Wind speed vertical extrapolation model validation under uncertainty

As the wind energy sector evolves, accurate predictions of complex, nonlinear phenomena are becoming increasingly crucial. While rigorous model validation frameworks are common in high-stakes disciplines like defense and safety, their adoption in wind energy has been limited. This manuscript addresses this gap by introducing a comprehensive validation framework tailored for wind energy systems. This framework integrates both aleatoric (natural variability) and epistemic (knowledge-based) uncertainties. This dual consideration allows for a more nuanced understanding of model performance, especially under varying experimental conditions. The validation framework is applied to a meteorological measurement dataset using a logarithmic vertical extrapolation model. We present a set of numerical tests that validation metrics should satisfy and compare several metrics, providing insights into the relative strengths and applicability.