Laserfiche WebLink
<br />37 <br /> <br />rapid increase of the ice crystal concentration near cloud edge argues against <br /> <br />any multiplication process, as does the relative uniformity in concentration <br /> <br />beyond that edge. In particular, the Hallett-Mossop mechanism was not <br /> <br />operating in the region where the ice crystal concentration was seen to <br /> <br />increase, because in that region the droplet sizes were still smaller <br /> <br />than 23 ~m and there were no riming objects. Furthermore, where the high ice <br /> <br />concentration originated the temperature was always colder than -8C. Multi- <br /> <br />plication through droplet shattering and splintering requires rather large <br /> <br />droplets (Hobbs and Alkez~.;reeny, 1968) \olhich were not present in the regions <br /> <br />where the ice crystal increase was observed. Another possible multiplication <br /> <br />mechanism is through crystal fragmentation (Hobbs and Farber, 1972; Vardiman <br /> <br />and Grant, 19?4). HO\olever, high concentrations are already present while <br /> <br />the crystals are quite small. Fragmentation may occur in these clouds, but <br /> <br />does not seem a likely explanation for the behavior \ole observed in cases such <br /> <br />as 29 December because a continuous process should occur rather than a rapid <br /> <br />process only near the cloud edge. <br /> <br />It is of interest to consider which nucleation modes could account for <br /> <br />the observed ice crystal development. Deposition nucleation should depend <br /> <br />primarily on ice supersaturation, and it would be surprising if true deposi- <br /> <br />tion nucleation sho\oled a sudden increase in activity at water saturation. <br /> <br />Also, deposition nucleation is probably the easiest to measure, and the <br /> <br />filter samples we collected indicated concentrations far too low to account <br /> <br />for the observed ice crystal concentrations. Immersion freezing is a <br /> <br />possibility, as is condensation-freezing. Conventional estimates of the <br /> <br />contact nucleation rate suggest that it would be too slow, but these <br /> <br />estimates may be considerably in error for this particular situation (as <br /> <br />discussed in Cooper and Vali, 1976). A condensation-freezing process seems <br /> <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I <br />I' <br />