Observed structure of mesoscale convective systems and implications for large-scale heating

Mesoscale convective cloud systems that produce large amounts of rain in mid-latitudes and most of the rain in the tropics consist of a combination of convective and stratiform cloud. The convective regions contain numerous deep cells that are often but not always arranged in lines. The stratiform region is an outgrowth of the convective ensemble; it lies adjacent to the convective region and is seeded by ice particles detrained from convective towers. Sometimes it lies to the rear of a propagating convective line, while on other occasions it surrounds the convection. The heating of the large-scale environment by a mesoscale convective system is affected by both the convective and stratiform regions. Although processes such as melting and radiation are important, the net heating by a system is dominated by condensation and evaporation associated with vertical air motions. This paper reviews recent observational evidence regarding the mean vertical motion profiles in the convective and stratiform regions of mesoscale convective systems and the implications of these mean motions for the vertical distribution of heating of the large-scale environment. In both the convective and the stratiform regions, vertical motions have been determined by various techniques, including composites of rawinsonde and aircraft wind data, single- and dual-Doppler precipitation radar analyses, and wind-profiler observations. In stratiform regions, these data consistently show mean vertical velocity that is upward in the upper troposphere and downward in the lower troposphere. The level separating upward from downward motion is located from 0 to 2 km above the 0°C level, depending on location within the stratiform region. Diagnostic calculations show that these vertical motion profiles imply heating of the upper troposphere and cooling of lower levels by stratiform-region processes. Data on vertical motions in the convective regions are less consistent from case to case. These data sometimes indicate that the mean vertical velocity in convective regions is maximum in the lower troposphere. In other cases, the data indicate a maximum in the high troposphere. Diagnostic calculations show that heating profiles diagnosed from these two types of profile are quite different, the first having a maximum of heating in the lower troposphere, while the second has a maximum in the middle troposphere. Although it is difficult to determine whether or not the differences in estimates arise from different types of observations, analysis methods or sampling strategies, it seems likely that they stem from differences in the large-scale environment of the different mesoscale systems. The ubiquitous occurrence of stratiform regions in mesoscale convective systems and hurricanes together with their consistent heating profiles, which systematically concentrate heating in upper levels while cooling lower levels, are a major consideration in evaluating the interaction of mesoscale systems with the large-scale environment. However, the consistency of the stratiform profiles from case to case indicates that the variability of net (convective plus stratiform) heating profiles from case to case lies primarily in the variation of the convective-region profiles from one case to another. It is suggested that future work should focus on the convective-region vertical profiles of vertical motion and heating and on the large-scale environmental factors that may control the variation of these profiles from case to case.