While current models for wind energy forecasting are often tailored to individual wind turbines (WTs), excluding smaller or planned installations, the new application follows a unified and scalable approach. Using a machine learning model, forecasts are generated for all European WTs according to the same principle, regardless of whether real production data are available or parameters have been custom-defined. This enables, for the first time, a flexible and cross-turbine assessment of wind energy potential.

Another key focus is on user accessibility and openness. Unlike many existing commercial systems, this platform is also designed for smaller stakeholders and users without deep technical expertise. The results are easy to interpret and include transparent error metrics. In addition, the web application integrates a crowdsourcing function that allows users to share their own production data. Over time, this will create a growing European database that continuously improves forecast quality and supports the broader advancement of renewable energy.

Open Access to the Webapp

On the homepage of the web application, a map displays the wind turbines in Europe (from zoom level 11 onwards). Users can select a combination of WT and time step of interest, for which the forecast is then generated. The forecast for the next six days is shown when “Show Entire Forecast” is clicked. Users can also modify the parameters of the selected turbines, download the forecast, and share their own production data to improve the model. A more detailed description, as well as a link to the GitHub repository, is also available within the web application. An initial loading time of approximately twelve hours is normal, as it is required to download weather data in the background.