Historical: This dataset is based on the ERA5 reanalysis and retrieved from the CDS. No bias
adjustment has been applied. Averages for NUTS0 (country) and NUTS2 (sub-country) areas have been computed.

Seasonal Forecasts: This dataset is currently available for ECMWF System 5, Météo-France ARPEGE System 7 and Met Office GloSEA5 System 14. Seasonal Forecasts are retrieved from the CDS and are subsequently bias adjusted via quantile mapping. The forecast datasets are updated monthly when the raw forecast of each system is available. Averages for NUTS0 (country) are also provided.

Projections: This dataset is available for eleven regional climate model (RCM)/global climate model (GCM) combinations for the EURO-CORDEX project (Table 1) and for two Representative Concentration Pathways (RCP4.5 and RCP8.5). Historical forcing is used from 1979-2005, and RCP forcing from 2006-2100. The original EURO-CORDEX simulations are retrieved from the Earth System Grid Federation (ESGF) portal and are subsequently bias adjusted by applying the Cumulative Distribution Function transform (CDFt) method. Further processing involved interpolation to a standard 0.25° latitude/longitude grid, and calculation of country and regional
averages.

Historical: data on a standard 0.25° latitude/longitude grid for the C3S Energy domain, as well as
NUTS0 and NUTS2 averages. They will eventually be available via the C3S Energy/CDS demonstrator.

Seasonal Forecasts: the bias adjusted hindcast and forecast data will eventually be available via the C3S Energy/CDS demonstrator.

Projections: EURO-CORDEX Indicator available via the Earth System Grid Federation (ESGF) portal.

Historical: Gridded historical data are available as NetCDF files. NUTS0 and NUTS2 are available as CSV files.

Seasonal Forecasts: Gridded seasonal forecast data are available as NetCDF files. NUTS0 are available as CSV files.

Projections: Gridded projection data are available as NetCDF files. NUTS0 and NUTS2 are available as CSV files.

copernicus-support@ecmwf.int

Historical: 1-hour time step as accumulated value. Averages available for daily, monthly, seasonal (DJF, MAM, JJA, SON) and annual are also computed for NUTS0 and NUTS2 mean values.

Seasonal Forecasts: 24-hour time step as accumulated value. Averages for monthly, seasonal (DJF, MAM, JJA, SON) and annual are also computed for NUTS0 mean values.

Projections: 24-hour time step as accumulated value. Daily NUTS0 and NUTS2 averages are also computed.

Historical: 0.25° by 0.25° latitude/longitude. Data are also available as averages for more than 30 European countries and 350 NUTS2 areas.

Seasonal Forecasts: 1° by 1° latitude/longitude and at country (NUTS0) scale. The underlying forecast systems run at different resolutions. Data are also available as averages for more than 30 European countries as NUTS0 areas.

Projections: 0.25° by 0.25° latitude/longitude. Data are also available as averages for more than 30 European countries and ca. 350 NUTS2 areas.

Historical: ERA5

Seasonal Forecasts:

1. ECMWF seasonal forecast system 5 technical description:
2. Météo-France seasonal forecast system 7 technical description:
3. Met Office

Projections: original data come from CORDEX simulations available via the ESGF portal.

Historical: original gridded data come from the ERA5 Reanalysis

Seasonal Forecasts: original data is available from the C3S Climate Data Store. See here for further details of the seasonal forecast models used.

Projections: original data come from CORDEX simulations available via the ESGF portal.

Historical: data were retrieved and post-processed using python scripts.

Seasonal Forecasts: data were processed using standard open source python libraries, including numpy, scipy, matplotlib, and the SciTools packages iris and cartopy.

Projection: data were processed using standard open source programmes, including Fortran, R and CDO.

Historical: ERA5 data have been compared to previously used ERA-Interim for consistency.

Seasonal Forecasts: forecast skill, such as measured by anomaly correlation, is performed and will be provided via the C3S Energy demonstrator. Maps representing the bias for different time- periods for each system are also provided, as well as graphs comparing hindcast dataset before
and after bias-correction.

Projections: a "health check" procedure was applied on bias adjusted data to verify:

• that the bias adjusted projections had similar seasonal means over the reference periods; by mapping the absolute differences, it was found that they are an order of
  magnitude smaller than the variables variations. 
•  that the changes between end of current and past centuries were similar with or without bias adjustment. We found that while the general patterns were quite similar, changes amplitudes can be modified by the CDFt method.