GLAST v2.0: A lake-specific daily surface water temperature dataset (1981–2100) integrating recent extremes and CMIP6 projections

Comprehensive long-term insights into lake thermal dynamics—spanning both historical evolution and future trajectories—are critical for assessing climate change impacts on freshwater ecosystems. Unlike gridded surface temperature products that average over land and water, lake-specific datasets like GLAST offer superior precision by resolving thermal dynamics for 92,245 individual lakes worldwide. However, the previous version (v1.0, https://zenodo.org/records/8322038) was constrained by a historical record ending in 2020 and reliance on older CMIP5-based forcing. Here, we introduce GLAST v2.0, which overcomes these limitations by extending the historical reconstruction and integrating latest-generation projections. Using the FLake model calibrated against satellite observations, we extended historical simulations (driven by ERA5-Land) to 1981–2025, thereby capturing recent extreme warming events. Future projections (2015–2100) were upgraded to the ISIMIP3b (CMIP6) protocol under SSP1-2.6, SSP3-7.0, and SSP5-8.5 scenarios. Acknowledging the inherent differences between reanalysis and ESM forcing, we intentionally retain the 2015–2025 overlap period to allow users to quantify discontinuities and apply tailored bias corrections. Extensive validation against independent observations confirms the dataset's robust performance in capturing interannual variability and recent warming trends. GLAST v2.0 provides a vital, high-resolution resource for assessing lake thermal evolution under the latest climate narratives.