Energy Meteorology

Wind and Solar Power Prediction

In recent years, the research field of energy meteorology has become a major focal point of interest. The ever growing share of renewable energies calls for increasingly reliable and comprehensive power forecasts, mainly to secure the reliability of the electrical grid and realisations on the stock market. The highest potential for improvement of such forecasts lies in the Numerical Weather Prediction.

Within two project collaborations EoCoE-II and HAF, the predictability of wind and clouds in the day-ahead forecast is investigated. Probabilistic weather predictions in form of ensembles are developed with the Weather Research and Forecasting Model (WRF). The distinctive feature here is the ensemble size. While operational weather centres are restricted to approximately 50 members, the one of the flagship code of EoCoE-II, ESIAS-met, can perform exascale ensemble forecasting, which can be extended to 1000 ensemble members.

Cloud scenarios and characteristics are best observed by satellites. The use of satellite information in prognostic models, resulting in the improvement of the short-term prediction of cloud movement, is an (as yet) unresolved issue. The potential benefit of combining the exascale ensemble forecast with satellite observations is analysed in the field on nonlinear data assimilation (Particle Filter). Furthermore, the use of Big Data Analytics methods, e.g. supervised machine learning, is investigated. The demanding computational effort is overcome by utilising the supercomputers at JSC.