The New European Wind Atlas (NEWA project) will be the based on 30 years of simulations with the Weather Research and Forecasting (WRF) model and the additional microscale downscaling. The broader European domain is partitioned into a set of 10 partially overlapping tiles that reach a horizontal resolution of 3 km and are blended into a single domain. Additionally, a probabilistic wind atlas assessment seeks to quantify wind resource uncertainties by developing an ensemble of 1-year WRF

more »

... ons over the complete NEWA domain with different model setups (e.g. Figure 1). An initial ensemble of 47 members was performed for a few domains. This allowed first, for designing an optimal model configuration for the mesoscale wind atlas production run, and second, for identifying a group of model configurations that contributed with large spread to the probabilistic assessment. The parameters that were found to have the largest effect on model spread were the choice of land surface model, the PBL scheme or the surface layer scheme. We try to quantify model sensitivity with several metrics that allow to evaluate in space and time the maximum range of differences in the ensemble of simulations. Additionally, projection in the space of empirical orthogonal functions allows for an objective classification of wind regions and quantification of model spread. The reduced set of model configurations that contribute most to model spread were used to obtain the probabilistic ensemble over the complete NEWA domain and one representative year. The output of the ensemble simulations, i.e. ensemble mean and spread among other metrics, is used to create a map displaying the uncertainty in the wind resource estimate at each geographical location.