Seasonal predictions of energy-relevant climate variables through Euro-Atlantic Teleconnections

Abstract

The goal of this analysis is the better understanding of how the large-scale atmospheric patterns affect the renewable resources over Europe and to investigate to what extent the dynamical predictions of the large-scale variability might be used to formulate empirical prediction of local climate conditions (relevant for the energy sector). The increasing integration of renewable energy into the power mix is making the electricity supply more vulnerable to climate variability, therefore increasing the need for skillful weather and climate predictions. Forecasting seasonal variations of energy relevant climate variables can help the transition to renewable energy and the entire energy industry to make better informed decision-making. At seasonal timescale climate variability can be described by recurring and persistent, large-scale patterns of atmospheric pressure and circulation anomalies that interest vast geographical areas. The main patterns of the North Atlantic region (Euro Atlantic Teleconnections, EATCs) drive variations in the surface climate over Europe. We analyze reanalysis dataset ERA5 and the multi-system seasonal forecast service provided by Copernicus Climate Change Service (C3S). We found that the observed EATC indices are strongly correlated with surface variables. However, the observed relationship between EATC patterns and surface impacts is not accurately reproduced by seasonal prediction systems. This opens the door to employ hybrid dynamical-statistical methods. The idea consists in combining the dynamical seasonal predictions of EATC indices with the observed relationship between EATCs and surface variables. We reconstructed the surface anomalies for multiple seasonal prediction systems and benchmarked these hybrid forecasts with the direct variable forecasts from the systems and also with the climatology. The analysis suggests that hybrid methodology can bring several improvements to the predictions of energy relevant Essential Climate Variables.