In this study, CTEI is extensively investigated using large-eddy simulations (LESs). Experiments with BR are designed so as not to have any source of turbulent kinetic energy production except for entrainment due to evaporative cooling. A total of 147 (49x3) LESs are performed. Three marine stratocumulus cases (FIRE, ASTEX, DYCOMS-II) are also simulated for comparison with the BR experiments.
The results of the BR experiments do show the hypothesized positive feedback, and the Randall-Deardorff stability criterion is confirmed. When CTEI takes place in the BR experiments, entrainment develops spontaneously and leads to cloud dissipation. Rapid cloud dissipation, i.e., within several hours, is simulated in some cases, but some clouds subject to CTEI do not fully evaporate in a ten-hour simulation.
In further experiments, a hypothesized dependence of the results on the liquid water mixing ratio is confirmed. As expected, with a typical stratocumulus liquid water mixing ratio, the CTEI feedback is weak.
A transition from stable to unstable conditions, followed by cloud break-up, is simulated for one observationally based stratocumulus case. In simulations based on other observationally based cases, breakup does not occur even though the criterion for instability is satisfied. We conclude that cloud dissipation can be prevented by sufficiently strong cloud-building processes.