Response of Tradewind Cumuli to Large-Scale Processes

The two-dimensional slab-symmetric numerical cloud model used by Soong and Ogura (1973) for studying the evolution of an isolated cumulus cloud is extended to investigate the statistical properties of cumulus clouds which would be generated under a given large-scale forcing composed of the horizontal advection of temperature and water vapor mixing ratio, vertical velocity, sea surface temperature and radiative cooling. Random disturbances of small amplitude are introduced into the model at low levels to provide random motion for cloud formation. The model is applied to a case of suppressed weather conditions during BOMEX for the period 22–23 June 1969 when a nearly steady state prevailed. The composited temperature and mixing ratio profiles of these two days are used as initial conditions and the time-independent large-scale forcing terms estimated from the observations are applied to the model. The result of numerical integration shows that a number of small clouds start developing after 1 h. Some of them decay quickly, but some of them develop and reach the tradewind inversion. After a few hours of simulation, the vertical profiles of the horizontally averaged temperature and moisture are found to deviate only slightly from the observed profiles, indicating that the large-scale effect and the feedback effects of clouds on temperature and mixing ratio reach an equilibrium state. The three major components of the cloud feedback effect, i.e., condensation, evaporation and vertical fluxes associated with the clouds, are determined from the model output. The vertical profiles of vertical heat and moisture fluxes in the subcloud layer in the model are found to be in general agreement with the observations. Sensitivity tests of the model are made for different magnitudes of the large-scale vertical velocity. The most striking result is that the temperature and humidity in the cloud layer below the inversion do not change significantly in spite of a relatively large variation in height and intensity of the trade-wind inversion. This may indicate that cumulus clouds respond quickly to the large-scale forcing and adjust their own transport properties to maintain the observed large-scale thermodynamic fields whose variation has a much longer time scale. Sensitivity tests on varying sea surface temperature indicate that a ±1°C change in the sea surface temperature does not change the height of the inversion significantly during a 6 h simulation period. Another simulation shows that a tradewind inversion can develop rapidly from an initial sounding without an inversion if the large-scale downward motion is fairly large.