Predictive Bias Potential in ERA5 and MERRA-2 Reanalysis Datasets in Wind Energy Resource Assessment

Presentation Description: Public reanalysis datasets are commonly used to correlate short on-site data measurement campaigns to long-term trends. Of these datasets, ERA5 and MERRA-2 are most commonly used globally. Geographical trends have emerged over the past 10 years that indicate the potential for divergence between the two datasets, which may artificially bias resulting long-term wind correlation performed as part of a preconstruction energy assessment. This study will examine global cases to indicate the relative predictive bias with each dataset and attempt to draw conclusions about spatial behaviors.