13 th International Conference on Atmospheric Electricity, August 13-18, 2007, Beijing, China The Study of Charge Structure Sensitivity in Simulated Thunderstorms * Rumjana Mitzeva 1 , Boryana Tsenova 2 , Rachel Albrecht 3 , Walt Petersen 4 1. Faculty of Physics, University of Sofia, Bulgaria 2. National Institute of Hydrology and Meteorology, BAS, Sofia, Bulgaria 3. University of Sao Paulo, Brazil 4. University of Alabama, Huntsville, Alabama, USA ABSTRACT: A small isolated thunderstorm is simulated by two one-dimensional cloud models using four different parameterizations of the non-inductive mechanism, based on laboratory data. The analyses show that during the early (pre-lightning) stage of thunderstorm electrification different charge distributions may be obtained using one and the same parameterization for non-inductive charge transfer in the different numerical models. 1. INTRODUCTION In the scientific community there is a general consensus that the non-inductive mechanism plays the major role in thunderstorm electrification. This mechanism is based on laboratory experiments (Reynolds et al. 1957, Takahashi (1978), Jayaratne et al. (1983), Saunders et al,1991 and others) showing that substantial charge is transferred during rebounding collisions between ice crystals and graupel pellets growing by the accretion of supercooled cloud droplets. However the various laboratory measurements (Takahashi, 1978, Saunders et al., 1991, Brooks et al., 1997) show different dependence on the magnitude and sign of separated charge as a function of Liquid Water Content (LWC) or effective water content and in-cloud temperatures. Using various schemes for the numerical simulation of cloud electrification, studies by Solomon and Baker (1993), Brooks et al. (1997), Scavuzzo et al. (1998), Helsdon et al. (2001), Mansell et al., (2005) and others, show that the non-inductive mechanism can make a significant contribution to the dipole or tripole structure of simulated clouds during thunderstorm development. However analysis of the numerical simulations provides conflicting conclusions. For example using Takahashi’1978 data Scavuzzo et al., 1998 calculated a tripolar charge structure, while the model simulations by Altaratz et al, 2005 developed a dipole. The electrification simulated in the Helsdon et al., 2001 thunderstorm is slower when the Saunders et al., 1991 laboratory data were incorporated in the model in comparison with the incorporation of Takahashi (1978) data, which is in contrast to the results presented in Altaratz et al.,2005. Since the electrification depends on microphysical and dynamical properties one possible reason for the discrepancy between the conclusions is that different thunderstorm cases were simulated. However the discrepancy may be due to the different manner of incorporation of laboratory data because of the specific features of numerical models. The aim of the present paper is to study what the most significant influences are on model-simulated charge structures during the growth stages of thunderstorms as a function of both model type and non-inductive charge parameterization. * Correspondence to: Rumjana Mitzeva, Faculty of Physics, University of Sofia, Sofia -1164, Bulgaria Email: rumypm@phys.uni-sofia.bg