Extremes in European wind and photovoltaic power production in 2050 identifiable by weather patterns

Wind and photovoltaic (PV) power production vary across time and space. We aim at a first systematic assessment of anomalies in PV and wind power production associated with different synoptic-scale weather patterns with a kilometre-scale resolution for all of Europe. To that end, we have developed the University of Cologne’s Renewable Energy Model (UoC-REM). UoC-REM simulates the hourly PV and wind power production using the reanalysis data set COSMO-REA6 with a horizontal resolution of 6 km. The installed capacities of PV and wind power for 2050 are taken from gridded scenario data. The output of UoC-REM is paired with data from a classification of 29 synoptic weather patterns identified and provided by the German Weather Service. Our results reveal substantial spatio-temporal differences in PV and wind power production depending on the weather pattern. We group the PV and wind power production from individual weather patterns into three groups to facilitate a composite analysis underlining similarities for the PV and wind power production across the weather patterns. These are (1) the group with anomalously high wind power production that almost always produces above average total production, primarily associated with patterns related to westerly winds, (2) the group of moderate PV plus wind power production with weather patterns causing mostly mild anomalies in both PV and wind power, but also ‘dark doldrums’ with simultaneously low wind and PV power production during blocking high pressure systems, e.g., South-Shifted Westerly, and (3) the group with high PV power production paired with below average wind power that is mostly associated with anti-cyclonic weather patterns. We identify that the lowest 10-day production event of PV plus wind power occurs during the pattern Anticyclonic South-Easterly from Group 3, with a reduction by 41% compared to the average hourly total production for Europe. On the contrary, the highest 10-day production event occurs during the pattern Cyclonic North-Westerly from Group 1, with an increase by 18%. Our results suggest that identifying the weather pattern can be used as a quick estimate of the overall production anomaly in PV and wind power production. It would allow to monitor and issue warnings of weather conditions that pose a risk to a future energy system that relies on more weather-dependent renewable sources, without the need to perform operational simulations of the expected power production. Future work will focus on an in-depth analysis of extreme events in renewable power production considering the full spatial and temporal resolution of the new dataset of UoC-REM.