Abstract
Simple theoretical models of evaporation and sensitive high-resolution Doppler radars are used to study the precipitation and velocity structures resulting from evaporation at the base of layers of ice particles.
The models show that most ice particles will evaporate completely within a 2 km depth of fall, with the depth being dependent primarily upon particle size and relative humidity. Vertical gradients of reflectivity factor of 20–30 dBZ per 750 m are predicted for relative humidities <75% for a non-rimed collection of particles with a distribution of sizes. The evaporative cooling produces a destabilized layer, the depth and intensity of which are most dependent upon the relative humidity and precipitation characteristics. A dynamical model shows that downdrafts of at least 6 m s−1 penetrating to a depth of 2 km can be produced by evaporation. The intensity and penetration depth of the downdrafts depend primarily on the ambient lapse rate of temperature.
The magnitudes of vertical gradients of reflectivity factor predicted by the models were seen in radar observations. On one occasion the base of the precipitation layer lowered with time at 200 m h−1, in excellent agreement with the calculations. Updrafts and downdrafts of 1-3 m s−1 were observed in the region of intense vertical gradients of reflectivity by a vertically pointing Doppler radar. These motions perturbed the precipitation field such that the downdrafts were in downward extending appendages called “stalactites” and updrafts in the holes between.
Doppler radar observations are presented of stalactites and convective motion fields which were associated with a uniformly generated precipitation layer and with trails from cellular generators at cloud top. The motions associated with the trails appeared to be better organized and to have greater vertical extent than those associated with the uniformly generated layer. In the latter case, the largest scales varied from 500 m to 1.5 km with several preferred scales at smaller wavelengths which suggested a tendency toward a breakdown of the stalactite associated motion fields. In the former case the scales were more uniform at 600–900 m.
An evolution of a stalactite layer was observed in the case of the trails. Initially, when the trails entered the top of the dry layer with a small angle of incidence, the stalactites were simply extensions of the trails and the updrafts and downdrafts appeared to follow the stalactites. The vertical extent of these perturbations at this point was 1.5-2 km. The shear at the top of the layer increased with time and the trails became more horizontal. Precipitation tended to be carried out of the trails by downdrafts resulting from locally enhanced chilling by evaporation. This resulted in more vertical stalactites and more vertical updrafts and downdrafts. When the trials became horizontal, they were much more disuse and the whole layer was destabilized. However, the destabilization was not sufficient to result in significant convection. The increase in shear was due to a decrease in the mean horizontal wind at the top of the dry layer which in turn was deduced to be the result of a transfer of energy to the perturbations, probably making them more intense. It appears that this interaction between the perturbations and the mean flow resulted in more intense but shorter lived stalactite associated motion fields than otherwise would have occurred.
