CELESTE-Clim: high-resolution climate projections for hydrological applications

Large-scale outputs from General Circulation and Earth System Models provide high-resolution climate projections for hydrological applications, including streamflow simulation, drought monitoring, and flood risk assessment. These models are essential tools for assessing climate responses under future greenhouse gas emission scenarios. Therefore, their projections often exhibit model-dependent biases that must be corrected before application in regional or local studies. In this work, we provide a dataset derived from an ensemble of eight CMIP6 climate models (CMCC-ESM2, EC-Earth3, EC-Earth3-CC, EC-Earth3-Veg, GFDL-ESM4, MPI-ESM1-2-HR, MRI-ESM2-0, and TaiESM1) for the SSP2-4.5 and SSP5-8.5 scenarios. The dataset covers the periods 1985-2014 (historical) and 2015-2100 (future), including daily variables (precipitation, tmin, tmax) and annual indicators derived from these variables. Bias correction was performed using Empirical Quantile Mapping (EQM), applied to ten climate indices, four related to precipitation, and six to temperature, for thirteen representative catchments located in Argentina, Brazil, Colombia, Uruguay, and France. Also, we provide a morphometric characterization for all study catchments. These areas cover a wide range of hydroclimatic regimes, land use, geology/soil types, and spatial scales.