Developing a wind and solar power data model for Europe with high spatial-temporal resolution

Abstract

This paper describes a wind and solar power production model for Europe based on the numerical weather prediction model COSMO-EU. The COSMO-EU model has hourly time resolution and a spatial resolution of 7 km x 7 km for Europe. The model is validated against power production information from the system operators in Denmark, Germany and Spain. Mean Average Error (MAE) (hourly error averaged for a year) relative to the wind installed capacity is in the range 4.9% -5.9% for wind power production and 2.4%-5.5% for PV (photovoltaic) power production. Root Mean Square Error is in the range 6.2%-7.6% and 4.5%-9.3% for wind and PV power production respectively. The results are compared with similar modelling based on wind and radiation data from the NCEP reanalysis model. This model has six hourly time resolution for wind resources and daily resolution for radiation data. Modelling of wind power production in Denmark, German and Spain has a MAE in the range 5.6%-8.5% and solar PV production 4.9%-6.4% for the NCEP reanalysis model.