The urban wind island from a three-dimensional perspective

The Urban Wind Island (UWI), a small but persistent positive mean boundary-layer wind anomaly over the city as a whole, has previously been revealed using a simplified conceptual model of the convective atmospheric boundary layer. This study extends the UWI research into less idealised cases by using the three-dimensional WRF mesoscale model for Amsterdam (The Netherlands) and its surroundings, at 500 m grid spacing. Two summers of forecast results for in total 173 days are used to identify whether the UWI persists in a three-dimensional modelling environment, and which conditions are optimal for its formation and persistence. In order to focus only on wind modified by surface processes, large-scale influences which modify wind speed, such as frontal passages, are identified and eliminated from the dataset. We then find that a positive UWI is present roughly half the time, with an order of magnitude that is similar to the previous work ($\sim$ 0.2–0.5 ${\mathrm{ms}}^{-1}$). In addition we find an evening UWI that is caused by the delayed onset of the transition from an unstable to a stable or a neutral boundary layer in the urban area, while the rural area is already stable and calm.