A Copula-Based Bivariate Bias Correction Method for Super-Resolved Physics-Based Wind Speed Data

Inverter-based resources (IBRs) data retrieved from physics-based models, such as wind speed, generally suffer from statistical biases in distribution. These biases result from various factors such as simplifications in the model physics, inaccuracies in input data or uncertainties in model parameterizations. Thus, we propose a copula-based bivariate bias correction method for wind speed from physics-based models to improve their reliability and enhance our understanding of downscaled physics-based model data. The proposed statistical method is applied on hourly wind speed data at 10m and 100m heights in Argonne, IL. We compare the method with the commonly applied benchmark models in the literature. The proposed copula-based method outperforms benchmark models and reduces the bias up to 20%.