19 August 2013
Kilometre-scale versions of the Met Office Unified Model are a key component of our forecast system, providing vital information about small-scale weather phenomena such as severe storms. To improve our forecasts we need to understand how processes in the atmosphere operate on these scales. While we do lots of research in house we are increasingly collaborating with other organisations, including academia.
The development of convective storms can depend very much on small scale variations in the atmosphere. These variations in winds, temperature and humidity can be placed into two categories - those linked to the surface topography, such as hills and coastlines, and those originating from larger scale structures in the atmosphere, such as fronts. Forecast accuracy can depend on the way we represent small-scale processes like convection, and turbulence in the model.
Representation of these small scale features in our forecasting model is improved if the resolution is increased by having more detailed surface data and a higher resolution representation of the atmosphere itself. To understand and improve representation of these small scale features, we are working to improve key parts of the model, in particular describing processes in the boundary layer of the Earth's atmosphere, turbulence and cloud microphysics.
This involves rerunning the model at different resolutions to replicate and examine how the model forecasts weather events of particular interest. This helps to get a detailed understanding the model's behaviour and performance alongside a more statistical view of many similar events. In addition, there is an increasing use of an ensemble forecast approach - running the model many times with slightly different starting points or formulations - to gain insight into the physical mechanisms, the nature of the predictability of local weather and how to present kilometre-scale model output probabilistically.
We are also interested in improving model representation of other weather elements such as wind and fog. Future work will investigate the benefits of very high-resolution grids of around 100 m, forecasting for urban areas and coupling high-resolution atmospheric models with ocean, wave and hydrology models which may affect the representation of important feedbacks in the combined system.
The MetOffice@Reading works closely with colleagues in academia particularly with one of our academic partners, the Department of Meteorology at the University of Reading, as well as with colleagues in Exeter. The Meteorology Department at the University of Reading has a reputation for outstanding research and has a long history of collaborating with the Met Office.
We have been involved in several field projects with Reading and other universities for example CSIP (the Convective Storms Initiation Project - looking at convective storms) and are currently involved with several projects including DIAMET (DIAbatic influences on Mesoscale structures in ExTratropical storms - looking at raining weather systems), Advanced Climate Technology Urban Atmospheric Laboratory (ACTUAL - urban meteorology), the Convective Precipitation Experiment (COPE - based in South West England looking at the lifecycle of convective clouds) and the Dynamical and Microphysical Evolution of Convective Storms (DYMECS - a statistical look at the structures of showers.)
As part of DYMECS, the University of Reading radar group is looking at the properties of convective storms using the Chilbolton research radar. The Met Office benefits from access to data from the radar along with its interpretation by several world leading experts in radar meteorology. In turn, university scientists benefit from access to the Met Office Unified Model and the expertise in its use and interpretation within the MetOffice@Reading group. This aids them in the meteorological interpretation of their results and enables their work to have a more immediate impact on improving weather forecasts.
Work on convective-scale modelling within the Mesoscale Modelling group at Reading goes hand in hand with the Advanced Nowcasting Research group at Reading which is involved in high-resolution data assimilation and using novel observations. Last year, the group delivered the Nowcasting Demonstration Project to produce short-range forecasts every hour on a 1.5km grid covering much of England and Wales. This used advanced 4D-Var data assimilation to get the best possible start to each forecast. Its capabilities were showcased for the London 2012 Olympic Games and it proved to be very successful.