Trusted artificial intelligence for environmental assessments: An explainable high-precision model with multi-source big data

Abstract

Environmental assessments are critical for ensuring the sustainable development of human civilization. The integration of artificial intelligence (AI) in these assessments has shown great promise, yet the "black box" nature of AI models often undermines trust due to the lack of transparency in their decision-making processes, even when these models demonstrate high accuracy. To address this challenge, we evaluated the performance of a transformer model against other AI approaches, utilizing extensive multivariate and spatiotemporal environmental datasets encompassing both natural and anthropogenic indicators. We further explored the application of saliency maps as a novel explainability tool in multi-source AI-driven environmental assessments, enabling the identification of individual indicators' contributions to the model's predictions. We find that the transformer model outperforms others, achieving an accuracy of about 98% and an area under the receiver operating characteristic curve (AUC) of 0.891. Regionally, the environmental assessment values are predominantly classified as level II or III in the central and southwestern study areas, level IV in the northern region, and level V in the western region. Through explainability analysis, we identify that water hardness, total dissolved solids, and arsenic concentrations are the most influential indicators in the model. Our AI-driven environmental assessment model is accurate and explainable, offering actionable insights for targeted environmental management. Furthermore, this study advances the application of AI in environmental science by presenting a robust, explainable model that bridges the gap between machine learning and environmental governance, enhancing both understanding and trust in AI-assisted environmental assessments.

Keywords: Explainable AI; Intelligent environmental assessment; Multi-source data; Transformer.

Graphical abstract

Fig. 1

The technical flowchart of explainable deep learning (transformer) and big data for environmental risk assessment.

Fig. 2

The study area (a) and its hydrogeological section map (b).

Fig. 3

High-precision and explainability analysis results of deep learning and environmental multi-source big data. a, Workflow network structure of environmental risk assessment based on deep learning. b, The results of the environment risk assessment using deep learning (transformer). c, The contribution of indicators (or importance of indicators) of the model based on explainable AI.

Fig. 4

Assessment results of natural and man-made important indicators. a, Groundwater depth. b, Recharge. c, Aquifer media. d, Pollution source location. e, Soil index. f, Terrain. g, Distribution diagram of aquifer permeability coefficient. h, Groundwater level contour. i, The toxicity effect distribution of total water hardness (THW) exceeds the standard. j, The toxicity effect distribution of total dissolved solids (TDS) exceeds the standard. k, THW pollution diffusion prediction. l, TDS pollution diffusion prediction. m, The toxicity effect distribution of excessive As. n, The toxicity effect distribution of exceeding NH₃-N. o, As pollution diffusion prediction. p, NH₃-N pollution diffusion prediction.