User-oriented evaluation framework for regional reanalyses

Global reanalyses are known for more than 25 years and are widely used for monitoring climate variability and change, for academic purposes and commercial applications. Meanwhile, they are a fundamental part of operational climate services such as the Copernicus Climate Change Service (C3S). Recently, also regional reanalyses that provide higher spatial and temporal resolution became popular. Examples of exploitation of regional reanalyses include applications in the energy sector. Of special interest are variables such as wind speed at hub heights of modern wind turbines (offshore and onshore), near surface temperature and surface solar irradiance. Moreover, regional reanalyses are potentially useful for climate applications dealing with extremes such as wind storms, regional temperature anomalies including heat waves and droughts. Here we show results from an evaluation approach that is used for monitoring the production process of the regional reanalysis COSMO-REA6 Generation 2 (R6G2), currently produced by Germany’s national meteorological Service (Deutscher Wetterdienst, DWD). Numerous datasets are used as observational reference, based either on in-situ measurements from stations and towers or satellite-borne remote sensing technology. Different products such as ERA5, COSMO-REA6 and HoKliSim-De are used as benchmark. The technical infrastructure for the standardised evaluation and monitoring is provided by the MAVIS system which builds on Freva (Free Evaluation System Framework, https://doi.org/10.5334/jors.253). One user requirement is the availability of uncertainty information. There has been evidence that ensemble information is required for the representation of individual extreme events such as wind storms. The capability of the MAVIS system and requirements for application of the evaluation framework to ensemble-based regional reanalyses will be discussed.