Lightning is an impressive meteorological phenomenon, both in terms of its physical manifestation and its destructive power. There are over 2000 thunderstorms globally at one time, producing at least 100 lightning strikes to earth per second. Between 1959 and 1990, lightning was responsible for $35 million in damage, 93 deaths, and more than 250 injuries in the United States, becoming the second leading cause of weather-related fatalities (NLSI, 2001). Currently, no lightning forecasting system is commonly available, other than predicting by observing the trend on real-time lightning displays. An understanding of the association between storm intensity and the initial lighting activity may enable lightning forecasts, as well as having the potential to improve thunderstorm forecasts.

For many years, scientists have studied the electrical processes of thunderstorms in order to determine the mechanisms that cause cloud-to-ground (CG) strikes. These mechanisms have been the topic of much debate; but in recent years, the theory that ice particle interactions cause electrification seems to have prevailed. In order for lightning to occur, large ice particles (graupel or small hail) must collide with small ice particles above the freezing level, resulting in a net negative charge for the large particle and a net positive charge for the small particle. However, the storms updraft and gravity separates the large particles from the small ones. The gravitational separation results in a charge separation, with negative charge lower and positive charge higher. In order for collisions to occur, large graupel particles need to interact with smaller ice particles. Updrafts provide the vertical velocity needed.

Reflectivities above the freezing level are likely to be associated with graupel particles at these higher altitudes. Dye et. al. 1986 stated that this was because the significant numbers of ice particles are not nucleated until about -15 degrees Celsius, adiabatic liquid content reaches a maximum near this altitude, and the maximum in the ice crystal diffusional growth rate occurs near this temperature. Dye et. al. 1986 also suggested that if this is the case then the onset of electrification will occur at different radar reflectivities (reflectivity being the measure of the radar echo intensity as discussed in Stephens, 1994) in different geographic regions due to the various concentrations and distributions of ice particle sizes and concentrations within clouds. Many past electrification studies were not able to observe lightning and radar reflectivity inside convective storms due to instrumentation limitations. Such limitations include not being able to see the altitude at which lightning occurs and not having the lightning mapping capability in conjunction with research radars that detect the microphysical properties of clouds. The 2000 Severe Thunderstorm Electrification and Precipitation Study (STEPS), conducted in eastern Colorado and western Kansas in May-July 2000, collected several comprehensive radar and lightning data sets using the New Mexico Institute of Mining and Technology's Lightning Mapping Array (LMA) and the National Center for Atmospheric Research's S-band polarimetric (S-Pol) radar. This makes it easier to investigate the relationships of lightning to storm structure and evolution, and enhance the understanding of electrification and lightning in severe storms on the High Plains (STEPS, 2000). More information on the STEPS field project can be found at http://www.mmm.ucar.edu/pdas/steps-science.htm.

Previous lightning research has been done on small isolated thunderstorms that form over the mountains of southwestern United States, and above sea breeze convergence zones in southeastern coastal areas (Krehbiel, 1986). The area of eastern Colorado and western Kansas is unique because CG lightning flashes often lower the positive charge to ground instead of the more usual occurrence of negative charge to ground being lowered. The occurrence of positive (+) CG strikes are closely linked to the occurrence of tornadoes, while a change of polarity has been found in severe storms. Little documentation currently exists on initial lightning activity, this being the first lightning that a storm produces. The purpose of this paper is to document several cases collected during STEPS to determine a relationship between initial lightning strikes and radar reflectivity. Questions such as whether the initial lightning activity was intracloud (IC) or CG, the altitude at which it originated, what reflectivity at which initial lightning activity occurred, how long it was after the first 25 dBZ radar echo, and how much lightning activity there was in the thunderstorm can be answered. Radar, the National Lightning Detection Network (NLDN), and the LMA, make it possible to investigate these questions in greater depth. This paper is a study of 14 storms. It discusses the instrumentation used to determine initial lightning, and compares storms with no lightning activity to those storms that did display lightning. It also compares thunderstorms with only IC to those with both IC and CG lightning, and thunderstorms that produced -CG lightning to those that produced +CG lightning. A summary of results is given and suggestions for future work are made. The individual storms are discussed in more depth in an appendix.