Balanced mesoscale motion and stratified turbulence forced by convection

Numerical experiments are carried out with a large-eddy simulation model to investigate the production of stratified turbulence and gravity wave energy in the atmospheric mesoscale by deep convection. Relatively long integrations (typically about 2.5 days) are carried out, so that statistical equilibrium might be achieved for an atmosphere subject to fixed cooling and a ground surface maintained at constant temperature. The main simulations may be considered as representing convective activity in a cold airstream passing over warm seas. The transfer of updraught kinetic energy into quasi-horizontal rotational energy, and its subsequent cascade to larger scales, is examined in the context of the production of balanced motion and stratified turbulence theory. Experiments are carried out to examine the effects of the Coriolis force, a wave-damping stratosphere, boundary-layer vertical wind shear, and precipitation. An inverse mesoscale energy cascade is observed in experiments both with and without background rotation (the Coriolis effect), although weaker in the latter case. Mean boundary-layer vorticity is found to be the principal source of the horizontal rotational energy created by deep convection, although the dipole character of the vorticity thereby produced frustrates the inverse energy cascade. At the level of convective detrainment, potential-vorticity anomalies have a small monopolar element in the rotating case and this implies a direct forcing of energy at large scales. It also promotes a more efficient inverse energy cascade through long-range influence. The inverse energy cascade arises primarily because of the quasi-two-dimensional, divergence-free nature of the flow at scales larger than the convective forcing, which is maintained by the stable stratification. Even in the absence of background rotation balanced flow arises, especially in the upper troposphere and in the decay phase of the simulations when the surface fluxes and atmospheric cooling driving the convection are removed. In these cases potential-vorticity inversion can be quite successful in determining the other field variables, with contributions from both geostrophic and cyclostrophic balance. Suppression of precipitation (i.e. removing the liquid-water phase) increases the net convective mass transfer through the elimination of convective downdraughts, which leads to a higher efficiency of balanced-flow energy production and upscale energy transfer. Intense 'meso-vortices', with warm core structure, form spontaneously after about 2.5 days integration in these runs.