Experiences with a 100m versions of the Unified Model over an urban area

With the recent increases in resolution operational weather forecast models,for example the UKV, the Met Office UK 1.5km configuration of the Unified Model (UM), can start to resolve urban areas and give Urban Heat Island signatures. At the same time very high resolution versions of the UM (down to 100m gridlength) are beginning used for various research activities (i.e. looking at cold pooling in valleys and convection). In this paper we look at some preliminary experiments with 100m gridlength UM configurations over London to compare with observations from the ClearfLo and ACTUAL observational campaigns. These experiments have been carried out using a similar configuration of the model to the 1.5km model (including 100m orography and land surface data).

Two clear sky daytime cases with convective boundary layers will be presented. As would be expected the model surface temperature responds to the detailed surface information. It is shown that, when large enough domains are used, the 100m model explicity represents overturning in the boundary layer in quantitative agreement with lidar vertical velocity measurements. An analysis of the resolved and unresolved fluxes in these cases will be presented and compared to the representation in a 1.5km and 500m model. Other comparisons of small scale features with observations will also be presented.