Method for partitioning permeability parameters of underground aquifer of watershed

Abstract

The invention discloses a drainage basin underground aquifer permeability parameter zoning method, which belongs to the technical field of hydrogeology, and comprises the following steps: collecting water level recovery data of a confined aquifer and an underwater aquifer through an artificial excitation test, and respectively inputting corresponding hydrogeology parameter inversion models to calculate permeability parameters of the confined aquifer; permeability parameters of the phreatic aquifer are determined; calculating a rainfall infiltration replenishment coefficient based on the water level change of the phreatic aquifer before and after rainfall and the rainfall amount; finally, the permeability parameters of the sampling points and the corresponding space coordinate features are fused for clustering analysis, partitions of all the permeability parameters of the underground aquifer of the drainage basin are generated through spatial interpolation, and the permeability parameters comprise the permeability coefficient, the water storage rate, the specific yield and the rainfall infiltration supply coefficient. According to the method, the problem of simulation errors of a numerical model on groundwater supply, runoff, drainage and other processes caused by traditional parameter homogenization can be solved, and the fine management quality of water resources is improved.

Description

Technical Field

[0001] This invention belongs to the field of hydrogeology and specifically relates to a method for zoning permeability parameters of underground aquifers in a watershed. Background Technology

[0002] With the intensification of global climate change, extreme weather events are occurring more frequently. Drought and floods remain two major problems facing comprehensive watershed management, and also pose significant challenges to the refined development, utilization, management, and protection of water resources. Constructing watershed hydrological cycle models under changing environments is an important means to achieve refined management and protection of watershed water resources. Numerous surface water, groundwater, and surface-groundwater coupled numerical simulation models have been developed and applied both domestically and internationally, providing quantitative basis for addressing drought, floods, and other problems.

[0003] The permeability parameters of aquifers in a watershed are crucial parameters affecting the accuracy of watershed hydrological models, and are closely related to groundwater recharge, runoff, and discharge processes. In recent years, influenced by natural and human activities, watershed aquifers have become highly heterogeneous and anisotropic. However, directly assigning empirical values ​​of aquifer permeability parameters as initial parameters to watershed hydrological models remains a common practice. This "parameter homogenization" can lead to significant errors in the model's simulation of groundwater recharge, runoff, and discharge processes, impacting the quality of refined water resource management. Summary of the Invention

[0004] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for zoning the permeability parameters of a watershed aquifer, which solves the problem that the traditional "parameter homogenization" causes errors in the model's simulation of groundwater recharge, runoff, and discharge processes, thus affecting the quality of refined water resource management.

[0005] To solve the above-mentioned technical problems, the present invention is implemented using the following technical solution:

[0006] This invention provides a method for zoning the permeability parameters of underground aquifers in a watershed, comprising:

[0007] Within the target watershed, water level recovery data for confined and unconfined aquifers were collected through artificial induction tests. This data was then input into corresponding hydrogeological parameter inversion models to calculate the permeability coefficient and storage rate of the confined aquifers, as well as the permeability coefficient, storage rate, and specific yield of the unconfined aquifers. Based on the changes in unconfined aquifer water level and precipitation before and after rainfall, the precipitation infiltration recharge coefficient was calculated. Cluster analysis was performed by integrating the permeability parameters of the sampling points with their corresponding spatial coordinate characteristics, and spatial interpolation was used to generate partitions for each permeability parameter of the watershed's underground aquifers. These permeability parameters include the permeability coefficient, storage rate, specific yield, and precipitation infiltration recharge coefficient.

[0008] The aforementioned method for zoning watershed aquifer permeability parameters, wherein the collection of water level recovery data for confined and unconfined aquifers within the target watershed through artificially induced tests includes:

[0009] Groundwater monitoring wells are installed in the confined aquifer and unconfined aquifer of the target watershed. A certain volume of water is injected into the well to artificially induce water level changes. Pressure level gauges are used to collect data on the recovery process of the water level in the well from its highest point to its initial state. The collection frequency is set according to the water level recovery rate; the faster the rate, the higher the collection frequency.

[0010] The aforementioned method for zoning watershed aquifer permeability parameters involves inputting water level recovery data from confined aquifers into a corresponding hydrogeological parameter inversion model to calculate the permeability coefficient and storage capacity of the confined aquifers, including:

[0011] The water level recovery data of the confined aquifer is input into the hydrogeological parameter inversion model of the confined aquifer to calculate the permeability coefficient K of the confined aquifer in the area where each groundwater monitoring well is located. cj and the water storage rate S of confined aquifer scj ;

[0012] The water level recovery data of the unconfined aquifer is input into the corresponding hydrogeological parameter inversion model to calculate the permeability coefficient, storage rate, and specific yield of the unconfined aquifer, including:

[0013] The water level recovery data of the unconfined aquifer is input into the hydrogeological parameter inversion model of the unconfined aquifer to calculate the permeability coefficient K of the unconfined aquifer in the area where each groundwater monitoring well is located. ui , water storage rate S of unconfined aquifer sui and the yield of unconfined aquifers μ i .

[0014] The aforementioned method for zoning groundwater aquifer permeability parameters in a watershed, wherein the calculation of the precipitation infiltration recharge coefficient based on the change in unconfined aquifer water level and precipitation before and after precipitation includes:

[0015] The groundwater level H before precipitation was obtained from a groundwater monitoring well in an unconfined aquifer. 1i And the highest groundwater level H after the precipitation ended. 2i The daily precipitation or the cumulative value of precipitation over several consecutive days is obtained from the nearest rainfall monitoring station selected from the groundwater monitoring well of the unconfined aquifer as the precipitation amount P. 0i Based on the groundwater level H before precipitation 1i The highest groundwater level H after the precipitation ends 2i and precipitation P 0i Calculate the precipitation infiltration recharge coefficient α i :

[0016] i {1,…,N u},

[0017] In the formula, N u This represents the total number of groundwater monitoring wells in the unconfined aquifer of the basin.

[0018] The aforementioned method for zoning permeability parameters of a watershed aquifer involves performing cluster analysis on the permeability parameters of the fused sampling points and their corresponding spatial coordinate features, and generating zoning for each permeability parameter of the watershed aquifer through spatial interpolation, including:

[0019] The target watershed is divided into fixed grids, a unified projection coordinate system is established, and grid points within the boundaries are retained. The permeability parameters and corresponding spatial coordinates of sampling points at different layers are read and standardized to obtain standardized permeability parameter features and corresponding spatial coordinate features. These standardized permeability parameter features and their corresponding spatial coordinate features are then weighted and fused using spatial weighting factors to construct feature vectors for clustering input. For each permeability parameter feature, K-Means clustering is performed on each candidate cluster number by setting a search range for the number of clusters. The silhouette coefficient is calculated, and the cluster number corresponding to the largest value is selected as the optimal value to generate the final cluster label. For each permeability parameter's final cluster label, the nearest n neighbor sampling points of each grid cell are retrieved using KD-Tree. The cluster label with the largest weight is determined based on the inverse distance weighting method. The cluster label is then interpolated and mapped to the entire grid to obtain the partitioning of each permeability parameter in the watershed.

[0020] The aforementioned method for zoning watershed aquifer permeability parameters, wherein the fixed grid subdivision of the target watershed, the unified projection coordinate system, and the retention of grid points within the boundary include: reading the target watershed boundary vector file, uniformly projecting it to the CGCS2000 coordinate system, generating a regular two-dimensional grid within the watershed boundary, and retaining grid points within the boundary by determining whether each grid point is located within the boundary polygon.

[0021] The aforementioned method for zoning permeability parameters of groundwater aquifers in a watershed involves reading the permeability parameters and corresponding spatial coordinates of sampling points at different strata, and then performing standardization processing to obtain the standardized permeability parameter characteristics and corresponding spatial coordinate characteristics, including:

[0022] Permeability parameters at sampling points Including: the permeability coefficient K of the confined aquifer in the watershed cj , water storage rate S of confined aquifer scj , Permeability coefficient K of unconfined aquifer ui , water storage rate S of unconfined aquifer suiWater supply degree μ i and precipitation infiltration recharge coefficient α i For 6 permeability parameters The calculation formulas for standardization are as follows:

[0023] ,

[0024] In the formula, for Standardized permeability parameter characteristics; Let j be the value of the t-th permeability parameter at the i-th sampling point in the unconfined aquifer or the j-th sampling point in the confined aquifer. {1,…,N c}, N c The total number of groundwater monitoring wells in the confined aquifer, where t is an integer and ranges from 1 to 6; and These are the mean and standard deviation of the t-th permeability parameter at all sampling points in its respective layer;

[0025] The spatial coordinates of the i-th sampling point in the unconfined aquifer or the j-th sampling point in the confined aquifer ( The calculation formula for standardization is:

[0026] , ,

[0027] In the formula, and The spatial coordinates of the i-th sampling point in the unconfined aquifer or the j-th sampling point in the confined aquifer ( Standardized spatial characteristics Spatial coordinates ( Spatial coordinates of all sampling points in the corresponding layer ( The mean of ) Spatial coordinates ( Spatial coordinates of all sampling points in the corresponding layer ( The standard deviation of ).

[0028] The aforementioned method for zoning permeability parameters of a watershed aquifer, wherein the standardized permeability parameter features are weighted and fused with their corresponding spatial coordinate features using spatial weighting factors to construct a feature vector for clustering input, includes:

[0029] Feature vectors constructed from the characteristics of each permeability parameter for clustering input The expression is:

[0030] ,

[0031] In the formula, This is a preset spatial weighting factor.

[0032] The aforementioned method for zoning permeability parameters of a watershed aquifer, wherein the feature vectors constructed for each permeability parameter feature are used to perform K-Means clustering on each candidate cluster number by setting a search range for the number of clusters, calculating the silhouette coefficient, and selecting the cluster number corresponding to the largest value as the optimal value, to generate the final cluster label, includes:

[0033] For each permeability parameter feature, a feature vector is constructed. By setting the search range for the number of clusters, the K-Means algorithm is used to cluster the feature vectors constructed for each permeability parameter feature for each candidate cluster number, and the corresponding clustering results are obtained. The silhouette coefficient of each clustering result is calculated, and the cluster number corresponding to the largest silhouette coefficient is selected as the optimal number of clusters. K-Means clustering is performed again based on the optimal number of clusters to generate the final cluster labels corresponding to each permeability parameter.

[0034] The aforementioned method for partitioning watershed aquifer permeability parameters includes, for each permeability parameter, using a KD-Tree algorithm to retrieve the n nearest neighbor sampling points of each grid cell, determining the cluster label with the maximum weight based on the inverse distance weighting method, and interpolating and mapping the cluster labels to the entire grid to obtain the partitioning of each watershed permeability parameter, including:

[0035] For each permeability parameter, a KD-Tree spatial index is constructed based on the spatial coordinates of the sampling points to construct the final cluster labels. Each grid cell in the target watershed is traversed, and its nearest n neighbor sampling points are retrieved to obtain the cluster labels of the neighbors and the spatial distance between the neighbors and the cell. The weight of each neighbor sampling point label is calculated using the inverse distance weighting method, and the neighbor cluster label with the largest weight is assigned to the corresponding grid cell. All grid cell labels are integrated and diffused to the target watershed grid to obtain the partitions of each permeability parameter in the watershed.

[0036] Compared with the prior art, the beneficial effects achieved by the present invention are as follows:

[0037] This invention addresses the problem of deviations in the simulation of groundwater recharge, runoff, and discharge processes caused by traditional "parameter homogenization," which affects the quality of refined water resource management. By determining permeability parameters such as precipitation infiltration recharge coefficient, permeability coefficient, storage rate, and water yield, this invention divides each permeability parameter in confined aquifers and unconfined aquifers in the watershed into different zones.

[0038] The watershed aquifer permeability parameter zoning method of the present invention integrates permeability parameter characteristics and spatial coordinate characteristics for cluster analysis. It takes into account both the numerical differences of the parameters themselves and the spatial distribution patterns, avoiding the zoning bias caused by relying solely on attributes or a single spatial dimension, and making the zoning results more consistent with actual geological conditions.

[0039] The watershed aquifer permeability parameter zoning method of the present invention diffuses the clustering results of discrete sampling points to the entire watershed grid through spatial interpolation, realizing continuous zoning from point to surface, solving the problem of spatial information loss caused by sparse sampling points. The resulting zoning of each permeability parameter can accurately reflect the spatial distribution differences of permeability parameters in different layers of the watershed, providing an intuitive and practical reference for watershed groundwater management.

[0040] The watershed aquifer permeability parameter zoning method of the present invention can provide "real and detailed" parameter input for watershed hydrogeological models, ultimately serving the precise management of water resources and the effective response to extreme disasters.

[0041] This invention delineates the permeability parameters of aquifers within a watershed. Essentially, it overcomes the limitations of the "homogeneity assumption" in traditional hydrological models through "spatial refinement," ultimately shifting from "empirical management" to "quantitative and precise regulation." This provides core data support for addressing climate change and ensuring watershed water resource security. Permeability parameter zoning is a quantitative means of solving the problem of aquifer heterogeneity, upgrading traditional "black box" empirical models to "transparent" physical mechanism models, thereby improving the scientific basis of water resource management and water ecosystem protection decisions. Attached Figure Description

[0042] Figure 1 This is a schematic diagram of a method for zoning permeability parameters of a watershed aquifer according to Embodiment 1 of the present invention. Detailed Implementation

[0043] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments and specific features in the embodiments are detailed descriptions of the technical solution of the present application, rather than limitations thereof. In the absence of conflict, the embodiments and technical features in the embodiments can be combined with each other.

[0044] In this article, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article generally indicates that the preceding and following related objects have an "or" relationship.

[0045] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0046] Example 1:

[0047] In response to the current lack of sufficient emphasis on utilizing massive amounts of observational data such as precipitation, surface water, and groundwater to calculate and delineate the permeability parameters of a watershed's groundwater aquifers (e.g., precipitation infiltration recharge coefficient, permeability coefficient, storage rate, and specific yield), this embodiment provides a method for zoning watershed groundwater aquifer permeability parameters, such as... Figure 1 This is a flowchart illustrating a method for zoning watershed aquifer permeability parameters according to Embodiment 1 of the present invention. This flowchart merely shows the logical sequence of the method described in this embodiment; however, in other possible embodiments of the present invention, different methods may be used, provided there are no conflicts. Figure 1 Complete the steps shown or described in the order indicated.

[0048] S1: Within the target watershed, collect water level recovery data for confined aquifers and unconfined aquifers through artificial induced tests;

[0049] S2: Input the water level recovery data of confined aquifers and unconfined aquifers into the corresponding hydrogeological parameter inversion models to calculate the permeability coefficient and storage rate of confined aquifers, as well as the permeability coefficient, storage rate and specific yield of unconfined aquifers.

[0050] S3: Calculate the precipitation infiltration recharge coefficient based on the changes in the water level of the unconfined aquifer before and after precipitation and the amount of precipitation;

[0051] S4: Cluster analysis is performed on the permeability parameters of the fusion sampling points and their corresponding spatial coordinate features, and spatial interpolation is used to generate partitions of each permeability parameter of the watershed's underground aquifer. The permeability parameters include permeability coefficient, storage rate, specific yield, and precipitation infiltration recharge coefficient.

[0052] Step S1 includes:

[0053] Groundwater monitoring wells are set up in the confined aquifer and unconfined aquifer of the watershed. Water level changes are artificially induced in the groundwater monitoring wells by injecting a certain volume of water instantaneously. Pressure level gauges are used to automatically collect data on the process of the water level in the groundwater monitoring well gradually decreasing from the highest point to the initial position. The sampling frequency of the water level decrease process is set according to the water level recovery rate. The faster the water level decrease rate, the higher the sampling frequency.

[0054] In one specific embodiment, if the water level drop rate is >5cm / min, it is considered rapid recovery, and the sampling frequency is set to 5 times / 1 second; when the water level drop rate drops to 3-5cm / min, it is considered medium-speed recovery, and the sampling frequency is adjusted to 1 time / 1 second; when the water level drop rate drops to 1-3cm / min, it is considered slow recovery, and the sampling frequency is adjusted to 1 time / 10 seconds; when the water level drop rate is <1cm / min, the sampling frequency is adjusted to 1 time / 1 minute; when the water level recovers to the initial position or tends to stabilize for more than 10 minutes, the sampling is stopped.

[0055] Rapid water level drops contain a wealth of crucial information related to aquifer permeability within a short period, requiring high-frequency acquisition to avoid data loss. Conversely, slow water level drops release information over a longer period, allowing for a lower acquisition frequency to reduce redundant data. This dynamic acquisition strategy maximizes the retention of key permeability parameter information during water level decline while avoiding data redundancy, providing high-quality data support for subsequent aquifer parameter retrieval and preventing parameter calculation errors caused by low-frequency acquisition.

[0056] Step S2 includes:

[0057] S21: Calculate the permeability coefficient K of the confined aquifer within the influence radius of each groundwater monitoring well (i.e., the influence range of artificially induced water level changes in the confined aquifer) using the hydrogeological parameter inversion model of the confined aquifer. cj and the water storage rate S of confined aquifer scj ;

[0058] The influence radius of a groundwater monitoring well is the distance at which the water level change at a certain distance is less than a set threshold. In this embodiment, the set threshold is 0.005m. The influence radius of a groundwater monitoring well can be determined by using this groundwater monitoring well as the central well and other observation wells arranged along the dominant seepage direction of the aquifer to monitor water level changes, or by using an empirical formula derived from an inversion model.

[0059] In this embodiment, the Kipp model is selected as the inversion model for hydrogeological parameters of confined aquifers.

[0060] S22: Calculate the permeability coefficient K of the unconfined aquifer within the influence radius of each groundwater monitoring well (i.e., the influence range of artificially induced water level changes in the unconfined aquifer) using the hydrogeological parameter inversion model of the unconfined aquifer. ui , water storage rate S of unconfined aquifer sui and the yield of unconfined aquifers μ i ;

[0061] In this embodiment, the SHB model is selected as the inversion model for hydrogeological parameters of unconfined aquifers.

[0062] Step S3 includes:

[0063] The response of unconfined aquifer water level to precipitation infiltration is time-dependent: the infiltration rate is rapid in the initial stage of precipitation, and the water level rises quickly; as soil moisture content becomes saturated, the infiltration rate slows down, and the water level rise becomes more gradual. Setting a sampling frequency of 30 minutes / time or higher, such as 15 minutes / time, can accurately capture the groundwater level H before precipitation. 1i The dynamic rise of water level during precipitation, and the highest groundwater level H after precipitation ends. 2i .

[0064] H 1i It needs to be the "stable water level before precipitation": that is, before the start of precipitation, the groundwater level is in a natural dynamic equilibrium. Natural dynamic equilibrium means there is no significant evaporation, extraction, or lagging effect of previous precipitation. Only at this time can the water level value accurately reflect the "initial benchmark" of precipitation infiltration. If H 1i Including previous water level fluctuations, such as the drop caused by evaporation in the previous days, the change in the unconfined aquifer water level ΔH is overestimated, ΔH = H 2i -H 1i .

[0065] H 2i The value must be "the highest water level that rose after the rainfall ended": at this point, the infiltration of rainfall has essentially stopped, meaning the soil has reached saturation or the rainfall has ended, and the water level no longer rises. Only then can the change in the unconfined aquifer water level ΔH truly represent the rise in water level caused by rainfall infiltration. If H is recorded before the water level reaches its peak... 2i This will lead to an underestimation of ΔH.

[0066] In this embodiment, the sampling frequency of the pressure level gauge in the groundwater monitoring well of the unconfined aquifer is set to 30 minutes / time to record the fluctuations in the groundwater level of the unconfined aquifer caused by natural precipitation; the rainfall monitoring station closest to the groundwater monitoring well is investigated, and the daily rainfall or the cumulative value of rainfall over multiple consecutive days is recorded as the precipitation amount P. 0i The groundwater level H before precipitation was obtained based on the duration of precipitation. 1i And the highest groundwater level H after the precipitation ended. 2i The groundwater infiltration recharge coefficient α of the unconfined aquifer at the location of the groundwater monitoring well is calculated using the following formula. i :

[0067] i {1,…,N u},

[0068] In the formula, N u This represents the total number of groundwater monitoring wells in the unconfined aquifer of the basin.

[0069] Step S4 includes:

[0070] S41: Perform fixed grid subdivision on the target watershed, unify the projected coordinate system, and retain grid points within the boundary;

[0071] S42: Read six permeability parameters (permeability coefficient and storage rate of confined aquifer, permeability coefficient, storage rate and specific yield of unconfined aquifer, and precipitation infiltration recharge coefficient) and their corresponding spatial coordinates from sampling points at different strata, and perform standardization processing to obtain the standardized permeability parameter characteristics and spatial coordinate characteristics.

[0072] S43: The standardized permeability parameter features are weighted and fused with the corresponding spatial coordinate features through spatial weighting factors to construct a feature vector for clustering input;

[0073] Each time, only one permeability parameter feature is selected and fused with its corresponding spatial feature. The six permeability parameter features are fused with their corresponding spatial coordinate features using spatial weighting factors to construct a feature vector for clustering input. After clustering and full-grid mapping in steps S44 and S45, the partitioning results for each permeability parameter are obtained.

[0074] S44: For each permeability parameter feature, the feature vector is constructed. By setting the cluster number search range, K-Means clustering is performed on each candidate cluster number. The silhouette coefficient is calculated and the cluster number corresponding to the largest value is selected as the best value to generate the final cluster label.

[0075] S45: For the final clustering labels of each permeability parameter, use KD-Tree to retrieve the nearest n neighbor sampling points of each grid cell, determine the clustering label with the maximum weight based on the inverse distance weighting method, and interpolate and map the clustering labels to the entire grid to obtain the partitioned raster of each permeability parameter in the watershed.

[0076] Step S41: Basic preparations: Mesh generation and spatial unification

[0077] A fixed mesh is generated for the target watershed, and the target watershed boundary vector file (shp) is read and a unified projected coordinate system (CGCS2000, EPSG:4490) is established. A regular two-dimensional mesh is generated within the target watershed boundary. For each mesh point, it is determined whether it falls within a polygon, and only internal points are retained for subsequent interpolation and zoning analysis. In one embodiment, the regular two-dimensional mesh is 200×200.

[0078] Step S41 unifies the spatial reference to avoid spatial coordinate deviations caused by differences in coordinate systems, such as mismatch between sampling point coordinates and grid coordinates; the grid, as a "spatial carrier," anchors the subsequent zoning results to fixed spatial units, facilitating the connection with watershed hydrological models that use grids as computational units.

[0079] Step S42: Data standardization: Eliminating the influence of dimensions and scale

[0080] For 6 permeability parameters (K) c S sc K u S su Standardize the values ​​of (μ, α) and the coordinates of the sampling point (x, y) respectively:

[0081] Read the attribute data and spatial coordinates of all sampling points. Extracting permeability parameters from sampling points and spatial coordinates ( The permeability coefficient K of the confined aquifer in the watershed cj Water storage rate S of confined aquifers in the watershed scj , Permeability coefficient K of unconfined aquifer ui , water storage rate S of unconfined aquifer sui Water supply degree μ i and precipitation infiltration recharge coefficient α i These 6 permeability parameters and spatial coordinates ( )standardization:

[0082] ,

[0083] In the formula, for Standardized permeability parameter characteristics; Let j be the value of the t-th permeability parameter at the i-th sampling point in the unconfined aquifer or the j-th sampling point in the confined aquifer. {1,…,N c}, N c The total number of groundwater monitoring wells in the confined aquifer, where t is an integer and ranges from 1 to 6; and denoted as the mean and standard deviation of the t-th permeability parameter at all sampling points in its respective layer.

[0084] , ,

[0085] In the formula, and The spatial coordinates of the i-th sampling point in the unconfined aquifer or the j-th sampling point in the confined aquifer ( Standardized spatial characteristics Spatial coordinates ( Spatial coordinates of all sampling points in the corresponding layer ( The mean of ) Spatial coordinates ( Spatial coordinates of all sampling points in the corresponding layer ( The standard deviation of ).

[0086] Step S42 eliminates the dimensional differences between different parameters, such as the permeability coefficient being in units of m / d while the specific yield has no unit, ensuring that the influence weight of each parameter on the results is fair during clustering, and preventing a certain parameter from dominating the clustering due to its large numerical range; after coordinate standardization, the spatial coordinate features representing distance can participate in the fusion with the permeability parameter features on the same order of magnitude, laying the foundation for subsequent joint clustering of "parameter attributes combined with spatial coordinates".

[0087] Step S43 performs feature fusion: taking into account both parameter attributes and spatial coordinates.

[0088] After standardizing the permeability parameter features and spatial coordinate features of all sampling points, each permeability parameter is selected and its corresponding spatial coordinate features are weighted and fused to construct an input feature vector for clustering. This process is performed separately for each of the six permeability parameters, resulting in independent clustering partitions. For each sampling point of each permeability parameter, its feature vector... For the standardized permeability parameter features and corresponding spatial coordinate features of the i-th or j-th sampling point of the t-th permeability parameter, a spatial weighting factor is used. The result after weighted fusion:

[0089] ,

[0090] In the formula, This is a preset spatial weighting factor.

[0091] Step S43 combines "parameter similarity" with "spatial proximity": points with similar parameters, such as high permeability areas, are more likely to be classified into the same area if they are spatially close; conversely, even if the parameters are similar, points that are spatially distant (such as different sub-basins) may not be classified into the same area due to differences in spatial proximity. They belong to different regions due to adjustments. If the spatial weighting factor... =0, clustering only by parameter attributes, that is, ignoring spatial coordinates, may lead to spatial fragmentation of the same partition; if A larger value indicates a higher weight for spatial coordinates, resulting in greater spatial continuity of the partition and conforming to the spatial autocorrelation of aquifer parameters.

[0092] Step S44: Perform objective clustering: Determine the optimal number of partitions

[0093] For each feature vector constructed for the permeability parameter features, a search range for the number of clusters is set. For each candidate cluster number, the K-Means algorithm is used to cluster the feature vectors constructed for each permeability parameter features to obtain the corresponding clustering results.

[0094] Calculate the silhouette coefficient for each clustering result, and select the number of clusters corresponding to the maximum silhouette coefficient as the optimal number of clusters;

[0095] In this embodiment, the search range for the number of clusters is set to 2~7. K-Means clustering is performed on each cluster number, and the feature vector obtained by fusing the standardized permeability parameter features and coordinate features is used. As input, For the cluster label of sampling point i or j, Cluster Label The cluster centers are then used to cluster the input feature vectors. The optimization objectives for the permeability parameters of confined aquifers and unconfined aquifers are as follows:

[0096] Optimization objectives for permeability parameters of confined aquifers: ,

[0097] Optimization objectives for permeability parameters of unconfined aquifers: ,

[0098] In this embodiment, the number of sampling points N c N represents the total number of groundwater monitoring wells in confined aquifers, which is also the total number of sampling points in confined aquifers. u This represents the total number of groundwater monitoring wells in the unconfined aquifer, which is also the total number of sampling points in the unconfined aquifer.

[0099] Further calculate the silhouette coefficient for each sample.

[0100] ,

[0101] In the formula, This represents the average distance from sampling point i or j to other samples within the same cluster under the t-th permeability parameter. This represents the average distance from sampling point i or j to all samples from the nearest other clusters under the t-th permeability parameter; ultimately, it is based on the silhouette coefficient of each sample. Take the average value to obtain the average silhouette coefficient corresponding to the number of clusters. The optimal number of clusters is selected based on the highest average silhouette coefficient, and the corresponding final cluster labels are generated.

[0102] The silhouette coefficient assesses clustering quality through "intra-cluster compactness" and "inter-cluster separation": the closer the coefficient is to 1, the more consistent the parameter characteristics within the partitions and the more significant the differences between the partitions, resulting in more reliable results. The feature vectors in step S3 serve as input data for the K-Means clustering algorithm, providing sample attribute information and spatial coordinate information for cluster analysis, which is the foundation for achieving clustering based on permeability parameters and spatial coordinate features.

[0103] Step S45 performs spatial interpolation: from sampling point to the entire watershed.

[0104] For each permeability parameter, a KD-Tree spatial index is constructed based on the spatial coordinates of the sampling points to construct the final cluster label. Each grid cell in the target watershed is traversed, and its nearest n neighbor sampling points are retrieved to obtain the cluster labels of the neighbors and the spatial distance between the neighbors and the cell.

[0105] The weights of the labels of each neighbor sampling point are calculated using the inverse distance weighting (IDW) method, where the weights are inversely proportional to the distance. The neighbor clustering label with the highest weight is assigned to the corresponding grid cell. The corresponding grid cell refers to the grid cell being processed when the weights of the nearest n neighbor sampling points of a certain grid cell are calculated and the label with the highest weight is determined.

[0106] All grid cell labels are integrated and diffused to the target watershed grid to obtain partitions for each permeability parameter of the watershed. The specific process is as follows:

[0107] For each grid cell Using a KD-Tree, the nearest n neighboring sampling points (KNN) are retrieved among all sampling points, and the clustering labels corresponding to these n neighboring sampling points are read. These n neighbors and grid cells The distance is The inverse distance weighted method (IDW) is used to calculate the distance between each neighbor m and the current grid cell, based on the inverse distance. weight Weights of the neighbors of n sampling points according to cluster category Perform grouped statistics: grouping items into the same cluster category The weights of all the neighbors are summed to obtain the category for the current cell. Total weight Compare the total weights of all categories. Assign the cluster category label with the highest total weight to the current grid cell. This enables spatial diffusion and interpolation of cluster labels.

[0108] ,

[0109] , ,

[0110] In the formula, For the t-th permeability parameter, each grid cell The weights of the neighbors of the m-th sampling point, where the weights are related to the distance from the neighbor to the current grid cell. The weight is inversely proportional to the distance; the closer the distance, the greater the weight. It is the current grid cell The index variables of the neighbors of the n sample points, The value range is 1 to n; For a small positive number, in this embodiment, The value is 10 −9 To avoid division by zero errors; This represents the clustering category under the t-th permeability parameter. In the current grid cell The total weighted score of the surrounding area is obtained by weighted summation, where the weighting factor is the weight of each neighbor. Only when the neighbor's cluster label With category Only when they are equal will they participate in the weighted summation; this is determined by the indicator function. The value is either 1 or 0, where 1 indicates that the category belongs to the category and 0 indicates that the category does not belong to the category. Indicates the current grid cell The final cluster label is the cluster category under the t-th permeability parameter, which has the highest weighted score. By selecting the cluster category with the highest weight, the partition label of the grid cell is determined. By analyzing all cluster categories The total weights are compared, and the cluster label with the highest score is selected.

[0111] The above method extends the partition labels to all grid cells throughout the entire watershed, forming partitioned raster data.

[0112] Step S45 expands the clustering labels of discrete sampling points from only tens or hundreds of points to thousands or tens of thousands of pixels in the entire watershed grid, achieving spatial continuity of the partitioning results; KNN-IDW takes into account both the "spatial proximity" principle with high weight of nearby points and the "majority principle" principle that neighbors in the same region are more likely to belong to the same class, avoiding spatial jumps in the interpolation results.

[0113] The final output of the partitioned raster can be directly used as the parameter input for the watershed hydrological model, solving the problem of simulation deviations in groundwater recharge, runoff and discharge processes caused by traditional "parameter homogenization", which affects the quality of refined water resource management.

[0114] The zoning scheme in step S4, from basic preparation to final interpolation forming a closed loop, essentially transforms discrete sampling information into continuous and reliable zoning results through scientific standardization, objective clustering, and reasonable interpolation. This zoning result can intuitively reflect the spatial differentiation patterns of permeability parameters within the watershed, providing a basis for decision-making regarding groundwater development and utilization, and pollution control. Specific groundwater development and utilization decisions include source site selection, and pollution control decisions include strict control of pollution sources in high-permeability areas. The output permeability parameter zoning grid can also directly support parameter assignment in the watershed hydrological model, improving the model's simulation accuracy of groundwater recharge, runoff, and discharge processes, providing crucial spatial basis for refined water resource management. Specific refined water resource management applications include identifying water supply areas during drought periods and assessing infiltration regulation during flood seasons.

[0115] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0116] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0117] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0118] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0119] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the technical principles of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method of zonally determining parameters of permeability of an underground aquifer of a river basin, characterized in that, The method comprises the following steps: Collecting water level recovery data of confined aquifer and phreatic aquifer in the target basin through artificial stimulation test; Inputting the water level recovery data of the confined aquifer and the phreatic aquifer into corresponding hydrogeological parameter inversion models respectively to calculate the permeability coefficient and water storage rate of the confined aquifer, and the permeability coefficient, water storage rate and specific yield of the phreatic aquifer; Calculating the precipitation infiltration recharge coefficient according to the water level change of the phreatic aquifer and the precipitation before and after the precipitation; Fusing the permeability parameters of the sampling points and the corresponding spatial coordinate features for cluster analysis, and generating the partitions of each permeability parameter of the underground aquifer in the basin through spatial interpolation, wherein the permeability parameters include the permeability coefficient, the water storage rate, the specific yield and the precipitation infiltration recharge coefficient.

2. The method of claim 1, wherein, The method of collecting water level recovery data of confined aquifer and phreatic aquifer in the target basin through artificial stimulation test comprises the following steps: Arranging groundwater monitoring wells in the confined aquifer and the phreatic aquifer of the target basin, artificially stimulating water level change by instantaneously injecting a certain volume of water into the well, collecting the recovery process data of the water level in the well from the highest point to the initial state by using a pressure type water level gauge, and setting the collection frequency according to the water level recovery rate, wherein the faster the rate is, the higher the collection frequency is.

3. The method for zonally parameterizing the permeability of the aquifer underlying a drainage basin according to claim 1 or 2, characterized in that, Inputting the water level recovery data of the confined aquifer into the corresponding hydrogeological parameter inversion model to calculate the permeability coefficient and water storage rate of the confined aquifer, which comprises the following steps: The water level recovery data of the confined aquifer is input into a hydrogeological parameter inversion model of the confined aquifer to calculate the permeability coefficient K of the confined aquifer in the area where each groundwater monitoring well is located cj and the storage rate S of the confined aquifer scj ; Inputting the water level recovery data of the phreatic aquifer into the corresponding hydrogeological parameter inversion model to calculate the permeability coefficient, water storage rate and specific yield of the phreatic aquifer, which comprises the following steps: The water level recovery data of the phreatic aquifer is input into a phreatic aquifer hydrogeological parameter inversion model to calculate the phreatic aquifer permeability coefficient K of the area where each groundwater monitoring well is located ui , the phreatic aquifer water storage rate S sui , and the phreatic aquifer yield μ i .

4. The method of claim 3, wherein, The method of calculating the precipitation infiltration recharge coefficient according to the water level change of the phreatic aquifer and the precipitation before and after the precipitation comprises the following steps: Obtaining the groundwater level value H before precipitation from a phreatic aquifer groundwater monitoring well 1i and the highest groundwater level value H after the groundwater rises after precipitation ends 2i , and obtaining the single-day precipitation or the cumulative value of continuous multi-day precipitation as the secondary precipitation P from a rainfall monitoring station selected near the phreatic aquifer groundwater monitoring well 0i ; According to the groundwater level value H before precipitation 1i , the highest groundwater level value H after the groundwater rises after precipitation ends 2i and the secondary precipitation P 0i Calculate the precipitation infiltration recharge coefficient α i : , i {1,...,N u}, In the formula, N u is the total number of monitoring wells for groundwater in the phreatic aquifer of the river basin.

5. The method of claim 4, wherein, The method of fusing the permeability parameters of the sampling points and the corresponding spatial coordinate features for cluster analysis, and generating the partitions of each permeability parameter of the underground aquifer in the basin through spatial interpolation comprises the following steps: Carrying out fixed grid subdivision on the target basin, unifying the projection coordinate system and retaining the grid points within the boundary; Reading the permeability parameters of the sampling points at different horizons and the corresponding spatial coordinates, respectively, standardizing the permeability parameters and the corresponding spatial coordinates to obtain the standardized permeability parameter features and the corresponding spatial coordinate features; Fusing the standardized permeability parameter features and the corresponding spatial coordinate features through spatial weight factors to construct feature vectors for cluster input; For the feature vectors constructed by the permeability parameter features, setting a cluster number search range, performing K-Means clustering on each candidate cluster number, calculating the silhouette coefficient and selecting the maximum cluster number as the best value to generate the final cluster label; For the final cluster label of each permeability parameter, using KD-Tree to search the nearest n neighbor sampling points of each grid pixel, determining the cluster label with the maximum weight based on the inverse distance weighted method, and mapping the cluster label through interpolation to the full grid to obtain the partitions of each permeability parameter in the basin.

6. The method of claim 5, wherein, The method of carrying out fixed grid subdivision on the target basin, unifying the projection coordinate system and retaining the grid points within the boundary comprises the following steps: Reading the target drainage basin boundary vector file, projecting it to the CGCS2000 coordinate system, generating a regular two-dimensional grid within the drainage basin boundary range, and retaining the grid points within the boundary by judging whether each grid point is located within the boundary polygon.

7. The method of claim 6, wherein, The reading of the permeability parameters of different layers and the corresponding spatial coordinates of the sampling points is standardized to obtain the standardized permeability parameter characteristics and the corresponding spatial coordinate characteristics, including: The permeability parameters of different layers and sampling points including: the permeability coefficient K of the basin confined aquifer cj , the water storage rate S of the confined aquifer scj , the permeability coefficient K of the phreatic aquifer ui , the water storage rate S of the phreatic aquifer sui , the water supply degree μ i , and the precipitation infiltration recharge coefficient α i , six kinds of permeability parameters The calculation formula for standardizing each of the six kinds of permeability parameters is: , In the formula, is the normalized permeability parameter characteristics; is the value of the tth permeability parameter at the ith sampling point in a phreatic aquifer or the jth sampling point in a confined aquifer, j {1,…,N c}, N c is the total number of groundwater monitoring wells in the confined aquifer, and t is an integer ranging from 1 to 6; and are the mean value and the standard deviation of the tth permeability parameter at all sampling points in the corresponding layer, respectively. The spatial coordinates of the i-th sampling point of the phreatic aquifer or the j-th sampling point of the confined aquifer are normalized by the following formula: ) ， ， In the formula, and The spatial coordinates of the i-th sampling point in the unconfined aquifer or the j-th sampling point in the confined aquifer ( Standardized spatial characteristics Spatial coordinates ( Spatial coordinates of all sampling points in the corresponding layer ( The mean of ) Spatial coordinates ( Spatial coordinates of all sampling points in the corresponding layer ( The standard deviation of ).

8. The method of claim 7, wherein, The standardized permeability parameter characteristics and the corresponding spatial coordinate characteristics are weighted and fused by a spatial weight factor to construct a feature vector for clustering input, including: Feature vectors for clustering input constructed from permeability parameter characteristics The expression is: ， In the formula, is a predetermined spatial weight factor.

9. The method of claim 8, wherein, The feature vectors constructed for each permeability parameter characteristic are searched for a clustering number, K-Means clustering is performed for each candidate clustering number, the silhouette coefficient is calculated and the clustering number corresponding to the maximum is selected as the best value, and the final clustering label is generated, including: The feature vectors constructed for each permeability parameter characteristic are searched for a clustering number, K-Means clustering is performed for each candidate clustering number, the silhouette coefficient is calculated and the clustering number corresponding to the maximum is selected as the best value, and the final clustering label is generated, including: The silhouette coefficient of each clustering result is calculated, and the clustering number corresponding to the maximum silhouette coefficient is selected as the best clustering number. According to the best clustering number, K-Means clustering is performed again to generate the final clustering label corresponding to each permeability parameter.

10. The method of claim 9, wherein, The final clustering label of each permeability parameter is used to retrieve the nearest n neighbor sampling points of each grid pixel using KD-Tree, and the clustering label with the maximum weight is determined based on the inverse distance weighting method, and the clustering label is interpolated and mapped to the full grid to obtain the partition of each permeability parameter in the drainage basin, including: The final clustering label of each permeability parameter is used to retrieve the nearest n neighbor sampling points of each grid pixel using KD-Tree, and the clustering label with the maximum weight is determined based on the inverse distance weighting method, and the clustering label is interpolated and mapped to the full grid to obtain the partition of each permeability parameter in the drainage basin, including: The weight of each neighbor sampling point label is calculated by the inverse distance weighting method, and the neighbor clustering label with the maximum weight is assigned to the corresponding grid pixel. All grid pixel labels are integrated and diffused to the target drainage basin grid to obtain the partition of each permeability parameter in the drainage basin.

Images