Exploring the effects of climate change and urban policies on lake water quality using remote sensing and explainable artificial intelligence

Climate change and urban policies significantly impact lake water quality, especially in rapidly developing regions. Existing impact assessments are limited by the duration and resolution of water quality data, while longterm impacts and contributions of climate and socioeconomic drivers have rarely been quantified. In this study, we utilized remote sensing imagery to obtain high spatiotemporal resolution data and Explainable Artificial Intelligence (XAI) to reveal how the comprehensive impacts vary over time. We obtained Trophic State Index (TSI) data between 2000 and 2022 from remote sensing imagery. XAI was employed to elucidate the intricate connections between TSI and the multifaceted factors that encompass climate and socioeconomic dynamics. Six typical lakes in East China were selected for this study. Results showed that the TSI of these lakes had generally decreased over the past two decades, being typically higher in the littoral zones compared to the center. Moreover, higher temperatures increased lake TSI. Meanwhile, higher rainfall increased TSI through increased lake nutrient inputs from surrounding watersheds where agricultural and industrial activities are intensive. It was also found that economic development impacts on lake TSI shifted around 2013. Economic development caused severe negative effects on lake water quality before 2013, while this phenomenon decreased thereafter. This is attributed to the implementation of a series of green transformation policies in the region. This study demonstrates the long-term contribution of climate change and urban policies in driving lake water quality variations, and provides implications for policy makers to establish positive economic, social, and environmental linkages.