Dynamic traceability and intelligent prediction system for heavy metal pollution of underground water

Abstract

The invention discloses an underground water heavy metal pollution dynamic traceability and intelligent prediction system, which comprises an intelligent monitoring and data acquisition module, and is characterized in that the intelligent monitoring and data acquisition module is used for monitoring the heavy metal concentration and the isotope ratio in real time; the multi-dimensional data establishment and intelligent modeling module is used for quantifying the pollution source contribution probability and dynamically updating a migration model; and the pollution migration simulation-visualization-early warning integration module is used for simulating heavy metal migration paths under different hydrogeological conditions, presenting a spatio-temporal evolution prediction result of pollutant concentration, and rapidly triggering a pollution risk alarm based on a preset threshold value. According to the method, the problems of low monitoring frequency, high data uncertainty and poor model coupling degree in the prior art are solved, the accuracy and timeliness of traceability and prediction can be improved, and environmental risk early warning and restoration decision making are supported.

Description

Technical Field

[0001] This invention relates to the field of environmental monitoring and pollution remediation technology, and in particular to a dynamic source tracing and intelligent prediction system for heavy metal pollution in groundwater. Background Technology

[0002] With the acceleration of industrialization and urbanization, and the increasing intensification of agriculture, soil and groundwater pollution has become a global problem threatening ecological environment security and public health. Groundwater, as a core freshwater resource for maintaining production, daily life, and ecosystem stability, faces a particularly severe pollution situation. The "National Soil Pollution Status Survey Bulletin" shows that the overall exceedance rate of soil pollution sites in my country reached 16.1%, with heavy metals such as cadmium and lead, as well as polycyclic aromatic hydrocarbons (PAHs), being the main pollutants. The "2023 China Ecological Environment Status Bulletin" indicates that 22.2% of the national groundwater environmental quality assessment sites were classified as Class V, with pollution problems being particularly prominent in economically developed regions such as the North China Plain and the Yangtze River Delta. Heavy metal pollution in groundwater exhibits significant multi-source and complex characteristics. Globally, human activities such as non-ferrous metal smelting, chemical production, traffic emissions, and the application of fertilizers and pesticides, combined with rock weathering under natural geological conditions, cause heavy metals such as Cd, Hg, As, Pb, and Mn to enter the groundwater system through infiltration and leaching, severely weakening the utilization value of groundwater. This type of pollution is characterized by its insidious nature, persistence, and bioaccumulation. It can induce diseases such as cancer and nervous system damage in humans through drinking water and skin contact, and it can also destroy aquatic biodiversity and amplify ecological risks through the food chain. Difficulties in sampling and complex geological conditions mean that there are still technological gaps in areas such as precise characterization of pollution, quantitative analysis of pollution sources, and the tracking of diffusion and transfer. Against this backdrop, my country has successively issued policies such as the "Soil Pollution Prevention and Control Action Plan" and the "Groundwater Management Regulations," highlighting the urgency of pollution prevention and control. Precise and efficient pollution source tracing technologies have become a key support for pollution control, responsibility allocation, and refined environmental decision-making.

[0003] Since the 1970s, source tracing technologies for soil and groundwater pollution have gradually developed and improved. From early source inventory methods such as emission inventories and material flow analysis, to the application of high-precision detection technologies like inductively coupled plasma mass spectrometry (ICP-MS) at the end of the 20th century, and further to the inclusion of statistical models, isotope tracing, and solute transport inversion methods, the source tracing technology system has been continuously enriched. Emerging methods such as geophysical exploration, remote sensing, and machine learning have provided innovative technical pathways for pollution source analysis. However, source tracing technology in my country is still in its initial stage. Current methods largely rely on historical data and field surveys, with limited resolution and mostly remaining at the qualitative identification level. They lack a systematic real-time monitoring system, making it difficult to accurately determine the contribution of pollution sources and analyze migration paths. At the same time, single-medium, single-technology source tracing and prediction models face challenges such as difficult sampling, insufficient data accuracy, high costs, and difficulties in technology integration when dealing with sites with complex pollution causes and highly heterogeneous environmental media. There is an urgent need to build a high-precision source tracing and trend prediction system integrating multiple technologies to provide scientific support for the precise prevention and control of soil and groundwater pollution. Summary of the Invention

[0004] The purpose of this invention is to provide a dynamic source tracing and intelligent prediction system for heavy metal pollution in groundwater, addressing the aforementioned problems. This system aims to solve the issues of monitoring lag, high data uncertainty, and poor model prediction reliability in existing technologies, thereby achieving near real-time response, accurate source tracing, and dynamic prediction of pollution events.

[0005] To achieve the above-mentioned objectives, the technical solution adopted by this invention is as follows: According to one aspect of the present invention, a dynamic source tracing and intelligent prediction system for heavy metal pollution in groundwater is provided, comprising: An intelligent monitoring and data acquisition module is used to monitor heavy metal concentration and isotope ratio in real time. A multidimensional data establishment and intelligent modeling module is used to quantify the contribution probability of pollution sources and dynamically update the migration model; The pollution migration simulation-visualization-early warning integrated module is used to simulate the migration path of heavy metals under different hydrogeological conditions, present the spatiotemporal evolution prediction results of pollutant concentration, and quickly trigger pollution risk alarms based on preset thresholds.

[0006] Preferably, the intelligent monitoring and data acquisition module includes a multi-parameter sensor unit, an automatic sampling unit, a self-cleaning and self-calibrating unit, and a remote data transmission unit; The multi-parameter sensor unit is used to collect in-situ data of heavy metals in groundwater; The automatic sampling unit is used to automatically collect, enrich, and pre-treat water samples; The self-cleaning and self-calibrating unit is used to clean the multi-parameter sensor unit; The data remote transmission unit is used for remotely transmitting monitoring data.

[0007] Preferably, the multi-parameter sensor unit includes an electrochemical sensor and an optical sensor, which are used for in-situ real-time monitoring of heavy metal elements.

[0008] Preferably, the automatic sampling unit includes a microfluidic chip and a solid-phase extraction device, which are used for automatic collection, enrichment and pretreatment of water samples.

[0009] Preferably, the self-cleaning and self-calibrating unit includes an ultrasonic device and an automatic standard solution injection device, which are used to clean the multi-parameter sensor unit.

[0010] Preferably, the data remote transmission unit includes an NB-IoT module or a LoRa module, which is used to transmit monitoring data.

[0011] Preferably, the multidimensional data establishment and intelligent modeling module includes a Bayesian probability model and a data assimilation unit. The Bayesian probability model is used to output the contribution probability distribution of pollution sources, and the data assimilation unit is used to couple the output of the Bayesian probability model with the Visual MODFLOW model.

[0012] Preferably, the pollution migration simulation-visualization-early warning integrated module includes a pollution migration simulation unit, a pollution visualization unit, and an early warning unit; The pollution migration simulation unit is used to simulate the migration paths of heavy metals under different hydrogeological conditions. The pollution visualization unit is used to present the migration trajectory of the pollution plume; The early warning unit is used to establish a multi-level early warning mechanism and send alarm information according to the multi-level early warning mechanism.

[0013] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention captures the dynamic evolution of pollution events through isotope monitoring and real-time pollutant data streams, enabling timely early warning and emergency response, and achieving a leap from static source tracing to dynamic tracking.

[0014] 2. This invention employs microfluidic chip and solid-phase extraction integrated pretreatment technology to achieve automatic collection, enrichment and standardized pretreatment of water samples, providing high-quality samples for laboratory isotope analysis, effectively reducing errors caused by manual operation, ensuring the accuracy of subsequent source tracing analysis data, and providing standardized sample preparation technology support for pollution source tracing.

[0015] 3. This invention constructs a key heavy metal isotope fingerprint database, combines it with a Bayesian model to quantify the contribution probability and uncertainty of each pollution source, and achieves cross-model collaboration with Visual MODFLOW through a data assimilation interface, enabling the combination of source tracing analysis and groundwater migration, thereby enhancing the scientific rigor, accuracy, visualization, and spatial resolution capabilities of pollution source analysis.

[0016] 4. This invention provides environmental managers with richer and more intuitive decision-making information through probabilistic visualization results and scenario simulation functions, changing the previous decision-making model that relied on single, deterministic simulation results. It enhances decision support capabilities by enabling visualized tracking of pollution risks and timely risk warnings.

[0017] 5. The system architecture of this invention is modular, and can flexibly add or remove monitoring points, adjust monitoring indicators (extendable to other organic pollutants), or replace / upgrade the backend physical model according to the needs of different sites, thus having good scalability and applicability. Attached Figure Description

[0018] Figure 1 This is a framework diagram of the groundwater heavy metal pollution dynamic tracing and intelligent prediction system of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and preferred embodiments. However, it should be noted that many details listed in the specification are merely to provide the reader with a thorough understanding of one or more aspects of the invention, and these aspects of the invention can be achieved even without these specific details.

[0020] Please see Figure 1 This invention provides a dynamic source tracing and intelligent prediction system for heavy metal pollution in groundwater, the technical solution of which is as follows: The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system includes an intelligent monitoring and data acquisition module, a multi-dimensional data establishment and intelligent modeling module, and a pollution migration simulation-visualization-early warning integrated module.

[0021] The intelligent monitoring and data acquisition module includes a multi-parameter sensor unit, an automatic sampling unit, a self-cleaning and self-calibrating unit, and a remote data transmission unit. The multi-parameter sensor unit integrates an electrochemical sensor and an optical sensor. The electrochemical sensor is an anodic stripping voltammetry sensor, and the optical sensor is a laser-induced breakdown spectroscopy sensor. The integrated electrochemical and optical sensors enable in-situ real-time detection of multiple heavy metals such as Mn, Pb, Cd, Cr, and Hg, with detection limits down to the μg / L level. The automatic sampling unit integrates a microfluidic chip and a solid-phase extraction subunit, capable of automatically collecting, enriching, and preprocessing water samples according to preset programs or remote commands, providing standardized samples for subsequent laboratory isotope analysis. The self-cleaning and self-calibrating unit is equipped with an ultrasonic cleaning subunit and an automatic standard solution injection subunit, effectively ensuring the stability and detection accuracy of the sensor during long-term operation. The remote data transmission unit uses low-power wide-area network technologies such as NB-IoT or LoRa to achieve real-time transmission of monitoring data and remote monitoring of equipment operation status, providing technical support for dynamic tracking and emergency response to heavy metal pollution.

[0022] The multidimensional data establishment and intelligent modeling module includes a Bayesian probabilistic model and a data assimilation unit. The Bayesian probabilistic model uses the Markov chain Monte Carlo algorithm for parameter estimation and uncertainty quantification, outputting the contribution probability distribution of each pollution source. The data assimilation unit couples the output of the Bayesian probabilistic model with the Visual MODFLOW model through dedicated middleware, enabling data exchange and dynamic parameter correction.

[0023] The multidimensional data establishment and intelligent modeling module constructs a system for data storage, model analysis, and cross-system collaboration. A pollution source fingerprint database is built as the foundational data support for source tracing analysis, storing heavy metal isotope characteristic data of various potential pollution sources, specifically including key fingerprint information such as δ55Mn, δ207Pb, δ114Cd, δ52Cr, and δ202Hg, providing a benchmark for pollution source identification. The Bayesian probabilistic source tracing model uses the Markov Chain Monte Carlo (MCMC) algorithm to construct a multi-source hybrid model. By integrating monitoring data and fingerprint database information, it outputs the contribution probability and uncertainty of each pollution source, improving the scientific rigor and reliability of the source tracing results. The data assimilation interface provides technical support for collaborative model operation. Through dedicated middleware, it enables efficient data exchange and parameter transfer between the Bayesian model and Visual MODFLOW, supporting dynamic updates of model parameters. This allows source tracing analysis to be combined with processes such as groundwater migration, further enhancing the spatial resolution capability of pollution source tracing.

[0024] The pollution migration simulation-visualization-early warning integrated module is used to accurately simulate the migration paths of heavy metals under different hydrogeological conditions, intuitively present the spatiotemporal evolution prediction results of pollutant concentrations, and quickly trigger pollution risk alarms based on preset thresholds. The module includes a pollution migration simulation unit, a real-time data dashboard, a 3D visualization engine, and an intelligent early warning unit. The pollution migration simulation unit uses real-time values ​​of pollution factors (including pollutant concentration, initial diffusion intensity, and other key indicators) output by a Bayesian model as its core input parameters. It automatically triggers preset model call commands, calls the Visual MODFLOW numerical simulation model, and converts the aforementioned real-time values ​​according to a preset format before inputting them into its corresponding parameter interface. Based on the input real-time parameters and combined with preset hydrogeological parameters of the study area (including aquifer permeability coefficient, effective porosity, attenuation coefficient, dispersion, groundwater velocity, and flow direction), the Visual MODFLOW numerical simulation model performs real-time dynamic simulation of the pollution plume diffusion scenario and predicts short-term and medium-to-long-term diffusion trends. Finally, it outputs results such as the spatial distribution range of the pollution plume at different time points, changes in concentration gradients, and the evolution of diffusion boundaries, providing accurate data support for pollution prevention and control decisions and the evaluation of remediation scheme effectiveness. The real-time data dashboard visually presents the spatiotemporal dynamic changes of water quality parameters at each monitoring point through charts. The 3D visualization engine, based on Visual MODFLOW's post-processing engine, statistically processes multiple sets of simulation results to display the migration path of the pollution plume in 3D. The boundary of the pollution plume is no longer a single line, but rather probability isosurfaces represented by different colors or transparency (such as 95%, 75%, and 50% confidence intervals), intuitively demonstrating the uncertainty of the prediction. The intelligent early warning unit establishes a multi-level early warning mechanism, which can automatically trigger corresponding alarm information based on the pollution migration prediction results, achieving timely response to pollution risk warnings.

[0025] Example Taking an industrial park as an example, the groundwater flows from north to south. There are three companies within the park: A (an electroplating plant, a potential source of Pb and Cr pollution), B (a battery factory, a potential source of Pb and Mn pollution), and C (a mining and smelting plant, a potential source of Mn pollution). Downstream from the park is a residential area. Long-term monitoring and risk warning of heavy metal pollution in the groundwater of this area are now necessary.

[0026] Implementation steps: 1. System Deployment and Initialization Ten intelligent sensing and data acquisition subsystems (monitoring wells) were deployed at the park boundary, downstream of enterprises, and upstream of residential areas. For three target enterprises (A, B, and C), the system collected samples of pollution sources such as wastewater from their discharge outlets and leachate from their solid waste dumps. These samples were then sent to a laboratory for MC-ICP-MS analysis to determine the stable isotope ratios of Mn, Pb, and Cr. For example, the measured δ²⁰⁶Pb ratio for plant A was 0.835±0.002, and for plant B it was 0.880±0.003; the measured δ⁵⁵Mn ratio for plant B was 0.945±0.001, and for plant C it was 1.147±0.002. Based on the test results, a site-specific "pollutant source isotope fingerprint database" was constructed and entered into the cloud server as prior information for the Bayesian model. At the same time, relying on the field investigation data and existing hydrogeological survey data (including basic parameters such as stratigraphic structure, permeability, dispersion, attenuation coefficient, distribution coefficient, effective porosity, groundwater level dynamics, initial pollutant concentration distribution, and pollution source characteristics), a basic Visual MODFLOW groundwater flow and solute transport model was constructed.

[0027] 2. Long-term operation and dynamic monitoring After the system starts automatic operation mode, each monitoring point automatically uploads heavy metal concentration and conventional water chemical parameters every hour. The isotope automatic sampling module collects water samples at a frequency of 24 hours / time. Maintenance personnel collect the collected samples and send them for testing every week. The cloud server continuously receives the above monitoring data, and the Bayesian model on it runs synchronously in real time to analyze the pollution status of each monitoring point. Under normal operating conditions, the model output results show that the contribution rate of each pollution source is less than 5%, and the pollution concentration is at the background level.

[0028] 3. Emergency Response and Intelligent Source Tracing One day, the online sensor data of monitoring well No. 3, located downstream of Plant A and Plant B, showed a sudden surge in Pb concentration from a background value of 5 μg / L to 80 μg / L, while Mn concentration also showed an abnormal increase. This abnormal signal triggered the system's automatic emergency response mechanism: on the one hand, the monitoring frequency of monitoring well No. 3 was increased to once every 10 minutes to enhance real-time tracking of pollution dynamics; on the other hand, the isotope sampling module of the monitoring well was immediately instructed to perform emergency sampling. After systematic analysis of the recovered emergency samples and regular samples, the laboratory obtained a δ206Pb value of 0.878 during this pollution incident. Based on the fusion analysis of concentration monitoring data and isotope ratio data using a Bayesian source tracing model, it was found that this δ206Pb highly matched the characteristic isotope fingerprint of Plant B (the battery factory). The model ultimately output the source tracing result: Plant B was the main contributor to this pollution incident, with a contribution probability of 92% ± 5%.

[0029] 4. Dynamic forecasting and risk warning After receiving the source tracing results, the Bayesian-MODFLOW dynamic coupling module automatically updates the MODFLOW model, embedding the instantaneous pollution source strength matching the monitored concentration into the grid cell corresponding to Plant B. Based on the updated model, the system performs 100 Monte Carlo simulations to simulate and predict the migration and diffusion of the pollution plume over the next 90 days. Subsequently, the visualization and early warning subsystem generates a probabilistic 3D pollution plume animation. The results show that, under the current hydrogeological conditions, the pollution plume is likely to reach the downstream residential area boundary within 65-75 days, with a probability of 80%. Since this risk probability exceeds the preset 50% threshold, the system immediately pushes early warning information to environmental protection management personnel and their mobile terminals, clearly indicating a high risk of pollution diffusion and recommending timely emergency response measures.

[0030] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A dynamic source tracing and intelligent prediction system for heavy metal pollution in groundwater, characterized in that, include: An intelligent monitoring and data acquisition module is used to monitor heavy metal concentration and isotope ratio in real time. A multidimensional data establishment and intelligent modeling module is used to quantify the contribution probability of pollution sources and dynamically update the migration model; The pollution migration simulation-visualization-early warning integrated module is used to simulate the migration path of heavy metals under different hydrogeological conditions, present the spatiotemporal evolution prediction results of pollutant concentration, and quickly trigger pollution risk alarms based on preset thresholds.

2. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 1, characterized in that: The intelligent monitoring and data acquisition module includes a multi-parameter sensor unit, an automatic sampling unit, a self-cleaning and self-calibration unit, and a remote data transmission unit. The multi-parameter sensor unit is used to collect in-situ data of heavy metals in groundwater; The automatic sampling unit is used to automatically collect, enrich, and pre-treat water samples; The self-cleaning and self-calibrating unit is used to clean the multi-parameter sensor unit; The data remote transmission unit is used for remotely transmitting monitoring data.

3. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 2, characterized in that: The multi-parameter sensor unit includes an electrochemical sensor and an optical sensor, which are used for in-situ real-time monitoring of heavy metal elements.

4. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 2, characterized in that: The automatic sampling unit includes a microfluidic chip and a solid-phase extraction device, which are used for the automatic collection, enrichment and pretreatment of water samples.

5. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 2, characterized in that: The self-cleaning and self-calibrating unit includes an ultrasonic device and an automatic standard solution injection device, which are used to clean the multi-parameter sensor unit.

6. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 2, characterized in that: The data remote transmission unit includes an NB-IoT module or a LoRa module, which is used to transmit monitoring data.

7. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 1, characterized in that: The multidimensional data establishment and intelligent modeling module includes a Bayesian probability model and a data assimilation unit. The Bayesian probability model is used to output the contribution probability distribution of pollution sources, and the data assimilation unit is used to couple the output of the Bayesian probability model with the Visual MODFLOW model.

8. The groundwater heavy metal pollution dynamic source tracing and intelligent prediction system according to claim 1, characterized in that: The pollution migration simulation-visualization-early warning integrated module includes a pollution migration simulation unit, a pollution visualization unit, and an early warning unit; The pollution migration simulation unit is used to simulate the migration paths of heavy metals under different hydrogeological conditions. The pollution visualization unit is used to present the migration trajectory of the pollution plume; The early warning unit is used to establish a multi-level early warning mechanism and send alarm information according to the multi-level early warning mechanism.

Images