1. ABOUT THE DATASET --------------------------------------------------------------------------------------- Title: Sub-seasonal forecasts of European electricity demand, wind power and solar power generation Creator(s): P.L.M. Gonzalez, H.C. Bloomfield, D.J. Brayshaw, A. Charlton-Perez Organisation(s): University of Reading, National Centre for Atmospheric Science Source(s): * This dataset uses meteorological data from two hindcast ensembles from modeling centers ECMWF (European Centre for Medium-range Weather Forecasting, ECMWF ENS-ER, https://www.ecmwf.int/en/forecasts/documentation-and-support/extended-range-forecasts, Owens and Hewson (2018) ) and NCEP (National Center for Environmental Prediction, CFS version 2, Saha et al. 2014) that are part of the S2S dataset (Sub-seasonal to Seasonal prediction, http://s2sprediction.net/). They were obtained from the ECMWF S2S data portal (https://apps.ecmwf.int/datasets/data/s2s-realtime-instantaneous-accum-ecmf/). Saha, Suranjana, and Coauthors, 2014: The NCEP Climate Forecast System Version 2. J. Climate, 27, 2185-2208 . DOI: 10.1175/JCLI-D-12-00823.1 Owens, R. G., & Hewson, T. D., 2018: ECMWF forecast user guide. http://dx.doi.org/10.21957/m1cs7h * This dataset uses meteorological data from the ERA5 reanalysis (Hersbach et al., 2020) available from https://cds.climate.copernicus.eu/cdsapp#!/home. Hersbach, H., Bell, B., Berrisford, P., Hirahara, S., Horányi, A., Muñoz-Sabater, J., Nicolas, J., Peubey, C., Radu, R., Schepers, D., et al.: The ERA5 global reanalysis, Quarterly Journal of the Royal Meteorological Society, 146, 1999–2049, 2020. https://doi.org/10.1002/qj.3803 Publication Year: 2020 Description: Sub-seasonal forecasts of daily country-level European electricity demand, wind power and solar power generation, along with the driving meteorological variables, from two S2S (subseasonal-to-seasonal) reforecast systems and lead times extending to 44 days. The matching ERA5-derived variables are also provided to facilitate verification analyses. Cite as: Gonzalez, Bloomfield, Brayshaw, Charlton-Perez (2020): Sub-seasonal forecasts of European electricity demand, wind power and solar power generation. University of Reading. Dataset. http://dx.doi.org/10.17864/1947.275 Related publication: Bloomfield, H.C., Brayshaw, D.J., Gonzalez, P.L.M. & Charlton-Perez, A.J. (2020). Sub-seasonal forecasts of demand, wind power and solar power generation for 28 European Countries. Submitted to Earth System Science Data. Contact: Paula LM Gonzalez (p.gonzalez@reading.ac.uk) 2. TERMS OF USE --------------------------------------------------------------------------------------- Copyright 2020, University of Reading. This dataset is licensed by the rights-holder(s) under a Creative Commons Attribution 4.0 International Licence: https://creativecommons.org/licenses/by/4.0/. 3. PROJECT AND FUNDING INFORMATION --------------------------------------------------------------------------------------- Title: Sub-seasonal to seasonal forecasting for energy (S2S4E) Dates: 1st December 2017 - 31st December 2020 Funding organisation: Horizon 2020 Grant no.: 776787 This project received funding from the Horizon 2020 programme under the grant agreement number 776787. 4. CONTENTS --------------------------------------------------------------------------------------- The dataset consists of nine daily variables, each of which is presented for the two mentioned S2S reforecast sets and the matching ERA5-derived parameters. The nine variables are: * country level electricity demand, * country level 2m temperature, * country level solar power, * country level surface solar radiation, * country level wind power, * European gridded 100m wind speed, * country level demand-net-wind, * country level demand-net-solar, * country level demand-net-renewables. They can be found in the relevant subdirectories. The contents of each folder are described below. Sample reading scripts for python, R and MATLAB are also included in a separate directory. <<< country_daily_demand >>> ecmwf_s2s_coun_demand_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). References can be found here: https://ec.europa.eu/eurostat/statistics-explained/index.php/Glossary:Country_codes - date: array of validity dates depending on the forecast start year, start day and lead time - demand_forecast: ECMWF forecasts of daily country level electricity demand in GW, for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - demand_rea: Matching country level electricity demand derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_demand_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for electricity demand derived from the NCEP CFSv2 system. <<< country_daily_2m_temperature >>> ecmwf_s2s_coun_t2m_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - t2m_forecast: ECMWF forecasts of daily country level 2m temperature in deg C, for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - t2m_rea: Matching country level 2m temperature derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_t2m_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for 2m temperature derived from the NCEP CFSv2 system. <<< country_daily_solar_power >>> ecmwf_s2s_coun_solar_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - solar_forecast: ECMWF forecasts of daily country level solar power generation in GW, for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - solar_rea: Matching country level solar power derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_solar_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for solar power generation derived from the NCEP CFSv2 system. <<< country_daily_surface_solar_radiation >>> ecmwf_s2s_coun_ssrd_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - ssrd_forecast: ECMWF forecasts of daily country level surface solar radiation en W/m^2, for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - ssrd_rea: Matching country level surface solar radiation derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_ssrd_fcst_w_VI_and_rea_DAILY_stmon##.nc - Same as above but for surface solar radiation derived from the NCEP CFSv2 system. <<< country_daily_wind_power >>> ecmwf_s2s_coun_windpower_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - windpower_forecast: ECMWF forecasts of daily country level wind power generation in GW, for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - windpower_rea: Matching country level wind power derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_windpower_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for wind power generation derived from the NCEP CFSv2 system. <<< EU_domain_daily_wind_speed >>> (This variable is split in 4 directories, according to the calendar months of the forecast starts: Jan-Mar, Apr-Jun, Jul-Sep, Oct-Dec) ecmwf_s2s_100m_wind_spd_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - wind_forecast: ECMWF forecasts of daily gridded 100m wind speed in m/s for a domain covering Europe, for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - wind_rea: Matching country level 100m wind speed derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_100m_wind_spd_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for 100m wind speed derived from the NCEP CFSv2 system. <<< country_daily_demand_net_wind >>> ecmwf_s2s_coun_demand_net_wind_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - dnw_forecast: ECMWF forecasts of daily country level demand-net-wind in GW (remaining electricity demand after subtraction of wind power generation), for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - dnw_rea: Matching country level demand-net-wind derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_demand_net_wind_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for demand-net-wind derived from the NCEP CFSv2 system. <<< country_daily_demand_net_solar >>> ecmwf_s2s_coun_demand_net_solar_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - dns_forecast: ECMWF forecasts of daily country level demand-net-solar in GW (remaining electricity demand after subtraction of solar power generation), for starts in the specified calendar month, and depending on: ensemble member, start year, start day and lead time. - dns_rea: Matching country level demand-net-solar derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_demand_net_solar_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for demand-net-solar derived from the NCEP CFSv2 system. <<< country_daily_demand_net_renewables >>> ecmwf_s2s_coun_demand_net_renewables_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Where ## is a 2-digit number ranging from 1 to 12, corresponding to a calendar month. Inside each of these NetCDF files, there are 4 arrays of data: - country_codes: list of 2-character strings describing the 28 European countries for which products were calculated (see list below). - date: array of validity dates depending on the forecast start year, start day and lead time - dnr_forecast: ECMWF forecasts of daily country level demand-net-renewables in GW (remaining electricity demand after subtraction of solar and wind power generations), for starts in the specified month, and depending on: ensemble member, start year, start day and lead time. - dnr_rea: Matching country level demand-net-renewables derived from ERA5, with structure depending on: start year, start day and lead time. ncep_s2s_coun_demand_net_renewables_fcst_w_VI_and_rea_DAILY_stmon_##.nc - Same as above but for demand-net-renewables derived from the NCEP CFSv2 system. Country codes: 'AT': Asutria 'BE': Belgium 'BG': Bulgaria 'HR': Croatia 'CZ': Czech Republic 'DK': Denmark 'FI': Finland 'FR': France 'DE': Germany 'GR': Greece 'HU': Hungary 'IE': Ireland 'IT': Italy 'LV': Latvia 'LT': Lithuania 'LU': Luxembourg 'ME': Montenegro 'NL': Neatherlands 'NO': Norway 'PL': Poland 'PT': Portugal 'RO': Romania 'SK': Slovakia 'SI': Slovenia 'ES': Spain 'SE': Sweden 'CH': Switzerland 'UK': United Kindgom Time periods: * The ECMWF reforecasts include starts for years in the period 1996-2015 and is initialized twice weekly (on Mondays and Thursdays). The dates of the model starts match all the Mondays and Thursdays in 2016. * The NCEP CFSv2 reforecasts include starts for years in the period 1999-2010 and is initialized daily. 5. METHOD and PROCESSING --------------------------------------------------------------------------------------- For all three of the models described below the countries modelled are: Austria, Belgium, Bulgaria, Croatia, Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Ireland, Italy, Latvia, Lithuania, Luxembourg, Montenegro, Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland, United Kingdom. Full details of the models can be found in Bloomfield et al.(2020b). A brief description of each will now follow. Demand model A time series of daily mean electricity demand for 28 European countries was derived from 2m temperature (T2m) data from the ERA5 reanalysis and from the S2S reforecasts. A lead-dependent bias correction with variance inflation (Doblas-Reyes et al. 2005) was implemented fore the reforecast data before the demand conversion. The demand model was developed on a daily resolution using a multiple-linear regression model where the possible model inputs were the day of the week, heating degree days (HDD) and cooling degree days (CDD). The model was optimised to choose the best set of parameters to minimise the Akaike information criteria. To retan only weather-dependent demand, the effects of the day of the week were ignored. The model was trained on data from the ENTSOe transparency platform from 2016-2017. Wind Power model A time series of daily wind power capacity factor for the mentioned European countries was derived from 100m wind speed data from the ERA5 reanalysis and the S2S reforecasts. In the later case, only daily midnight wind speeds were available and a consistent ERA5 dataset was derived to match it. The model was developed by selecting the optimal wind turbine that would be present in each re-analysis grid box based on the climatology of the 100m wind speeds. The wind power generation in each grid box was then calculated and weighted by the location of the turbines (taken from thewindpower.net database) in order to get country-aggregate wind power capacity factor data. Solar Power model A time series of daily solar power capacity factor for the European countries was derived from 2m temperatures and surface short wave radiation data from the ERA5 reanalysis and the S2S reforecasts. In the case of ERA5, the daily radiation values were aggregated from the hourly outputs to match the reforecasts accumulation step. The solar power model is an empirical model developed from the methods used in Evans and Florschuetz 1977 and Bett and Thornton (2016). References: Doblas-Reyes, F. J., Hagedorn, R., and Palmer, T.: The rationale behind the success of multi-model ensembles in seasonal forecasting—II. Calibration and combination, Tellus A: Dynamic Meteorology and Oceanography, 57, 234–252, 2005. Evans, D.L. and L.W. Florschuetz (1977). Cost studies on terrestrial photovoltaic power systems with sunlight concentration. Solar Energy, 19, 255-262. Bett, P. E., & Thornton, H. E. (2016). The climatological relationships between wind and solar energy supply in Britain. Renewable Energy, vol 87 pt 1, 96-110. DOI: 10.1016/j.renene.2015.10.006. Bloomfield, H. C., Brayshaw, D. J., & Charlton‐Perez, A. J. (2020). Characterizing the winter meteorological drivers of the European electricity system using targeted circulation types. Meteorological Applications, 27(1). DOI:10.1002/met.1858 Bloomfield, H.C., Brayshaw, D.J., Gonzalez, P.L.M. & Charlton-Perez, A.J. (2020). Sub-seasonal forecasts of demand, wind power and solar power generation for 28 European Countries. Submitted to Earth System Science Data.