No abstract

Abstract. Further laboratory measurements of charge transfer between ice crystals and riming graupel pellets, which are thought to be associated with the electrification processes within thunderstorms, have been carried out in the University of Manchester Institute of Science and Technology cloud chamber. In experiments with clouds in the temperature range -6 øC to -26 øC, the supercooled droplet spectrum has been extended to larger droplet sizes, above 60/xm maximum diameter, representative of the broadest spectrum observed in some thunderstorm cloud charging regions. The results indicate that at temperatures from -6 øC to -18øC, broadening the droplet spectrum leads to negative graupel charging at higher values of cloud effective liquid water content than has been reported in previous laboratory studies. The significance of the result is that in order to ensure that laboratory experiments simulate as closely as possible the thunderstorm cloud microphysical environment, attention must be paid to the spectrum of droplets used. Two mechanisms of charge transfer that may account for this behavior are discussed, the relative growth rate theory and the surface splinter theory, and both are found to be compatible with the results on the assumption that the larger droplets lead to a reduction in the rate of vapor deposition to the timing surface. Analysis of the implications of these results to thunderstorm electrification requires more details of the evolution of droplet spectra in thunderclouds, their spatial and temporal development and location relative to observed regions of electrification.

Avila and Caranti [1994] measured electric charge transferred when 100 /zm frozen ice spheres collide with a timing target. This work is relevant to thunderstorm electrification caused by the interactions of ice crystals and small graupel with falling graupel pellets. Avila and Caranti note that charges are transferred in two modes, one in which ice fragments from the timing target carry charge away and one in which no fragments are observed. We argue here that there is no requirement for fracturing of a rimer surface or fragment production in thunderstorm electrification processes involving the collisions of ice crystals with a timing target as studied by Jayaratne et al., [1983] and Saunders et al., [1991]. Furthermore, the results to be discussed here lead to the conclusion that ice crystals impacting with frost growths on a timer are unable to remove these surface features. A consequence of this analysis is that temperature gradients along features on the rime surface, as invoked by Avila and Caranti, are not necessarily the cause of crystal/graupel charging in thunderstorms. The fragmentation of rime was studied by Griggs and Choularton [1986] who used glass beads of diameter 485 #m to simulate ice particles impacting a rime target. They used rime densities appropriate to small graupel pellets in thunderstorms, from 0.15 to 0.5 gcm -3, corresponding to temperatures of-15 ø to -3 øC. At -3 øC the velocity required to break the rime was around 70 m s -•, while at -15øC the velocity needed was around 30 m s -•. It was concluded that rime breakup during collisions with ice particles in thunderstorms is unlikely. The authors also studied the conditions required to break ice crystals. Dendritic crystals were the most fragile; they noted that a 3-mm-long dendrite growing on a substrate could be broken off by an airstream at only 2 m s -•. Any vapor deposits on the riming target studied by Avila and Caranti would only have been tens of microns in length, but there is the possibility that the charged fragments they observed originated as vapor-grown deposits (frost) on the rime surface and were removed by the impact of 100-/tm ice spheres at 5 m s -•. Also of relevance to this debate is the work of Jayaratne and Griggs [1991] who blew fragments of ice off a timed target with a high-speed airstream. They formed a rime deposit on the front o f a rod target then directed an air jet at the rime with the rime deposit upwind, downwind, or at 90 ø to the air jet, while noting any charge transfer to the target. When the air jet impinged on the front or the side of the Copyright 1996 by the American Geophysical Union. Paper number 95JD03066. 0148-0227/96/95JD-03066502.00 rime, charge transfer was observed when the rime broke off; however, when a downwind deposit broke off, no charge transfer was observed. These results suggest that charge transfer probably occurs when the separated time fragment rubs against the target during its removal but not during the fracture process. In the charge transfer experiments of Jayaratne et...

Experiments were conducted with a wind tunnel in a cold room, in order to investigate the influence of the cloud-droplet spectrum on the charges transferred when individual ice spheres collided with a fixed artificial graupel pellet growing by riming. The experiments were carried out with ice spheres of about 100 pm in diameter, impact velocities around 4 m s-', temperatures between -10 "C and -30 "C and effective water contents representative of real clouds. Two different cloud-droplet spectra were used. One had more than 30% of the droplets with sizes greater than 13 pm, and the other had more than 50% of the droplets greater than that. The new results show that the size distribution of the droplets is very important to the sign of electric charge transferred. The target graupel charged positively over all the temperature range covered when the smaller-droplet spectrum was used, but negatively at temperatures below -18 "C for the larger-droplet spectrum. These results show the importance of droplet sizes to thunderstorm charging.

Measurements of charge transfer during individual collisions between ice spheres and a target growing from vapor deposition have been performed in order to establish the importance of fragmentation in the charging process. It is found that when the target is a few degrees colder than the environment, the charge pulses have mixed signs but the number of positive ones becomes dominant as the target temperature decreases. This behavior is reproduced in collisions with dielectric particles and a target in the same growth conditions. Reversal in the temperature gradient reversed the sign of the charge transfer. Direct evidence of fragments produced during collisions suggests that they are responsible for the measured charge transfer.