Multi-indicator water-quality prediction in mining areas using a feature-tokenizer transformer with spatiotemporal features
Environmental Research · 2026-09 · 306 · 125210
DOI: 10.1016/j.envres.2026.125210 ↗Kurz gefasst
Fachartikel in der Zeitschrift Environmental Research, 2026-09. Verfasst von Zihan Liu und 6 weiteren Autor:innen. Beteiligte Einrichtung: College of Earth Science and Technology, Shandong University of Science and Technology, Qingdao, 266590, PR China; State Key Laboratory of Disaster P…. Hauptthemen laut MEDLINE: Mining, Water Quality, Environmental Monitoring. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 42 Literaturangaben. Gefördert von: National Natural Science Foundation of China.
Abstract (Originalsprache)
Mining activities can produce persistent and spatially heterogeneous impacts on water quality, but stable multi-indicator prediction across large-scale mining areas remains challenging. We developed a Feature-Tokenizer Transformer (FT-Transformer) multi-task framework for nationwide mining-area water-quality prediction by integrating 55,744 monitoring records with spatial, mining-related, temporal, categorical, and interpolated monthly mean air temperature. The model jointly predicted pH, dissolved oxygen (DO), ammonium nitrogen (NH4-N), and permanganate index (CODMn), and was evaluated using 5 × 3 nested cross-validation against Elastic Net, Random Forest, and XGBoost. The FT-Transformer achieved a mean coefficient of determination (mean R2) of 0.790 ± 0.008, with R2 values of 0.756, 0.756, 0.826, and 0.820 for pH, DO, NH4-N, and CODMn, respectively. It achieved the best overall performance and outperformed the benchmark models for pH, NH4-N, and CODMn, whereas Random Forest showed a slight advantage for DO. Multi-task ablation indicated that shared representations improved joint prediction, especially for NH4-N and CODMn. Response, attribution, subgroup, and scenario analyses showed indicator-specific and context-dependent model behavior: temperature was most consistently associated with DO, while mining-stage and regional background variables contributed more strongly to NH4-N and CODMn. The proposed framework supports multi-indicator water-quality prediction, model applicability screening, and monitoring-priority assessment in heterogeneous mining areas.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Substanzen
Autor:innen und Institutionen
- College of Earth Science and Technology, Shandong University of Science and Technology, Qingdao, 266590, PR China; State Key Laboratory of Disaster Prevention and Ecology Protection in Open-pit Coal Mines, Shandong Univ…
- Institute of Marine Science and Technology, Shandong University, Qingdao, 266237, PR China.
- School of Mathematics and Physics, The University of Queensland, St Lucia, 4072, Australia.
- College of Mathematics and Systems Science, Shandong University of Science and Technology, Qingdao, 266590, PR China.
Förderung
- National Natural Science Foundation of China
Zitieren
APA
Liu, Z., Sui, X., Lyu, X., An, X., Wu, J., Lu, Y., & Zhang, W. (2026). Multi-indicator water-quality prediction in mining areas using a feature-tokenizer transformer with spatiotemporal features. Environmental Research, 306, 125210. https://doi.org/10.1016/j.envres.2026.125210
Vancouver
Liu Z, Sui X, Lyu X, An X, Wu J, Lu Y, et al. Multi-indicator water-quality prediction in mining areas using a feature-tokenizer transformer with spatiotemporal features. Environmental Research. 2026;306:125210. doi:10.1016/j.envres.2026.125210
BibTeX
@article{Liu2026,
author = {Liu, Zihan and Sui, Xiang and Lyu, Xianzhou and An, Xiaobiao and Wu, Jinran and Lu, Yuchen and Zhang, Weijie},
title = {Multi-indicator water-quality prediction in mining areas using a feature-tokenizer transformer with spatiotemporal features},
journal = {Environmental Research},
year = {2026},
volume = {306},
pages = {125210},
doi = {10.1016/j.envres.2026.125210}
}
EUVIMED