The sensitivity of a stratocumulus transition:: Model simulations of the ASTEX first Lagrangian

In this study, a one-dimensional ensemble-average model is used to simulate the Atlantic Stratocumulus Transition Experiment first Lagrangian, where the same airmass was followed from the subtropical high pressure region en route towards the trade wind region. The airmass experiences increasing sea-surface temperature and a change from subsidence to weak ascent on its way south. This causes the marine boundary layer (MBL) to grow and the cloud deck to change from a solid stratocumulus deck to more broken stratocumulus clouds with cumulus clouds developing beneath, and reaching up into the stratocumulus clouds. A control run is analyzed and compared in detail with the observations. Both a statistical evaluation and a more subjective evaluation are performed, where both establish confidence in the model performance. The model captures the MBL growth and the cloud liquid water, as well as the drizzle flux, is well predicted by the model. A sensitivity study was performed with the objective of examining the MBL and the cloud response to external and internal 'forces'. The results show that, if drizzle formation is not allowed, unrealistically high cloud liquid water mixing ratios are predicted. Even though the drizzle flux is very small, it is still important for the water budget of the MBL and for the boundary-layer dynamics. We also found that the sea-surface temperature increase is more important for the increasing cloud top height than the synoptic-scale divergence fields. However, the synoptic-scale subsidence is crucial during the first day, when the sea-surface temperature was constant, in keeping the cloud top at a constant height. Drizzle evaporation below the cloud base seems to be important for below-cloud condensation. The drizzle predictions are significantly altered when the prescribed cloud droplet and/or drizzle drop numbers are altered.