Impact of model resolution on the representation of the wind speed field: An example from the United Kingdom

The United Kingdom (UK) is experiencing rapid growth in wind farm development, primarily in the south as well as in offshore regions. As such, reliable long‐term wind statistics are integral in planning future development. Given the scarcity of in situ data, atmospheric reanalyses are beginning to be used for this purpose. However, most reanalyses have resolutions of ∼50 km or lower – scales too coarse to capture many topographic and coastal effects. To address this, dynamical downscaling has been used to obtain higher‐resolution data. However, it is unclear how the downscaling process impacts the temporal and spatial representation of the wind field. Here a set of reanalysis and analysis datasets with a common lineage and with horizontal resolutions ranging from ∼75 to ∼9 km are used to investigate the impact of model resolution on the representation of the spatial and temporal variability in the wind field in the UK. To assess this impact, the decorrelation length and temporal scale – DCLS and DCTS, respectively – are used to characterize spatial and temporal variability. The results can be classified into two categories – resolution‐dependent and resolution‐independent features. Resolution‐dependent results suggest that even resolutions of ∼30 km do not capture the wind field's variability, especially in coastal and mountainous regions. However, resolution‐independent results suggest that there are regions, such as central Ireland, over the oceans, and southeastern England, where the 30 km resolution captures the wind fields' variability. Results also support the historic choice of wind farm development in the UK, namely, the southern/southwestern tip of England as well as off the southern shore. The elevation (m) for the (a) ERA‐I, 75 km; (b) eERA5, 62 km; (c) ERA5, 31 km, and (d) ECOA, 9 km datasets.