The clouds and radiation group is responsible for improving the representation of solar and infrared (thermal) radiation as well as prediction of cloud and precipitation in both forecast and climate models.
The most important issue faced by the team is the representation of clouds on scales smaller than the model resolution, where unresolved physical processes and variations in conditions must be taken into account. This forms a significant amount of the work undertaken by the group.

Cloud microphysics and radiation parametrizations used at the Met Office are based firmly on observational evidence and physical theory. For example, cloud physics parametrizations are developed using aircraft, satellite, balloon and radar observations in conjunction with the latest theoretical ideas.
Simplified modelling tools, such as the single column model, are also used for parametrization development and help determine the numerical stability of parametrizations. The group works to draw together the varied information from these data sources to produce robust and simple representations of the relevant processes.

The important task of evaluation of the model predictions for cloud and radiation forms a large part of the external collaboration with non-Met Office groups. A number of novel ways to infer cloud properties from satellite or radar and lidar data can be used to evaluate the predicted structure of clouds. Similarly, data from the model may be used to simulate observations, such as radiances. Evaluation of the operational model is also conducted via regular consultation with forecasters.

The representation of radiation in the model is evaluated against detailed line-by-line radiation codes. The full solution of the three-dimensional radiative transfer problem, even in the absence of clouds, is computationally expensive and too slow for an atmospheric model to use. This means that approximation techniques need to be developed in order to speed up the scheme. Such techniques include the assumption that radiation can be modelled as vertical fluxes in model columns, across a number of frequency bands, rather than individual lines.

Find out more about the large-scale cloud and precipitation parametrization.

Find out more about the radiation parametrization.