Search

Wind energy resource assessment typically requires numerical modeling at fine resolutions, which is computationally expensive for multi‐year timescales. Increasingly, unsupervised machine learning techniques are used to identify representative weather patterns that can help simulate long‐term behavior. Here we develop a novel wind energy workflow that combines the weather patterns from unsupervised clustering with a numerical weather prediction model (WRF) to obtain efficient long‐term predictions of wind farm power and downstream wakes, which provide a good approximation to full WRF simulations at vastly reduced computational cost. We use ERA5 reanalysis data and compare clustering on low altitude pressure and wind velocity, a more relevant variable for wind resource assessment. We also compare varying domain sizes for the clustering. A WRF simulation is run at each cluster center and the results aggregated into a long‐term prediction using a novel post‐processing technique. We consider two case study regions and show that our long‐term predictions achieve good agreement with a year of WRF simulations in 2% of the computational time. Moreover, clustering over a Europe‐wide domain produces good agreement for predicting wind farm power output, but clustering over smaller domains is required for downstream wake predictions which agree with the year of WRF simulations. Our approach facilitates multi‐year predictions of power output and downstream farm wakes, by providing a fast, reliable, and flexible methodology applicable to any global region. Moreover, this constitutes the first tool to help mitigate effects of wind energy loss downstream of wind farms. Plain Language Summary: With the need to transition to renewable energy, wind energy is becoming increasingly important. However, the build‐out of wind farms must be carefully planned, not least because entire wind farms extracting energy from the atmosphere cause large scale wind velocity deficits downstream. Wind energy resource assessment typically relies on numerical models, but these are computationally expensive to run for the multi‐year predictions required for sustainable development. It is therefore common to use a smaller set of weather conditions that are representative of a longer period. Here, we develop such a workflow based on identifying characteristic weather patterns using a clustering technique from the machine learning literature. Based on the characteristic patterns identified, we perform a small number of short numerical model experiments, one for each weather pattern. By applying a novel post‐processing technique to combine the outputs from these experiments, we obtain predictions of wind farm power and downstream wakes over the original multi‐year period, using just 2% of the computation time required by a purely numerical modeling approach. Our workflow thus constitutes a fast, reliable, and flexible methodology that is applicable to any global region, which can be used for the sustainable development of wind energy. Key Points: Fast, reliable and flexible method for long‐term wind energy resource assessment using only six short‐term simulationsEvaluation of suitable variables and domain sizes for deriving weather patternsWeather prediction model is combined with clustering method for energy prediction [ABSTRACT FROM AUTHOR]