Decadal changes in water transparency of lakes in the middle and lower reaches of the Yangtze River: Trends and implications

• Landsat imagery and machine learning were used to estimate lake Zsd with an R² of 0.712. • From 2013–2023, Zsd showed strong seasonality; 11 of 24 lakes had declining transparency trends. • Only one lake showed an increasing Zsd trend over the 10-year monitoring period. • Ecological restoration efforts like embankment removal had no measurable effect on lake transparency. • The study highlights the value of remote sensing and ML in long-term lake water clarity assessment. In the past few decades, freshwater lakes in the middle and lower reaches of the Yangtze River have experienced significant ecological degradation, with many lakes transitioning from oligotrophic to eutrophic states. In response to this issue, China has implemented some ecological restoration measures in the Yangtze River basin to improve water quality, but the results of those measures on water transparency are unsure. This study integrates long-term Landsat satellite imagery with a machine learning model to estimate Secchi disk depth (Zsd), a critical indicator of lake transparency and ecological health. The model achieved an average absolute error of 13.4 cm, a root mean square error of 17.9 cm, and a coefficient of determination (R 2 ) of 0.712.From 2013 to 2023, Zsd maps revealed significant seasonal fluctuations in transparency. Among the 24 monitored lakes, one exhibited an upward trend, while 11 showed declining trends. The analysis also identified a positive correlation between Zsd and water levels in Poyang and Dongting Lakes, suggesting that higher water levels contribute to higher transparency. However, short-term ecological measures, such as returning embankments to lakes, did not result in measurable improvements in Zsd. This study highlights the importance of systematic monitoring of lake transparency to assess ecosystem health. It demonstrates the potential of integrating remote sensing and machine learning for effective water management and provides a foundation for evaluating the long-term outcomes of ecological restoration policies.