The flight tracks were aligned with the mean depth of the cloud layer, in order to interpret along-track changes as orographically-induced changes, assuming that the large-scale flow and stability are in steady state. On all days the upstream flow is stably stratified (Ns^2>0 at all levels), but small variations in this stability cause a significant difference in the cloud vertical structure.
On some days the flow was in quasi-steady state and the precipitation production entirely confined to the upwind side of the mountain ridge. Snowfall was generally light in these cases, and stability low due to cold air aloft. Updrafts and echo outlines appeared remarkably cumuliform. Supercooled water was present in the updrafts, even at temperatures around -15 to -20ºC, with some drops around 50 micron in diameter. Glaciation tended to occur rapidly on the upslope side, and in the cases examined here very little liquid water was present in updrafts closer to the ridge crest. On the downslope side snow particles were advected 10-20 km before they were entirely sublimated.
On other days a deep precipitation system was advected into the region, but even then shallow orographically-generated snow could generally be clearly separated. Snowfall was generally heavier and more widespread in these cases, and N2 higher. Echoes were deeper and more stratified, and updrafts weaker, with evidence of vertically-propagating gravity waves.
In this talk an example of each type is illustrated, and then some simple calculations of advective time scale, observed and theoretically expected precipitation growth/decay timescales are presented. Inferences are drawn about precipitation efficiency and the relative distribution of snowfall on upslope and downslope sides of the ridge.