To produce new parametrisation schemes that will enable convective cloud processes to be represented more accurately within atmospheric general circulation models used for weather and climate prediction.


Very high resolution cloud-resolving numerical models will be validated against observational data and against
other models. They will be used to provide detailed 3-dimensional data, concerning both the evolution of
convective cloud systems and their ensemble characteristics, for evaluating and improving parametrisation
schemes for general circulation models.
For the observational case studies at least four classes of convective cloud system will be selected from the
following: North Atlantic stratocumulus, North Sea cumulus, convective clouds during a cold-air outbreak over
the extratropical ocean, continental European thunderstorms, precipitating convective cloud systems over the
tropical western Pacific.The initial conditions, the large-scale forcing and the diagnostics to be used in
validating the cloud-resolving models will be defined as precisely as possible.
Each cloud-resolving modelling group will run the same set of basic experiments to provide cloud-system
realisations for two or more of the observational cases selected, grouped into two work packages:
non-precipitating convective clouds (stratocumulus and fair-weather cumulus), precipitating convective clouds
(cumulonimbus). Results will be exchanged amongst the participants, and each cloud-resolving modelling group
will work closely with at least one of the groups using general circulation models.
A joint workshop will assess the results obtained from the
cloud-resolving models and will draw conclusions
concerning the robustness of individual models and the error bars applicable to their results. Common features
diagnosed from the models will include: condensation, precipitation, radiation, and the transports of heat,
moisture and momentum. Some preliminary comparisons will be made of the cloud-model output with the
characteristics of parametrisations used in single-column versions of general circulation models; deficiencies will
be highlighted and areas for improvement proposed.
The general circulation modelling groups, working closely with the cloud-scale modelling groups, will conduct
a more detailed assessment of the cloud-system realisations and develop and test modified versions of their
parametrisation schemes.
Finally the bulk of the cloud-resolving model data will be systematically compared against the single-column
general circulation model results to identify remaining shortcomings.