How does a single-cell shower differ from a multi-cell thunderstorm, or even a 'supercell'?
All are formed within an unstable environment (see "Stable and unstable air masses", and all require the following to be in place: (i) Instability through a reasonable depth of the troposphere; preferably (but NOT necessarily) extending above the freezing level; (ii) sufficient moisture to sustain the cloud-building process - medium level dryness will often kill shower formation unless low-level inflow of moisture is substantial; (iii) a trigger action - i.e. something to kick the whole process into life by lifting the parcel that goes on to grow into a moderate depth cumulus cloud, or a well-developed 'supercell' complex.
Once these conditions are met, then consideration of things like shear, CAPE, helicity, etc., are needed as follows:- (for definitions, see the Glossary, and in particular for helicity, see "What is helicity?")
Single-cell showers: the 'classic' growth/decay model of a Cumulus cloud , whereby a single moist convective cell develops in an airmass that is moderately unstable (CAPE values ~ 100 J/kg), provided of course that there is sufficient depth of moisture and there is an initial trigger action. When the updraught and the precipitation downdraught occupy virtually the same atmospheric column (there is little or no vertical relative wind shear to tilt the cloud), the downdraught quickly swamps the updraught - the shower soon decays (perhaps lasting only a matter of minutes - the cloud would last longer though), yielding small amounts of rain/snow. However, when there is a change of wind speed with height (but little directional change), the updraught column is tilted forward, and the resultant precipitation downdraught is held clear of the downdraught, allowing greater development and moderate intensity showers occur. The cold downdraught though soon swamps the inflow of surface air, cutting off the updraught and the shower decays after about 20 to 30 minutes. These events would be typical of Polar Maritime airmasses.
Multi-cell thunderstorms: Whenever wind shear is present in an unstable atmosphere, the developing convective clouds will be tilted to a greater or lesser extent. As seen above (single-cell showers), when only the wind speed changes, then short-lived, non-propagating showers are produced. However, given *both* change of wind speed and direction with height (relative to the storm motion), and sufficiently high CAPE (> ~ 250 J/kg), then the precipitation downdraught is skewed well to the side of the storm updraught, and does not interfere with it - allowing that storm cell to develop its full potential - other necessary factors (e.g. sufficient moisture) being in place. In addition, the downdraught will hit the surface and spread horizontally as a cold density current (gust front). At some point, this will meet the low-level inflow, and a new 'daughter' cell (see the Glossary) may be initiated which may grow into a full-scale storm cell in its own right. This usually (but not always) occurs to the right of the cloud motion, and the whole storm complex appears then to move to the right .. in fact the daughter cells take over from each successive parent to produce this effect. Large Cumulonimbus (Cb) clouds are produced with these processes; each cell lasting at least half-an-hour, and depending upon external forcing agents (e.g. coastal convergence, synoptic troughs, orographic lifting), the storm complexes may last for several hours.
Supercell thunderstorms: Although in some respects, 'supercell' storms can be regarded as a special (and intense) case of the multi-cell storm, there are important differences as well. The environment is still sheared in the vertical, indeed markedly so in the lower layers, and daughter cells are produced. However, a key distinguishing element between supercell and non-supercell events is the presence of a rotating updraught. CAPE values for supercell events will typically be ~1000 J/kg or more, and helicity will also be high - hence the tendency to rotation of the storm complex, and its individual elements. It is thought unlikely, for example, that giant hail would be possible unless the updraught were enhanced by the presence of rotation within the system.
The overall storm motion may be quite small (e.g. Wokingham storm), with the spawned cells forming close to the base of the parent cloud - often several daughter cells coinciding - these form an almost self-perpetuating system lasting several hours. These mechanisms produce the most severe late spring / summertime thunderstorms with local intense rainfall leading to flooding, plus occurrence of hail, possible tornadoes etc. (Note however that slow displacement of such storms should not be assumed - results from North America show displacements in excess of 40 knots / 74 km/hr ). Potential instability at medium levels (circa 500 hPa / 5 to 6km) is also required, as is an initial inhibiting factor (warm / dry air capping surface based instability) to allow the 'loaded gun' effect to build up.