Assessing variables of regional reanalysis data sets relevant for modelling small-scale renewable energy systems

Data that allow the characterization of short-term and inter-annual variability of renewable energy resources availability are becoming highly valuable for energy system modellers. Global reanalysis data provide long time series of records without gaps and full spatial coverage at no cost for the final user. However, these exhibit coarse spatial resolution. The COSMO-REA6 and COSMOS-REA2 regional reanalysis for Europe overcome this limitation by increasing the resolution of the reanalysis to six and two kilometres respectively. This work presents an assessment of solar radiation and wind speed variables of these data sets that were available to the public in January 2018. This assessment is performed using data of the Bavarian agro-meteorological network and the Czech Hydrometeorological Institute. Eight accuracy indicators for hourly data in the period 1995-2015 are calculated. Widely used indicators such as the Pearson's correlation coefficient in some cases reach values above 0.82 for wind speeds and above 0.92 for global horizontal irradiance. The mean bias error is consistently better than +/- 9.3 W/m(2) for the full set of irradiance data, +/- 25.1 W/m(2) for only day-time irradiance data and is, with a few exceptions, lower than +/- 1 m/s for the wind speed data. (C) 2018 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).