The electrification of hail

Abstract

Reynolds, Brook and Gourley (1957) derived a value of 5 x 10(^-4) e.s.u. for the charge separated per crystal collision when a simulated hailstone rotated in a cloud consisting of ice crystals together with supercooled water droplets of some 5μ size. From this estimate they concluded that the electrification of thunderclouds could be explained in terras of the electrification of hail by impacting ice crystals. Latham and Mason (196I B) performed similar experiments in the absence of water droplets and obtained an estimate which was 5 orders of magnitude less than Reynolds' value. They also measured the electrification of an iced probe by supercooled water droplets shattering on it and derived a value of 4 x 10(^-6) e.s.u. for the mean charge separated per droplet diameter range 40 - 100 μ. They concluded from these experiments that the charge separated by Ice crystal impacts was not sufficient to explain thunderstorm electrification. They proposed instead that the droplet shattering mechanism offered a satisfactory explanation of the magnitude of the charge separated in thunderclouds. It was the purpose of this thesis to investigate the results of these two authorities and in particular to seek an explanation for the large discrepancy between their results on ice crystal impactions. In this laboratory similar experiments to Reynolds, Brook and Gourley were performed, and it was concluded that the results could be explained qualitatively in terms of the temperature gradient theory, but quantitatively the charge separated was larger than predicted by the theory. Experiments similar to those of Latham and Mason on crystal impacts were performed. The quantity of charge separated per crystal collision and how It depended on the sign and magnitude of the measured temperature difference between the iced probe and the crystals, the presence of impurities in the probe, and the impact velocity of the ice crystals was determined. An estimate of 2.5 X 10(^-7) e.s.u. was obtained for an impact velocity of 20 m sec (^-1) and measured temperature difference of 10 C. This was 50 times greater than the value found by Latham and Mason but it was shown that the two values could be reconciled. It was shown that they could also be reconciled with the previous value of 5 x 10(^-6) e.s.u. It was further observed that the charge separated per crystal collision increased markedly as the impact velocity increased. Apparatus was built which enabled stable streams of uniformly sized uncharged water droplets in the radius range 50 - 150μ, to be produced. Smaller droplets down to 30μ radius were produced in unstable streams. Droplets were made to encounter a rotating iced robe connected to an electrometer. It was found that appreciable quantities of charge were separated only for the larger droplets. If the droplets were above about 0 C they charged the probe positively and if they were supercooled they charged it negatively. The quantity of negative charge separated decreased as the degree of supercooling increased. The maximum mean charge separated for a 150μ radius droplet was 10(^-5) e.s.u. and for a 90μ radius droplet was 4 x 10(^-7) e.s.u. It was concluded that the charge was separated by the droplets splashing. Droplets which were in the process of freezing were allowed to encounter the probe. The droplets always charged the probe negatively and mean charges of up to 2 x 10(^-4) e.s.u. per 150μ. radius droplet were separated. The charge separated appeared to be proportional to he cube of the droplet radius. Although the results were not directly comparable with the results of Latham and Mason, it was considered that a similar charge separation mechanism was operative, and that it could be explained more readily by the Workman - Reynolds effect than by the temperature gradient theory.