Method of separating elastic and aperture elastic deformations and associated apparatus

Abstract

The invention discloses a method for separating elastic deformation and pore elastic deformation and related equipment, and relates to the field of geodesy, and the method comprises the steps: taking a preset number of InSAR earth surface deformation time sequences as an initial mass center, and employing a spatial clustering algorithm to obtain a plurality of clusters; for each cluster, determining the category of the cluster based on the average value and spatial distribution characteristics of the InSAR earth surface deformation time sequence in the cluster; if the category of the cluster is the pore elastic deformation dominant category, obtaining a pore elastic deformation dominant deformation time sequence according to the InSAR earth surface deformation time sequence in the cluster; if the category of the cluster is a seasonal pore elastic deformation and elastic load deformation aliasing category, performing principal component analysis on the InSAR earth surface deformation time sequence in the cluster to obtain an elastic load deformation sequence and a seasonal pore elastic deformation sequence; according to external data, separation is proved to be successful. According to the method, the elastic deformation and the hole elastic deformation can be effectively separated from the InSAR earth surface deformation.

Description

Technical Field

[0001] This application relates to the field of geodesy, and in particular to methods and related equipment for separating elastic and porosity elastic deformations. Background Technology

[0002] Interferometric Synthetic Aperture Radar (InSAR) has become an important tool for monitoring large-scale, high-precision surface deformation, especially in the field of hydrogeology, where it is widely used to observe land subsidence and rebound caused by groundwater extraction or recharge. These deformation signals are of crucial value for understanding aquifer system behavior, assessing geological hazard risks, and managing groundwater resources. In related technologies, surface deformation data of aquifer systems obtained through InSAR monitoring are typically processed and analyzed as a single, unified deformation field. Conventional methods directly correlate these total deformations with changes in groundwater head statistically, or input them into simple compaction models to estimate aquifer parameters or invert changes in groundwater storage. However, from a physical perspective, surface deformation caused by groundwater changes mainly comprises two distinct components: elastic load deformation and pore elastic deformation. Current technical solutions have failed to effectively separate these two physical mechanisms of deformation from the total deformation field observed by InSAR. This shortcoming prevents the full realization of the application potential of InSAR technology in hydrogeology. It is neither able to accurately correlate changes in hydrological loads nor to directly and accurately obtain key information reflecting the compaction state of aquifers from deformation data. Summary of the Invention

[0003] The purpose of this application is to provide a method and related equipment for separating elastic and porosimetric deformations, which can effectively separate elastic and porosimetric deformations from InSAR surface deformations.

[0004] To achieve the above objectives, this application provides the following solution: In a first aspect, this application provides a method for separating elastic and porosimetric deformations from InSAR surface deformations, comprising: S1. Acquire the InSAR surface deformation dataset and external data for the target area; the InSAR surface deformation dataset includes multiple InSAR surface deformation time series. S2. Using a preset number of InSAR surface deformation time series as initial centroids, a spatial clustering algorithm is used to spatially cluster multiple InSAR surface deformation time series to obtain multiple clusters; S3. For each of the multiple clusters, extract the spatial distribution characteristics of the InSAR surface deformation time series within the cluster, obtain the average deformation time series based on the average value of the InSAR surface deformation time series within the cluster, and determine the category to which the cluster belongs based on the average deformation time series and the spatial distribution characteristics. S4. If the cluster belongs to the category of porosity elastic deformation-dominated, then the deformation time series dominated by porosity elastic deformation is obtained based on the InSAR surface deformation time series within the cluster. S5. If the cluster belongs to the category of mixed seasonal porosity elastic deformation and elastic load deformation, then perform principal component analysis on the InSAR surface deformation time series within the cluster, and obtain the elastic load deformation sequence and seasonal porosity elastic deformation sequence based on the principal component analysis results. S6. Calculate the Pearson correlation coefficients between the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series and the corresponding external data. If each Pearson correlation coefficient is greater than the corresponding set threshold, the separation is successful; otherwise, return to S2 and perform spatial clustering again until the separation is successful.

[0005] Secondly, this application provides a system for separating elastic and porosimetric deformations from InSAR surface deformations, comprising: The data acquisition module is used to acquire InSAR surface deformation datasets and external data for the target area; the InSAR surface deformation dataset includes multiple InSAR surface deformation time series. The clustering module is used to take a preset number of InSAR surface deformation time series as initial centroids and use a spatial clustering algorithm to spatially cluster multiple InSAR surface deformation time series to obtain multiple clusters; The determination module is used to extract the spatial distribution characteristics of the InSAR surface deformation time series within each of the multiple clusters, obtain the average deformation time series based on the average value of the InSAR surface deformation time series within the cluster, and determine the category to which the cluster belongs based on the average deformation time series and the spatial distribution characteristics. The pore elastic deformation acquisition module is used to obtain the deformation time series dominated by pore elastic deformation based on the InSAR surface deformation time series within the cluster if the cluster belongs to the category dominated by pore elastic deformation. The separation module is used to perform principal component analysis on the InSAR surface deformation time series within the cluster if the cluster belongs to the category of mixed seasonal porosity elastic deformation and elastic load deformation, and to obtain the elastic load deformation sequence and seasonal porosity elastic deformation sequence based on the principal component analysis results. The verification module is used to calculate the Pearson correlation coefficients of the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series with the corresponding external data. If each Pearson correlation coefficient is greater than the corresponding set threshold, it proves that the separation is successful; otherwise, spatial clustering is performed again until the separation is successful.

[0006] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described above for separating elastic and porosimetric deformations from InSAR surface deformation.

[0007] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method and related equipment for separating elastic and porosity elastic deformation. First, based on spatial clustering of InSAR time series, InSAR surface deformation time series with similar evolutionary behaviors are automatically divided into different clusters. The clusters are initially classified into two categories: porosity elastic deformation-dominated and a mixture of seasonal porosity elastic deformation and elastic loading deformation. Second, for the clusters in the mixed category of seasonal porosity elastic deformation and elastic loading deformation, principal component analysis is used to reconstruct the elastic loading deformation sequence and the seasonal porosity elastic deformation sequence. Furthermore, by using the average deformation time series within each cluster as a representative signal, pixel-level random noise is effectively suppressed, significantly improving the signal-to-noise ratio. Finally, to determine the physical reliability and robustness of the separation results, this application sets external data verification as a mandatory closed-loop step, evaluates the correlation between the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series and the corresponding external data, and uses this feedback to optimize the process, ultimately achieving accurate separation of the elastic load deformation series and the seasonal pore elastic deformation series. After separation, the direct compaction response signal can be obtained through the elastic load deformation series, and the compaction / rebound state of the aquifer system can be quantitatively reflected through the seasonal pore elastic deformation series. This allows for more accurate correlation with changes in groundwater head, which can then be used for aquifer parameter inversion and geomechanical behavior analysis. Attached Figure Description

[0008] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 A flowchart illustrating a method for separating elastic and porosimetric deformations from InSAR surface deformation provided in this application; Figure 2 A schematic diagram of the clustering results obtained by the k-means clustering algorithm provided in this application; Figure 3 A schematic diagram of the pure elastic load deformation sequence and the seasonal pore elastic deformation sequence provided for this application; Figure 4 A schematic diagram of the temporal characteristics of the first principal component provided in this application; Figure 5 The first principal component score diagram provided for this application; Figure 6 A schematic diagram comparing the pore elastic deformation and groundwater head observation data provided in this application; Figure 7 A schematic diagram of the technical route for a method to separate elastic and porosimetric deformation from InSAR surface deformation provided in this application; Figure 8 A schematic diagram of the functional modules of the system for separating elastic and porosimetric deformation from InSAR surface deformation provided in this application; Figure 9 This is a schematic diagram of the structure of a computer device provided in this application. Detailed Implementation

[0010] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0011] To make the objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0012] like Figure 1 As shown, this application provides a method for separating elastic and porosimetric deformations from InSAR surface deformation. The core of this method lies in utilizing the physical differences in the temporal and spatial characteristics of elastic and porosimetric deformations to achieve automated separation through a "clustering-discrimination-purification-verification" process. This method includes the following steps S1 to S6: S1. Acquire the InSAR surface deformation dataset and external data for the target area; the InSAR surface deformation dataset includes multiple InSAR surface deformation time series. After acquiring the InSAR surface deformation dataset for the target area, perform atmospheric correction and leveling preprocessing to unify it into an absolute deformation field. Simultaneously collect two sets of external data: regional terrestrial water storage change data, from hydrological models or satellite gravity observation data; and groundwater head observation data.

[0013] S2. Using a predetermined number of InSAR surface deformation time series as initial centroids, a spatial clustering algorithm is employed to spatially cluster multiple InSAR surface deformation time series, resulting in multiple clusters. The morphological features include overall trends and seasonality. This application utilizes unsupervised machine learning, specifically K-means clustering, to spatially cluster multiple InSAR surface deformation time series. This clustering is not based on simple geographical or geometric features, but rather on time series features that reflect the physical mechanisms of deformation, such as trends and seasonality, thus making the classification results more accurate.

[0014] Specifically, the spatial clustering algorithm used is the K-means algorithm. The K-means algorithm is a classic unsupervised machine learning method. The clustering process of the K-means algorithm is as follows: A predetermined number of InSAR surface deformation time series are used as initial centroids. The Euclidean distance between each InSAR surface deformation time series in the target area's InSAR surface deformation dataset (excluding the initial centroid) and the initial centroid is calculated. Each InSAR surface deformation time series is then assigned to a cluster based on the Euclidean distance. The average value of the InSAR surface deformation time series in each cluster is calculated, and the centroid is updated based on the average value until the centroid no longer changes, at which point the iteration stops. The predetermined number of centroids can be determined based on statistical criteria, such as the elbow rule or prior geological knowledge.

[0015] The core objective of the K-means algorithm is to group deformation points with similar temporal evolution behavior into the same cluster. Specifically, it automatically divides a set of data points into K clusters, ensuring that data points within the same cluster are similar to each other, while data points in different clusters differ significantly. The algorithm iteratively optimizes to minimize the sum of squared errors within each cluster. This application uses the following formula to determine the cluster category: .

[0016] in, Let be the number of clusters, and i be the index of the cluster. For the first Clusters, For the first InSAR surface deformation time series within a cluster For the first The center of each cluster, It is a Euclidean distance.

[0017] S3. For each of the multiple clusters, extract the spatial distribution characteristics of the InSAR surface deformation time series within the cluster, obtain the average deformation time series based on the average value of the InSAR surface deformation time series within the cluster, and determine the category to which the cluster belongs based on the average deformation time series and the spatial distribution characteristics.

[0018] The pore elastic deformation-dominated type exhibits a significant long-term settlement or rebound trend over time, with a relatively high rate and minimal seasonal fluctuations. Spatially, its distribution highly coincides with the confined zones of major aquifers, concentrated in areas of intense groundwater extraction. Secondly, the mixed type of seasonal pore elastic deformation and elastic load deformation, based on a small long-term trend over time, is superimposed with seasonal fluctuations that coincide with seasonal changes in groundwater head or terrestrial water storage. Its spatial distribution is usually more discrete, commonly found in hydrologically sensitive areas.

[0019] S4. If the cluster belongs to the category of porosity elastic deformation-dominated, then the deformation time series dominated by porosity elastic deformation is obtained based on the InSAR surface deformation time series within the cluster.

[0020] Specifically, when the cluster belongs to the category of porosity-elastic deformation-dominated, the InSAR surface deformation time series within the cluster is a porosity-elastic deformation-dominated deformation time series.

[0021] S5. If the cluster belongs to the category of mixed seasonal porosity elastic deformation and elastic load deformation, then perform principal component analysis on the InSAR surface deformation time series within the cluster, and obtain the elastic load deformation sequence and seasonal porosity elastic deformation sequence based on the principal component analysis results.

[0022] Principal Component Analysis (PCA) is a commonly used data dimensionality reduction technique. It projects the original high-dimensional data into a lower-dimensional space through linear transformation.

[0023] S6. Calculate the Pearson correlation coefficients between the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series and the corresponding external data. If each Pearson correlation coefficient is greater than the corresponding set threshold, the separation is successful; otherwise, return to S2 and perform spatial clustering again until the separation is successful.

[0024] The elastic load deformation mentioned above refers to the instantaneous, reversible elastic bending of the underlying solid Earth's framework caused by changes in the mass load of surface water, such as lakes and reservoirs, or groundwater bodies. It primarily reflects increases or decreases in regional hydrological load. Pore elastic deformation, on the other hand, originates from changes in the pore fluid pressure of an aquifer, i.e., changes in hydraulic head, leading to compression or expansion of the porous media framework. This deformation directly characterizes the compaction or rebound process of the aquifer itself and is closely related to the spatiotemporal variations of groundwater head.

[0025] In view of this, this application can effectively separate the elastic deformation component caused by hydrological mass load from the pore elastic deformation component caused by changes in aquifer pore pressure in the InSAR surface deformation data of the target area. Furthermore, after separation, the changes in regional surface water and groundwater storage, i.e., hydrological load, can be constrained and inverted through the elastic load deformation sequence to obtain a direct compaction response signal. Simultaneously, the pore elastic deformation sequence can directly and quantitatively reflect the compaction / rebound state of the aquifer system, enabling more accurate correlation with groundwater head changes, thus facilitating aquifer parameter inversion and geomechanical behavior analysis. This application will overcome the limitations of existing methods that confuse the two components from a technical perspective, significantly improving the physical clarity, interpretation accuracy, and practical value of InSAR surface deformation data in hydrological and geomechanical applications.

[0026] After classifying the clusters, this application employs principal component analysis and harmonic analysis to separate and purify the mixed categories of seasonal hole elastic deformation and elastic load deformation in order to obtain pure elastic load deformation sequences and seasonal hole elastic deformation sequences. This filters out intra-cluster noise and extracts pure physical components, improving the quality of the separated signals and the reliability of subsequent applications.

[0027] The following section will elaborate on the steps in S5 above, which involve performing principal component analysis on the InSAR surface deformation time series within the cluster and obtaining the elastic load deformation sequence and the seasonal porosity elastic deformation sequence based on the principal component analysis results.

[0028] One feasible approach is that, in S5 above, principal component analysis is performed on the InSAR surface deformation time series within the cluster, and the elastic load deformation sequence and seasonal porosity elastic deformation sequence are obtained based on the principal component analysis results, specifically including the following S51 to S53. Wherein: S51, for each InSAR surface deformation time series in the cluster, perform the following operation: The InSAR surface deformation time series of the first target is fitted using the least squares method. Based on the fitting result, the trend term in the InSAR surface deformation time series of the first target is subtracted to obtain the InSAR surface deformation time series of the second target.

[0029] Wherein, the first target InSAR surface deformation time series is any InSAR surface deformation time series. The first target InSAR surface deformation time series is fitted using the following formula: .

[0030] in, The first target is the InSAR surface deformation time series. This refers to the a-th epoch in the InSAR surface deformation time series of the first target. For the intercept term, For linear speed, This represents the trend term of the least squares method.

[0031] S52, perform principal component analysis on the second target InSAR surface deformation time series corresponding to the InSAR surface deformation time series, and retain the first principal component to obtain the principal component analysis results corresponding to the InSAR surface deformation time series.

[0032] S53, Based on the principal component analysis results, the elastic load deformation sequence and the seasonal porosity elastic deformation sequence are obtained. The principal component analysis results include the time feature vector of the first principal component and the score vector of the first principal component.

[0033] The first principal component score vector shows alternating positive and negative values. Deformation points with positive scores in the first principal component are reconstructed as elastic load deformation sequences, while deformation points with negative scores in the first principal component are reconstructed as seasonal pore elastic deformation sequences.

[0034] This application uses principal component analysis to obtain the elastic load deformation sequence and the seasonal orifice elastic deformation sequence. Separation can be achieved solely based on the statistical characteristics of the observed data, i.e., the covariance structure, without prior knowledge of the specific mixing coefficients of the elastic load deformation sequence and the seasonal orifice elastic deformation sequence, demonstrating strong robustness. Furthermore, this application retains the first principal component, i.e., the most important signal, while automatically filtering out other components representing noise, local anomalies, or minor processes, significantly improving the signal-to-noise ratio. In addition, this application uses the spatial differentiation of the first principal component score vector as the separation criterion. This criterion has a clear mathematical definition and objective calculation, ensuring the repeatability of the decoupling process and the verifiability of the results.

[0035] In S53 above, the steps of obtaining the elastic load deformation sequence and the seasonal porosity elastic deformation sequence based on the principal component analysis results include S531 to S532. Wherein: S531, reconstruct the seasonal porosity elastic deformation sequence based on the time feature vector of the first principal component and the positive value in the score vector of the first principal component. Multiply the time feature vector of the first principal component by the positive value in the score vector of the first principal component to reconstruct the seasonal porosity elastic deformation sequence.

[0036] S532, reconstruct the elastic load deformation sequence based on the time feature vector of the first principal component and the negative values ​​in the score vector of the first principal component.

[0037] The representation of the reconstructed seasonal porosity elastic deformation sequence and elastic load deformation sequence can be illustrated by the following example: For instance, the second target InSAR surface deformation time series corresponding to the InSAR surface deformation time series is... Principal component analysis is then performed on the above matrix to obtain the time feature vector [1,2,3,4,5] and the score vector [-3,4,-5,6] of the first principal component. At this point, the time feature vector and the score vector of the first principal component only show the vector dimensions and do not represent the actual calculation results of the above matrix.

[0038] At this point, the seasonal porosity elastic deformation sequence for the time feature vector [1,2,3,4,5] of the first principal component and the score vector [-3,4,-5,6] of the first principal component is: .

[0039] The elastic load deformation sequence is as follows: .

[0040] This application observes high-frequency noise in the reconstructed seasonal orifice elastic deformation sequence and the reconstructed elastic load deformation sequence from the time characteristics of the first principal component. To avoid high-frequency noise, harmonic analysis needs to be performed on the reconstructed seasonal orifice elastic deformation sequence and the reconstructed elastic load deformation sequence to retain the main seasonal signal, filter out the trend term and noise, and obtain a pure seasonal orifice elastic deformation sequence and elastic load deformation sequence. The following formula is used to perform harmonic analysis on the initial sequence of seasonal orifice elastic deformation; .

[0041] in, This is the initial sequence of seasonal pore elastic deformation. This is the b-th epoch in the initial sequence of seasonal porosity elastic deformation. For the intercept term, For linear speed, For the trend term of harmonic analysis, and For seasonal amplitude, Let k represent noise, and k be the number of seasonal terms.

[0042] The steps for reconstructing the elastic load deformation sequence and performing harmonic analysis on the reconstructed seasonal orifice elastic deformation sequence in this application are consistent. Taking the reconstruction of the seasonal orifice elastic deformation sequence as an example, in S531 above, the steps for reconstructing the seasonal orifice elastic deformation sequence based on the positive values ​​in the time feature vector of the first principal component and the score vector of the first principal component include: The initial sequence of seasonal pore elastic deformation is reconstructed based on the time feature vector of the first principal component and the positive values ​​in the score vector of the first principal component.

[0043] Harmonic analysis is performed on the initial sequence of seasonal pore elastic deformation. Based on the harmonic analysis results, the trend term and noise in the initial sequence of seasonal pore elastic deformation are subtracted to obtain the seasonal pore elastic deformation sequence.

[0044] This application proposes a set of discrimination criteria based on external data and closes the loop between the discrimination criteria and the iterative optimization process, ensuring the robustness and reliability of the results. Specifically, because groundwater head observation data and seasonal pore elastic deformation are highly correlated, and terrestrial water storage change data and elastic load deformation are highly correlated, groundwater head observation data and terrestrial water storage change data are selected as external data. That is, the external data includes groundwater head observation data and terrestrial water storage change data. In S1 above, the external data corresponding to the deformation time series dominated by pore elastic deformation is groundwater head observation data; the external data corresponding to the seasonal pore elastic deformation sequence is groundwater head observation data; and the external data corresponding to the elastic load deformation sequence is terrestrial water storage change data.

[0045] In S6 above, the Pearson correlation coefficients of the deformation time series dominated by the pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series with the corresponding external data are calculated respectively. If the Pearson correlation coefficient is greater than a set threshold, it proves that the separation is successful. Specifically, this includes S61 to S62: S61, perform harmonic analysis on the groundwater head observation data, and determine the trend term and seasonal term of the groundwater head observation data based on the harmonic analysis results.

[0046] S62, calculate the first Pearson correlation coefficient between the deformation time series dominated by pore elastic deformation and the trend term of the groundwater head observation data, the second Pearson correlation coefficient between the seasonal pore elastic deformation series and the seasonal term of the groundwater head observation data, and the third Pearson correlation coefficient between the elastic load deformation series and the terrestrial water storage change data. If the first Pearson correlation coefficient is greater than a first set threshold, the second Pearson correlation coefficient is greater than a second set threshold, and the third Pearson correlation coefficient is greater than a third set threshold, then the separation is successful.

[0047] The deformation time series dominated by pore elastic deformation is compared with the trend term of groundwater head observation data, and the first Pearson correlation coefficient is calculated. A strong correlation indicates that the deformation time series dominated by pore elastic deformation has been successfully separated. The seasonal pore elastic deformation series is compared with the seasonal term of groundwater head observation data, and the second Pearson correlation coefficient is calculated. A strong correlation indicates that the seasonal pore elastic deformation series has been successfully separated. Correlation analysis is performed on the elastic load deformation series with terrestrial water storage change data, such as satellite gravity observations and hydrological model data. A high Pearson correlation coefficient indicates that the elastic load deformation series has been successfully separated. If the verification fails, the clustering or discriminant parameters can be adjusted for iterative optimization until the separation results meet physical consistency.

[0048] Through the above process, this application realizes the automatic separation of physically meaningful elastic load deformation sequence and pore elastic deformation sequence from hybrid InSAR surface deformation dataset. It also innovatively processes the trend term and seasonal term of the pore elastic deformation sequence separately, namely the deformation time series dominated by pore elastic deformation and the seasonal pore elastic deformation sequence, thus ensuring the reliability and verifiability of the separation results.

[0049] This application achieves precise decoupling of the physical mechanism, resolving the fundamental confusion problem in existing technologies. It significantly improves the accuracy and reliability of the physical interpretation of deformation signals. A complete "separation-verification" closed loop is established, ensuring the physical consistency and verifiability of the results. An automated and robust separation framework is provided, enhancing the universality of this application.

[0050] This application takes the groundwater variation zone between 122°W and 118°W longitude and 34°N and 40°N latitude as an example, referring to... Figure 1 The overall flowchart shown first executes S1: acquiring the InSAR surface deformation time series for May 2019, 2020, 2020, 2021, and 2021 for the groundwater change zone from 122°W to 118°W and 34°N to 40°N, as well as groundwater head observation data and terrestrial water storage change data as verification benchmarks.

[0051] S2, based on the morphological characteristics of InSAR surface deformation time series from May 2019, 2020, 2020, 2021, and 2021, uses spatial clustering algorithms such as K-means to automatically divide the entire deformation field into several spatial clusters with similar evolutionary behaviors. The clustering results are shown below. Figure 2 As shown, the red and green areas correspond to different types of clusters.

[0052] S3 in Figure 2Based on this, a physical category was determined for each cluster. By analyzing the changing trends and spatial distribution continuity of the InSAR surface deformation time series within each cluster, the clusters were classified as either "pore elastic deformation-dominated" or "seasonal pore elastic deformation and elastic load deformation mixed". Figure 2 The red areas correspond to clusters dominated by pore elastic deformation, while the green areas correspond to clusters of mixed seasonal pore elastic deformation and elastic load deformation.

[0053] For clusters identified as "dominantly pore elastic deformation", execute S4 and directly use the average deformation time series of the cluster as the output pore elastic deformation component.

[0054] For clusters classified as "mixed seasonal porosity elastic deformation and elastic load deformation," S5 separation is performed. First, all InSAR surface deformation time series within this cluster are detrended preprocessed. Then, principal component analysis is performed on the detrended InSAR surface deformation time series, and the first principal component is extracted. This principal component contains two key elements: the temporal feature vector of the first principal component, shown in... Figure 3 The data clearly presents the periodic evolution of the time feature vector of the first principal component in May 2019, 2020, May 2020, 2021, and May 2021; its first principal component score vector is shown in... Figure 4 The values ​​exhibit significant positive and negative phenomena in the [-300, 200] space. Based on Figure 4 The first principal component score vector shown is used to reconstruct the elastic load deformation sequence and the seasonal pore elastic deformation sequence, respectively. This yields pure elastic load deformation sequences and seasonal pore elastic deformation sequences for May 2019, 2020, 2020, 2021, and 2021. Their typical forms are as follows: Figure 5 The diagrams are shown on the left and right. Blue represents the elastic load deformation sequence, and orange represents the seasonal pore elastic deformation sequence.

[0055] S6 is the physical verification and iterative optimization step. The deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series obtained from steps S4 and S5 are compared with the corresponding external data. Specifically, the comparison between the pore elastic deformation series and groundwater head observation data in May 2019, 2020, May 2020, 2021, and May 2021 is shown below. Figure 6 As shown, the pore elastic deformation sequence and the groundwater head observation data show a basically consistent trend and a strong correlation, thus verifying the results. In summary, the complete technical route of this application is summarized in... Figure 7 middle.

[0056] This application also provides a system for separating elastic and porosimetric deformations from InSAR surface deformation to implement the method described above. The solution provided by this system is similar to the solution described in the above method; therefore, specific limitations in one or more system embodiments for separating elastic and porosimetric deformations from InSAR surface deformation provided below can be found in the limitations of the method for separating elastic and porosimetric deformations from InSAR surface deformation described above, and will not be repeated here.

[0057] In one exemplary embodiment, such as Figure 8 As shown, a system for separating elastic and porosity elastic deformation from InSAR surface deformation includes: The data acquisition module 801 is used to acquire the InSAR surface deformation dataset and external data of the target area; the InSAR surface deformation dataset includes multiple InSAR surface deformation time series.

[0058] Clustering module 802 is used to take a preset number of InSAR surface deformation time series as initial centroids and use a spatial clustering algorithm to spatially cluster multiple InSAR surface deformation time series to obtain multiple clusters.

[0059] The determination module 803 is used to extract the spatial distribution characteristics of the InSAR surface deformation time series within each of the plurality of clusters, obtain the average deformation time series based on the average value of the InSAR surface deformation time series within the cluster, and determine the category to which the cluster belongs based on the average deformation time series and the spatial distribution characteristics.

[0060] The pore elastic deformation acquisition module 804 is used to obtain the deformation time series dominated by pore elastic deformation based on the InSAR surface deformation time series within the cluster if the cluster belongs to the category dominated by pore elastic deformation.

[0061] The separation module 805 is used to perform principal component analysis on the InSAR surface deformation time series within the cluster if the cluster belongs to the category of mixed seasonal porosity elastic deformation and elastic load deformation, and to obtain the elastic load deformation sequence and seasonal porosity elastic deformation sequence based on the principal component analysis results.

[0062] The verification module 806 is used to calculate the Pearson correlation coefficients of the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series with the corresponding external data. If each Pearson correlation coefficient is greater than the corresponding set threshold, it proves that the separation is successful; otherwise, spatial clustering is performed again until the separation is successful.

[0063] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 9 As shown, the computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores InSAR surface deformation datasets and external data. The I / O interfaces are used for information exchange between the processor and external devices. The communication interface is used for communication with external terminals via a network connection. When executed by the processor, the computer program implements a method for separating elastic and porosimetric deformations from InSAR surface deformations.

[0064] Those skilled in the art will understand that Figure 9 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0065] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0066] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for separating elastic and porosity elastic deformation from InSAR surface deformation, characterized in that, The method includes: S1. Acquire the InSAR surface deformation dataset and external data for the target area; the InSAR surface deformation dataset includes multiple InSAR surface deformation time series. S2. Using a preset number of InSAR surface deformation time series as initial centroids, a spatial clustering algorithm is used to spatially cluster multiple InSAR surface deformation time series to obtain multiple clusters; S3. For each of the multiple clusters, extract the spatial distribution characteristics of the InSAR surface deformation time series within the cluster, obtain the average deformation time series based on the average value of the InSAR surface deformation time series within the cluster, and determine the category to which the cluster belongs based on the average deformation time series and the spatial distribution characteristics. S4. If the cluster belongs to the category of porosity elastic deformation-dominated, then the deformation time series dominated by porosity elastic deformation is obtained based on the InSAR surface deformation time series within the cluster. S5. If the cluster belongs to the category of mixed seasonal porosity elastic deformation and elastic load deformation, then perform principal component analysis on the InSAR surface deformation time series within the cluster, and obtain the elastic load deformation sequence and seasonal porosity elastic deformation sequence based on the principal component analysis results. S6. Calculate the Pearson correlation coefficients between the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series and the corresponding external data. If each Pearson correlation coefficient is greater than the corresponding set threshold, the separation is successful; otherwise, return to S2 and perform spatial clustering again until the separation is successful.

2. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 1, characterized in that, The cluster category is determined using the following formula: ： in, Let be the number of clusters, and i be the index of the cluster. For the first Clusters, For the first InSAR surface deformation time series within a cluster For the first The centroid of a cluster, It is a Euclidean distance.

3. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 1, characterized in that, The steps of performing principal component analysis on the InSAR surface deformation time series within the cluster, and obtaining the elastic load deformation sequence and the seasonal porosity elastic deformation sequence based on the principal component analysis results, include: For each InSAR surface deformation time series within the cluster, perform the following operations: The InSAR surface deformation time series of the first target is fitted using the least squares method. Based on the fitting result, the trend term in the InSAR surface deformation time series of the first target is subtracted to obtain the InSAR surface deformation time series of the second target. Wherein, the first target InSAR surface deformation time series is any InSAR surface deformation time series in the InSAR surface deformation time series; Principal component analysis was performed on the second target InSAR surface deformation time series corresponding to the InSAR surface deformation time series, and the first principal component was retained to obtain the principal component analysis results corresponding to the InSAR surface deformation time series. Based on the principal component analysis results, the elastic load deformation sequence and the seasonal pore elastic deformation sequence were obtained.

4. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 3, characterized in that, The following formula is used to fit the InSAR surface deformation time series of the first target: ； in, The first target is the InSAR surface deformation time series. This refers to the a-th epoch in the InSAR surface deformation time series of the first target. For the intercept term, For linear speed, This represents the trend term of the least squares method.

5. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 3, characterized in that, The principal component analysis results include the temporal feature vector of the first principal component; The steps for obtaining the elastic load deformation sequence and the seasonal porosity elastic deformation sequence based on the principal component analysis results include: Based on the time feature vector of the first principal component and the positive values ​​in the score vector of the first principal component, the seasonal porosity elastic deformation sequence is reconstructed. The elastic load deformation sequence is reconstructed based on the time feature vector of the first principal component and the negative values ​​in the score vector of the first principal component.

6. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 3, characterized in that, The steps for reconstructing the seasonal porosity elastic deformation sequence based on the time feature vector and the positive values ​​in the score vector of the first principal component include: Based on the positive values ​​in the time feature vector and the score vector of the first principal component, the initial sequence of seasonal pore elastic deformation is reconstructed. Harmonic analysis is performed on the initial sequence of seasonal pore elastic deformation. Based on the harmonic analysis results, the trend term and noise in the initial sequence of seasonal pore elastic deformation are subtracted to obtain the seasonal pore elastic deformation sequence.

7. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 1, characterized in that, The following formula is used to perform harmonic analysis on the initial sequence of seasonal pore elastic deformation; ； in, This is the initial sequence of seasonal pore elastic deformation. This is the b-th epoch in the initial sequence of seasonal porosity elastic deformation. For the intercept term, For linear speed, For the trend term of harmonic analysis, and For seasonal amplitude, Let k represent noise, and k be the number of seasonal terms.

8. The method for separating elastic and porosity elastic deformation from InSAR surface deformation according to claim 1, characterized in that, The external data includes groundwater head observation data and terrestrial water storage change data; the external data corresponding to the deformation time series dominated by pore elastic deformation is groundwater head observation data; the external data corresponding to the seasonal pore elastic deformation series is groundwater head observation data; and the external data corresponding to the elastic load deformation series is terrestrial water storage change data. Calculate the Pearson correlation coefficients between the deformation time series dominated by the pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series, and the corresponding external data. If the Pearson correlation coefficient is greater than a set threshold, the separation is considered successful. Specifically, this includes: Harmonic analysis was performed on the groundwater head observation data, and the trend term and seasonal term of the groundwater head observation data were determined based on the harmonic analysis results. Calculate the first Pearson correlation coefficient between the deformation time series dominated by pore elastic deformation and the trend term of the groundwater head observation data, the second Pearson correlation coefficient between the seasonal pore elastic deformation series and the seasonal term of the groundwater head observation data, and the third Pearson correlation coefficient between the elastic load deformation series and the terrestrial water storage change data; If the first Pearson correlation coefficient is greater than the first set threshold, the second Pearson correlation coefficient is greater than the second set threshold, and the third Pearson correlation coefficient is greater than the third set threshold, then the separation is successful.

9. A system for separating elastic and porosimetric deformations from InSAR surface deformations, employing the method for separating elastic and porosimetric deformations from InSAR surface deformations according to any one of claims 1-8, characterized in that, The system includes: The data acquisition module is used to acquire InSAR surface deformation datasets and external data for the target area; the InSAR surface deformation dataset includes multiple InSAR surface deformation time series. The clustering module is used to take a preset number of InSAR surface deformation time series as initial centroids and use a spatial clustering algorithm to spatially cluster multiple InSAR surface deformation time series to obtain multiple clusters; The determination module is used to extract the spatial distribution characteristics of the InSAR surface deformation time series within each of the multiple clusters, obtain the average deformation time series based on the average value of the InSAR surface deformation time series within the cluster, and determine the category to which the cluster belongs based on the average deformation time series and the spatial distribution characteristics. The pore elastic deformation acquisition module is used to obtain the deformation time series dominated by pore elastic deformation based on the InSAR surface deformation time series within the cluster if the cluster belongs to the category dominated by pore elastic deformation. The separation module is used to perform principal component analysis on the InSAR surface deformation time series within the cluster if the cluster belongs to the category of mixed seasonal porosity elastic deformation and elastic load deformation, and to obtain the elastic load deformation sequence and seasonal porosity elastic deformation sequence based on the principal component analysis results. The verification module is used to calculate the Pearson correlation coefficients of the deformation time series dominated by pore elastic deformation, the seasonal pore elastic deformation series, and the elastic load deformation series with the corresponding external data. If each Pearson correlation coefficient is greater than the corresponding set threshold, it proves that the separation is successful; otherwise, spatial clustering is performed again until the separation is successful.

10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement the method for separating elastic and porosimetric deformations from InSAR surface deformations as described in any one of claims 1-8.

Images