The use of 100 m scale NWP models to understand differences between different measures of mixing height in a morning growing clear convective boundary layer over London

Abstract

AbstractAn important practical problem is to understand the depth to which pollutants will be mixed in the boundary layer over cities. This is particularly important shortly after sunrise on clear days when the mixed layer is often shallow and there may be large emissions from rush‐hour traffic. The 100 m scale NWP models currently under development have the potential to represent the boundary‐layer structure over cities on neighbourhood scales for air quality and other applications. Here we describe work with Met Office Unified Model (MetUM) configurations with grid lengths of 100, 50 and 25 m to represent the morning growth of a clear convective boundary layer over London compared to observations. All three models explicitly represent the convective overturning although it is not as well resolved at 100 m and this is a particular problem early in the day. It is found that in both the observations and in the models that the depth to which aerosol is mixed up into the model from the surface is lower than the turbulence mixed‐layer depth as measured from the vertical velocity variance. We present an analysis in the 25 m model that shows that the reason for the discrepancy is that the aerosol is transported upwards from the surface by convective plumes whereas further up there is wave motion which does not transport aerosol. The signature for these two layers can be seen both in the 25 m model and in the observations. We discuss the representation of these effects in the coarser‐resolution 100 and 55 m models and also possible mechanisms for the generation of the waves.