A large-scale condensation scheme, able to treat separately both atmospheric cloud condensate and precipitation content in a prognostic way, has been implemented and validated in Météo-France's operational global model, ARPEGE. The proposed scheme can be used for climate simulations and short-range numerical weather prediction, although it was originally designed for the future variational assimilation of cloud and precipitation observations.
The main originalities of the scheme, compared with other existing schemes having a similar moderate level of complexity, lie in the inclusion of a prognostic variable for rain and snow content, and in the use of a simple semi-Lagrangian treatment of the fall of precipitation. The calculations of large-scale condensation/evapora tion and cloud fraction are based on probability-density functions, and the parametrized microphysical processes that involve precipitation are autoconversion, collection, and evaporation/sublimation.
Various observations, which include satellite data from METEOSAT and from the Defense Meteorological Satellite Program's Special Sensor Microwave Imager, have been used for validating the cloud scheme within three-dimensional ARPEGE simulations at operational resolution for cases from the Fronts and Atlantic Storm-Track Experiment. These ARPEGE simulations have also been compared with 10 km runs obtained with the Met Office's Unified Model and with the French Méso-NH research model. In addition, cloud radar, ceilometer, and lidar observations from the Atmospheric Radiation Measurement project have been utilized for validating the simulation of a synoptic winter cloud system over the southern Great Plains in the USA. The behaviour of the scheme was then assessed at a coarser resolution, with a particular focus on the zonal-mean radiative budget of the earth and the zonal-mean cloud cover.
Finally, the question of the sensitivity of the results from the new scheme to various parameters has been addressed, including the time step and the specification of the fall velocities for rain and snow. Copyright © 2002 Royal Meteorological Society.