Groundwater resource comprehensive evaluation method and system based on data analysis

Abstract

This invention discloses a data analysis-based comprehensive evaluation method and system for groundwater resources, relating to the field of groundwater resource monitoring technology. By collecting relevant groundwater resource data from the area to be evaluated, a groundwater resource evaluation index system is constructed, including a health index, resilience index, and ecosystem service index. A game theory-based weighting method is used to determine the weight coefficients of each index. Based on the determined weight coefficients, a comprehensive groundwater evaluation index for the area to be evaluated is calculated. Finally, groundwater resources are classified according to the obtained comprehensive evaluation index. This invention overcomes the limitations of traditional technologies that focus only on a single dimension of water quality or quantity, achieving a comprehensive characterization of the groundwater system from its living state, disturbance resistance capacity, and socio-ecological contributions. It provides a systematic, dynamic, complete, and intelligent comprehensive evaluation solution for groundwater resources, offering a scientific basis for sustainable groundwater resource management.

Description

Technical Field

[0001] This invention relates to the field of groundwater resource monitoring technology, and in particular to a method and system for comprehensive evaluation of groundwater resources based on data analysis. Background Technology

[0002] Groundwater resources, as a vital strategic water resource, play an irreplaceable role in ensuring domestic water supply for urban and rural residents, supporting industrial and agricultural production, and maintaining ecosystem balance. However, with intensifying climate change and increasing human activity, groundwater resources are facing multiple pressures, including water quantity depletion, water quality degradation, and ecological function decline. Conducting scientific and comprehensive groundwater resource assessments has become a key technical means to support sustainable water resource management, land spatial planning, and ecological environmental protection.

[0003] Traditional groundwater resource assessment technologies have mainly gone through three stages of development: Phase 1: Single-Factor Evaluation Technology. This technology, based on normative documents such as the "Groundwater Quality Standard" (GB / T14848-2017), compares various water quality monitoring indicators one by one, determining the overall water quality category based on the worst indicator level. This method is simple to operate and standardized, and is widely used in the routine water quality compliance evaluation of drinking water sources. However, this technology suffers from a conservative "one-vote veto" flaw, failing to reflect the overall condition of the water body and completely ignoring water quantity factors and spatial heterogeneity.

[0004] The second stage: Comprehensive index evaluation techniques. To overcome the limitations of single-factor evaluation, researchers developed the comprehensive index method, which characterizes groundwater quality by calculating the weighted or geometrical mean of multiple indicators. Furthermore, fuzzy mathematics comprehensive evaluation methods were introduced, constructing membership functions to describe the "fuzzy" boundaries of water quality indicators with respect to various standards and assigning different weights to each indicator, thus more accurately reflecting the water quality status. Some studies also incorporated GIS technology to achieve spatial visualization of the evaluation results. However, these techniques still primarily focus on water quality assessment, with insufficient consideration for water quantity, sustainability, and ecosystem services.

[0005] The third stage: Multivariate statistical and machine learning evaluation techniques. In recent years, with the development of big data and artificial intelligence technologies, researchers have begun to use multivariate statistical and machine learning methods such as factor analysis, TOPSIS, neural networks, and support vector machines for groundwater evaluation. For example, some studies use high-precision hydrogeological parameters output from numerical simulations as input to drive machine learning models for groundwater potential evaluation, achieving a prediction accuracy of over 85%. Other studies integrate multi-source satellite remote sensing data with interpretable artificial intelligence to invert the spatiotemporal changes of large-scale groundwater reserves and attribute the impacts of climate change and human activities.

[0006] While the development of existing technologies has improved the accuracy and efficiency of assessments to some extent, it still focuses on a single dimension of water quality or quantity assessment. Groundwater systems, as complex living organisms, should be evaluated across three interconnected dimensions: their health status, resilience to disturbances, and ecosystem services provided to society. Current technologies lack a systematic approach that integrates these three dimensions into a unified evaluation framework. Therefore, this paper proposes a comprehensive groundwater resource evaluation method and system based on data analysis. Summary of the Invention

[0007] The main objective of this invention is to provide a comprehensive evaluation method and system for groundwater resources based on data analysis, which can effectively solve the problems in the background art.

[0008] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A data-driven comprehensive evaluation method for groundwater resources includes the following steps: S1: Collect relevant data on groundwater resources in the area to be evaluated, including water quality monitoring data, water quantity monitoring data, hydrogeological parameters, meteorological data, land use data, socio-economic data, and ecological data, and perform quality control and normalization preprocessing on the collected data; S2: Based on the collected multi-source data, construct a groundwater resource evaluation index system, wherein the evaluation index includes at least the health index H(x), the resilience index R(x), and the ecosystem service index S(x); S3: The weighting coefficients α, β, and γ of the health index H(x), the resilience index R(x), and the ecosystem service index S(x) are determined by using a game theory-based combination weighting method. S4: Based on the determined weighting coefficients, use the formula Calculate the comprehensive evaluation index of groundwater in the area to be evaluated. , where α+β+γ=1; S5: Classify groundwater resources according to the obtained comprehensive evaluation index and output the evaluation results.

[0009] Furthermore, in step S2, The health index H(x) is calculated comprehensively based on the groundwater's water quality health, water quantity health, metabolic function, and self-purification capacity, and is defined as follows: H1 is the water quality health index, H2 is the water quantity health index, H3 is the metabolic function index, H4 is the self-purification capacity index, and ω is the water quality health index. i λ is the weight of the i-th index, λ is the pollution decay coefficient, and t is the pollution duration. The resilience index R(x) is calculated based on the groundwater's absorption capacity, recovery capacity, and adaptability, and is defined as follows: ;R a R is the water absorption capacity index. r R is the recovery capacity index. e For adaptability index; The service index S(x) is calculated comprehensively based on the value of water supply services, ecological maintenance, geological environment stability, climate regulation, and cultural services, and is defined as follows: V k For the value of the k-th type of service, V k，max The value of the k-th class is the maximum value in the region, λ. k Weights for service value types.

[0010] Furthermore, the water quality health index H1 is estimated using a cloud model: ; Ex represents the expectation of the cloud model, En represents entropy, He represents hyperentropy, and x represents the expected value of the cloud model. std σ represents the water quality standard value, and σ represents the standard deviation. The water health index H2 is defined as follows: Q avail Q represents the amount of exploitable resources. demand For water demand, h represents the actual groundwater level. eco The water level is the ecological warning level; min() is the minimum value operation. The metabolic function index H3 is defined as follows: ; S t S is the current water chemical entropy, S0 is the background entropy value, and S max p is the maximum entropy value. k The relative abundance of the k-th hydrochemical type; The self-cleaning capacity index H4 is estimated using a fuzzy neural network: ; For the fuzzy membership degree of the input variable, is the consequent parameter of the fuzzy rule, and J is the number of fuzzy rules.

[0011] Furthermore, the water absorption index R a Defined as: V buffer As an adjustable storage capacity, V total For total storage capacity, C represents the annual change rate of pollutant concentration. C max This is the maximum permissible rate of change; The recovery capacity index R r Defined as: T reference For reference recovery time, T recoveryRecovery time is calculated as follows: L is the characteristic length of the aquifer, and D e Where θ is the effective diffusion coefficient, θ is the porosity, C0 is the initial concentration, and C eq To balance the concentration, C t The target concentration; The adaptability index R e Defined as: K c η is the hydraulic connectivity index, M is the aquifer thickness, D is the burial depth, and η is the heterogeneity index.

[0012] Furthermore, the water supply service value V1 is defined as: Q represents the water supply, and P represents the water volume. shadow For the shadow price, η loss Water loss rate; The ecological maintenance value V2 is defined as follows: B is the base current contribution rate, A wetland For wetland areas that rely on groundwater, U energy The monetary value of energy per unit area is defined as: , τ i Let R be the energy conversion rate of the i-th type of input. i Let i be the input amount for the i-th example; The geological environment stability value V3 is defined as follows: C prevention For full protection cost, k is the curve steepness coefficient, h is the actual water level, and h crit This is the critical water level; The climate regulation value V4 is defined as follows: C ω The volumetric heat capacity of water. T represents the temperature adjustment range, A represents the area of ​​the affected region, and P represents the area of ​​the affected region. carbon For carbon sink prices; The cultural service value V5 is defined as follows: WTP existence WTP is a platform for demonstrating willingness to pay value. bequest Willingness to pay for the bequest value, based on the formula: Let Y be the income level, X be the socio-demographic variable, and a, b, and c be the fitting coefficients.

[0013] Furthermore, in step S3, the use of game theory combinatorial weighting method includes: S31: The subjective weights ω of each index are calculated using the triangular fuzzy number hierarchical analysis method. sub : , , l f mf u f ; represents the triangular fuzzy number judgment of the f-th expert, θ f Here, F represents the confidence level, and F represents the number of experts. S32: Calculate the objective weights ω of each index using the improved CRITIC method. obj : r jk Let σ be the correlation coefficient between the j-th and k-th indicators. j Let j be the standard deviation of the j-th indicator; S33: Construct a game theory combinatorial weighting model and solve for the combinatorial coefficients. Specifically: Construct the weight matrix W=[ω sub T ω obj T ], where T is the matrix transpose; Construct the Gram matrix G=W T W; Solve the linear system of equations Gθ=b, where b=W T w T w=(ω) sub +ω obj ) / 2; Sub-step 5.4: Normalize the combination coefficients and calculate the final combination weights: θ * =θ / ∑θ,ω * =θ1 * ω sub +θ2 * ω obj .

[0014] Furthermore, in step S5, the classification of groundwater resources based on the obtained comprehensive evaluation index includes: S51: Construct a comprehensive evaluation grading standard to classify the health level of groundwater systems into five levels from high to low, specifically: When the comprehensive evaluation index ranges from [0.85 to 1.00], the system is in a healthy state, with a health level of I. When the comprehensive evaluation index ranges from [0.70, 0.85), the system is in a sub-healthy state, with a health level of II. When the comprehensive evaluation index ranges from [0.55, 0.70), the system is in a risky state, and its health level is Level III. When the comprehensive evaluation index ranges from [0.45, 0.55), the system is in a damaged state, and its health level is IV. When the comprehensive evaluation index ranges from [0.00, 0.45), the system is in a state of collapse and its health level is V. S52: Fuzzy membership functions are used to handle the boundaries of each health level. Specifically: c g Let σ be the standard central value for the g-th health level. g is the width parameter for the g-th health level.

[0015] Furthermore, it also includes: Constructing a 3D radar map for feature analysis: ; Calculate the balance index of each of the evaluation indicators: ,in, ; Based on the balance index, strengths and weaknesses are identified, and the identification rules are as follows: Advantages: ; Shortcomings: .

[0016] Furthermore, it also includes: conducting sensitivity analysis, including local sensitivity analysis and global sensitivity analysis; wherein: The local sensitivity analysis uses the finite difference method to calculate the local sensitivity coefficient S of the i-th evaluation index on the j-th spatial unit. ij : x ij Let be the value of the i-th evaluation index on the j-th spatial unit; The global sensitivity analysis uses the Sobol' method to calculate the first-order exponent S. i And the total effect index S Ti : x ～i Let E be the evaluation index excluding the i-th one, E be the conditional expectation, and Var be the total variance.

[0017] A comprehensive groundwater resource evaluation system based on data analysis includes: The data acquisition and preprocessing module is used to collect multi-source data of the evaluation area, including water quality monitoring data, water quantity monitoring data, hydrogeological parameters, meteorological data, land use data, socio-economic data and ecological data, and to perform quality control and normalization preprocessing on the collected data. The health index calculation module is connected to the data acquisition and preprocessing module and is used to construct the groundwater health index H(x) based on the preprocessed data. The health index H(x) is calculated comprehensively based on four dimensions: water quality health, water quantity health, metabolic function, and self-purification capacity. The resilience index calculation module is connected to the data acquisition and preprocessing module and is used to construct the groundwater resilience index R(x) based on the preprocessed data. The resilience index R(x) is calculated comprehensively based on three dimensions: absorption capacity, recovery capacity and adaptability. The service index calculation module is connected to the data acquisition and preprocessing module and is used to construct the ecosystem service index S(x) based on the preprocessed data. The service index S(x) is calculated comprehensively based on five dimensions: water supply service value, ecological maintenance value, geological environment stability value, climate regulation value, and cultural service value. The weight determination module is connected to the health index calculation module, the resilience index calculation module, and the service index calculation module, respectively, and is used to determine the weight coefficients α, β, and γ of the health index H(x), the resilience index R(x), and the service index S(x) using a game theory combined weighting method. The comprehensive evaluation index calculation module is connected to the health index calculation module, resilience index calculation module, service index calculation module, and weight determination module, respectively, and is used to calculate the index based on the formula: Calculate the comprehensive evaluation index of groundwater, where α+β+γ=1; The grading and output module is connected to the comprehensive evaluation index calculation module and is used to grade groundwater resources according to the comprehensive evaluation index and output the evaluation results. The visualization analysis module is connected to the health index calculation module, resilience index calculation module, service index calculation module and comprehensive evaluation index calculation module respectively. It is used to generate a three-dimensional radar chart for feature analysis, calculate the balance index and identify the strength and weakness dimensions. The sensitivity analysis module is connected to both the data acquisition and preprocessing module and the comprehensive evaluation index calculation module. It is used to perform local and global sensitivity analyses. The global sensitivity analysis employs the Sobol' method to calculate the first-order exponent S. i And the total effect index S Ti .

[0018] The present invention has the following beneficial effects: Compared with existing technologies, this solution breaks through the limitations of traditional technologies that only focus on a single dimension of water quality or quantity. For the first time, it incorporates the three dimensions of "health-resilience-service" into a unified evaluation framework. By constructing a three-dimensional coupled model, it achieves a comprehensive characterization of the groundwater system from "living state-disturbance resistance-social and ecological contribution", which can more comprehensively reflect the real condition of the groundwater system.

[0019] Compared with existing technologies, this approach constructs a three-dimensional resilience index that includes absorption capacity, recovery capacity, and adaptability. In comparative tests simulating pollution event impacts, it can quantitatively characterize the resilience differences among different aquifer systems. This quantitative result provides crucial decision-making support for groundwater risk management.

[0020] Compared with existing technologies, this scheme realizes the explicit accounting of the implicit value of groundwater by introducing the ecosystem service index, providing a more complete value basis for the design of ecological compensation mechanisms and land spatial planning. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating the data analysis-based comprehensive evaluation method for groundwater resources according to the present invention. Figure 2 This is a schematic diagram of the structure of the groundwater resource comprehensive evaluation system based on data analysis of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0023] Example 1: This invention provides a comprehensive evaluation method for groundwater resources based on data analysis. (See also...) Figure 1 The flowchart shown below uses a coastal city in eastern China as an example. This city has an area of ​​approximately 1200 km². 2 The groundwater type is mainly Quaternary pore water, with an aquifer thickness of 20-50m and an average annual precipitation of 650mm. In recent years, due to over-exploitation and seawater intrusion, groundwater resources are facing serious threats. The implementation process of this invention is described in detail, including the following stages and steps: Phase 1: System Initialization and Parameter Setting Step 101: Evaluation Zoning and Grid Division The study area was discretized into a 1km × 1km grid, resulting in a total of 1200 grid cells I × J = 40 × 30 = 1200 cells. Each grid cell serves as an independent evaluation unit, denoted as Ω. ij , where i = 1, 2, ..., 40; j = 1, 2, ..., 30.

[0024] Step 102: Evaluation of time series settings The evaluation period is from 2015 to 2024, with a time step of 1 year and a total of N=10 time nodes, denoted as T={t0, t1, ..., t9}, where t0 corresponds to 2015 and t9 corresponds to 2024.

[0025] Step 103: System Parameter Initialization Configure the following system parameters: The initial weights for each dimension of the health index are as follows: w1=0.35, w2=0.30, w3=0.20, w4=0.15; The initial weights for each type of service index are as follows: λ1=0.40, λ2=0.25, λ3=0.15, λ4=0.10, λ5=0.10; Initial weights for the comprehensive evaluation dimensions, such as: α=0.40, β=0.30, γ=0.30; Pollution attenuation coefficient, such as λ=0.05 (determined by referring to a table based on the type of water-containing medium). Phase Two: Data Acquisition and Preprocessing Step 201: Multi-source data acquisition The data acquisition and preprocessing module collects data from multiple sources through the following methods:

[0026] Step 202: Data Quality Control Outlier detection and missing value handling were performed on the collected data: Outlier detection: The 3σ principle was used to review data with a mean of 3σ. A total of 87 outliers were identified, of which 21 were removed after verification, and 66 were corrected.

[0027] Missing value handling: Linear interpolation is used for missing values ​​in time series data, and Kriging interpolation is used for missing values ​​in spatial data. Interpolation accuracy is verified through cross-validation and R... 2 =0.88.

[0028] Step 203: Data Normalization Map all indicator data to the interval [0, 1].

[0029] Phase 3: Calculation of the Health Index H(x) Step 301: Calculation of Water Quality Health Index H1 (using cloud model) With grid cell Ω 15，20 For example, there are 3 monitoring wells in this grid, and the average water quality value in 2024 is used:

[0030] Calculate cloud model parameters using the TDS metric as an example: Expected value: Ex = 850; Entropy En = (2 / π) 1 / 2 ×1 / n×∑∣x k -Ex∣=1.253×28.5=35.7; Hyperentropy He = 0.1 × En = 3.57; Generate random numbers ε ~ N(0, 1), and take ε = 0.63: En′=En+He×ε=35.7+3.57×0.63=37.95.

[0031] Calculate membership degree: μ(x) j =exp[-(850-850)] 2 / (2×37.95) 2 ]=1; Calculate the exponential contribution: μ(x) j )×exp[-(850-1000) 2 / (2×100) 2 ] = exp(-1.125) = 0.325; Repeatedly calculate all indicators and take the average: H1=0.58.

[0032] Step 302: Calculation of Water Health Index H2 This grid cell data: Extractable resources Q avail =12.5×10 4 m 3 / a; Water demand Q demand =15.8×10 4 m 3 / a; The actual water level is h = 8.2m; Ecological warning water level h eco =6.5m (determined based on the distribution of vegetation roots); Calculate the water ratio: Q avail / Q demand =12.5 / 15.8=0.79; Calculate the water level coefficient: min(1, h / h) eco ) = min(1, 8.2 / 6.5) = 1; Water health index: H2=0.79×1=0.79.

[0033] Step 303: Calculation of Metabolic Function Index H3 According to the hydrochemical type classification (Shukarev classification), the hydrochemical type distribution of this grid cell is as follows:

[0034] Calculate the current water chemical entropy: S t =-∑p k lnp k=-(0.45ln0.45+0.30ln0.30+0.15ln0.15+0.10ln0.10)=1.21; Background entropy value S0 = 0.95 (based on 2015 data) Maximum entropy S max =ln4=1.39; Metabolic function index: H3=exp[-(1.21-0.95) / (1.39-0.95)]=exp(-0.59)=0.55.

[0035] Step 304: Calculation of self-cleaning capacity index H4 (fuzzy neural network) Construct a fuzzy neural network with 4 inputs and 1 output: Input: pH=7.2, Eh=215mV, TOC=2.8mg / L, microbial activity=0.65; Number of fuzzy rules: J=5 4 =625; Taking rule number 1 as an example (low pH, low Eh, low TOC, low microbial activity): Calculate the membership degree of each input: ; Rule activation strength: α1 = 0.38 × 0.94 × 0.73 × 0.61 = 0.159; The normalized activation strength and corresponding output are calculated for all 625 rules, and finally: H4=0.62.

[0036] Step 305: Comprehensive Health Index Health index: H(x) = (0.58) 0.35 ×0.79 0.30 ×0.55 0.20 ×0.62 0.15 )×e -0.05×9 =0.62×0.64=0.40.

[0037] Phase 4: Calculation of the resilience index R(x) Step 401: Absorption Capacity R a calculate This grid cell data: Adjustable storage capacity V buffer =18.5×10 4 m 3 (Within the range of water level fluctuations); Total storage capacity V total =42.3×10 4 m 3 ; The annual change rate of pollutant concentration ΔC = 3.2 mg / (L) a); Maximum permissible rate of change ΔC max =8.5mg / (L a); Absorption capacity index: R a =18.5 / 42.3×(1-3.2 / 8.5)=0.437×(1-0.376)=0.273.

[0038] Step 402: Recovery Ability R r calculate This grid cell data: Characteristic length of aquifer L = 350m (variogram sill value); Effective diffusion coefficient D e =2.3×10 -4 m 2 / s; Porosity θ = 0.28; Initial concentration C0 = 180 mg / L; equilibrium concentration C eq =120mg / L; Target concentration C t =150 mg / L; Recovery time: T recovery = 350 2 / (2.3×10 -4 ) × 0.28 / π 2 ×ln[(180-120) / (150-120)]=121.5 days; Reference recovery time T reference =150 days; Recovery Capacity Index: R r =exp(-121.5 / 150)=exp(-0.81)=0.445.

[0039] Step 403: Adaptability R e calculate This grid cell data: Hydraulic connectivity index K c =0.52 (based on tracer experiments); Aquifer thickness M=28m; Burial depth D=15 m; The average permeability coefficient is 8.5 m / d; Standard deviation of permeability σ K =3.4 m / d; Heterogeneity index: η = 3.4 / 8.5 = 0.4; Adaptability Index: R e = (0.52 × 28) / (15 × 0.4) = 14.56 / 6 = 2.43 > 1 R e =1.0.

[0040] Step 404: Overall Resilience Index: .

[0041] Phase 5: Calculation of Service Index S(x) Step 501: Water Supply Service Value V1 This grid cell data: Water supply Q = 8.5 × 10 4 m 3 / a; Shadow price P shadow =3.2 yuan / m 3 (Alternative water source method); Water loss rate η loss =0.12.

[0042] Water supply service value: V1 = 8.5 × 10 4 ×3.2×(1-0.12)=23.94×10 4 Yuan=239,400 Yuan Step 502: Ecological Maintenance Value V2 This grid cell data: Base current contribution rate B = 0.35 (digital filtering method) Area A of wetlands dependent on groundwater wetland =0.12km 2 =12hm 2 ; Energy conversion rate τ1 = 1.05 × 10 5 sej / J (solar energy); Solar energy input R1 = 5.2 × 10 13 J / hm 2 ; Energy / Currency ratio: 1.2 × 10 12 sej / yuan; Energy per unit area: Monetary value U emergy = (1.05 × 10 5 ×5.2×10 13 ) / (1.2×10 12 =4.55 × 10 6 Yuan / hm 2; Ecological maintenance value: V² = 0.35 × 12 × 4.55 × 10 6 =19.11×10 6 Yuan = 19.11 million yuan.

[0043] Step 503: Geological Environment Stability Value V3 This grid cell data: Full protection cost C prevention =8.5 million yuan (recharge project); The curve steepness coefficient k = 0.8; The actual water level h = 8.2 m; Critical water level h crit =5.5m (critical value for seawater intrusion); Geological environment stability value: V3 = 850 / (1 + e) -0.8×(8.2-5.5) )=850 / 1.115=7.62 million yuan.

[0044] Step 504: Climate Regulation Value V4 This grid cell data: Volumetric heat capacity C w =4.18×10 6 J / (m 3 K); Temperature adjustment range ΔT = 1.8K (remote sensing inversion); The area affected is A = 0.8 km² 2 =8×10 5 m 2 ; Carbon sink price P carbon =65 yuan / t, equivalent to 1.63 × 10 -8 Yuan / J; Climate regulation value: V4 = 4.18 × 10 6 ×1.8×8×10 5 ×1.63×10 -8 =98.2×10 3 Yuan = 98,200 yuan.

[0045] Step 505: Cultural Service Value V5 Conduct a conditional valuation survey with a sample size of n=450: WTP has the willingness to pay value existence =85 yuan / person / year; Willingness to Pay Bequest Value (WTP) bequest =62 yuan / person / year; The number of beneficiaries is N=3200; Cultural service value: V5 = (85 + 62) × 3200 = 147 × 3200 = 47.04 × 10 4 Yuan = 470,400 yuan.

[0046] Step 506: Comprehensive Service Index Maximum value of each service (region-wide statistics): V 1，max =652,000 yuan; V 2，max =38.5 million yuan; V 3，max =12.4 million yuan; V 4，max =215,000 yuan; V 5，max =896,000 yuan; Normalized value: V1 / V 1，max =23.94 / 65.2=0.367; V2 / V 2，max =1911 / 3850=0.496; V3 / V 3，max =762 / 1240=0.615; V4 / V 4，max =9.82 / 21.5=0.457 V5 / V 5，max =47.04 / 89.6=0.525.

[0047] Service index: S(x) = 0.40 × 0.367 + 0.25 × 0.496 + 0.15 × 0.615 + 0.10 × 0.457 + 0.10 × 0.525 = 0.462.

[0048] Phase 6: Weight Determination Step 601: Subjective weight calculation (triangular fuzzy number AHP) Five experts were invited to conduct pairwise comparisons across three dimensions: health, resilience, and service. Expert 1's judgment matrix (triangular fuzzy number) is as follows:

[0049] Calculate the fuzzy weights for expert 1: W 1，H =(1+1+2, 1+2+3, 1+3+4) / (1+1+2+1 / 3+1+1 / 3+1 / 4+1 / 3+1, ...) =(4, 6, 8) / (5.58, 9.33, 14.5) =(0.72, 0.64, 0.55); Deblurring: defuzzify=(0.72+4×0.64+0.55) / 6=0.64.

[0050] Similarly, expert 1's W 1，R =0.23, w 1，S =0.13.

[0051] Based on the combined opinions of 5 experts (confidence levels of 0.9, 0.8, 0.9, 0.7, and 0.8 respectively): w sub =(0.42, 0.33, 0.25).

[0052] Step 602: Objective Weight Calculation (Improved CRITIC Method) Based on the health, resilience, and service index values ​​of 1200 grids across the entire region, as shown below:

[0053] Information content: C H =0.18 × 1.00 = 0.18; C R =0.21 × 0.87 = 0.183; C S =0.15 × 0.91 = 0.137; Objective weight: w obj =(0.18 / 0.5, 0.183 / 0.5, 0.137 / 0.5)=(0.36, 0.37, 0.27).

[0054] Step 603: Game Theory Combinatorial Weighting Construct the weight matrix: W = [w sub T w obj T ]= ; Calculate the Gram matrix: G = W T W= = ; Calculate the average weight w' = (w sub +w obj ) / 2=(0.39, 0.35, 0.26).

[0055] Calculate the right-hand vector: b = W T w' T = = ; Solve the system of linear equations: =

[0056] Solving for θ, we get: θ1 = 0.52, θ2 = 0.48; Normalization: θ =(0.52, 0.48); Final combined weights: w =0.52×(0.42, 0.33, 0.25)+0.48×(0.36, 0.37, 0.27)=(0.39, 0.35, 0.26)=(0.39, 0.35, 0.26); That is: α=0.39, β=0.35, γ=0.26.

[0057] Phase 7: Comprehensive Evaluation and Grading Step 701: Calculate the comprehensive evaluation index GWR HRS =0.39×0.40+0.35×0.49+0.26×0.462=0.156+0.172+0.120=0.448 Step 702: Level Classification According to the evaluation grading standards, GWR HRS =0.448∈[0.40, 0.55), belonging to level IV (poor).

[0058] Verification using fuzzy membership functions: Level I: c1=0.925, σ1=0.05, μ1=exp(-(0.448-0.925)) 2 / (2×0.05 2 ))≈0; Level II: c²=0.775, σ²=0.05, μ²≈0; Level III: c3=0.625, σ3=0.05, μ3≈0; Level IV: c4=0.475, σ4=0.05, μ4=exp(-(0.448-0.475)) 2 / 0.005)=exp(-0.1458)=0.86; Level V: c5=0.20, σ5=0.10, μ5=exp(-(0.448-0.20)) 2 / 0.02)=exp(-3.075)=0.046.

[0059] The maximum membership degree is 0.86, corresponding to level IV, which is consistent with direct determination.

[0060] Step 703: Radar Chart Analysis 3D radar image feature vector: V radar =(0.40, 0.49, 0.46); Balance calculation:

[0061] The balance degree B = 0.917 > 0.9, indicating a high degree of balance.

[0062] Advantage Dimension: Resilience R=0.49; Weakness dimension: Health H=0.40.

[0063] Phase 8: Sensitivity Analysis Step 801: Local Sensitivity Analysis Taking water level h as an example, in grid cell Ω 15，20 Raise the water level by 10% (from 8.2m to 9.02m) and recalculate the comprehensive evaluation index: ΔGWR = 0.468 - 0.448 = 0.020; Local sensitivity coefficient: S h，15，20 =0.020 / 0.82=0.0244per0.1m.

[0064] Step 802: Global Sensitivity Analysis (Sobol's Method) Using Monte Carlo sampling, with a sample size N=1000 and an input variable m=5, the calculation results are as follows:

[0065] Conclusion: Water level is the main controlling factor, and the interaction between extraction volume and permeability coefficient is significant and cannot be ignored.

[0066] Example 2: This invention provides a data analysis-based comprehensive evaluation system for groundwater resources, used to implement the above-mentioned method. (See also...) Figure 2 The system architecture diagram shown includes: The data acquisition and preprocessing module is used to collect multi-source data of the evaluation area, including water quality monitoring data, water quantity monitoring data, hydrogeological parameters, meteorological data, land use data, socio-economic data and ecological data, and to perform quality control and normalization preprocessing on the collected data. The health index calculation module is connected to the data acquisition and preprocessing module. It is used to construct the groundwater health index H(x) based on the preprocessed data. The health index H(x) is calculated comprehensively based on four dimensions: water quality health, water quantity health, metabolic function and self-purification capacity. The resilience index calculation module is connected to the data acquisition and preprocessing module. It is used to construct the groundwater resilience index R(x) based on the preprocessed data. The resilience index R(x) is calculated comprehensively based on three dimensions: absorption capacity, recovery capacity and adaptability. The service index calculation module is connected to the data acquisition and preprocessing module. It is used to construct the ecosystem service index S(x) based on the preprocessed data. The service index S(x) is calculated comprehensively based on five dimensions: water supply service value, ecological maintenance value, geological environment stability value, climate regulation value, and cultural service value. The weight determination module is connected to the health index calculation module, the resilience index calculation module, and the service index calculation module, respectively. It is used to determine the weight coefficients α, β, and γ of the health index H(x), the resilience index R(x), and the service index S(x) using a game theory combination weighting method. The comprehensive evaluation index calculation module is connected to the health index calculation module, resilience index calculation module, service index calculation module, and weight determination module, respectively, and is used to calculate the index based on the formula: Calculate the comprehensive evaluation index of groundwater, where α+β+γ=1; The grading and output module is connected to the comprehensive evaluation index calculation module. It is used to grade groundwater resources according to the comprehensive evaluation index and output the evaluation results. The visualization analysis module is connected to the health index calculation module, resilience index calculation module, service index calculation module, and comprehensive evaluation index calculation module, respectively. It is used to generate a 3D radar chart for feature analysis, calculate the balance index, and identify the strengths and weaknesses. The sensitivity analysis module is connected to both the data acquisition and preprocessing module and the comprehensive evaluation index calculation module. It is used for local and global sensitivity analysis. The global sensitivity analysis employs the Sobol' method to calculate the first-order exponent S. i And the total effect index S Ti .

[0067] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of this invention is defined by the appended claims and their equivalents.

Claims

1. A comprehensive evaluation method for groundwater resources based on data analysis, characterized in that, Includes the following steps: S1: Collect relevant data on groundwater resources in the area to be evaluated, including water quality monitoring data, water quantity monitoring data, hydrogeological parameters, meteorological data, land use data, socio-economic data, and ecological data, and perform quality control and normalization preprocessing on the collected data; S2: Based on the collected multi-source data, construct a groundwater resource evaluation index system, wherein the evaluation index includes at least the health index H(x), the resilience index R(x), and the ecosystem service index S(x); S3: The weighting coefficients α, β, and γ of the health index H(x), the resilience index R(x), and the ecosystem service index S(x) are determined by using a game theory-based combination weighting method. S4: Based on the determined weighting coefficients, use the formula Calculate the comprehensive evaluation index of groundwater in the area to be evaluated. , where α+β+γ=1; S5: Classify groundwater resources according to the obtained comprehensive evaluation index and output the evaluation results.

2. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 1, characterized in that, In step S2, The health index H(x) is calculated comprehensively based on the groundwater's water quality health, water quantity health, metabolic function, and self-purification capacity, and is defined as follows: H1 is the water quality health index, H2 is the water quantity health index, H3 is the metabolic function index, H4 is the self-purification capacity index, and ω is the water quality health index. i λ is the weight of the i-th index, λ is the pollution decay coefficient, and t is the pollution duration. The resilience index R(x) is calculated based on the groundwater's absorption capacity, recovery capacity, and adaptability, and is defined as follows: ;R a R is the water absorption capacity index. r R is the recovery capacity index. e For adaptability index; The service index S(x) is calculated comprehensively based on the value of water supply services, ecological maintenance, geological environment stability, climate regulation, and cultural services, and is defined as follows: V k For the value of the k-th type of service, V k，max The value of the k-th class is the maximum value in the region, λ. k Weights for service value types.

3. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 2, characterized in that, The water quality health index H1 was estimated using a cloud model. ; Ex represents the expectation of the cloud model, En represents entropy, He represents hyperentropy, and x represents the expected value of the cloud model. std σ represents the water quality standard value, and σ represents the standard deviation. The water health index H2 is defined as follows: Q avail Q represents the amount of exploitable resources. demand For water demand, h represents the actual groundwater level. eco The water level is the ecological warning level; min() is the minimum value operation. The metabolic function index H3 is defined as follows: ; S t S is the current water chemical entropy, S0 is the background entropy value, and S max p is the maximum entropy value. k The relative abundance of the k-th hydrochemical type; The self-cleaning capacity index H4 is estimated using a fuzzy neural network: ; For the fuzzy membership degree of the input variable, is the consequent parameter of the fuzzy rule, and J is the number of fuzzy rules.

4. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 2, characterized in that, The water absorption capacity index R a Defined as: V buffer As an adjustable storage capacity, V total For total storage capacity, C represents the annual change rate of pollutant concentration. C max This is the maximum permissible rate of change; The recovery capacity index R r Defined as: T reference For reference recovery time, T recovery Recovery time is calculated as follows: L is the characteristic length of the aquifer, and D e Where θ is the effective diffusion coefficient, θ is the porosity, C0 is the initial concentration, and C eq To balance the concentration, C t The target concentration; The adaptability index R e Defined as: K c η is the hydraulic connectivity index, M is the aquifer thickness, D is the burial depth, and η is the heterogeneity index.

5. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 2, characterized in that, The water supply service value V1 is defined as follows: Q represents the water supply, and P represents the water volume. shadow For the shadow price, η loss Water loss rate; The ecological maintenance value V2 is defined as follows: B is the base current contribution rate, A wetland For wetland areas that rely on groundwater, U energy The monetary value of energy per unit area is defined as: , τ i Let R be the energy conversion rate of the i-th type of input. i Let i be the input amount for the i-th example; The geological environment stability value V3 is defined as follows: C prevention For full protection cost, k is the curve steepness coefficient, h is the actual water level, and h crit This is the critical water level; The climate regulation value V4 is defined as follows: C ω The volumetric heat capacity of water. T represents the temperature adjustment range, A represents the area of ​​the affected region, and P represents the area of ​​the affected region. carbon For carbon sink prices; The cultural service value V5 is defined as follows: WTP existence WTP is a platform for demonstrating willingness to pay value. bequest Willingness to pay for the bequest value, based on the formula: Let Y be the income level, X be the socio-demographic variable, and a, b, and c be the fitting coefficients.

6. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 1, characterized in that, In step S3, the use of game theory combinatorial weighting method includes: S31: The subjective weights ω of each index are calculated using the triangular fuzzy number hierarchical analysis method. sub : , , l f m f u f ; represents the triangular fuzzy number judgment of the f-th expert, θ f Here, F represents the confidence level, and F represents the number of experts. S32: Calculate the objective weights ω of each index using the improved CRITIC method. obj : r jk Let σ be the correlation coefficient between the j-th and k-th indicators. j Let j be the standard deviation of the j-th indicator; S33: Construct a game theory combinatorial weighting model and solve for the combinatorial coefficients. Specifically: Construct the weight matrix W=[ω sub T ω obj T ], where T is the matrix transpose; Construct the Gram matrix G=W T W; Solve the linear system of equations Gθ=b, where b=W T w T w=(ω) sub +ω obj ) / 2; Sub-step 5.4: Normalize the combination coefficients and calculate the final combination weights: θ * =θ / ∑θ,ω * =θ1 * ω sub +θ2 * ω obj .

7. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 1, characterized in that, In step S5, the classification of groundwater resources based on the obtained comprehensive evaluation index includes: S51: Construct a comprehensive evaluation grading standard to classify the health level of groundwater systems into five levels from high to low, specifically: When the comprehensive evaluation index ranges from [0.85 to 1.00], the system is in a healthy state, with a health level of I. When the comprehensive evaluation index ranges from [0.70, 0.85), the system is in a sub-healthy state, with a health level of II. When the comprehensive evaluation index ranges from [0.55, 0.70), the system is in a risky state, and its health level is Level III. When the comprehensive evaluation index ranges from [0.45, 0.55), the system is in a damaged state, and its health level is IV. When the comprehensive evaluation index ranges from [0.00, 0.45), the system is in a state of collapse and its health level is V. S52: Fuzzy membership functions are used to handle the boundaries of each health level. Specifically: c g Let σ be the standard central value for the g-th health level. g is the width parameter for the g-th health level.

8. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 1, characterized in that, Also includes: Constructing a 3D radar map for feature analysis: ; Calculate the balance index of each of the aforementioned evaluation indicators: ,in, ; Based on the balance index, strengths and weaknesses are identified, and the identification rules are as follows: Advantages: ; Shortcomings: .

9. The comprehensive evaluation method for groundwater resources based on data analysis according to claim 1, characterized in that, Also includes: Sensitivity analysis was conducted, including local sensitivity analysis and global sensitivity analysis; among which: The local sensitivity analysis uses the finite difference method to calculate the local sensitivity coefficient S of the i-th evaluation index on the j-th spatial unit. ij : x ij Let be the value of the i-th evaluation index on the j-th spatial unit; The global sensitivity analysis uses the Sobol' method to calculate the first-order exponent S. i and total effect index S Ti : x ～i Let E be the evaluation index excluding the i-th one, E be the conditional expectation, and Var be the total variance.

10. A comprehensive groundwater resource evaluation system based on data analysis, applied to the method described in any one of claims 1-9, characterized in that, include: The data acquisition and preprocessing module is used to collect multi-source data of the evaluation area, including water quality monitoring data, water quantity monitoring data, hydrogeological parameters, meteorological data, land use data, socio-economic data and ecological data, and to perform quality control and normalization preprocessing on the collected data. The health index calculation module is connected to the data acquisition and preprocessing module and is used to construct the groundwater health index H(x) based on the preprocessed data. The health index H(x) is calculated comprehensively based on four dimensions: water quality health, water quantity health, metabolic function, and self-purification capacity. The resilience index calculation module is connected to the data acquisition and preprocessing module and is used to construct the groundwater resilience index R(x) based on the preprocessed data. The resilience index R(x) is calculated comprehensively based on three dimensions: absorption capacity, recovery capacity and adaptability. The service index calculation module is connected to the data acquisition and preprocessing module and is used to construct the ecosystem service index S(x) based on the preprocessed data. The service index S(x) is calculated comprehensively based on five dimensions: water supply service value, ecological maintenance value, geological environment stability value, climate regulation value, and cultural service value. The weight determination module is connected to the health index calculation module, the resilience index calculation module, and the service index calculation module, respectively, and is used to determine the weight coefficients α, β, and γ of the health index H(x), the resilience index R(x), and the service index S(x) using a game theory combined weighting method. The comprehensive evaluation index calculation module is connected to the health index calculation module, resilience index calculation module, service index calculation module, and weight determination module, respectively, and is used to calculate the index based on the formula: Calculate the comprehensive evaluation index of groundwater, where α+β+γ=1; The grading and output module is connected to the comprehensive evaluation index calculation module and is used to grade groundwater resources according to the comprehensive evaluation index and output the evaluation results. The visualization analysis module is connected to the health index calculation module, resilience index calculation module, service index calculation module and comprehensive evaluation index calculation module respectively. It is used to generate a three-dimensional radar chart for feature analysis, calculate the balance index and identify the strength and weakness dimensions. The sensitivity analysis module is connected to both the data acquisition and preprocessing module and the comprehensive evaluation index calculation module. It is used to perform local and global sensitivity analyses. The global sensitivity analysis employs the Sobol' method to calculate the first-order exponent S. i and total effect index S Ti .

Images