TURBULENT TRANSPORT IN FIELDS OF WARM CUMULUS CLOUDS

Simulations of fields of warm (ice-free) cumulus clouds are made using a large-eddy model based on observational soundings. Comparisons betw een numerical and experimental results show good agreement. The cumulu s clouds produced by the model are similar to those observed, with clo ud base, cloud top and cloud scales all in good agreement. At cloud to p, vertical velocities are found to be +/-5 m s(-1) and liquid-water m ixing ratios have a maximum of 1.6 g kg(-1), which is consistent with the observations. By considering fractional cloud cover and averaging over time, the profiles of the turbulent kinetic energy and fluxes ave raged over the horizontal level and in-cloud are derived. Much of the turbulent energy and transport in the boundary layer is found to be pr oduced in cloud. The dominant regions of turbulent kinetic energy resu lt from the strong surface heating and the two in-cloud buoyancy sourc es of latent heating through condensation at cloud base and evaporativ e cooling following entrainment at cloud top. This is supported by the upward transport of energy at mid-cloud levels and downward at cloud- top. The terms in the turbulent kinetic-energy budget are derived, and once again buoyant production is found to be dominant in both the clo ud and the horizontal averages. The vertical turbulent fluxes of heat, moisture and liquid are analysed and found to vary in accordance with the turbulent processes driving the boundary layer. The penetration o f the clouds above the main inversion shows that the clouds were effec tive in coupling the surface to the base of the free troposphere by tr ansporting the heat and moisture upwards. Overall, the kinetic-energy profiles and turbulent fluxes in the cumulus case-study, both averaged over the layer and the clouds, show good agreement with the observati ons. This demonstrates the validity of the large-eddy simulation model as a tool for studying cumulus clouds at greater vertical resolution than is possible by current observations, and provides a basis for der iving more accurate parametrization schemes of convective cloud-capped boundary layers.