Heat, moisture, and momentum budgets of isolated deep midlatitude and tropical convective clouds as diagnosed from three-dimensional model output. Part II: Sensitivity to ice phase and small changes in ambient shear strengthand low-level moisture supply

This project uses a three-dimensional anelastic cloud model with a simple i ce phase parameterization to evaluate the feedback of isolated deep convect ive clouds over a horizontal scale comparable to one grid cell in typical m esoscale numerical weather prediction models. A more specific focus in this paper is the sensitivity of the feedback to modest changes in the initial vertical wind shear intensity and low-level moisture supply, as well as to the ice phase. Two parallel sets of comparative simulations are run for a quasi-steady sev ere Oklahoma supercell thunderstorm in strong vertical wind shear versus a weaker, less persistent, and narrower tropical Atlantic cumulonimbus with a slowly decaying and pulsating updraft in much weaker shear. The horizontal Reynolds averaging approach of Anthes is adopted to diagnose the budgets f or heat, moisture, and horizontal momentum. Several similarities and differ ences between the midlatitude and tropical control experiments were delinea ted in Part I. The main findings of the sensitivity study are described bel ow. The midlatitude storm evolves to maturity somewhat later (earlier) for stro nger (weaker) shear, though with little effect on peak updraft speed or bas ic storm structure. Quantitatively, the convection is more sensitive to moi sture supply changes, although basic structure is again preserved. With inc reased moisture the peak updraft speed increases by similar to 15% and the apparent heating and drying amplitudes by similar to 40%, and vice versa fo r the drier run. The vertical eddy fluxes are the main modulating factors. Without ice the peak updraft is similar to 10% weaker, though with no syste matic effect on downdraft speed, the later stages show gradual weakening in contrast to the quasi-steady control case, and the apparent heating and dr ying amplitudes are similar to 25% lower due to decreased condensation and also (for heat) the absence of any latent heat release by glaciation. The tropical cumulonimbus is for the most part less sensitive to shear inte nsity than its midlatitude counterpart. The pulsations are weaker in strong er shear and vice versa, but varying the shear has no systematic effect on either downdraft intensity or updraft evolution, affecting the budgets to a modest degree chiefly through the vertical eddy transport profiles. Omitti ng ice also affects the tropical cumulonimbus less than the midlatitude sup ercell storm, only slightly affecting updraft speed and the various budgets , especially for momentum. However, the tropical cumulonimbus is much more sensitive to moisture suppl y than the midlatitude supercell. The updraft is almost 25% weaker in the d ry run and similar to 45% stronger with slower decay and stronger pulsation s in the moist run, which also produces a deeper cloud with less downshear tilt and a more extensive anvil. Apparent heating and drying amplitudes are roughly doubled in the moist run and halved in the dry run, modulated main ly by condensation and vertical eddy transport amplitudes. The momentum bud get is also notably sensitive to moisture supply, especially in the moist v ariation, in which the upper-level horizontal pressure gradient force promo tes the enhanced anvil blowoff and reduced cloud tilt.