{"title":"Trusted artificial intelligence for environmental assessments: An explainable high-precision model with multi-source big data","authors":"","doi":"10.1016/j.ese.2024.100479","DOIUrl":null,"url":null,"abstract":"<div><p>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.</p></div>","PeriodicalId":34434,"journal":{"name":"Environmental Science and Ecotechnology","volume":null,"pages":null},"PeriodicalIF":14.0000,"publicationDate":"2024-08-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2666498424000930/pdfft?md5=910b7c0d9e34a403e422cad4c29a7fe1&pid=1-s2.0-S2666498424000930-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Environmental Science and Ecotechnology","FirstCategoryId":"93","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2666498424000930","RegionNum":1,"RegionCategory":"环境科学与生态学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENVIRONMENTAL SCIENCES","Score":null,"Total":0}
引用次数: 0
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.
环境评估对于确保人类文明的可持续发展至关重要。将人工智能(AI)整合到这些评估中已显示出巨大的前景,但人工智能模型的 "黑箱 "性质往往会因为其决策过程缺乏透明度而破坏信任,即使这些模型表现出很高的准确性也是如此。为了应对这一挑战,我们利用包含自然和人为指标的大量多变量和时空环境数据集,评估了变压器模型与其他人工智能方法的性能。我们进一步探索了显著性地图作为一种新型可解释性工具在多源人工智能驱动的环境评估中的应用,从而能够识别单个指标对模型预测的贡献。我们发现,变压器模型优于其他模型,准确率达到约 98%,接收器工作特征曲线下面积 (AUC) 为 0.891。从区域来看,中部和西南部研究区域的环境评估值主要分为 II 级或 III 级,北部区域为 IV 级,西部区域为 V 级。通过可解释性分析,我们发现水硬度、溶解性总固体和砷浓度是模型中影响最大的指标。我们的人工智能驱动环境评估模型准确且可解释,为有针对性的环境管理提供了可操作的见解。此外,本研究还提出了一个稳健、可解释的模型,弥合了机器学习与环境治理之间的差距,增强了人们对人工智能辅助环境评估的理解和信任,从而推动了人工智能在环境科学中的应用。
期刊介绍:
Environmental Science & Ecotechnology (ESE) is an international, open-access journal publishing original research in environmental science, engineering, ecotechnology, and related fields. Authors publishing in ESE can immediately, permanently, and freely share their work. They have license options and retain copyright. Published by Elsevier, ESE is co-organized by the Chinese Society for Environmental Sciences, Harbin Institute of Technology, and the Chinese Research Academy of Environmental Sciences, under the supervision of the China Association for Science and Technology.