SEPTEMBER 2005 M A C GORMAN ET AL. 2583 The Electrical Structure of Two Supercell Storms during STEPS DONALD R. MACGORMAN AND W. DAVID RUST NOAA/National Severe Storms Laboratory, and Cooperative Institute for Mesoscale Meteorological Studies, Norman, Oklahoma PAUL KREHBIEL AND WILLIAM RISON New Mexico Institute of Mining and Technology, Socorro, New Mexico ERIC BRUNING Cooperative Institute for Mesoscale Meteorological Studies, and School of Meteorology, University of Oklahoma, Norman, Oklahoma KYLE WIENS Department of Atmospheric Science, Colorado State University, Fort Collins, Colorado (Manuscript received 30 July 2004, in final form 1 March 2005) ABSTRACT Balloon soundings were made through two supercell storms during the Severe Thunderstorm Electrifi- cation and Precipitation Study (STEPS) in summer 2000. Instruments measured the vector electric field, temperature, pressure, relative humidity, and balloon location. For the first time, soundings penetrated both the strong updraft and the rainy downdraft region of the same supercell storm. In both storms, the strong updraft had fewer vertically separated charge regions than found near the rainy downdraft, and the up- draft’s lowest charge was elevated higher, its bottom being near the 40-dBZ boundary of the weak-echo vault. The simpler, elevated charge structure is consistent with the noninductive graupel–ice mechanism dominating charge generation in updrafts. In the weak-echo vault, the amount of frozen precipitation and the time for particle interactions are too small for significant charging. Inductive charging mechanisms and lightning may contribute to the additional charge regions found at lower altitudes outside the updraft. Lightning mapping showed that the in-cloud channels of a positive ground flash could be in any one of the three vertically separated positive charge regions found outside the updraft, but were in the middle region, at 6–8 km MSL, for most positive ground flashes. The observations are consistent with the electrical structure of these storms having been inverted in polarity from that of most storms elsewhere. It is hy- pothesized that the observed inverted-polarity cloud flashes and positive ground flashes were caused by inverted-polarity storm structure, possibly due to a larger than usual rime accretion rate for graupel in a strong updraft. 1. Introduction was observed instead. These measurements, made at the ground, tended to be dominated by charge in the The distribution of charge is one of the fundamental lower part of storms. electrical properties of thunderstorms and was the sub- By 1950, the dominant paradigm of the thunderstorm ject of some of the earliest investigations of electricity. charge distribution was that it typically consists of three The earliest measurements were made in the eigh- vertically stacked charge regions: a large upper region teenth century by Benjamin Franklin and others, who of positive charge, a large middle region of negative demonstrated that negative charge usually was present charge, and a sporadic lower region containing a lesser in thunderstorms, though sometimes positive charge amount of positive charge (e.g., Wilson 1920; Simpson and Robinson 1941; Williams 1989). It was usually un- Corresponding author address: Don MacGorman, University of derstood that this described only the overall, gross Oklahoma/CIMMS, 100 E. Boyd, Rm. 1110, Norman, OK 73019. character of the distribution. Several years later, theory E-mail: [email protected] and airborne measurements indicated that a negative © 2005 American Meteorological Society Unauthenticated | Downloaded 09/29/21 02:31 PM UTC MWR2994 2584 MONTHLY WEATHER REVIEW VOLUME 133 screening layer charge typically forms on the upper evated higher than usual. Marshall et al. (1995) also cloud boundary, due to the discontinuity in electrical found a tendency for the height of the lowest charge in conductivity between cloud and clear air (e.g., Von- the strong updraft to increase with increasing updraft negut et al. 1962; Marshall and Rust 1991). speed. Outside the strong updrafts of supercell storms A more detailed analysis of electric-field soundings (as for other types of storms), the electric-field obser- through thunderstorms suggested that this paradigm vations showed more vertically stacked charge regions was too simple (Rust and Marshall 1996). A synthesis than in the strong updraft and showed charge extending of many electric-field soundings (Stolzenburg et al. down to lower altitudes. 1998b) suggested that the previous paradigm might be All published supercell soundings of the electric field sufficient for regions of strong updraft speeds (which thus far have consisted of only a single ascending Stolzenburg et al. defined as Ն10 m sϪ1). However, in sounding into either the strong updraft or weak updraft regions of weak updraft, the charge distribution was of each storm. None has included multiple soundings of more complex, with additional charge regions typically the same supercell storm to look at the change in elec- occurring below the lowest height of charge in the trical structure with time, and none has sampled the strong updraft core. Similar complexity of electrical rainy downdraft region and low-level precipitation structure was observed in the stratiform precipitation core. (The free ascent rate of the balloons is 3–5msϪ1 region of mesoscale convective systems (Stolzenburg et just after launch. Regions with downdraft speeds larger al. 1994, 1998b). than this, such as the rainy downdraft, cannot be The electrical properties of supercell storms have sampled by ascending balloons, but only by instruments been of interest since at least the 1950s [e.g., see review descending from above.) For none has a global posi- by MacGorman (1993)]. Lightning flash rates tend to tioning system (GPS) been available to provide more be much larger in supercell storms than in ordinary accurate tracking of balloons. Also, in no published su- isolated thunderstorms, and other unusual electrical percell case have polarimetric radar data been acquired phenomena have been reported for supercell storms or lightning flashes been mapped in three dimensions (e.g., Vonnegut and Weyer 1966). One issue, raised by for the storms into which soundings were flown. Rust et al. (1981) and others, is whether the electrical Data to begin addressing these issues were obtained properties of supercell storms represent a simple scal- during summer 2000 from the Severe Thunderstorm ing with size of the properties of normal thunderstorms, Electrification and Precipitation Study (STEPS) field or whether there are any basic differences in the elec- program (Lang et al. 2004) in Kansas, Colorado, and trical structure and evolution of supercell storms. Nebraska. The present study analyzes a total of four MacGorman et al. (1989) observed that the mesocy- balloon-borne electric-field soundings of two supercell clone region of a supercell storm had few, if any, cloud- storms that occurred during STEPS. One storm oc- to-ground flashes, though total flash rates were larger curred during the evening of 29–30 June, and the other, than in most other types of isolated storms. Ground during the evening of 5–6 July. flash rates increased as the mesocyclone weakened and the precipitation at middle levels of the mesocyclone 2. Instrumentation and analysis techniques descended. MacGorman et al. hypothesized that the lowest charge region in the updraft was elevated higher One of the primary sources of data for this study was than usual by the very large updraft speeds and rotation a balloon-borne electric-field meter of the type that has of the supercell storm and that this caused fewer been used for more than two decades (Winn et al. 1978; ground flashes than usual. The subsequent increase of MacGorman and Rust 1998, 127–131). Balloons were ground flash rates, they suggested, was caused by the launched into storms from a mobile laboratory, as de- descent of the main negative charge on precipitation scribed by Rust (1989) and Rust and Marshall (1989). and the formation of a small positive charge in precipi- The location of the balloon and simultaneous measure- tation below the negative charge. ments of temperature, pressure, humidity, and wind Simulations of supercell storms (Ziegler and were provided by radiosondes. These radiosondes were MacGorman 1994; Mansell 2000) showed elevated modified by the National Center for Atmospheric Re- charge structure in the updraft and showed charge re- search (NCAR) to have full GPS capability, and so gions occurring lower in the precipitation core than in provided more reliable tracking inside storms than pro- the updraft, consistent with the hypothesis. Similarly, vided by previously available radiosondes. Electric- from electric-field soundings, Marshall et al. (1995) and field data were processed to give the three-dimensional Stolzenburg et al. (1998a) found that charge in the electric-field vector as a function of time and location strong updrafts of supercell storms did tend to be el- along the balloon track, a new capability. The balloons Unauthenticated | Downloaded 09/29/21 02:31 PM UTC SEPTEMBER 2005 M A C GORMAN ET AL. 2585 used in STEPS were made of 15-␮m (0.6-mil) thick the horizontal components is zero. The technique tends plastic film, instead of the latex that had been used to indicate spurious charge regions as the balloon previously, so that the balloons would be more resistant moves toward or away from relatively small charge re- to hail damage and would