HERA-WB: a multi-decadal, sub-daily water balance dataset for Europe (1951-2020)

Freshwater availability shapes ecosystems and human societies. Anthropogenic activities, including land use and climate change, have altered the hydrological cycle. We introduce HERA-WB (Hydrological European ReAnalysis – Water Balance), a spin-off of the HERA high-resolution hydrological reanalysis for Europe generated with the OS LISFLOOD hydrological model. The model is forced by bias-corrected, downscaled ERA5-Land climate reanalysis and incorporates dynamic maps of land use, water demand, and reservoirs to represent human influence. From this simulation we derive 17 hydrological variables characterising blue and green water fluxes across 2,214 European catchments. The dataset spans the period 1951-2020, with a 6-hourly temporal resolution. We assess the performance of the reanalysis against benchmark datasets from four key hydrological variables: river discharge, surface soil moisture, snow water equivalent, and actual evaporation. We further propose a tiered classification of the variables according to the strength of their independent evaluation, to guide users on the confidence warranted for each. HERA-WB combines a temporal coverage and resolution rarely available together for a dataset of this kind. It highlights the capabilities of global hydrological models like OS LISFLOOD to provide skilful simulations beyond river discharge. By providing internally consistent, co-located estimates of blue and green water fluxes, the dataset can support the development and benchmarking of machine-learning hydrological models as constraints or as targets for evaluating process consistency. Furthermore, the spatial and temporal coverages of HERA-WB enable the analysis of dynamics governing changes in freshwater availability, informing continental to local water policies.