Spatiotemporal prediction for groundwater heavy metal contamination using Soft-DTW-based clustering and graph neural
Yong He1, Zi-Long Duan2, Xiang-Hong Ding3
1Central South University, Changsha, 410083, China; Key Laboratory of Metallogenic Prediction of Nonferrous Metals and Geological Environment Monitoring (Central South University), Ministry of Education, Changsha, 410083, China; School of Geosciences and Info-Physics, Central South University, Changsha, 410083, China; Key Laboratory of Nonferrous and Geological Hazard Detection, Changsha, 410083, China.
A new Soft Dynamic Time Warping clustering-based Graph Neural Network (SDCGNN) accurately predicts groundwater heavy metal contamination zones. This approach improves monitoring and remediation strategies for contaminated sites.
Area of Science:
- Environmental Science
- Hydrogeology
- Data Science
Background:
- Groundwater heavy metal contamination poses risks to ecosystems and human health.
- Accurate spatiotemporal prediction is crucial for effective site management.
- Heterogeneity and complex correlations challenge existing prediction models.
Purpose of the Study:
- To develop an advanced model for identifying groundwater contamination zones.
- To improve the accuracy of heavy metal contaminant spatiotemporal prediction.
- To enhance monitoring optimization and remediation decision-making.
Main Methods:
- Proposed a Soft Dynamic Time Warping clustering-based Graph Neural Network (SDCGNN).
- Utilized Soft Dynamic Time Warping (Soft-DTW) for temporal pattern clustering.
- Implemented a hierarchical local-global graph fusion framework for multi-scale modeling.
Main Results:
- SDCGNN achieved high accuracy with MAE of 0.213 mg/L and MAPE of 5.51%.
- Outperformed baseline models by 49.9% (MAE) and 61.4% (MAPE).
- Accurately predicted contamination in source, plume, and attenuation zones, capturing transport dynamics.
Conclusions:
- The SDCGNN model offers a promising approach for groundwater heavy metal contamination prediction.
- Zone-aware modeling enhances understanding of contaminant behavior.
- The findings facilitate optimized site remediation strategies and monitoring.
