046] The groundwater well water level observation model is trained based on historical groundwater well water level time series observation data carrying abnormal time labels of groundwater well water level changes.

[0047] Specifically, in the embodiment of the present invention, the groundwater storage and change monitoring and early warning system has two functions, one is the groundwater storage volume change (groundwater storage and change) monitoring function realized by the groundwater storage and change monitoring module 1, and the other is the groundwater well water level early warning function realized by the groundwater level early warning module 2.

[0048] For the groundwater storage and change monitoring module 1, when implementing the corresponding functions, it is first necessary to obtain satellite radar data and gravity satellite data of the target area. Satellite radar data refers to the relevant data of the target area collected by satellites, radars, etc., which is usually InSAR data. Gravity satellite data refers to the relevant data of the target area collected by the GRACE gravity satellite, which is usually satellite gravity data.

[0049] By combining satellite radar data with gravity satellite data, it is possible to monitor the changes in groundwater storage in the target area. Since the spatial resolution of the change in total terrestrial water reserves obtained by inverting gravity satellite data is usually 350km×350km, the spatial resolution of the change in groundwater storage obtained is low. The spatial resolution of satellite radar data is usually 50m-200m, so the spatial resolution of the change in groundwater storage obtained by inverting gravity satellite data can be constrained by satellite radar data, thereby achieving high-resolution monitoring of groundwater storage changes in the target area, which can increase the spatial resolution of groundwater storage changes to 50km×50km.

[0050] For the groundwater level early warning module 2, when realizing its function, it is necessary to first obtain the groundwater well water level time series observation data of the target area, which can be obtained by arranging the data obtained by observing the water level in the groundwater wells in the target area at different times through a water level observation device in chronological order. The water level observation device can be selected as needed and is not specifically limited here.

[0051] Then, the groundwater well water level time series observation data is input into the groundwater well water level observation model, so that the groundwater well water level change in the target area can be judged abnormally, and the groundwater well water level observation model can output the judgment result. The judgment result can include two situations: the groundwater well water level change is normal and abnormal. When the judgment result is that the groundwater well water level change is abnormal, it means that the groundwater well water level in the target area needs to be warned, that is, the groundwater well water level warning information of the target area can be issued through the groundwater level warning module 2. The groundwater well water level warning information can be the current value of the groundwater well water level and the difference between the current value and the normal value, so that the person receiving the information can grasp the water level of the groundwater well in the target area, so as to take measures to deal with it in time.

[0052] Among them, the groundwater well water level observation model can be constructed based on a neural network model, and can be trained by historical groundwater well water level time series observation data carrying abnormal time labels of groundwater well water level changes. In an embodiment of the present invention, the groundwater well water level observation model can also be called a GNET deep learning model.

[0053] For example, the historical groundwater well water level time series observation data can be input into the initial model to obtain the abnormal prediction time of the groundwater well water level change output by the initial model, and then the loss function of the initial model is calculated by the abnormal prediction time of the groundwater well water level change and the abnormal time label of the groundwater well water level change, and the model parameters of the initial model are updated based on the loss function. The historical groundwater well water level time series observation data is replaced, and the above process is repeated until the loss function converges and the training is completed. Finally, the groundwater well water level observation model is obtained.

[0054] The groundwater storage change monitoring and early warning system provided in the embodiment of the present invention includes: a groundwater storage change monitoring module and a groundwater level early warning module. The groundwater storage change monitoring module can be used to perform high-resolution monitoring of groundwater storage changes in the target area. The groundwater level early warning module is used to warn the groundwater well water level in the target area and send early warning information, which plays a role in early warning of excessive exploitation and depletion of groundwater, so that timely measures can be taken to deal with it.

[0055] On the basis of the above-mentioned embodiments, in the groundwater storage and change monitoring and early warning system provided in the embodiments of the present invention, the groundwater storage and change monitoring module comprises:

[0056] A ground subsidence data determination unit, configured to determine ground subsidence data based on the satellite radar data by using a synthetic aperture radar interferometry technique; the ground subsidence data includes a ground subsidence area and a ground subsidence amount;

[0057] An initial change amount determination unit, used for inverting and obtaining an initial change amount of groundwater storage in the land subsidence area based on the gravity satellite data;

[0058] A monitoring unit is used to monitor the change of the groundwater storage capacity based on the correlation between the initial change amount and the ground subsidence amount.

[0059] Specifically, in the embodiment of the present invention, the groundwater storage change monitoring module may include a ground subsidence data determination unit, an initial change amount determination unit, and a monitoring unit. Among them:

[0060] The ground subsidence data determination unit can be used to determine the ground subsidence data based on satellite radar data using synthetic aperture radar interferometry technology. Among them, synthetic aperture radar interferometry (InSAR) technology is a technology that uses radar satellite measurements to map millimeter-level displacements on the earth's surface. It takes into account the constant changes on the earth's surface and can be measured at night and in any weather conditions. Through InSAR technology, the ground subsidence data of the target area can be determined. The ground subsidence data can be characterized by a digital elevation model (DEM), and the ground subsidence data can include the ground subsidence area and the ground subsidence amount.

[0061] The initial change determination unit can be used to invert the initial change of groundwater storage in the land subsidence area through gravity satellite data. The process of obtaining the initial change can be to first invert the total land water storage change through gravity satellite data, and then determine the initial change of groundwater storage in the land subsidence area through the total land water storage change.

[0062] The monitoring unit can be used to monitor the change of groundwater storage through the correlation between the initial change of groundwater storage in the ground subsidence area and the ground subsidence. Here, the correlation between the two types of data with different spatial resolutions can be first constructed, that is, the correlation between the initial change of groundwater storage in the ground subsidence area and the ground subsidence is determined. Among them, the spatial resolution of the initial change is lower than the spatial resolution of the ground subsidence.

[0063] This correlation can be characterized by the corresponding relationship between the two in the same ground subsidence area. Through this correlation, the initial change can be spatially downscaled, and then the high-resolution change of groundwater storage can be determined, thereby realizing the monitoring of groundwater storage changes.

[0064] In the embodiment of the present invention, the groundwater storage change monitoring module monitors the change of groundwater storage capacity through the correlation between the initial change of groundwater storage capacity in the ground subsidence area and the ground subsidence, which can improve the monitoring accuracy.

[0065] On the basis of the above-mentioned embodiment, in the groundwater storage and change monitoring and early warning system provided in the embodiment of the present invention, the ground subsidence data determination unit is used to:

[0066] Focusing the satellite radar data to obtain single-view complex data;

[0067] Based on the single-view complex data, the flat ground interference map is generated by using auxiliary digital elevation model or geometric removal;

[0068] Based on the flat-earth interference pattern, the ground subsidence data is determined.

[0069] Specifically, in an embodiment of the present invention, when determining the ground subsidence data, the ground subsidence data determination unit may first perform data focusing on the satellite radar data to obtain single look complex data. Data focusing is data conversion. Since the satellite radar data is usually in RAW format, the ω-k algorithm of F. Rocca may be used to convert the RAW format data into single look complex (SLC) data. It is understandable that the single look complex data may be a master-slave image pair.

[0070] Then, the flat ground interference map can be generated based on the single-view complex data by using an auxiliary digital elevation model (DEM) or geometric removal. Prior to this, the baseline of the single-view complex data can be estimated first, and then the registration of the single-view complex data can be achieved.

[0071] Here, the flat earth interferogram can be generated by the intensity map of the single-view complex data, the synthetic phase and the slant distance generated by the DEM. The flat earth interferogram can be a multi-view flat earth interferogram. There are two ways to implement this process, one is to use an auxiliary DEM, and the other is to use a method without an auxiliary DEM.

[0072] For high-resolution SAR or terrain with large undulations, auxiliary DEM is usually used. The distance difference between the ground point and the sensor can be calculated by the phase difference between the master and slave images in the single-view complex data. The fringes of the interference image reflect the terrain or displacement information, similar to the contour lines, and the final output is the flattened interference image.

[0073] For other cases, the initial interferogram is usually obtained without auxiliary DEM. It is also necessary to subtract the relevant terrain (or ellipsoid) and constant phase components from the initial interferogram by geometric removal to obtain the flat earth interferogram.

[0074] Finally, the ground subsidence data can be determined through the flat ground interferogram. In this process, the flat ground interferogram can be adaptively filtered and coherently generated, phase unwrapped, orbit redefined, and phase conversion height and geocoding can be performed in sequence to obtain DEM. The DEM can be used to represent the ground subsidence data.

[0075] Among them, adaptive filtering and phase generation refers to filtering the flat-ground interference pattern to remove the phase noise caused by the flat-ground interference, and generating the coherence map of the interference (used to describe the phase quality) and the intensity map of the filtered main image at the same time.

[0076] Since the interference phase can only be modulo 2π, as long as the phase change exceeds 2π, the phase will restart and cycle. Phase unwrapping is to unwrap the phase after de-flattening and filtering to solve the problem of 2π ambiguity. According to the specific data characteristics, phase editing can be further performed to correct the errors of phase unwrapping.

[0077] Orbit redefinition refers to correcting the satellite orbit and phase offset, calculating orbit refinement and phase offset, and eliminating possible slope phases. This step is critical to whether the unwrapped phase can be correctly converted into elevation or deformation values. This step must be done whether generating DEM or deformation results.

[0078] Elevation conversion and geocoding,This step is to convert the actual phase that has been absolutely calibrated and unwrapped,combined with the synthetic phase, into DEM and geocode it,which is the map projection.

[0079] In the embodiment of the present invention, in the process of determining the ground subsidence data, an auxiliary digital elevation model or a geometric removal method can be used to generate a flat ground interference map, which can make the ground subsidence data more accurate.

[0080] On the basis of the above-mentioned embodiment, in the groundwater storage change monitoring and early warning system provided in the embodiment of the present invention, the initial change amount determination unit is used to:

[0081] Based on the gravity satellite data, the change in total land water storage is obtained by inversion;

[0082] The initial change is determined based on the total terrestrial water storage change and the GLDAS assimilation system data.

[0083] Specifically, in the embodiment of the present invention, when determining the initial change amount, the initial change amount determination unit may first invert the total land water storage change amount based on the gravity satellite data. In this process, the density change of the earth's surface may be calculated first, and then the density change may be converted into the change of equivalent water height, so as to invert the distribution of land water storage.

[0084] Then, the initial change can be determined based on the change in total terrestrial water reserves and the GLDAS assimilation system data. The VIC model in the GLDAS assimilation system is a semi-distributed macro-hydrological model. Using the GLDAS VIC model, the moisture content of soil at a depth of 0 to 200 cm, as well as the properties of canopy water and surface water can be simulated. Therefore, the GLDAS assimilation system data can be the GLDAS hydrological model data. Furthermore, by deducting the GLDAS hydrological model data from the change in total terrestrial water reserves, the initial change in groundwater storage in the land subsidence area can be obtained.

[0085] In the embodiment of the present invention, a method for determining the initial variation is provided, and the GLDAS assimilation system data is introduced, so that the result of the initial variation can be made more accurate.

[0086] On the basis of the above-mentioned embodiment, in the groundwater storage and change monitoring and early warning system provided in the embodiment of the present invention, the gravity satellite data includes Level 2 spherical harmonic coefficient gravity field model data;

[0087] Accordingly, the groundwater storage and change monitoring module also includes a data processing module, which is used to:

[0088] Performing data preprocessing and data postprocessing on the Level 2 spherical harmonic coefficient gravity field model data;

[0089] The data preprocessing includes static field subtraction processing and low-order term replacement;

[0090] The data post-processing includes Gaussian smoothing and decorrelation filtering.

[0091] Specifically, in the embodiment of the present invention, the gravity satellite data may include Level 2 spherical harmonic coefficient gravity field model data and Mascon data.

[0092] Among them, for the spherical harmonic coefficient gravity field model data, the data processing module in the groundwater storage and change monitoring module can be used to perform data preprocessing and data post-processing.

[0093] Data preprocessing can include static field subtraction and low-order term replacement. Since the original gravity field consists of two parts, the static field and the time-varying field, after subtracting the static field, the part of the gravity field that changes with time can be obtained. At the same time, since the satellite cannot measure the change of the center of the earth, the first-order term coefficient in the spherical harmonic coefficient is 0, so it needs to be replaced according to the corresponding technical documentation. At the same time, the accuracy of the measured earth flattening may be low, resulting in C 20 The accuracy of this item is low, and it is necessary to use the C measured by Satellite Laser Ranging (SLR). 20Here, the technical documents may include TN-13a, TN-14, etc.

[0094] Data post-processing can include Gaussian smoothing and decorrelation filtering. When performing Gaussian filtering, you can first construct a Gaussian function kernel to filter the spherical harmonic coefficients and reduce the weight of high-order coefficients to achieve the purpose of suppressing high-order errors. When constructing a Gaussian function kernel, you need to select a suitable filter radius, which is selected according to the target area. The Gaussian function kernel is as follows:

[0095]

[0096]

[0097] Where r is the distance between two points on the earth’s surface, which is called the Gaussian filter smoothing radius; a is the radius of the earth, W is the weight factor, W0 is the 0th order weight factor, W1 is the 1st order weight factor, and W l+1 is the l+1th order weight factor, and b is the bias term.

[0098] Decorrelation filtering can be a destriping operation, which can be implemented by Swenson destriping filtering. Swenson destriping filtering is a decorrelation filter based on a sliding window polynomial. Among them, since there is no obvious correlation between the low-order spherical harmonic coefficients, no decorrelation filtering operation is performed on them. There is a strong correlation between the high-order spherical harmonic coefficients, that is, when the order is greater than n, the correlation of the spherical harmonic coefficients is eliminated by filtering. Since Swenson destriping filtering takes into account the different parities of different orders at the same order, the filtered spherical harmonic coefficients can be written as:

[0099]

[0100] Among them, l is the order of the spherical harmonic coefficient of the time-varying gravity field, m is the degree, w is the width of the sliding window, Λ is the Swenson convolution kernel parameter, and its calculation formula is:

[0101]

[0102] in,

[0103]

[0104] It can be understood that the calculation process of Λ is a polynomial least squares fitting process, and p is the degree of the fitting polynomial.

[0105] In the embodiment of the present invention, by performing data preprocessing and data postprocessing on the spherical harmonic coefficient gravity field model data, the difficulty of monitoring the change of groundwater storage capacity in the target area can be reduced and the monitoring accuracy can be improved.

[0106] On the basis of the above-mentioned embodiment, the groundwater storage and change monitoring and early warning system provided in the embodiment of the present invention, the groundwater level early warning module includes:

[0107] A feature extraction unit, used for inputting the groundwater well water level time series observation data into the feature extraction layer of the groundwater well water level observation model, and the feature extraction layer determines the spatiotemporal semantic features of the groundwater well water level change based on the attention mechanism;

[0108] A feature recovery unit, used for inputting the spatiotemporal semantic features into the feature recovery layer of the groundwater well water level observation model to obtain the groundwater well water level change information output by the feature recovery layer;

[0109] The decision unit is used to input the groundwater level change information into the groundwater threshold decision layer to obtain the judgment result output by the groundwater threshold decision layer.

[0110] Specifically, in an embodiment of the present invention, the groundwater level warning module may include a feature extraction unit, a feature recovery unit and a decision unit, and the groundwater well water level observation model may include a feature extraction layer, a feature recovery layer and a groundwater threshold decision layer.

[0111] Through the feature extraction unit, the groundwater well water level time series observation data can be input into the feature extraction layer of the groundwater well water level observation model, and the feature extraction layer determines the spatiotemporal semantic features of the groundwater well water level changes based on the attention mechanism.

[0112] The feature extraction layer may be a convolution layer, which may include a convolution module, a pooling module, and an activation module. The convolution module is used to perform a convolution operation on the candidate area using an attention mechanism through a convolution kernel, the pooling module is used to perform a pooling operation on the result of the convolution operation, and the activation module is used to perform an activation operation on the result of the pooling operation.

[0113] The groundwater well water level observation model may include one or more feature extraction layers. When multiple feature extraction layers are included, the convolution kernel size of the convolution module in each feature extraction layer may be the same or different, and is not specifically limited here. For example, there are three feature extraction layers, and the three feature extraction layers are connected in sequence, and the convolution kernel size of the convolution module in the previous feature extraction layer decreases in sequence. The convolution kernel size of the convolution module in the first feature extraction layer may be 512×512, the convolution kernel size of the convolution module in the second feature extraction layer may be 256×256, and the convolution kernel size of the convolution module in the third feature extraction layer may be 128×128.

[0114] When the convolution module performs convolution operations on the candidate areas through the convolution kernel, it can use a series of convolution sliding calculation methods to perform convolution operations on the time series of groundwater well water level time series observation data according to the convolution kernel size and the sliding step size, thereby obtaining the spatiotemporal semantic characteristics of the groundwater well water level changes.

[0115] The activation module needs to use activation functions such as Sigmoid function or ReLu function to transform the linear connection in the groundwater well water level observation model into a nonlinear relationship, so that the mapping of the input and output of the groundwater well water level observation model is more in line with the actual situation, and automatic decision-making can be made on the spatiotemporal semantic features extracted by the feature extraction layer.

[0116] In this groundwater well water level observation model, not only are several feature extraction layers set up to extract the spatiotemporal semantic features of groundwater well water level changes, but the spatiotemporal semantic features extracted by the previous feature extraction layers can also be calculated with the current feature extraction layer to make more effective use of shallow spatiotemporal semantic features and deep spatiotemporal semantic features, thereby improving computational efficiency.

[0117] Through the feature recovery unit, the spatiotemporal semantic features can be input into the feature recovery layer of the groundwater well water level observation model to obtain the groundwater well water level change information output by the feature recovery layer. Here, the feature recovery layer can also be a convolution layer, and the feature extraction layer can include a deconvolution module, a pooling module, and an activation module. The deconvolution module is used to perform a deconvolution operation on the spatiotemporal semantic features through a convolution kernel, the pooling module is used to perform a pooling operation on the result of the deconvolution operation, and the activation module is used to perform an activation operation on the result of the pooling operation.

[0118] The groundwater well water level observation model may include one or more feature recovery layers, and the number of feature recovery layers is equal to the number of feature extraction layers. When there are multiple feature recovery layers, the convolution kernel size of the deconvolution module in each feature recovery layer may be the same or different, and is not specifically limited here. For example, there are three feature recovery layers, and the three feature recovery layers are connected in sequence, and the convolution kernel size of the deconvolution module in the previous feature recovery layer decreases in sequence. The convolution kernel size of the deconvolution module in the first feature recovery layer may be 128×128, the convolution kernel size of the deconvolution module in the second feature recovery layer may be 256×256, and the convolution kernel size of the deconvolution module in the third feature recovery layer may be 512×512.

[0119] The structure and function of the pooling module and activation module in the feature recovery layer are consistent with those of the pooling module and activation module in the feature extraction layer, and will not be repeated here.

[0120] Through the decision-making unit, the groundwater level change information can be input into the groundwater threshold decision layer of the groundwater well water level observation model to obtain the judgment result output by the groundwater threshold decision layer. After the groundwater well water level observation model is completed, the groundwater threshold decision layer stores the groundwater level change threshold. Therefore, the groundwater threshold decision layer can compare the groundwater level change information with the groundwater level change threshold, and then make an abnormal judgment on the groundwater well water level change in the target area.

[0121] In an embodiment of the present invention, introducing the attention mechanism of the feature extraction layer in the groundwater well water level observation model can make the obtained spatiotemporal semantic features more accurate, improve the accuracy of the groundwater well water level change information, and thus improve the accuracy of the judgment results.

[0122] On the basis of the above-mentioned embodiments, in the groundwater storage and change monitoring and early warning system provided in the embodiments of the present invention, the attention mechanism includes a spatial attention mechanism and / or a channel attention mechanism.

[0123] Specifically, in an embodiment of the present invention, the attention mechanism introduced in the feature extraction layer may include at least one of a spatial attention mechanism and a temporal attention mechanism. The spatial attention mechanism can assist in obtaining spatiotemporal semantic features spatially, and the temporal attention mechanism can assist in obtaining spatiotemporal semantic features temporally.

[0124] The attention mechanism aims to tell the groundwater well water level observation model the key areas to learn. Through the attention mechanism, mathematical operations can be used to increase the weights of the groundwater well water level change areas, and the new weights obtained through the attention mechanism are combined with the convolutional layer in the channel dimension to improve the accuracy of the groundwater well water level change information.

[0125] The groundwater well water level observation model can use two attention mechanisms, namely spatial attention mechanism and channel attention mechanism. The channel attention mechanism can be expressed by the following formula:

[0126] M c (F)=δ(MLP(AvgPool(F))+MLP(MaxPool(F)))

[0127] Among them, δ is the activation function, F is the intermediate feature map, and MLP is a multi-layer perceptron with shared weights.

[0128] Different from the channel attention mechanism, the spatial attention mechanism focuses on the position information. First, it uses the maximum pooling and average pooling on the channel dimension to obtain two different feature layers, then merges the two feature layers and uses the convolution operation to generate the spatial attention distribution map. The calculation process of the spatial attention mechanism is shown in the following formula:

[0129] M c (F) = δ(f n*n ([AvgPool(f)]; MaxPool(F)])

[0130] Among them, f n*n It is an n*n convolution operation.

[0131] In the embodiments of the present invention, different attention mechanisms are introduced to assist in the extraction of spatiotemporal semantic features from different angles, thereby improving the extraction accuracy.

[0132] On the basis of the above-mentioned embodiment, in the groundwater storage and change monitoring and early warning system provided in the embodiment of the present invention, the groundwater level early warning module further includes a sending unit, the sending unit is communicatively connected with the terminal device in the target area, and the sending unit is used to:

[0133] The groundwater well water level warning information is sent to the terminal device.

[0134] Specifically, in the embodiment of the present invention, the groundwater level warning module further includes a sending unit, which can be communicatively connected with a terminal device in the target area, and the user holding the terminal device may be a user of all mobile devices in the target area.

[0135] Through the sending unit, groundwater well water level warning information can be sent to the terminal device, so that the user holding the terminal device can obtain the groundwater well water level warning information in time.

[0136] On the basis of the above-mentioned embodiments, in the groundwater storage and change monitoring and early warning system provided in the embodiments of the present invention, the sending unit is connected to the terminal device in the target area through 5G network communication, so that rapid transmission of information between the sending unit and the terminal device can be achieved.

[0137] On the basis of the above-mentioned embodiments, in the groundwater storage and change monitoring and early warning system provided in the embodiments of the present invention, the decision unit is specifically used for:

[0138] Based on the groundwater threshold decision layer, comparing the groundwater level change information with the groundwater level change threshold;

[0139] If the comparison result is that the groundwater level change information is greater than or equal to the groundwater level change threshold, then determining that the result of the judgment is that the groundwater well water level change is abnormal;

[0140] Otherwise, it is determined that the result of the judgment is that the change in the groundwater well water level is normal.

[0141] Specifically, in the embodiment of the present invention, the decision unit can compare the groundwater level change information with the groundwater level change threshold through the groundwater threshold decision layer, and judge the comparison result. If the comparison result is that the groundwater level change information is greater than or equal to the groundwater level change threshold, it can be determined that the result of the judgment is that the groundwater well water level change is abnormal. Otherwise, if the comparison result is that the groundwater level change information is less than the groundwater level change threshold, it can be determined that the result of the judgment is that the groundwater well water level change is normal.

[0142] In summary, the groundwater storage and change monitoring and early warning system provided in the embodiment of the present invention performs spatial downscaling at the functional level through the correlation between groundwater changes monitored by GRACE and surface subsidence monitored by InSAR technology, thereby improving the spatial resolution of groundwater inverted by GRACE data. In the process of data collection and broadcasting, a 5G wireless network transmission module is used for data transmission, which can take into account both data quality and speed, and can quickly respond to early warnings. In establishing a groundwater monitoring model, GNSS data, measured groundwater storage data, and InSAR data can be combined for comprehensive analysis to improve the resolution, accuracy, and reliability of GRACE gravity satellite monitoring of groundwater. In the process of data training and decision-making, several avoidance design schemes may appear, that is, the data source simultaneously uses high-resolution satellite remote sensing images for comprehensive training and decision-making. In the process of deep learning network model design, a similar processing level or faster processing speed can be achieved through a newer network structure.

[0143] The foregoing description of specific exemplary embodiments of the present invention is for the purpose of illustration and demonstration. These descriptions are not intended to limit the present invention to the precise form disclosed, and it is clear that many changes and variations can be made based on the above teachings. The purpose of selecting and describing the exemplary embodiments is to explain the specific principles of the present invention and its practical application, so that those skilled in the art can realize and utilize various different exemplary embodiments of the present invention and various different selections and changes. The scope of the present invention is intended to be limited by the claims and their equivalents.

Claims

1. A groundwater storage and change monitoring and early warning system, characterized in that: include: Groundwater storage and change monitoring module and groundwater well water level early warning module; among them, The groundwater storage change monitoring module is used to obtain satellite radar data and gravity satellite data of the target area, and constrain the gravity satellite data based on the satellite radar data to monitor the changes in groundwater storage in the target area; The groundwater well water level warning module is used to obtain the groundwater well water level time series observation data of the target area, and based on the groundwater well water level time series observation data and the groundwater well water level observation model, perform abnormal judgment on the groundwater well water level change in the target area, and if the result of the judgment is that the groundwater well water level change is abnormal, then issue the groundwater well water level warning information of the target area; The groundwater well water level observation model is trained based on historical groundwater well water level time series observation data that carries abnormal time labels of groundwater well water level changes; Wherein, the groundwater storage and change monitoring module includes: a ground subsidence data determining unit, configured to determine ground subsidence data based on the satellite radar data using a synthetic aperture radar interferometry technique; the ground subsidence data including a ground subsidence area and a ground subsidence amount; an initial change determination unit, configured to obtain, based on the gravity satellite data, an initial change in groundwater storage capacity in the land subsidence area by inversion; A monitoring unit is used to monitor the change of the groundwater storage capacity based on the correlation between the initial change amount and the ground subsidence amount.

2. The groundwater storage and change monitoring and early warning system according to claim 1 is characterized in that: The ground subsidence data determining unit is configured to: Focusing the satellite radar data to obtain single-view complex data; Based on the single-view complex data, a flat-earth interferogram is generated by using an auxiliary digital elevation model or geometric removal; The ground subsidence data is determined based on the flat earth interferogram.

3. The groundwater storage and change monitoring and early warning system according to claim 1 is characterized in that: The initial variation determination unit is configured to: Based on the gravity satellite data, the change in total land water storage is obtained by inversion; The initial change is determined based on the change in the total land water storage and GLDAS assimilation system data.

4. The groundwater storage and change monitoring and early warning system according to claim 1 is characterized in that: The gravity satellite data includes Level 2 spherical harmonic coefficient gravity field model data; Accordingly, the groundwater storage and change monitoring module further includes a data processing module for: Performing data preprocessing and data postprocessing on the Level 2 spherical harmonic coefficient gravity field model data; The data preprocessing includes static field deduction processing and low-order term replacement; The data post-processing includes Gaussian smoothing and decorrelation filtering.

5. The groundwater storage and change monitoring and early warning system according to any one of claims 1 to 4, characterized in that: The groundwater well water level early warning module includes: a feature extraction unit, configured to input the groundwater well water level time series observation data into a feature extraction layer of the groundwater well water level observation model, and the feature extraction layer determines the spatiotemporal semantic features of the groundwater well water level changes based on an attention mechanism; a feature recovery unit, configured to input the spatiotemporal semantic features into a feature recovery layer of the groundwater well water level observation model, and obtain the groundwater well water level change information output by the feature recovery layer; A decision-making unit is used to input the groundwater well water level change information into the groundwater threshold decision layer of the groundwater well water level observation model to obtain the judgment result output by the groundwater threshold decision layer.

6. The groundwater storage and change monitoring and early warning system according to claim 5 is characterized in that: The attention mechanism includes a spatial attention mechanism and / or a channel attention mechanism.

7. The groundwater storage and change monitoring and early warning system according to claim 5 is characterized in that: The groundwater well water level warning module further includes a sending unit, which is communicatively connected to a terminal device in the target area and is configured to: The groundwater well water level warning information is sent to the terminal device.

8. The groundwater storage and change monitoring and early warning system according to claim 7 is characterized in that: The sending unit is connected to the terminal device in the target area through 5G network communication.

9. The groundwater storage and change monitoring and early warning system according to claim 5 is characterized in that: The decision-making unit is specifically used for: Based on the groundwater threshold decision layer, the groundwater well water level change information is compared with the groundwater well water level change threshold; If the comparison result is that the groundwater well water level change information is greater than or equal to the groundwater well water level change threshold, then determining that the result of the judgment is that the groundwater well water level change is abnormal; Otherwise, it is determined that the result of the judgment is that the change in the groundwater well water level is normal.