BTW Darryl, another example of how running a model which utilizes the
"equations of motion" at a high resolution (i.e. higher than used by
government models) can produce more accurate thermal forecasts occurs
for the Central Valley (of California). Forecasts based on RUC/NAM
very often over-predict the thermal tops there, as many have noticed
probably including yourself. But RASP-predicted thermal tops are
typically much lower over both the northern Central Valley (Williams
RASP) and southern Central Valley (Panoche RASP). The reason is that
the 4km resolution "equations of motion" is better able to simulate
the mountain-valley circulation pattern, i.e. the upward motion over
the surrounding mountains and downward vertical motion over the
valley, than is the 12km resolution RUC/NAM, and that downward motion
over the valley "caps" thermals at a lower height than would otherwise
occur.

Jack