Long-term spatiotemporal heterogeneity and drivers of lake thermal dynamics in the Yangtze river basin

Lake surface water temperature (LSWT) governs physical and biogeochemical processes in aquatic environments, serving as an indicator of ecosystem responses to climate change. This study investigated the spatiotemporal evolution and driving mechanisms of LSWT in six representative large lakes across the upper, middle, and lower reaches of the Yangtze River Basin. A long-term (1988-2022) monthly LSWT dataset was reconstructed using the statistical mono-window algorithm in Google Earth Engine platform. Validation against in-situ hydrological data demonstrated high retrieval accuracy, with Nash-Sutcliffe efficiency coefficients exceeding 0.95 and root mean square errors below 1.65 °C. Results revealed a pervasive warming trend across the basin, averaging 0.439 °C/10a, with the most rapid warming observed in the middle reaches. Significant spatiotemporal heterogeneity was identified. Lakes in the upper and middle reaches exhibited a distinct seasonal hysteresis relative to air temperature, whereas shallow lakes in the lower reaches showed synchronous variations. Notably, a regional regime shift in thermal conditions occurred in the middle reaches (Dongting and Poyang Lakes) during 1996-1997. The interplay of cooling effects from river inflows and localized warming from urbanization shaped spatial LSWT patterns. Attribution analysis indicated that air temperature was the dominant driver, explaining over 60.5% of interannual variations. However, anthropogenic impacts were spatially distinct. The expansion of impervious surfaces significantly exacerbated warming in lakes, particularly during specific seasons. These findings disentangle the impacts of natural climate change and human activities, providing a scientific basis for sustainable lake management in complex river basins.