Whereas MCS evolution is a combination of multi-scale propagation and dissipation processes having contrary to conservative translation significant temporal and spatial variations.
From the practical view point the translation velocity can be considered as steady advection of radar reflectivity field. Suggested algorithms designed to estimate this velocity are classified in structure and object techniques. The calculating of spatial cross-correlation matrix between sequential gridded precipitation frames is the example of methods capturing mesoscale precipitation structure. The examples of object based techniques are cells identification and tracking algorithms. Translation velocity is obtained by averaging of individual convective cell's velocities. There were also developed several techniques which combine both object and structure approaches. Some of combined techniques apply a kind of thresholds (precipitation intensity, space cluster scales, keeping the shape convective patterns) providing selection or filtering of slowly and rapidly evolving precipitating elements (field fragments, frames). That helps to determine translation and, possibly, propagation/dissipation rate. For examination of various automatic and interactive translation velocity algorithms on its utility and restrictions there were estimated more then 300 MCS radar patterns.
As noted above some of object translation techniques can produce various cinematic characteristics of individual clusters such as the cluster center of mass velocity, severe convection core velocity, absolute maxima velocity etc. In this case magnitude and direction of ongoing cluster propagation are estimated from vectorial difference between one of representative cluster velocities and the translation velocity. Unfortunately, the propagation vector obtained by object vs. advection movement algorithms will describe only the mean evolution of one of preselected scale elements.
A range of MCS propagation forms and scales is available. Storm scale propagation (up to 2h and 50 km) appears as discrete multi- or continuous super-cell thunderstorm. A horizontal expansion of thunderstorm ensembles combines ongoing storm propagation and new objects development. The principal method to capture all of propagation and dissipation scales and forms simultaneously is to create MCS life-cycle matrix (LCM) that represent some kind of translating reference where precipitation histories of all pixels embedded in translating air mass are recorded by any way. One of robust, widely adopted in practice, LCM based method is 2D life-cycle maps overlaying of time-labeled radar frames i.e. more recent storms layer covers the previous one, as do it in Napoleon pastry cake. With some programming adjustment LCMs can be transformed into pseudo 3D life-cycle display where new propagation regions, ongoing and dissipation convection or convective/stratiform transformation are conventionally outlined. It is evident that radar derived propagation and dissipation rates are the remnants of previous intensification or damping of convective and stratiform updrafts and downdrafts. In order to detect the convective and mesoscale circulation one can be applied our unique Mean Wind Relative (MWR) reference analysis which is based on translation component subtracted from Doppler radial velocity. Guidance for key mesoscale inflow/outflow patterns is presented.
Based on translation velocity, storm movement, propagation and life-cycle methods a number of diagnostic and forecasting rules of thumb have been developed by authors for different types of MCS organization. Some of these rules are statistically validated for both extratropical and tropical mesoscale systems, and can be applied as probabilistic severe weather warning and nowcasting tool.