A method for quantitatively evaluating the effect of phytoremediation of heavy metals in karst wetland water body

Abstract

The present application belongs to the technical field of karst wetland ecological restoration and environmental assessment, and particularly relates to a quantitative evaluation method for heavy metal plant remediation effect of karst wetland water. The present application delimits the pollution risk area and importance grade by combining the hydrological flow direction and the location of pollution source through investigation on the karst wetland; sampling units are arranged in each area to collect karst water-sediment-aquatic plant samples; a multi-dimensional evaluation index system is constructed to calculate the index data of water pollution remediation, plant ecological adaptation, karst environment specificity, etc.; a quantitative evaluation model for remediation effect is constructed to calculate the evaluation results of each area, and then the overall remediation score is obtained by weighted summation to determine the remediation grade; a long-term risk prediction model based on deep learning is constructed to determine the long-term risk grade; and finally an evaluation report is output. The present application can realize quantitative evaluation of aquatic plant remediation effect and intelligent prediction of long-term risk, and provides basic support for karst wetland water heavy metal pollution remediation and ecological long-term control.

Description

Technical Field

[0001] This invention belongs to the field of karst wetland ecological restoration and environmental assessment technology, specifically involving a quantitative assessment method for the effect of phytoremediation of heavy metals in karst wetland water. It is applicable to the accurate detection, quantitative assessment, effect classification and judgment and long-term risk prediction of phytoremediation effect of heavy metal pollution in water of various karst landform wetlands. Background Technology

[0002] Karst wetlands are unique ecosystems in the karst regions of southwestern my country, possessing vital ecological functions such as water conservation, water purification, biodiversity maintenance, and carbon sequestration regulation. They are a core component of the ecological security barrier of southwestern China. Affected by human activities such as industrial wastewater discharge, agricultural non-point source pollution, livestock farming, and mining, most karst wetland water bodies currently suffer from varying degrees of heavy metal accumulation pollution. Heavy metals are characterized by persistent toxicity, non-degradability, and strong bioaccumulation, accumulating step-by-step through the water-sediment-aquatic plant-food chain. This not only damages the fragile ecological structure of karst wetlands but also threatens regional drinking water safety and sustainable ecological development.

[0003] Aquatic plant remediation technology has become a mainstream technique for the treatment of heavy metal pollution in karst wetlands due to its advantages such as low cost, eco-friendliness, no secondary pollution, and in-situ remediation capabilities. Through the adsorption, absorption, enrichment, and solidification of heavy metals by emergent, floating, and submerged aquatic plants, as well as the synergistic effects of rhizosphere microorganisms, the content of heavy metals in water and sediments can be effectively reduced, achieving the gradual restoration of wetland ecosystems. Accurate and scientific quantitative assessment of remediation effects is crucial for controlling remediation progress, optimizing remediation plans, and determining the remediation status.

[0004] Current assessment technologies for heavy metal remediation in wetland waters are mostly applicable to ordinary plain wetlands, and are largely based on static water bodies or slow hydrological exchange scenarios. They determine the remediation effect by measuring changes in total heavy metal content and plant growth data in water samples before and after remediation. This assessment method is not specific, has large errors, and is limited in scope. It is not suitable for the unique hydrological scenario of karst wetlands. As a special geographical unit, karst wetlands have fundamentally different hydrogeological conditions from ordinary plain wetlands. Their hydrological flow has multiple paths, and the probability of continuous pollution risk varies in different areas within the wetland. Existing assessment methods cannot comprehensively consider the impact of each area on the overall remediation effect and long-term risks, resulting in poor reference value for the assessment results.

[0005] Therefore, existing technologies urgently need an assessment method that can conduct multi-regional and multi-dimensional quantitative assessments of the unique hydrological scenarios of karst wetlands and effectively evaluate long-term risks. Summary of the Invention

[0006] The purpose of this invention is to provide a quantitative evaluation method for the remediation effect of heavy metal plants on karst wetland water bodies, in order to solve the technical problems raised in the background art.

[0007] This invention provides a quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water, comprising: Before carrying out aquatic plant remediation, a site survey of the target karst wetland water body is conducted to determine the hydrological flow direction and the location of pollution sources; based on the hydrological flow direction and the location of pollution sources, the distribution of pollution risk areas is determined, and the importance level of each pollution risk area is determined in combination with the measured original heavy metal pollution data; Sampling units were deployed in each pollution risk area. After the aquatic plant remediation was completed, the sampling units were used to obtain heavy metal pollution remediation data, plant growth data and karst environment data in each pollution risk area. A multi-dimensional assessment indicator system was constructed, and multi-dimensional assessment indicators were calculated for each pollution risk area. Core indicators for water pollution remediation were determined based on heavy metal pollution remediation data and original heavy metal pollution data. Plant ecological adaptation indicators were determined based on plant growth data. Karst environment-specific indicators were determined based on karst environment data and pre-collected original environmental data. A quantitative assessment model for remediation effectiveness was constructed. Based on core indicators for water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators, the quantitative assessment results of local remediation effectiveness under a multi-dimensional assessment indicator system were determined. According to the importance level of each pollution risk area, the quantitative assessment results of local remediation effectiveness of all pollution risk areas were weighted and summed to obtain the overall quantitative remediation score. The restoration level is determined based on the scoring range to which the overall quantitative restoration score belongs; Based on the site boundary of the target karst wetland water body, the distribution of pollution risk areas, the importance level of each pollution risk area, and the quantitative assessment results of local remediation effects, a remediation assessment feature sequence is constructed. The remediation assessment feature sequence is then input into a pre-trained long-term risk prediction model based on a deep learning model to output the long-term risk level. After outputting the long-term risk level, the final output is a complete assessment report that includes sampling data, indicator analysis results, comprehensive score, level determination results, risk prediction results, and optimization suggestions.

[0008] As a preferred approach, the distribution of pollution risk areas is determined based on hydrological flow direction and pollution source location, and the importance level of each pollution risk area is determined by combining the measured raw heavy metal pollution data, including: Based on the hydrological flow direction and the location of pollution sources, multiple pollution development paths within the wetland boundary are determined, as well as the water flow velocity in each section of the pollution development path; The distribution of pollution risk areas is determined based on the pollution development path and the water flow velocity in each section of the pollution development path; pollution risk areas are divided into high-risk areas, medium-risk areas and low-risk areas. The importance level of each pollution risk area is determined based on the distribution of pollution risk areas and the measured raw heavy metal pollution data. The importance level includes high importance, medium importance, slight importance, and low importance. The distribution of pollution risk areas is determined based on the pollution development path and the water flow velocity in each segment of the pollution development path, including: Areas located at the intersection of multiple pollution development paths and with water flow velocity less than the preset value are classified as high-risk areas due to the superposition of multiple heavy metal pollution paths, weak water flow pollution dispersion capacity, and high risk of heavy metal pollution deposition. For areas located at the intersection of multiple pollution development paths and with water flow velocity greater than the preset value, there is superposition of heavy metal pollution from multiple paths, but the water flow has a strong ability to disperse pollution, and the risk of heavy metal pollution deposition is moderate, so these areas are classified as medium-risk areas. Areas located on a single pollution development path and with water flow velocity less than the preset value will not experience the superposition of heavy metal pollution from multiple paths, but the water flow's ability to disperse pollution is weak, and these areas are classified as medium-risk areas. Areas located on a single pollution development path and with water flow velocity greater than a preset value are classified as low-risk areas because they do not experience the superposition of heavy metal pollution from multiple paths and have a strong ability to disperse pollution through water flow.

[0009] Preferably, the importance level of each pollution risk area is determined based on the distribution of pollution risk areas and the measured raw heavy metal pollution data, including: For high-risk areas, if the original heavy metal pollution data shows a severe exceedance of the standard, the importance level is marked as high importance; if the original heavy metal pollution data shows a slight exceedance of the standard, the importance level is marked as medium importance; and if the original heavy metal pollution data does not exceed the standard, the importance level is marked as low importance. For medium-risk areas, if the original heavy metal pollution data shows a severe exceedance of the standard, the importance level is marked as medium importance; if the original heavy metal pollution data shows a slight exceedance of the standard, the importance level is marked as low importance; if the original heavy metal pollution data does not exceed the standard, the importance level is marked as low importance. For low-risk areas, if the original heavy metal pollution data shows a severe exceedance of the standard, the importance level is marked as low importance; if the original heavy metal pollution data shows a slight exceedance of the standard or no exceedance, the importance level is marked as low importance.

[0010] As a preferred method, the methods for obtaining the measured raw heavy metal pollution data include: The pollution risk area is evenly divided into a preset number of grid cells. Heavy metal pollution data is collected in each grid cell. The maximum value among all sampled data in the pollution risk area is taken as the measured raw heavy metal pollution data.

[0011] Preferably, when arranging sampling units in each pollution risk area, the higher the importance level of the pollution risk area, the more grid units the pollution risk area is evenly divided into, the larger the area of ​​each grid unit is equal, and each grid unit needs to be arranged with sampling units. The final calculated heavy metal pollution remediation data and plant growth data are the average of the sampling data of all grid units.

[0012] Preferably, the sampling unit includes a water sampling unit, a sediment sampling unit, and an aquatic plant sampling unit; The core indicators for water pollution remediation include at least one of the following: water heavy metal removal rate, sediment heavy metal solidification rate, heavy metal bioaccumulation coefficient, heavy metal translocation coefficient, and pollution load reduction index. The plant ecological adaptation index data shall include at least one of the following: aquatic plant survival rate, biomass growth rate, community coverage, and root activity; The specific indicators for karst environments include at least one of the following: pH fluctuation range of water body, calcium ion concentration change rate, and proportion of heavy metals in karst bound state.

[0013] As a preferred method, the construction of a quantitative assessment model for repair effectiveness includes: By combining the analytic hierarchy process (AHP) with the entropy weight method, a combined weighting of three levels of evaluation indicators—core indicators for water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators—is employed to avoid subjective weight bias. This approach constructs a quantitative evaluation model for local remediation effectiveness and calculates the quantitative evaluation result S of the local remediation effect. Among these indicators, the core indicators for water pollution remediation account for 60%-65% of the weight, the plant ecological adaptation indicators account for 20%-25%, and the karst environment-specific indicators account for 10%-15%. The calculation formula for the quantitative evaluation model of local repair effect is as follows: The expression for the quantitative evaluation result S of the local repair effect is: S = Σ(Wi × Xi'); Where Wi is the combined weight of each indicator, and Xi' is the standardized score of each indicator; Indicator standardization processing methods: The standardized formula for positive indicators is: Xi'=(Xi-Xmin) / (Xmax-Xmin) The standardized formula for negative indicators is: Xi'=(Xmax-Xi) / (Xmax-Xmin) In the formula: Xi is the measured value of the index; Xmax and Xmin are the maximum and minimum values ​​of the index in the entire domain. After standardization, the value range of all indexes is unified to [0,1]. Based on the importance level of each pollution risk area and the quantitative assessment results of the local remediation effects of all pollution risk areas, a quantitative assessment model for the overall remediation effect is constructed, and the overall quantitative remediation score Sc is obtained by weighted summation. The calculation formula for the overall restoration effect quantitative assessment model is as follows: The expression for the overall quantitative repair score Sc is: Sc=Σ(w a,i ×S i ); Where Sc represents the overall quantitative repair score, w a,i S represents the weighting coefficient of the i-th pollution risk area; i This represents the quantitative assessment result of the local remediation effect in the i-th pollution risk area.

[0014] As a preferred option, the repair levels include Level 1 complete repair, Level 2 effective repair, Level 3 preliminary repair, and Level 4 ineffective repair; Level 1 complete restoration Sc≥0.85: The heavy metal content in water and sediment meets the standards, the vegetation enrichment is stable, the karst ecological environment is normal, and there is no risk of heavy metal rebound; Secondary effective remediation 0.70≤Sc<0.85: Heavy metal pollution is significantly reduced, plant growth is good, remediation effect is stable, and only slight pollution residues remain in some areas; Level 3 preliminary remediation 0.50≤Sc<0.70: Heavy metal pollution has been controlled to some extent, but the remediation stability is insufficient, and the plant configuration and remediation process need to be optimized. Level IV remediation ineffective (Sc < 0.50): poor heavy metal removal effect, poor plant compatibility, no improvement or even aggravation of pollution.

[0015] As a preferred approach, methods for constructing long-term risk prediction models include: Multiple wetland water body restoration cases were obtained, and restoration assessment feature sequences were extracted from each case and labeled with the true label Y of the long-term risk level. The labeled long-term risk level was divided according to the actual measurement, and the long-term risk level included Level 1 long-term effectiveness, Level 2 medium-term effectiveness, and Level 3 short-term effectiveness. Using the site boundary as a reference, add spatial location labels to each pollution risk zone according to the distribution of pollution risk zones; Based on the spatial location labels, importance levels, and quantitative assessment results of local remediation effects, a regional vector x is constructed for each pollution risk area. i ; Arrange the region vectors of all pollution risk areas in descending order of importance to form a remediation assessment feature sequence X, where X = {x1, x2, ..., x}. i}; The repair assessment feature sequence X is input into the encoder of the Transformer architecture to generate a contextual representation of the sequence. The decoder of the Transformer architecture generates a prediction result Y of the long-term risk level based on the true label of the long-term risk level and the contextual representation of the repair assessment feature sequence. output ; The Transformer architecture is trained using a training set constructed from multiple wetland water body restoration cases until the loss function converges to obtain a long-term risk prediction model.

[0016] As a preferred option, the evaluation report includes: The sampling data includes heavy metal pollution remediation data, plant growth data, and karst environment data for each pollution risk area; The results of the indicator analysis include core indicator data for water pollution remediation, plant ecological adaptation indicator data, and karst environment-specific indicator data for each pollution risk area; The comprehensive score includes a quantitative assessment of the local remediation effects in each pollution risk area and an overall quantitative remediation score for the entire target karst wetland water body. The grading results include the grading level determined based on the overall quantitative grading score; The risk prediction results include the long-term risk level output by the long-term risk prediction model; The optimization recommendations include optimized matching schemes for each pollution risk area when the long-term risks are significant; The optimized matching scheme is matched based on an optimized scheme library constructed from historical repair cases. The matching method includes: For pollution risk areas where the quantitative assessment of local remediation effects exceeds a preset threshold, feature vectors are constructed based on the core indicators of water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators for the pollution risk area. The similarity between the feature vectors of the pollution risk area and the feature vectors of historical successful cases in the optimization scheme library is calculated, and the optimization scheme adopted by the case with the highest similarity is selected for output. The output results include at least the plant variety optimization scheme, the community structure adjustment scheme, and the zonal enhanced remediation scheme.

[0017] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: Before carrying out aquatic plant restoration, this invention conducts a site survey of the target karst wetland water body to determine the hydrological flow direction and the location of pollution sources; based on the hydrological flow direction and the location of pollution sources, it determines the distribution of pollution risk areas and, combined with the measured original heavy metal pollution data, determines the importance level of each pollution risk area. This invention sets up an importance classification mechanism for the hydrological flow direction and pollution source location of the target karst wetland water body to comprehensively consider the degree of influence of different areas on the assessment of restoration effect and long-term risk assessment. Subsequently, sampling units were deployed in each pollution risk area. After aquatic plant remediation was completed, these units were used to acquire heavy metal pollution remediation data, plant growth data, and karst environment data for each pollution risk area. A multi-dimensional assessment index system was constructed. For each pollution risk area, core water pollution remediation index data were determined based on heavy metal pollution remediation data and original heavy metal pollution data. Plant ecological adaptation index data were determined based on plant growth data. Karst environment-specific index data were determined based on karst environment data and pre-collected original environmental data. A quantitative assessment model for remediation effectiveness was constructed to determine the quantitative assessment results of local remediation effectiveness under the multi-dimensional assessment index system. Based on the importance level of each pollution risk area, the quantitative assessment results of local remediation effectiveness in all pollution risk areas were weighted and summed to obtain an overall quantitative remediation score. The remediation level was determined based on the score range to which the overall quantitative remediation score belonged. The specific weight coefficient values ​​can be set according to actual needs. This invention also constructs a remediation assessment feature sequence based on the site boundary of the target karst wetland water body, the distribution of pollution risk areas, the importance level of each pollution risk area, and the quantitative assessment results of local remediation effects. The remediation assessment feature sequence is input into a pre-trained long-term risk prediction model based on a deep learning model to output a long-term risk level. After outputting the long-term risk level, a complete assessment report is finally output, including sampling data, indicator analysis results, comprehensive score, level determination results, risk prediction results, and optimization suggestions.

[0018] This invention is precisely adapted to the unique ecological environment of karst wetland water bodies, comprehensively quantifying the short-term effects of aquatic plant remediation and conducting long-term stability assessments. It accurately identifies remediation shortcomings and pollution rebound risks, providing precise data support for optimizing remediation technologies. The assessment process is simple to operate, relying on routine water quality, plant, and sediment testing, resulting in low cost and high efficiency. This invention can accurately determine the remediation effect status, serving as a valid reference and avoiding the waste of human, material, and financial resources caused by over-remediation or ineffective remediation, significantly reducing the operation and maintenance costs of karst wetland heavy metal remediation projects. The assessment system of this invention is standardized and highly versatile, adaptable to different types and pollution levels of karst wetland water bodies in Southwest China. It can be widely applied in scenarios such as acceptance of heavy metal pollution remediation projects in karst wetland water bodies, daily operation and maintenance monitoring, and remediation technology screening, possessing strong engineering promotion value and industry application prospects. Attached Figure Description

[0019] Figure 1 A flowchart illustrating a quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water. Figure 2 A three-tiered evaluation index system architecture diagram for the remediation effect of phytoremediation of heavy metals in karst wetland waters, which is a quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland waters. Detailed Implementation

[0020] It should be noted that: 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 of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0021] The term "and / or" simply describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0022] Combination Figure 1 This embodiment introduces a quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water, which includes: Step S1: Before carrying out aquatic plant restoration, conduct a site survey of the target karst wetland water body to determine the hydrological flow direction and the location of pollution sources; determine the distribution of pollution risk areas and the importance level of each pollution risk area based on the hydrological flow direction and the location of pollution sources.

[0023] This embodiment determines multiple pollution development paths within the wetland boundary based on hydrological flow direction and pollution source location, as well as the water flow velocity in each segment of the pollution development path; it determines the distribution of pollution risk areas based on the pollution development paths and the water flow velocity in each segment of the pollution development path; the pollution risk areas are divided into high-risk areas, medium-risk areas, and low-risk areas; and it determines the importance level of each pollution risk area based on the distribution of pollution risk areas and the measured original heavy metal pollution data, wherein the importance level includes high importance, medium importance, slight importance, and low importance.

[0024] The distribution of pollution risk areas is determined based on the pollution development path and the water flow velocity in each segment of the path, including: Areas located at the intersection of multiple pollution development paths with water flow velocities less than a preset value are classified as high-risk areas due to the presence of overlapping heavy metal pollution from multiple paths and weak water flow dispersion capabilities, resulting in a high risk of heavy metal pollution deposition; areas located at the intersection of multiple pollution development paths with water flow velocities greater than a preset value are classified as medium-risk areas due to the presence of overlapping heavy metal pollution from multiple paths but strong water flow dispersion capabilities, resulting in a moderate risk of heavy metal pollution deposition; areas located on a single pollution development path with water flow velocities less than a preset value are classified as medium-risk areas due to the absence of overlapping heavy metal pollution from multiple paths but weak water flow dispersion capabilities; and areas located on a single pollution development path with water flow velocities greater than a preset value are classified as low-risk areas due to the absence of overlapping heavy metal pollution from multiple paths and strong water flow dispersion capabilities. This invention first considers the pollution development path and water flow velocity to initially determine the risk level of pollution risk areas, and then further evaluates the actual importance level based on measured data.

[0025] The importance level of each pollution risk area is determined based on the distribution of pollution risk areas and the measured raw heavy metal pollution data, including: for high-risk areas, if the raw heavy metal pollution data shows severe exceedance, the importance level is marked as high; if the raw heavy metal pollution data shows slight exceedance, the importance level is marked as medium; if the raw heavy metal pollution data does not exceed the standard, the importance level is marked as low. For medium-risk areas, if the raw heavy metal pollution data shows severe exceedance, the importance level is marked as medium; if the raw heavy metal pollution data shows slight exceedance, the importance level is marked as low; if the raw heavy metal pollution data does not exceed the standard, the importance level is marked as low. For low-risk areas, if the raw heavy metal pollution data shows severe exceedance, the importance level is marked as low; if the raw heavy metal pollution data shows slight exceedance or does not exceed the standard, the importance level is marked as low. This invention determines the importance level of pollution risk areas, thereby ensuring that high-importance pollution risk areas that are particularly critical to the remediation effect will be given priority consideration in the subsequent quantitative assessment of the effect. The term "severe exceedance" means that the data exceeds the normal value by more than 50%, and "mild exceedance" means that the data exceeds the normal value but the exceedance is less than 50%.

[0026] The method for obtaining the measured raw heavy metal pollution data includes: uniformly dividing the pollution risk area into a predetermined number of grid cells, collecting heavy metal pollution data (concentration data of at least one heavy metal) in each grid cell, and using the maximum value among all sampled data within the pollution risk area as the measured raw heavy metal pollution data. This invention uses the maximum sampled value as the measured heavy metal pollution data to avoid misjudging the importance level of severely polluted areas as low.

[0027] Step S2: Deploy sampling units in each pollution risk area. After the aquatic plant remediation is completed, use the sampling units to obtain heavy metal pollution remediation data, plant growth data, and karst environment data for each pollution risk area.

[0028] When deploying sampling units in each pollution risk area, the higher the importance level of the pollution risk area, the more grid units the pollution risk area is evenly divided into. Each grid unit has an equal area, and each grid unit needs to be equipped with a sampling unit. The final calculated heavy metal pollution remediation data and plant growth data are the average of the sampling data of all grid units.

[0029] The sampling unit includes a water sampling unit, a sediment sampling unit, and an aquatic plant sampling unit. This invention simultaneously deploys these units at each sampling point to ensure fixed sampling locations and consistent timing, avoiding interference from hydrological fluctuations and geological differences in karst wetlands. The sampling time of this invention needs to be selected during the stable growth period of aquatic plants, avoiding periods of severe hydrological fluctuations such as heavy rain and flood seasons. Based on the deployed sampling points, this invention simultaneously collects three core samples: water, sediment, and aquatic plants. The samples are pre-processed and subjected to index testing, with all indicators using national standard testing methods. Water samples are collected from the surface 0-50cm depth of wetland water, sealed and protected from light, pre-treated to remove suspended particulate matter, and the content of dissolved heavy metals and total heavy metals in the water is detected. Simultaneously, the pH value, calcium ion concentration, bicarbonate concentration, turbidity, and other characteristic karst water quality indicators are also detected. Sediment samples were collected from the top 0-20cm layer of wetland sediments. Gravel, plant and animal remains, and other impurities were removed. The samples were then air-dried, ground, and sieved. The total and available heavy metal content of the sediments was analyzed, with the available content specifically tailored to the heavy metal activation characteristics of karst alkaline environments. Aquatic plant samples were collected from dominant aquatic plants used for remediation. These samples were divided into three parts: roots, stems, and leaves. The roots and stems were washed to remove surface water and sediment, dried, and pulverized. The heavy metal content of each part was analyzed, and enrichment amounts, translocation coefficients, and concentration coefficients were calculated.

[0030] Step S3: Construct a multi-dimensional assessment indicator system. Calculate multi-dimensional assessment indicators for each pollution risk area. Determine core indicators for water pollution remediation based on heavy metal pollution remediation data and original heavy metal pollution data. Determine plant ecological adaptation indicators based on plant growth data. Determine karst environment-specific indicators (such as...) based on karst environment data and pre-collected original environmental data. Figure 2 (As shown).

[0031] The core indicators for water pollution remediation include at least one of the following: water heavy metal removal rate, sediment heavy metal solidification rate, heavy metal bioaccumulation coefficient, heavy metal translocation coefficient, and pollution load reduction index. The expressions for the heavy metal removal rate in water, the heavy metal solidification rate in sediments, the heavy metal bioaccumulation coefficient, the heavy metal translocation coefficient, and the pollution load reduction index are as follows: Heavy metal removal rate in water = (C0 - C) t ) / C0; C0 and C t The values ​​represent the heavy metal concentrations in the water before and after aquatic plant remediation, respectively. Heavy metal solidification rate of sediment = 1 L / T, where L represents the amount of heavy metals leached after treatment and T represents the total amount of heavy metals in the sediment; Heavy metal bioaccumulation coefficient = C organism / C environment C organism C represents the concentration of heavy metals in an organism. environmen t represents the concentration of heavy metals in the environmental medium; Heavy metal transport coefficient = C shoot / C root C shoot C represents the concentration of heavy metals in the above-ground parts (stems and leaves) of a plant. root Indicates the concentration of heavy metals in plant roots; Pollution load reduction index = (L0 - L) t ) / L0, where L0 and L t The heavy metal pollution load in the water before and after aquatic plant remediation.

[0032] The plant ecological adaptation index data shall include at least one of the following: aquatic plant survival rate, biomass growth rate, community coverage, and root activity, which are used to determine the growth stability and ecological adaptability of the restoration plants. The expressions for aquatic plant survival rate, biomass growth rate, community coverage, and root activity are as follows: Aquatic plant survival rate = A t / A0, A0 and A t These represent the total number of plants planted at the initial stage of restoration and the number of plants that survived after restoration, respectively. Biomass growth rate = (w t - w0) / w0, w t w0 and w0 represent the average dry weight of aquatic plants after restoration and the average dry weight of aquatic plants before restoration, respectively.

[0033] Community coverage = S cover / S all S cover and S all These represent the vertical projection area of ​​the restored aquatic plant canopy and the total water area of ​​the karst wetland, respectively. Root activity = C / (m × t), where C represents the standard TTC reduction amount, m represents the fresh weight of the root tip, and t represents the reaction time.

[0034] The specific karst environment indicators include at least one of the following: pH fluctuation range of water body, calcium ion concentration change rate, and proportion of heavy metals in karst-bound state, which are used to correct the interference and deviation of the special karst environment on the remediation effect of heavy metals.

[0035] The formulas for calculating the pH fluctuation range, calcium ion concentration change rate, and proportion of heavy metals in karst form are as follows: pH fluctuation range of water body = pH max - pH min pH max and pH min These represent the maximum and minimum pH values ​​of the water at all sampling points in the pollution risk area during the monitoring period, respectively. Rate of change in calcium ion concentration = (C before - C after ) / C before C before and C after These represent the calcium ion concentrations in the water before and after aquatic plant remediation, respectively. The proportion of heavy metals in karst-bound state = P carb = M carb / M total ×100% where, M carb M represents the content of carbonate-bound heavy metals in sediments within this pollution risk area. total This represents the total amount of the heavy metal in the sediment.

[0036] Step S4: Construct a quantitative assessment model for remediation effectiveness. Based on the core indicator data for water pollution remediation, plant ecological adaptation indicator data, and karst environment-specific indicator data, determine the quantitative assessment results of local remediation effectiveness under a multi-dimensional assessment indicator system. Based on the importance level of each pollution risk area, weighted summation of the quantitative assessment results of local remediation effectiveness for all pollution risk areas is performed to obtain the overall quantitative remediation score.

[0037] By combining the analytic hierarchy process (AHP) with the entropy weight method, a combined weighting of three levels of evaluation indicators—core indicators for water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators—is implemented to avoid subjective weight bias. This constructs a quantitative evaluation model for local remediation effects and calculates the quantitative evaluation result S of the local remediation effect. As an exemplary embodiment, the core indicators for water pollution remediation account for 60%-65% of the weight, the plant ecological adaptation indicators account for 20%-25% of the weight, and the karst environment-specific indicators account for 10%-15% of the weight.

[0038] The calculation formula for the quantitative evaluation model of local repair effect is as follows: The expression for the quantitative evaluation result S of the local repair effect is: S = Σ(Wi × Xi'); Where Wi is the combined weight of each indicator, and Xi' is the standardized score of each indicator.

[0039] Standardization methods for indicators (eliminating dimensions): The standardized formula for the positive indicator (the larger the value, the better the repair effect) is: Xi'=(Xi-Xmin) / (Xmax-Xmin) The standardized formula for negative indicators (the larger the value, the worse the repair effect) is: Xi'=(Xmax-Xi) / (Xmax-Xmin) In the formula: Xi is the measured value of the index; Xmax and Xmin are the maximum and minimum values ​​of the index in the whole domain. After standardization, the value range of all indexes is unified to [0,1].

[0040] All indicators were standardized based on the ecological threshold of karst wetlands to eliminate differences in the dimensions of indicators across different dimensions. At the same time, the basic scores were corrected for deviations using karst environment-specific indicators to obtain accurate quantitative assessment results of local restoration effects.

[0041] A quantitative assessment model for overall remediation effectiveness is constructed based on the importance level of each pollution risk area and the quantitative assessment results of the local remediation effects of all pollution risk areas. The overall quantitative remediation score Sc is obtained by weighted summation. The weight coefficient of the pollution risk area is positively correlated with the importance level. The local remediation effect of high-importance pollution risk areas is particularly important for the assessment of the overall remediation effect.

[0042] The calculation formula for the overall restoration effect quantitative assessment model is as follows: The expression for the overall quantitative repair score Sc is: Sc=Σ(w a,i ×S i ); Where Sc represents the overall quantitative repair score, w a,i S represents the weighting coefficient of the i-th pollution risk area; i This represents the quantitative assessment result of the local remediation effect in the i-th pollution risk area.

[0043] Step S5: Determine the repair level based on the scoring range to which the overall quantitative repair score belongs.

[0044] The repair levels include Level 1: Complete Repair; Level 2: Effective Repair; Level 3: Preliminary Repair; and Level 4: Ineffective Repair.

[0045] Level 1 Complete Remediation (Sc≥0.85): Heavy metal content in water and sediments meets standards, plant enrichment is stable, karst ecological environment is normal, and there is no risk of heavy metal rebound. Level 2 Effective Remediation (0.70≤Sc<0.85): Heavy metal pollution is significantly reduced, plant growth is good, remediation effect is stable, and only slight pollution residue exists in some areas. Level 3 Preliminary Remediation (0.50≤Sc<0.70): Heavy metal pollution is under certain control, but remediation stability is insufficient, and plant configuration and remediation process need to be optimized. Level 4 Ineffective Remediation (Sc<0.50): Poor heavy metal removal effect, poor plant adaptability, and pollution is not improved or even aggravated.

[0046] Step S6: Construct a remediation assessment feature sequence based on the site boundary of the target karst wetland water body, the distribution of pollution risk areas, the importance level of each pollution risk area, and the quantitative assessment results of local remediation effects. Input the remediation assessment feature sequence into a pre-trained long-term risk prediction model based on a deep learning model to output the long-term risk level.

[0047] The long-term risk prediction model is constructed using the Transformer architecture, and the construction method of the long-term risk prediction model includes: Multiple wetland water body restoration cases were obtained. For each case, restoration assessment feature sequences were extracted and labeled with a true label Y for the long-term risk level. The long-term risk level was classified based on actual measurements, including Level 1 (long-term effective), Level 2 (medium-term effective), and Level 3 (short-term effective). The specific method for classifying the long-term risk level based on actual measurements included: after aquatic plant restoration, continuous monitoring was conducted to obtain the trend of restoration effect over time. If, within a preset monitoring period (e.g., within two years after aquatic plant restoration), the heavy metal pollution concentration remained consistently below the restoration target value, and the plant ecological adaptation index did not show a significant decline, the long-term risk level was determined to be Level 1 (long-term effective). If, during the monitoring period, the heavy metal concentration slightly increased but remained below the level before aquatic plant restoration, and the restoration effect showed a slow downward trend, the long-term risk level was determined to be Level 2 (medium-term effective). If the target was met in the short term after restoration, but the heavy metal concentration rapidly increased during the monitoring period, and the restoration effect of aquatic plants significantly declined or failed, the long-term risk level was determined to be Level 3 (short-term effective).

[0048] The method for extracting the remediation assessment feature sequence includes: using the site boundary as a reference, adding spatial location labels to each pollution risk area according to the distribution of pollution risk areas, and constructing a region vector x for each pollution risk area based on the spatial location labels, importance level, and quantitative assessment results of local remediation effects. iIn this system, spatial location labels are represented by coordinates, and importance levels are represented by category IDs, such as 1 / 2 / 3 / 4 representing high importance, medium importance, low importance, and low importance, respectively. The region vectors of all pollution risk areas are arranged in descending order of importance to form a remediation assessment feature sequence X, where X = {x1, x2, ..., x...}. i The repaired evaluation feature sequence X is input into the encoder of the Transformer architecture to generate a contextual representation of the sequence. The Transformer architecture is an existing deep learning model that excels at processing sequence data. The decoder of the Transformer architecture generates a prediction result of the long-term risk level based on the true label of the long-term risk level and the contextual representation of the repaired evaluation feature sequence. The expression of the long-term risk prediction model is as follows: Y output =Decoder(Y, H) encoder ) Among them, Y output Y represents the predicted result of the long-term risk level, and H represents the true label of the long-term risk level. encoder This represents the contextual representation of the repair evaluation feature sequence.

[0049] The long-term risk prediction model of this invention will output the long-term risk level based on the current stage of remediation results, and predict the risk of reverse release of heavy metals and pollution rebound in the next 1-2 years.

[0050] Step S7: Output a complete evaluation report including sampling data, indicator analysis results, comprehensive score, level determination results, risk prediction results, and optimization suggestions.

[0051] The sampling data includes heavy metal pollution remediation data, plant growth data, and karst environment data for each pollution risk area; the indicator analysis results include core water pollution remediation indicator data, plant ecological adaptation indicator data, and karst environment-specific indicator data for each pollution risk area; the comprehensive score includes quantitative assessment results of local remediation effects in each pollution risk area and overall quantitative remediation score for the entire target karst wetland water body; the grade determination results include the remediation grade determined based on the overall quantitative remediation score; the risk prediction results include the long-term risk grade output by the long-term risk prediction model; and the optimization suggestions include optimized matching schemes for each pollution risk area when the long-term risk is high. These optimized matching schemes are based on an optimization scheme library constructed from historical remediation cases. The matching process involves the following methods: For each pollution risk area, a feature vector is constructed based on the core indicators of water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators for that pollution risk area. The feature vector of the pollution risk area is then compared with the feature vectors of historical successful cases in the optimization scheme library (for example, cosine similarity can be used). The optimization scheme adopted by the case with the highest similarity is selected and output. The output results should include at least the following: plant variety optimization scheme (e.g., replacing with varieties with more developed root systems and longer-lasting effects), community structure adjustment scheme (e.g., increasing the vertical layers of emergent plants, floating-leaved plants, and submerged plants to improve community coverage and stability), and zonal reinforcement remediation scheme (e.g., increasing planting density).

[0052] This invention is precisely adapted to the unique ecological environment of karst wetlands, effectively solving the problems of weak targeting, large errors, and single-dimensionality in traditional assessment methods. It can comprehensively quantify the short-term effects and long-term stability of aquatic plant remediation, accurately identify remediation shortcomings and pollution rebound risks, and provide precise data support for the optimization of remediation technologies. The assessment process of this invention is standardized and easy to operate, relying on routine water quality, plant, and sediment testing, resulting in low assessment costs and high efficiency. This invention can accurately determine the remediation effect status, thus serving as a valid reference and avoiding the waste of human, material, and financial resources caused by over-remediation or ineffective remediation, significantly reducing the operation and maintenance costs of karst wetland heavy metal remediation projects. The assessment system of this invention is standardized and highly versatile, adaptable to different types and pollution levels of karst wetland water bodies in Southwest China. It can be widely applied to scenarios such as acceptance of karst wetland heavy metal pollution remediation projects, daily operation and maintenance monitoring, and remediation technology screening, possessing strong engineering promotion value and industry application prospects.

[0053] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention 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.

[0054] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. 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 illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0055] 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.

[0056] 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.

[0057] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water, characterized in that, include: Before carrying out aquatic plant restoration, a site survey of the target karst wetland water body should be conducted to determine the hydrological flow direction and the location of pollution sources. The distribution of pollution risk areas is determined based on hydrological flow direction and pollution source location, and the importance level of each pollution risk area is determined by combining the measured raw heavy metal pollution data. Sampling units were deployed in each pollution risk area. After the aquatic plant remediation was completed, the sampling units were used to obtain heavy metal pollution remediation data, plant growth data and karst environment data in each pollution risk area. A multi-dimensional assessment indicator system was constructed, and multi-dimensional assessment indicators were calculated for each pollution risk area. Core indicators for water pollution remediation were determined based on heavy metal pollution remediation data and original heavy metal pollution data. Plant ecological adaptation indicators were determined based on plant growth data. Karst environment-specific indicators were determined based on karst environment data and pre-collected original environmental data. A quantitative assessment model for remediation effects was constructed, and the quantitative assessment results of local remediation effects under a multi-dimensional assessment indicator system were determined based on core indicator data for water pollution remediation, plant ecological adaptation indicator data, and karst environment-specific indicator data. Based on the importance level of each pollution risk area, the quantitative assessment results of the local remediation effects of all pollution risk areas are weighted and summed to obtain the overall quantitative remediation score. The restoration level is determined based on the scoring range to which the overall quantitative restoration score belongs; Based on the site boundary of the target karst wetland water body, the distribution of pollution risk areas, the importance level of each pollution risk area, and the quantitative assessment results of local remediation effects, a remediation assessment feature sequence is constructed. The remediation assessment feature sequence is then input into a pre-trained long-term risk prediction model based on a deep learning model to output the long-term risk level. After outputting the long-term risk level, the final output is a complete assessment report that includes sampling data, indicator analysis results, comprehensive score, level determination results, risk prediction results, and optimization suggestions.

2. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, The distribution of pollution risk areas is determined based on hydrological flow direction and pollution source location, and the importance level of each pollution risk area is determined by combining the measured raw heavy metal pollution data, including: Based on the hydrological flow direction and the location of pollution sources, multiple pollution development paths within the wetland boundary are determined, as well as the water flow velocity in each section of the pollution development path; The distribution of pollution risk areas is determined based on the pollution development path and the water flow velocity in each section of the pollution development path; pollution risk areas are divided into high-risk areas, medium-risk areas and low-risk areas. The importance level of each pollution risk area is determined based on the distribution of pollution risk areas and the measured raw heavy metal pollution data. The importance level includes high importance, medium importance, slight importance, and low importance. The distribution of pollution risk areas is determined based on the pollution development path and the water flow velocity in each segment of the pollution development path, including: Areas located at the intersection of multiple pollution development paths and with water flow velocity less than the preset value are classified as high-risk areas due to the superposition of multiple heavy metal pollution paths, weak water flow pollution dispersion capacity, and high risk of heavy metal pollution deposition. For areas located at the intersection of multiple pollution development paths and with water flow velocity greater than the preset value, there is superposition of heavy metal pollution from multiple paths, but the water flow has a strong ability to disperse pollution, and the risk of heavy metal pollution deposition is moderate, so these areas are classified as medium-risk areas. Areas located on a single pollution development path and with water flow velocity less than the preset value will not experience the superposition of heavy metal pollution from multiple paths, but the water flow's ability to disperse pollution is weak, and these areas are classified as medium-risk areas. Areas located on a single pollution development path and with water flow velocity greater than a preset value are classified as low-risk areas because they do not experience the superposition of heavy metal pollution from multiple paths and have a strong ability to disperse pollution through water flow.

3. The method for quantitatively evaluating the effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 2, characterized in that, The importance level of each pollution risk area is determined based on the distribution of pollution risk areas and the measured raw heavy metal pollution data, including: For high-risk areas, if the original heavy metal pollution data shows a severe exceedance of the standard, the importance level is marked as high importance; if the original heavy metal pollution data shows a slight exceedance of the standard, the importance level is marked as medium importance; and if the original heavy metal pollution data does not exceed the standard, the importance level is marked as low importance. For medium-risk areas, if the original heavy metal pollution data shows a severe exceedance of the standard, the importance level is marked as medium importance; if the original heavy metal pollution data shows a slight exceedance of the standard, the importance level is marked as low importance; if the original heavy metal pollution data does not exceed the standard, the importance level is marked as low importance. For low-risk areas, if the original heavy metal pollution data shows a severe exceedance of the standard, the importance level is marked as low importance; if the original heavy metal pollution data shows a slight exceedance of the standard or no exceedance, the importance level is marked as low importance.

4. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, The methods for obtaining raw heavy metal pollution data from actual measurements include: The pollution risk area is evenly divided into a preset number of grid cells. Heavy metal pollution data is collected in each grid cell. The maximum value among all sampled data in the pollution risk area is taken as the measured raw heavy metal pollution data.

5. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, When deploying sampling units in each pollution risk area, the higher the importance level of the pollution risk area, the more grid units the pollution risk area is evenly divided into. Each grid unit has an equal area, and each grid unit needs to be equipped with a sampling unit. The final calculated heavy metal pollution remediation data and plant growth data are the average of the sampling data of all grid units.

6. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, The sampling unit includes a water sampling unit, a sediment sampling unit, and an aquatic plant sampling unit; The core indicators for water pollution remediation include at least one of the following: water heavy metal removal rate, sediment heavy metal solidification rate, heavy metal bioaccumulation coefficient, heavy metal translocation coefficient, and pollution load reduction index. The plant ecological adaptation index data shall include at least one of the following: aquatic plant survival rate, biomass growth rate, community coverage, and root activity; The specific indicators for karst environments include at least one of the following: pH fluctuation range of water body, calcium ion concentration change rate, and proportion of heavy metals in karst bound state.

7. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, Methods for constructing quantitative assessment models of repair effectiveness include: By combining the analytic hierarchy process (AHP) with the entropy weight method, a combined weighting of three levels of evaluation indicators—core indicators for water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators—is employed to avoid subjective weight bias. This approach constructs a quantitative evaluation model for local remediation effects, yielding the quantitative evaluation results. Among these, the core indicators for water pollution remediation account for 60%-65% of the weight, the plant ecological adaptation indicators account for 20%-25%, and the karst environment-specific indicators account for 10%-15%. The calculation formula for the quantitative evaluation model of local repair effect is as follows: The expression for the quantitative evaluation result S of the local repair effect is: S = Σ(Wi × Xi'); Where Wi is the combined weight of each indicator, and Xi' is the standardized score of each indicator; Indicator standardization processing methods: The standardized formula for positive indicators is: Xi'=(Xi-Xmin) / (Xmax-Xmin) The standardized formula for negative indicators is: Xi'=(Xmax-Xi) / (Xmax-Xmin) In the formula: Xi is the measured value of the index; Xmax and Xmin are the maximum and minimum values ​​of the index in the entire domain. After standardization, the value range of all indexes is unified to [0,1]. Based on the importance level of each pollution risk area and the quantitative assessment results of the local remediation effects of all pollution risk areas, a quantitative assessment model for the overall remediation effect is constructed, and the overall quantitative remediation score Sc is obtained by weighted summation. The calculation formula for the overall restoration effect quantitative assessment model is as follows: The expression for the overall quantitative repair score Sc is: Sc=Σ(w a,i ×S i ); Where Sc represents the overall quantitative repair score, w a,i S represents the weighting coefficient of the i-th pollution risk area; i This represents the quantitative assessment result of the local remediation effect in the i-th pollution risk area.

8. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, Repair levels are categorized as follows: Level 1: Complete Repair; Level 2: Effective Repair; Level 3: Preliminary Repair; Level 4: Ineffective Repair. Level 1 complete restoration Sc≥0.85: The heavy metal content in water and sediment meets the standards, the vegetation enrichment is stable, the karst ecological environment is normal, and there is no risk of heavy metal rebound; Secondary effective remediation 0.70≤Sc<0.85: Heavy metal pollution is significantly reduced, plant growth is good, remediation effect is stable, and only slight pollution residues remain in some areas; Level 3 preliminary remediation 0.50≤Sc<0.70: Heavy metal pollution has been controlled to some extent, but the remediation stability is insufficient, and the plant configuration and remediation process need to be optimized. Level IV remediation ineffective (Sc < 0.50): poor heavy metal removal effect, poor plant compatibility, no improvement or even aggravation of pollution.

9. The quantitative evaluation method for the remediation effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, Methods for constructing long-term risk prediction models include: Multiple wetland water body restoration cases were obtained. For each case, restoration assessment feature sequences were extracted and labeled with the true label Y of the long-term risk level. The labeled long-term risk level was divided according to the actual measurement, and the long-term risk level included Level 1 long-term effectiveness, Level 2 medium-term effectiveness, and Level 3 short-term effectiveness. Using the site boundary as a reference, add spatial location labels to each pollution risk zone according to the distribution of pollution risk zones; Based on the spatial location labels, importance levels, and quantitative assessment results of local remediation effects, a regional vector x is constructed for each pollution risk area. i ; Arrange the region vectors of all pollution risk areas in descending order of importance to form a remediation assessment feature sequence X, where X = {x1, x2, ..., x}. i }; The repair assessment feature sequence X is input into the encoder of the Transformer architecture to generate a contextual representation of the sequence. The decoder of the Transformer architecture generates a prediction result Y of the long-term risk level based on the true label of the long-term risk level and the contextual representation of the repair assessment feature sequence. output ; The Transformer architecture is trained using a training set constructed from multiple wetland water body restoration cases until the loss function converges to obtain a long-term risk prediction model.

10. The method for quantitatively evaluating the effect of phytoremediation of heavy metals in karst wetland water bodies according to claim 1, characterized in that, The assessment report states: The sampling data includes heavy metal pollution remediation data, plant growth data, and karst environment data for each pollution risk area; The results of the indicator analysis include core indicator data for water pollution remediation, plant ecological adaptation indicator data, and karst environment-specific indicator data for each pollution risk area; The comprehensive score includes a quantitative assessment of the local remediation effects in each pollution risk area and an overall quantitative remediation score for the entire target karst wetland water body. The grading results include the grading level determined based on the overall quantitative grading score; The risk prediction results include the long-term risk level output by the long-term risk prediction model; The optimization recommendations include optimized matching schemes for each pollution risk area when the long-term risks are significant; The optimized matching scheme is matched based on an optimized scheme library constructed from historical repair cases. The matching method includes: For pollution risk areas where the quantitative assessment of local remediation effects exceeds a preset threshold, feature vectors are constructed based on the core indicators of water pollution remediation, plant ecological adaptation indicators, and karst environment-specific indicators for the pollution risk area. The similarity between the feature vectors of the pollution risk area and the feature vectors of historical successful cases in the optimization scheme library is calculated, and the optimization scheme adopted by the case with the highest similarity is selected for output. The output results include at least the plant variety optimization scheme, the community structure adjustment scheme, and the zonal enhanced remediation scheme.

Images