It is no exaggeration to say that GATE observations were a turning point in the field. No longer would it be possible to seriously propose that tropical convection has a scale separation from the synoptic scale; sooner or later the ubiquitous presence of organized convection on the mesoscale would have to be dealt with.
In terms of fundamental knowledge of how large, meso, and small scale processes interact, GATE benefitted from long, expert scientific planning, from ample resources, and from the relative simplicity of its large scale environment. The observing array "caught" weather systems regularly, because the ITCZ was tightly constrained between colder ocean areas north and south, and because the easterly waves affecting the region were strong and regular in their passage from west Africa into the Atlantic. (By comparison, TOGA COARE had no such spatial gradients in SST, and the dominant disturbance was the 40-day intraseasonal oscillation.) The observing array in GATE enabled direct and rather accurate estimates of Q1 and Q2, the effects of organized convective systems on the larger-scale momentum fields, among other things. In recent years, it has become common for those using cloud resolving models to compare simulations against GATE datasets as a step toward both validation of their models, and for evaluating cumulus parameterization schemes.
The talk will briefly review the main accomplishments before, during, as well as after the GATE experiment, and their implications on the problem of cumulus parameterization.