First, the interplay of various forcing mechanisms in mountain cumulus formation is studied. For example, it is shown how the timing of cumulus initiation upwind from a ridge is controlled by the growth of the convective boundary layer. Moisture is transported upwards by CBL turbulence and then dynamically lifted by flow over the ridge. On the downstream side, the convergence of the thermally induced upslope flow with the large-scale flow is shown to be essential. This confirms the results of earlier studies of the 'lee-side convergence' mechanism of mountain cumulus formation.
Second, the performance of the model as a real-time forecasting tool is evaluated. Initial and boundary conditions are taken from MAPS (Mesoscale Analysis and Prediction System) analyses and forecasts. Predicted cumulus cloud patterns and onset times are compared to GOES data and visual observations.