CNN-based groundwater pollution source identification method and system under strong heterogeneous conditions

Abstract

The application discloses a groundwater pollution source identification method and system based on a convolutional neural network (CNN) under strong heterogeneous conditions, and relates to the technical field of groundwater environmental protection; the hydrogeological conditions are reasonably generalized, and a conceptual model of groundwater solute transport is constructed; a random sampling method is used to generate a plurality of groups of working conditions representing the emission intensity of the pollution source, and a permeability coefficient field of the aquifer is generated through sequential Gaussian simulation; based on the permeability coefficient field of the aquifer, a groundwater solute transport-two-domain reference model is constructed; the plurality of groups of working conditions are input into the groundwater solute transport-two-domain reference model one by one, thereby forming a sample of "model input-model response" and constituting a training sample set and a test sample set; a deep learning-CNN algorithm is used to construct a groundwater pollution source identification model based on the training sample set, and the model can accurately identify the position and emission intensity of the pollution source, effectively improves the pollution source identification efficiency and precision, and can provide a reliable solution for groundwater pollution treatment.

Description

Technical Field

[0001] This invention relates to the field of groundwater environmental protection technology, specifically to a method and system for identifying groundwater pollution sources under strongly heterogeneous conditions based on convolutional neural networks (CNNs). Background Technology

[0002] Due to the concealed and complex nature of groundwater pollution, traditional monitoring and investigation methods for obtaining information on groundwater pollution sources are not only inefficient but also costly. Accurately identifying the number, location coordinates, and release history of groundwater pollution sources is the primary task in groundwater pollution control and is crucial for subsequent groundwater environmental remediation. However, on the other hand, natural aquifer systems are highly heterogeneous and complex systems. When pollutants intrude into the aquifer, their migration patterns and spatiotemporal variations are significantly influenced by the heterogeneity of the aquifer medium. Traditional pollution source identification methods often fail to adequately consider the impact of spatial heterogeneity when addressing such problems, resulting in inaccurate simulation results that do not accurately reflect reality.

[0003] Groundwater pollution source identification is a process that uses existing prior data (including pollutant concentration and distribution, water level, water quality parameters, etc.) to deduce key details describing the groundwater pollution sources, such as the number of sources, their precise location coordinates, and release history. However, current traditional methods frequently require calling groundwater solute transport simulation models when identifying groundwater pollution sources. This process is not only computationally intensive and time-consuming, but also makes it difficult to ensure the efficiency of the identification process and the accuracy of the final results. Summary of the Invention

[0004] The purpose of this invention is to propose a method and system for identifying groundwater pollution sources under strongly heterogeneous conditions based on CNN, so as to cope with the complexity and heterogeneity of underground aquifer systems and achieve accurate identification of the location coordinates and emission intensity of groundwater pollution sources.

[0005] According to a first aspect of the present disclosure, a method for identifying groundwater pollution sources under strongly heterogeneous conditions based on CNN is provided, comprising the following steps:

[0006] After analyzing hydrogeological and groundwater pollution data, we reasonably generalize hydrogeological conditions and construct a conceptual model of groundwater solute transport to accurately describe the migration and movement of pollutants in groundwater.

[0007] In the conceptual model of groundwater solute transport, based on the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, a random sampling method is used to generate several sets of operating conditions characterizing the emission intensity of the pollution source, and a sequential Gaussian simulation is used to generate the aquifer permeability coefficient field to represent the polluted groundwater aquifer system under strong heterogeneity.

[0008] Based on the aquifer permeability coefficient field, a groundwater solute transport-dual-domain benchmark model is constructed to accurately reflect the physicochemical processes of pollutants in the groundwater aquifer system;

[0009] Several sets of operating conditions are input into the groundwater solute transport-dual-domain baseline model one by one to obtain the model response of the observation data of the observation well. The model response is the location coordinates of the pollution source and its emission intensity. This forms a sample of "model input-model response" and constitutes the training sample set and the test sample set.

[0010] A deep learning-CNN algorithm is used to build a groundwater pollution source identification model based on the training sample set, and the model is used to predict the location coordinates and emission intensity of the pollution source.

[0011] Furthermore, the simulated annealing method is used to optimize the Latin hypercube sampling algorithm SA-LHS, and several sets of working conditions are randomly sampled.

[0012] Furthermore, a groundwater solute transport-dual-domain baseline model was constructed using Modflow and MT3DMS programs.

[0013] Furthermore, the groundwater pollution source identification model includes an input layer, a convolutional layer, a pooling layer, a fully connected layer, and an output layer.

[0014] Furthermore, the convolutional layer scans and performs convolution operations on the output of the previous layer using convolutional kernels to extract local features and perform dimensionality reduction processing. The specific expression is as follows:

[0015]

[0016] In the formula, M is the j-th feature output of the l-th layer. j It is the j-th convolutional region of the l-th layer. It is the i-th feature output of the l-1 layer. It is the j-th weight value of the i-th convolutional kernel in the l-th layer. It is the i-th bias of the l-th layer, and * represents the convolution operation.

[0017] After the convolution operation, an activation function is used to enhance the nonlinearity of the extracted local features, improving the learning speed and applicability of the convolutional neural network, further optimizing the model and increasing computational speed. The ReLU function is generally used for nonlinear calculations.

[0018] ReLU = max(0, x)

[0019] In the formula, x is a time series.

[0020] Furthermore, the pooling layer compresses the extracted features to reduce dimensionality and improve computation speed. A commonly used pooling method is max pooling, whose operational model is as follows:

[0021]

[0022] In the formula, It is the value of the t-th neuron in the i-th feature of the l-th layer (t∈[(j﹣1)W,jW]), where W is the width of the pooling region. It represents the value of the neurons in the (l+1)th layer.

[0023] Furthermore, the fully connected layer acts as a classifier, integrating and classifying the outputs of the pooling layer and then mapping the results before outputting them. The mathematical model for this process is as follows:

[0024] O=f(b o +f v w o )

[0025] In the formula, b o Let f be the deviation vector. v For the eigenvector, w o The weight matrix is ​​used; through the fully connected layer step, the process of data from feature extraction to target recognition is realized.

[0026] According to a second aspect of the present disclosure, a groundwater pollution source identification system based on CNN under strongly heterogeneous conditions is provided, comprising:

[0027] The conceptual model building module analyzes hydrogeological and groundwater pollution-related data, reasonably generalizes hydrogeological conditions, and constructs a conceptual model of groundwater solute transport to accurately describe the migration and movement of pollutants in groundwater.

[0028] The working condition acquisition module, in the above groundwater solute transport conceptual model, generates several sets of working conditions characterizing the emission intensity of the pollution source based on the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, using a random sampling method, and generates the aquifer permeability coefficient field through sequential Gaussian simulation to represent the polluted groundwater aquifer system under strong heterogeneity conditions.

[0029] The benchmark model building module constructs a groundwater solute transport-dual-domain benchmark model based on the aquifer permeability coefficient field to accurately reflect the physicochemical processes of pollutants in the groundwater aquifer system.

[0030] The sample set acquisition module inputs several sets of working conditions one by one into the groundwater solute transport-dual-domain baseline model to obtain the model response of the observation data from the observation wells. The model response is the location coordinates of the pollution source and its emission intensity, thus forming a "model input-model response" sample and constituting the training sample set and the test sample set.

[0031] The prediction module uses a deep learning-CNN algorithm to build a groundwater pollution source identification model based on the training sample set, and uses the model to predict the location coordinates and emission intensity of the pollution source.

[0032] According to a third aspect of the present disclosure, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and running on the memory, wherein the processor executes the program to implement the CNN-based groundwater pollution source identification method under strongly heterogeneous conditions.

[0033] According to a fourth aspect of the present disclosure, a computer-readable storage medium is provided having a computer program stored thereon that, when executed by a processor, implements the CNN-based groundwater pollution source identification method under strongly heterogeneous conditions.

[0034] Compared with existing technologies, the above-mentioned technical solutions adopted in this invention have the following advantages: 1) This invention establishes a groundwater pollution source identification model based on CNN under strongly heterogeneous conditions. Compared with traditional groundwater pollution source identification methods, this model can accurately identify groundwater pollution sources under strongly heterogeneous conditions, effectively solve complex high-dimensional nonlinear problems, reduce the application of groundwater solute transport conceptual models, avoid cumbersome calculation processes, shorten the calculation cycle, and improve calculation efficiency.

[0035] 2) This invention solves the problems of accuracy and stability in groundwater pollution source identification. By leveraging the stronger learning ability of CNN, it grasps the complex input-output relationship in the groundwater pollution source identification model. Furthermore, it has good robustness and generalization ability to cope with the problem of groundwater pollution source identification under different types and intensities of pollution sources and complex, highly heterogeneous hydrogeological conditions, thereby ensuring the stability of the model.

[0036] 3) Based on prior data, this invention can accurately identify the location of pollution sources and their emission intensity by obtaining observation data from 7 observation wells for any one year. Attached Figure Description

[0037] The accompanying drawings, which form part of this application, are used to provide a further understanding of this application. The illustrative embodiments of this application and their descriptions are used to explain this application and do not constitute an undue limitation of this application.

[0038] Figure 1 This is an overview map of the instance area;

[0039] Figure 2 A schematic diagram of a two-domain model applicable to strongly heterogeneous conditions;

[0040] Figure 3 This is a schematic diagram of the CNN algorithm. Detailed Implementation

[0041] The present disclosure will be further described below with reference to the accompanying drawings and embodiments.

[0042] It should be noted that the following detailed descriptions are illustrative and intended to provide further explanation of this application. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0043] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments according to this application. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0044] It should be noted that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and systems according to various embodiments of this disclosure. It should be noted that each block in a flowchart or block diagram may represent a module, segment, or portion of code, which may include one or more executable instructions for implementing the logical functions specified in the various embodiments. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that shown in the drawings. For example, two consecutively represented blocks may actually be executed substantially in parallel, or they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the flowcharts and / or block diagrams, and combinations of blocks in the flowcharts and / or block diagrams, may be implemented using a dedicated hardware-based system that performs the specified functions or operations, or using a combination of dedicated hardware and computer instructions.

[0045] Example 1:

[0046] This embodiment provides a method for identifying groundwater pollution sources under strongly heterogeneous conditions based on CNN, including the following steps:

[0047] S1. After analyzing hydrogeological and groundwater pollution-related data, we reasonably generalize hydrogeological conditions and construct a conceptual model of groundwater solute transport to accurately describe the migration and movement of pollutants in groundwater.

[0048] S2. In the above groundwater solute transport conceptual model, based on the characteristics of the pollution source (pollutant type, pollutant concentration distribution, etc.) and the prior interval of the aquifer permeability coefficient (porosity, permeability coefficient, etc.) (as shown in Table 1), a random sampling method is used to generate several sets of working conditions characterizing the emission intensity of the pollution source, and a sequential Gaussian simulation is used to generate the aquifer permeability coefficient field to represent the polluted groundwater aquifer system under strong heterogeneity conditions.

[0049] In conjunction with the specific circumstances of the embodiments, such as Figure 1 As shown; it is assumed that the model boundary is set at a location where the impact of pollutant migration can be approximately ignored. Specifically, blue lines AB and CD are generalized as constant-head boundaries, and black lines AC and BD are generalized as impermeable boundaries. It is assumed that the permeability coefficient is constant within its corresponding grid to simplify calculations.

[0050] Table 1 Prior Intervals for Pollution Source Parameters

[0051]

[0052] S3. Based on the aquifer permeability coefficient field, a groundwater solute transport-dual-domain baseline model is constructed, such as... Figure 2 As shown, this accurately reflects the physicochemical processes of pollutants in the underground aquifer system;

[0053] The physicochemical parameters of this benchmark model are detailed in Table 2:

[0054] Table 2 Physicochemical parameters

[0055]

[0056] Based on the above information, a groundwater solute transport-two-domain baseline model was constructed and initialized using Modflow and MT3DMS programs.

[0057] S4. Input several sets of working conditions into the groundwater solute transport-dual-domain baseline model one by one to obtain the model response of the observation data of the observation well. The model response is the location coordinates of the pollution source and its emission intensity, thus forming a sample of "model input-model response" and constituting the training sample set and the test sample set.

[0058] The training sample size can be 350 for model training, and an additional 150 test samples are formed for model performance testing.

[0059] S5. A groundwater pollution source identification model is constructed based on the training sample set using the deep learning-CNN algorithm, such as... Figure 3 As shown, the model is used to predict the location coordinates and emission intensity of pollution sources.

[0060] It should be noted that the accuracy and reliability of the model were verified by using training and test samples to perform inversion prediction calculations.

[0061] The groundwater pollution source identification model constructed using the CNN algorithm demonstrates high accuracy in predicting the location coordinates of pollution sources and their emission intensity. Based on the test samples, the CNN-based model exhibits high accuracy in fitting emission history (Table 3) and identifying pollution source coordinates (Table 4). This indicates that the CNN-based model can accurately learn the complex input-output relationships in groundwater pollution source identification even under highly heterogeneous conditions, thereby obtaining crucial information about the pollution sources.

[0062] Table 3. Comparison of the average emission values ​​from 10 pollution identification model trials with actual emission values.

[0063]

[0064] Table 4. Comparison of the average coordinates of the pollution identification model after 10 iterations with the actual location coordinates.

[0065]

[0066] In summary, this invention effectively addresses the shortcomings of traditional groundwater pollution source identification models when handling complex datasets. It not only reduces computation time and computational load but also achieves accurate identification of groundwater pollution source coordinates and emission intensity under highly heterogeneous conditions while ensuring efficient computation. This innovative achievement not only provides a scientific basis for the precise treatment of groundwater pollution but also offers a reference for the identification and detection of pollution sources in other environmental fields. Furthermore, by applying computer algorithms and models interdisciplinaryly to solve complex environmental problems related to groundwater pollution, it provides strong technical support for the prevention and control of groundwater pollution and promotes technological innovation in this field.

[0067] Example 2:

[0068] This embodiment provides a CNN-based groundwater pollution source identification system under strongly heterogeneous conditions, including:

[0069] The conceptual model building module analyzes hydrogeological and groundwater pollution-related data, reasonably generalizes hydrogeological conditions, and constructs a conceptual model of groundwater solute transport to accurately describe the migration and movement of pollutants in groundwater.

[0070] The working condition acquisition module, in the above groundwater solute transport conceptual model, generates several sets of working conditions characterizing the emission intensity of the pollution source based on the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, using a random sampling method, and generates the aquifer permeability coefficient field through sequential Gaussian simulation to represent the polluted groundwater aquifer system under strong heterogeneity conditions.

[0071] The benchmark model building module constructs a groundwater solute transport-dual-domain benchmark model based on the aquifer permeability coefficient field to accurately reflect the physicochemical processes of pollutants in the groundwater aquifer system.

[0072] The sample set acquisition module inputs several sets of working conditions one by one into the groundwater solute transport-dual-domain baseline model to obtain the model response of the observation data from the observation wells. The model response is the location coordinates of the pollution source and its emission intensity, thus forming a "model input-model response" sample and constituting the training sample set and the test sample set.

[0073] The prediction module uses a deep learning-CNN algorithm to build a groundwater pollution source identification model based on the training sample set, and uses the model to predict the location coordinates and emission intensity of the pollution source.

[0074] Example 3:

[0075] An electronic device includes a memory, a processor, and a computer program stored in the memory and running thereon. When the processor executes the program, it implements the aforementioned CNN-based method for identifying groundwater pollution sources under highly heterogeneous conditions, comprising:

[0076] After analyzing hydrogeological and groundwater pollution data, we reasonably generalize hydrogeological conditions and construct a conceptual model of groundwater solute transport to accurately describe the migration and movement of pollutants in groundwater.

[0077] In the conceptual model of groundwater solute transport, based on the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, a random sampling method is used to generate several sets of operating conditions characterizing the emission intensity of the pollution source, and a sequential Gaussian simulation is used to generate the aquifer permeability coefficient field to represent the polluted groundwater aquifer system under strong heterogeneity.

[0078] Based on the aquifer permeability coefficient field, a groundwater solute transport-dual-domain benchmark model is constructed to accurately reflect the physicochemical processes of pollutants in the groundwater aquifer system;

[0079] Several sets of operating conditions are input into the groundwater solute transport-dual-domain baseline model one by one to obtain the model response of the observation data of the observation well. The model response is the location coordinates of the pollution source and its emission intensity. This forms a sample of "model input-model response" and constitutes the training sample set and the test sample set.

[0080] A deep learning-CNN algorithm is used to build a groundwater pollution source identification model based on the training sample set, and the model is used to predict the location coordinates and emission intensity of the pollution source.

[0081] Example 4:

[0082] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned CNN-based groundwater pollution source identification method under strongly heterogeneous conditions, comprising:

[0083] After analyzing hydrogeological and groundwater pollution data, we reasonably generalize hydrogeological conditions and construct a conceptual model of groundwater solute transport to accurately describe the migration and movement of pollutants in groundwater.

[0084] In the conceptual model of groundwater solute transport, based on the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, a random sampling method is used to generate several sets of operating conditions characterizing the emission intensity of the pollution source, and a sequential Gaussian simulation is used to generate the aquifer permeability coefficient field to represent the polluted groundwater aquifer system under strong heterogeneity.

[0085] Based on the aquifer permeability coefficient field, a groundwater solute transport-dual-domain benchmark model is constructed to accurately reflect the physicochemical processes of pollutants in the groundwater aquifer system;

[0086] Several sets of operating conditions are input into the groundwater solute transport-dual-domain baseline model one by one to obtain the model response of the observation data of the observation well. The model response is the location coordinates of the pollution source and its emission intensity. This forms a sample of "model input-model response" and constitutes the training sample set and the test sample set.

[0087] A deep learning-CNN algorithm is used to build a groundwater pollution source identification model based on the training sample set, and the model is used to predict the location coordinates and emission intensity of the pollution source.

[0088] Those skilled in the art will understand that the modules or steps described above can be implemented using general-purpose computer devices. Optionally, they can be implemented using computer-executable program code, which can then be stored in a storage device for execution by a computer device. Alternatively, they can be fabricated as separate integrated circuit modules, or multiple modules or steps can be fabricated as a single integrated circuit module. This disclosure is not limited to any particular combination of hardware and software.

[0089] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

[0090] While the specific embodiments of this disclosure have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of this disclosure. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of this disclosure are still within the scope of protection of this disclosure.

Claims

1. A CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions, characterized in that, The method comprises the following steps: After analyzing the hydrogeological and groundwater pollution related data, the hydrogeological conditions are reasonably generalized, and a groundwater solute transport conceptual model is constructed to accurately describe the migration and movement of pollutants in groundwater; In the groundwater solute transport conceptual model, according to the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, a random sampling method is used to generate a plurality of groups of working conditions representing the discharge intensity of the pollution source, and a sequential Gaussian simulation is used to generate an aquifer permeability coefficient field to represent a contaminated underground aquifer system under strong heterogeneous conditions; Based on the aquifer permeability coefficient field, a groundwater solute transport-bi-domain benchmark model is constructed to accurately reflect the physical and chemical processes of pollutants in the underground aquifer system; The plurality of groups of working conditions are input into the groundwater solute transport-bi-domain benchmark model one by one to obtain the model response of the observation data of the observation well, which is the position coordinates and discharge intensity of the pollution source, thereby forming a "model input-model response" sample and constituting a training sample set and a test sample set; A deep learning-CNN algorithm is used to construct a groundwater pollution source identification model based on the training sample set, and the model is used to predict the position coordinates and discharge intensity of the pollution source. 2.The CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions according to claim 1, characterized in that, The simulated annealing method is used to optimize the Latin hypercube sampling algorithm SA-LHS to randomly sample a plurality of groups of working conditions. 3.The CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions according to claim 1, characterized in that, The Modflow and MT3DMS programs are used to construct a groundwater solute transport-bi-domain benchmark model.

4. The CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions according to claim 1, characterized in that, The groundwater pollution source identification model comprises an input layer, a convolution layer, a pooling layer, a full connection layer and an output layer.

5. The CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions according to claim 4, characterized in that, The convolution layer scans and convolves the output results of the previous layer through a convolution kernel, extracts local features, and performs dimension reduction processing, and the specific expression is: In the formula, is the first layer of the first feature output, is the first layer of the first convolution region, is the first layer of the first feature output, is the first layer of the first convolution kernel of the first weight value, is the first layer of the first bias, represents a convolution operation; After the convolution operation, an activation function is used to enhance the nonlinearity of the extracted local features, so a ReLU function is used for non-linear calculation: In the formula, is a time series.

6. The CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions according to claim 4, characterized in that, The pooling layer compresses the extracted features to reduce the dimension, and the operation model is: wherein is the number of features in the layer, is the number of neurons in the layer, , is the width of the pooling region, is the number of neurons in the layer.

7. The CNN-based method for identifying groundwater pollution sources under strong heterogeneous conditions according to claim 4, characterized in that, The full connection layer serves as a classifier to integrate and classify the output results of the pooling layer and output the results after mapping, and the mathematical model of the specific process is: In the formula, is a bias vector, is a feature vector, is a weight matrix; through this step of the full connection layer, the process of data from feature extraction to target recognition is realized.

8. A groundwater pollution source identification system under strong heterogeneous conditions based on CNN, characterized in that, It comprises: The conceptual model establishment module analyzes the hydrogeological and groundwater pollution related data, reasonably generalizes the hydrogeological conditions, and constructs a groundwater solute transport conceptual model to accurately describe the migration and movement of pollutants in groundwater; The working condition acquisition module, in the groundwater solute transport conceptual model, generates a plurality of groups of working conditions representing the discharge intensity of the pollution source according to the characteristics of the pollution source and the prior interval of the aquifer permeability coefficient, and generates an aquifer permeability coefficient field through sequential Gaussian simulation to represent a contaminated underground aquifer system under strong heterogeneous conditions; The benchmark model establishment module, based on the aquifer permeability coefficient field, constructs a groundwater solute transport-bi-domain benchmark model to accurately reflect the physical and chemical processes of pollutants in the underground aquifer system; The sample set obtaining module inputs each group of working conditions into the groundwater solute transport-two-domain benchmark model one by one to obtain model responses of the observation data of the observation well, the model responses being position coordinates and discharge intensities of the pollution source, thereby forming a "model input-model response" sample and constituting the training sample set and the test sample set; The prediction module adopts a deep learning-CNN algorithm, constructs a groundwater pollution source identification model based on the training sample set, and uses the model to realize prediction of the position coordinates and the discharge intensity of the pollution source. 9.An electronic device comprising a memory, a processor, and a computer program stored on the memory, wherein the processor implements the CNN-based method for identifying a groundwater pollution source under strong heterogeneous conditions according to any one of claims 1-7 when executing the program. 10.A computer-readable storage medium having a computer program stored thereon, wherein the program implements the CNN-based method for identifying a groundwater pollution source under strong heterogeneous conditions according to any one of claims 1-7 when executed by a processor.

Images