JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 101,NO. D14, PAGES 18,961-18,977, AUGUST 27, 1996 Microphysical and electrical evolution of a Florida thunderstorm 1. Observations Jeffrey R. French, • John H. Helsdon, Andrew G. Detxvilcr, and Paul L. Smith Institute of Atmospheric Sciences, South Dakota School of Mines and Tcchnology• Rapid City Abstract. This study deals with the microphysical andelectrical evolution of a thunderstorm that occurred on August 9, 1991,during the Convection and Precipitationf Electrification (CAPE) Experiment in eastern Florida. During its approximately 1-hour lifetime, the storm waspene- tratedseveral times by the Institute of Atmospheric Sciences' T-28 aircraftat midlevels. It was also penetrated at low andmiddle-levels by a National Oceanographic and Atmospheric Ad- ministration (NOAA) P-3 andscanned by three radars, oneof whichhadmultiparameter capa- bilities, operated by theNational Center for Atmospheric Research. Two stages of the storm's evolution are analyzed herein during whichthe storm grew to produce precipitation andlight- ning. The first stage, sampled during thefirst T-28 penetration at 5.25 km (-3øC) andtheP-3 at 6.4km(-10øC), was characterized by a 2- to3-kin wide updraft (maximum 14m s '•) with cloud liquid water contents up to4 g m -3, low concentrations ofgraupel at-10øC, and small tomedium raindrops in concentrations ofless than 200 m -3 at-3øC. A downdraft region also existed that wasdevoid of cloud liquid water, but contained graupel up to 2 mm. Radardata(Zr)•) are consis- tent with a coalescence-dominated precipitation generation mechanism followed by transport of drops in theupdraft to heights with temperatures colder than-7øC, where freezing formed grau- pel that continued to grow by riming. Electrification during thisstage remained weak. The sec- ondstage, sampled during the second andthirdT-28 penetrations andthe second P-3 penetra- tion, wascharacterized at midlevels by a narrower updraft anda more diffuse, broad downdraft separated by a 1- to 2-km wide transition zone. The updraft continued to show significant cloud liquid water (•2 g m -3) with few precipitation particles, while the downdraft had very little cloud liquid with graupel in concentrations >1 e-•. The transition zone shared both updraft and down- draft characteristics. The increase in ice concentration wasaccompanied by a rapidincrease in the electrification ofthe cloud with peak electric fields reaching-20 kV m -• atT-28altitude and thedetection of lightning by ground-based sensors andpilotreport. As timeprogressed, precipi- tation particle concentrations reached several per liter at midlevels in both updrafts anddown- drafts.The observations areconsistent with electrification through a precipitation-based mecha- nisminvolving the development of the ice phase. 1. Introduction and large-scale separation to gravitational sorting of the particles. Questions dealing with the electrical nature of thtmderclouds This theory has been explored tbr virtually all types ofhydrome- have been puzzling atmospheric scientists for centuries, yet most teor collisions [e.g., Latham and Mason, 1962; Attfdetynattr and major advances in the field have come in only the last fewdec- Johnson, 1972; Brooks and Saunders, 1994].The noninductive ades. Despite these advances, many questions concerning charge hypothesis has been developed through a set of laboratory ex- structure, mechanisms for charge separation, and electrical el- periments dating back to Reynolds et al. [1957]. It has been fects on a thunderstorm's microphysical evolution remain largely shown [e.g., Reynolds et al., 1957; Church, 1966; Takahashi, um'esolved. 1978a; Jayaratne et al., 1983' Saunders et al., 1991] thatcharge It is unclear at this time exactly which mechanisms are pri- will betransferred during rebounding collisions between graupel marily responsible forthe electrification of thunderstorms. The and smaller icecrystals in thepresence of supercooled water. convective hypothesis, first proposed byGrenet [1947] and later Large-scale charge separation then results from gravitational by Vonnegut [1953], attributes charge separation to the motion of separation ofthe two different-sized hydrometeor species. A re- charged cloud particles and small ions in the cloud updraft and in view of these theories along with their strengths and wealo•esses the downdraft atits edges. The inductive hypothesis, originally inexplaining observed electrical characteristics of thunderstorms proposed by Elster and Geltel [1913] for raindrop/cloud droplet is given by Saunders [1993, 1995]. collisions, attributes charge transfer to the interaction of polar- ized hydrometeors in the presence of an ambientelectric field •Now at Department of Atmospheric Sciences, University of Wyoming, Laramie. Copyright 1996 bythe.tunerican Geophysical Union. Paper number 96JD01625. 0148-0227/96/96JD-01625 $09.00 Recently, several field investigations involving both remote and in situ measurements have been mounted to characterize the microphysical andelectrical state of convective clouds in orderto shed lightonthequestion of charge separation. It is noteworthy thatrelatively few of these investigations have studied subtropi- cal storms such as occur along the east coast of Florida. Studies of subtropical storms have involved radar analysis andground- based electric field data to infer the interior electrical structure of the thunderstorms [e.g., ,lacobson and Krider, 1976; Krehbiel, 18,961