Intended audience
Academy members from the power systems and meteorological areas, forecasting services providers, renewable plant operators and market traders, and widely TSOs, DSOs and regulators.
Key messages
- RES are sensitive to high-frequency and very localized variations of the atmospheric flow (wind guts, cloud shadows, fog occcurrence etc.)
- Refining parameterizations used in NWP models is the first step to improve RES prediction
- Getting rid of these parameterizations, for instance using Large-Eddy Simulations (LES), is a promising venue because it allows to resolve most of these small-scale fluctuations
- The spectacular growth in computing power using GPUs now enable LES based weather prediction below 100m resolution. This is fine enough to resolve the effects of wind turbines, low clouds and turbulence.
- Using ensemble simulations translates atmospheric state inherent unpredictability into a probabilistic view
- The spectacular growth in computing power using GPUs now enable LES based weather prediction below 100m resolution. This is fine enough to resolve the effects of wind turbines, low clouds and turbulence.
- A fourth webinar scheduled in April will complete this webinar on Advanced weather forecasting for RES applications and present how observations data can be used, together with measurements from RES plants, as input to dedicated RES forecasting models to predict the power output of these plants.
