Sentinel-3 OLCI images processed with the Atmospheric Correction for Optical Water Types, A4O [Hieronymi et al. in prep & 2023], and the water algorithm OLCI Neural Network Swarm, ONNS [Hieronymi et al., 2017]. ONNS derives inherent optical properties (IOPs) from which the concentrations of water constituents are estimated. In addition, the results of an Optical Water Type (OWT) classification based on A4O reflectances are provided [Bi and Hieronymi, 2024]. All available satellite data of a day for the region of interest are merged in a common grid at approximately original resolution. Information about the variables are given in the attached Additional Info. Version 2 of the data has the license and some metadata corrected. Please use and refer only to Version 2 (see link below).

Satellite remote sensing enables global monitoring of water quality in freshwater and marine ecosystems. However, consistent data quality is a challenge due to variations in the performance of used algorithms for different waters. In this exemplary dataset, we use a novel approach for atmospheric correction and retrieval for water quality characteristics in inland waters, coastal areas, and the open sea. Copernicus Sentinel-3 OLCI satellite images are processed with the Atmospheric Correction for Optical Water Types, A4O [Hieronymi et al. in prep & 2023], and the water algorithm OLCI Neural Network Swarm, ONNS [Hieronymi et al., 2017]. ONNS derives inherent optical properties (IOPs) from which the concentrations of water constituents are estimated. In addition, the results of an Optical Water Type (OWT) classification based on A4O reflectances are provided [Bi and Hieronymi, 2024]. All available satellite data of a day for the region of interest are merged in a common grid at approximately original resolution. An overview of the variables in the dataset can be found in the Additional Information; a detailed description of the contents and background, as well as an optical analysis of the waters, can be found in Hieronymi et al. [2025]. Version 2 of the dataset has the license and some metadata corrected. Data itself remains unchanged.

This datasets contains simulation output for the global hydrological models HydroPy and MPI-HM. Both used meteorological forcing from the GSWP3 dataset for the period 1979-2014 and a 50 years spinup period. The analysis of this simulations is published at https://doi.org/10.5194/gmd-2021-53 .

The data of this experiment have been used in (Hagemann et al., 2020). It comprise daily data of surface runoff and subsurface runoff (drainage) from JSBACH and MPI-HM and simulated daily discharges (river runoff). To generate river runoff, the Hydrological discharge (HD) model (Hagemann et al., 2020; Hagemann and Ho-Hagemann, 2021) was used that was operated at 5 arc minutes horizontal resolution. Different to the published version of HD model parameters (5.0) on Zenodo, an earlier version (4.0) of flow directions and model parameters has been used that is provided as an auxiliary data file. The HD model was set up over the European domain covering the land areas between -11°W to 69°E and 27°N to 72°N. First, the respective forcing data of surface and sub-surface runoff were interpolated to the HD model domain using conservative remapping. Then, daily discharges were simulated with the HD model for the period 1979-2009 (1999-2009 for HD5-MESCAN). In addition, daily discharges were analogously simulated using only JSBACH forcing with the global 0.5° version 1.10 of the HD model. The associated flow directions and model parameters of vs. 1.10 are provided as an auxiliary data file. The HD forcing data are: a) HD5-JSBACH In order to generate daily input fields of surface runoff and drainage, the land surface scheme JSBACH (vs. 3 + frozen soil physics; (Ekici et al., 2014)) was forced globally at 0.5° with daily atmospheric forcing data based on the Interim Re-Analysis of the European Centre for Medium-Range Weather Forecast (ERA-Interim; (Dee et al., 2011)). These forcing data are bias-corrected (see (Beer et al., 2014)) towards the so-called WATCH forcing data (WFD; (Weedon et al., 2011)) that have been generated in the EU project WATCH. b) HD5-MPIHM The MPI-M hydrology model MPI-HM (Stacke and Hagemann, 2012) was driven by daily WATCH forcing data based on ERA-Interim (WFDEI; (Weedon et al., 2014)) from 1979-2009 to generate daily input fields of surface runoff and drainage at global 0.5° resolution. c) HD5-MESCAN Six hourly data of surface runoff and drainage (variable name: percolation) were retrieved from the MESCAN-SURFEX regional surface reanalysis (Bazile et al., 2017) created in the EU project UERRA (Uncertainties in Ensembles of Regional ReAnalysis; www.uerra.eu). SURFEX (Masson et al., 2013) is a land surface platform that was driven by atmospheric forcing at 5.5 km. The forcing comprises 24h-precipitation, near-surface temperature and relative humidity analyzed by the MESCAN surface analysis system as well as radiative fluxes and wind downscaled at 5.5 km from the 3DVar re-analysis conducted with the HARMONIE system at 11 km (Ridal et al., 2017). The latter has been generated using six-hourly fields of the ERA-Interim reanalysis as boundary conditions and covers a domain comprising Europe and parts of the Atlantic, which is similar to the European domain of the Coordinated Downscaling Experiment (CORDEX) at 11 km.

This experiment comprises data that have been used in Hagemann et al. (submitted). It comprises daily data of surface runoff and subsurface runoff from HydroPy and simulated daily discharges (river runoff) of the HD model. The discharge data close the water cycle at the land-ocean interface so that the discharges can be used as lateral freshwater input for ocean models applied in the European region. a) HD5-ERA5 ERA5 is the fifth generation of atmospheric reanalysis (Hersbach et al., 2020) produced by the European Centre for Medium-Range Weather Forecasts (ECMWF). It provides hourly data on many atmospheric, land-surface, and sea-state parameters at about 31 km resolution. The global hydrology model HydroPy (Stacke and Hagemann, 2021) was driven by daily ERA5 forcing data from 1979-2018 to generate daily input fields of surface and subsurface runoff at the ERA5 resolution. It uses precipitation and 2m temperature directly from the ERA5 dataset. Furthermore, potential evapotranspiration (PET) was calculated from ERA5 data in a pre-processing step and used as an additional forcing for HydroPy. Here, we applied the Penman-Monteith equation to calculate a reference evapotranspiration following (Allen et al., 1998) that was improved by replacing the constant value for albedo with a distributed field from the LSP2 dataset (Hagemann, 2002). In order to initialize the storages in the HydroPy model and to avoid any drift during the actual simulation period, we conducted a 50-years spin-up simulation by repeatedly using year 1979 of the ERA5 dataset as forcing. To generate river runoff, the Hydrological discharge (HD) model (Hagemann et al., 2020; Hagemann and Ho-Hagemann, 2021) was used that was operated at 5 arc minutes horizontal resolution. The HD model was set up over the European domain covering the land areas between -11°W to 69°E and 27°N to 72°N. First, the forcing data of surface and sub-surface runoff simulated by HydroPy were interpolated to the HD model grid. Then, daily discharges were simulated with the HD model. b) HD5-EOBS The E-OBS dataset (Cornes et al., 2018) comprises several daily gridded surface variables at 0.1° and 0.25° resolution over Europe covering the area 25°N-71.5°N x 25°W-45°E. The dataset has been derived from station data collated by the ECA&D (European Climate Assessment & Dataset) initiative (Klein Tank et al., 2002; Klok and Klein Tank, 2009). In the present study, we use the best-guess fields of precipitation and 2m temperature of vs. 22 (EOBS22) at 0.1° resolution for the years 1950-2018. HydroPy was driven by daily EOBS22 data of temperature and precipitation at 0.1° resolution from 1950-2019. The potential evapotranspiration (PET) was calculated following the approach proposed by (Thornthwaite, 1948) including an average day length at a given location. As for HD5-ERA5, the forcing data of surface and sub-surface runoff simulated by HydroPy were first interpolated to the HD model grid. Then, daily discharges were simulated with the HD model. Main reference: Hagemann, S., Stacke, T. Complementing ERA5 and E-OBS with high-resolution river discharge over Europe. Oceanologia. Submitted.

The LUCAS LUC historical dataset consists of annual land use and land cover maps from 1950 to 2015. It is based on land cover data from the LANDMATE PFT dataset that was generated from ESA-CCI LC data. The ESA-CCI LC land cover classes are converted into 16 plant functional types and non-vegetated classes employing the method of Reinhart et al. (2021). The land use change information from the Land-Use Harmonization Data Set version 2 (LUH2 v2h, Hurtt et al. 2020) were imposed using the land use translator developed by Hoffmann et al. (2021). For each year, a map is provided that contains 16 fields. Each field holds the fraction the respective plant functional types and non-vegetated classes in the total grid cell (0-1). The LUCAS LUC dataset was constructed within the HICSS project LANDMATE and the WCRP flagship pilot study LUCAS to meet the requirements of downscaling experiments within EURO-CORDEX. Plant functional types and non-vegetated classes: 1 - Tropical broadleaf evergreen trees 2 - Tropical deciduous trees 3 - Temperate broadleaf evergreen trees 4 - Temperate deciduous trees 5 - Evergreen coniferous trees 6 - Deciduous coniferous trees 7 - Coniferous shrubs 8 - Deciduous shrubs 9 - C3 grass 10 - C4 grass 11 - Tundra 12 - Swamp 13 - Non-irrigated crops 14 - Irrigated crops 15 - Urban 16 - Bare

The LUCAS LUC future dataset consists of annual land use and land cover maps from 2016 to 2100. It is based on land cover data from the LANDMATE PFT dataset for the year 2015. The LANDMATE PFT consists of 16 plant functional types and non-vegetated classes that were converted from the ESA-CCI LC land cover data according to the method of Reinhart et al. (2021). The land use change information from the Land-Use Harmonization Data Set version 2 (LUH2 v2.1f, Hurtt et al. 2020) were imposed using the land use translator developed by Hoffmann et al. (2021). The projected land use change information was derived for different Shared Socioeconomic Pathways (SSPs) and Representative Concentration Pathways (RCPs) combinations used in the framework of the 6th phase of Coupled Modelling Intercomparison Project (CMIP6). For each year, a map is provided that contains 16 fields. Each field holds the fraction the respective plant functional types and non-vegetated classes in the total grid cell (0-1). The LUCAS LUC dataset was constructed within the HICSS project LANDMATE and the WCRP flagship pilot study LUCAS to meet the requirements of downscaling experiments within EURO-CORDEX. Plant functional types and non-vegetated classes: 1 - Tropical broadleaf evergreen trees 2 - Tropical deciduous trees 3 - Temperate broadleaf evergreen trees 4 - Temperate deciduous trees 5 - Evergreen coniferous trees 6 - Deciduous coniferous trees 7 - Coniferous shrubs 8 - Deciduous shrubs 9 - C3 grass 10 - C4 grass 11 - Tundra 12 - Swamp 13 - Non-irrigated crops 14 - Irrigated crops 15 - Urban 16 - Bare

The LANDMATE PFT dataset provides a land cover map for Europe for the year 2015 in 0.1° (~10km) and 0.018° (~2km) resolution. The dataset is based on land cover data of the ESA Climate Change Initiative (ESA-CCI, native resolution: 300m) which is translated into 16 plant functional types (PFTs) and non-vegetated classes employing the cross-walking procedure introduced by Reinhart et al. (2021). The translation is done under consideration of the Holdridge Life Zones (HLZs), a system, that classifies land areas based on bioclimatic properties. Through the HLZs, regional distinction of the individual PFT distribution can be achieved. The land cover information is given as fractions per grid cell where each fraction represents the area covered by the respective land cover within each grid cell (0-1). The dataset is available in two different horizontal resolutions, 0.1° (~10km) and 0.018° (~2km), whereby the land cover information is resampled using a fractional approach to achieve the desired resolution. The LANDMATE PFT dataset was carefully developed and designed to meet the present and future requirements of regional climate models and is therefore recommended to be used for regional climate modeling over the European Continent. The LANDMATE PFT dataset (0.1° resolution) serves as basemap for the historical and future land use and land cover dataset LUCAS LUC developed by Hoffmann et al. (2021).

The LANDMATE PFT dataset provides land cover maps for Europe for the years 1992-2015 in 0.1° (~10km) resolution and a land cover map for Europe for the year 2015 in 0.018° (~2km) resolution. Minor interpolation issues found in LANDMATE PFT Version 1.0 are fixed in LANDMATE PFT Version 1.1. The dataset is based on land cover data of the ESA Climate Change Initiative (ESA-CCI, native resolution: 300m) which is translated into 16 plant functional types (PFTs) and non-vegetated classes employing the cross-walking procedure introduced by Reinhart et al. (2022). The translation is done under consideration of the Holdridge Life Zones (HLZs), a system, that classifies land areas based on bioclimatic properties. Through the HLZs, regional distinction of the individual PFT distribution can be achieved. The land cover information is given as fractions per grid cell where each fraction represents the area covered by the respective land cover within each grid cell (0-1). A high-resolution (0.018°) version for the year 2015 is provided. For custom resolution or other time steps in the higher resolution please contact the authors.

The module allows for taking into account wind farms in atmospheric modelling via the wind farm parametrization by Fitch et al, 2012 in the regional climate model COSMO-CLM. Prerequisite is a wind farm mask file. Further details are given in the " Step-by-step implementation" document. Version 2.0: Update of wind farm parametrization