EnergAIze: Empirical statistical downscaling with EPISODES

For the purpose of the EnergAIze project (FFG grant No. FO999899927, AI for Green - Call 2022) empirical statistical downscaling was carried out via the EPISODES method (Kreienkamp et al., 2019, Wetter und Klima - Deutscher Wetterdienst - EPISODES). The EPISODES method was initialized by Deutscher Wetterdienst (DWD) and is under current development in cooperation of DWD and GeoSphere Austria. Empirical-statistical climate methods use statistical methods to analyze and predict climate behavior based on observational data. These methods focus on identifying patterns, correlations and trends in historical climate data (such as temperature and precipitation) in order to make predictions about future climate conditions. EPISODES establishes a statistical relationship between coarse-scale atmospheric patterns (comparable to large-scale weather patterns) and their regional effects using historical data. This relationship is applied to the future, taking future trends into account. Compared to RCMs (i.e. dynamical regional climate models), the method is simple and not very computationally intensive. Nevertheless, multivariate projections can be calculated for selected meteorological parameters, the output format of which is similar to that of RCMs. A prerequisite for high-quality climate projections with EPISODES is a long time series of observations that describe as many different weather situations from the past as possible. For EnergAIze, the CERRA reanalysis data set (5.5km resolution) serves as reference for the past and also provides the target grid for EPISODES. Instead of using the whole European domain an Alpine area was selected for ESD. Here you have access to a test data set, driven with ERA5 for the period 1991-2020, with the CERRA grid (Alpine section) as the target grid.