The resolution of operational Numerical Weather Prediction models has improved steadily over time, to the extent that horizontal resolution of about 10x10 km is now achievable. However, neither this or the next generation (which may improve resolution by a factor or 2 or 3) is sufficient to resolve small urban areas or, in practice, the variations that tend to occur within large urban areas. Until sufficient computer power becomes available, alternative methods are required for operational urban forecasting. Furthermore, for some applications, such as forecasting road surface conditions, it is questionable whether full 3D modelling will ever be appropriate.
While some features of urban meteorology require solution of the 3D problem, the broader features of the urban boundary layer can be described using a 1D model; indeed, most routinely applied air quality models require only 1D information. With this in mind, the U.K. Meteorological Office (UKMO) has developed a forecasting model based upon a 1D version of its NWP model (the Unified Model), forced by dynamical forcing from 3D NWP, modified in a simple way to take some account of local orography. Extensive modifications have been made to the surface exchange scheme to provide better urban (and rural) simulations, based upon a multiple tile approach driven by detailed land-use and orography data. Care has been taken to represent the 'urban' tile component realistically, both in terms of drag and surface heat and moisture exchange.
This paper describes the formulation of the model, in particular the modifications that have been made to deal with orography and urban areas which are often, in practice, an inhomogeneous mix of buildings and vegetation. Results from operational trials will be presented. The results will concentrate on the ability of the model to simulate realistic urban/rural contrasts, primarily in wind, temperature, humidity and stability that are absent from the driving NWP data. The results will be used to show that the model system can exhibit forecasting skill comparable to human forecasters.