Predictability is a concept that is discussed fairly regularly in mesoscale meteorology, but for which detailed and exhaustive studies are quite limited in number. Most discussions focus on one of several specific mesoscale phenomena that are believed to be more predictable than others, perhaps because of the influences of complex terrain or the highly helical nature of the flow, and thus forecasts of this phenomenon are argued to be predictable out to longer time periods. However, the quantitative studies needed to evaluate these arguments are lacking, even though they are needed greatly as meteorologists are asked to provide more and more accurate projections of forecast improvements to gain resources and demonstrate economic value. Predictability is arguably one of the most overlooked aspects of mesoscale meteorology and a strong case for the need of more and better studies on mesoscale predictability will be one aspect of this talk.

The history of mesoscale predictability also will be summarized and critically evaluated. This begins with the original studies of limited area models that suggested a greater inherent predictability in these models than global models. The reasons for this apparent improvement in predictability were attributed to the greater representation of terrain and the diurnal heating cycle. However, later studies illustrated convincingly that this improved predictability was due to the lateral boundaries being constrained, which limited error growth within the limited area domain. More recent studies on mesoscale predictability also will be examined, and some of the authors own recent unpublished work on the predictability of mesoscale convective systems will be shown as an example of the type of work that can be done and as a stimulus for discussion and (hopefully) vigorous debate.