Primary instrumentation in STEPS included the S-Pol radar, Colorado State University's CHILL Radar, the Goodland Doppler radar, the NLDN, an armored T-28 aircraft for storm penetrations, two mobile balloon sounding systems to characterize the storm environment, six mobile mesonets for the observation of weather beneath storms, the LMA to map the three- dimensional distribution of lightning, and the National Severe Storms Laboratory mobile balloon unit for electric field, temperature, and wind profiles within storms.

For this research, the NLDN, radar data, and the LMA were used to make CG lightning polarity, location, and reflectivity observations. Below is a brief description of these instruments.

National Lightning Detection Network (NLDN)
The NLDN consists of 108 ground-based remote sensing stations that monitor CG lightning activity across the United States using triangulation techniques. Within seconds of a lightning strike, the NLDN detects both +CG and negative (-)CG lightning, locates where each flash strikes the ground, and estimates its peak current. Using signal processing and global positioning systems with time synchronization, the NLDN is able to detect 95% of all CG strikes (NLDN, 2000). For the purposes of this research, the NLDN data were overlaid on Goodland Doppler radar data and were used to detect storms that exhibited CG lightning, and to determine whether or not the storm would be suitable for further analysis.

Lightning Mapping Array (LMA)
The LMA, which was first tested in 1998, maps the location of lightning in three dimensions. It allows one to study the initiation and development of lightning flashes and determine the initial CG lightning and IC flash rates for storms. The system operates by detecting radiation from lightning discharges in an unused VHF television channel (Krehbiel, 1999).

Figure 1 shows the locations of the LMA stations relative to the Doppler network. The three radars lie at the vertices of the blue-yellow triangle. The LMA, as seen in red, consists of 15 time-of arrival stations for locating the lightning. Ten of the stations are electric field sensors and field change instruments for measuring the overall charge structure of storms. The field sensors sample continuously at a 10 to 50 Hz rate, while the field change data sample at a 5 or 10 kHz rate. The other five stations are fast electric field change recording stations for detecting and locating CG and IC strikes.

Figure 1 : Lightning Mapping Array

The electric field data are time-tagged for accurate time synchronization between stations to allow the data to be located (STEPS, 2000). These data are then plotted every ten minutes on a display. For this study, the LMA was used to detect the initial strike, determine if it was CG or IC lightning, and to record the altitude at which it originated. Because many strikes can occur over 10 minutes, the data over the 10-minute interval were plotted on the Goodland radar scan. All strikes occurring during a radar scan were displayed. The only time strikes were not displayed was for 8 seconds between the beginning and end of a radar sequence. Overlaying lightning and radar data made it easier to tell if a storm was producing lightning. The altitude for the first data point was then recorded for plotting on the time-height diagrams (discussed in the following section) for analysis.

Radar
Radar data from the field project were taken from Goodland Doppler radar and the S-Pol research radar. Parameters such as velocity, reflectivity, and differential properties are derived from these instruments, providing a picture of what is going on inside a storm and the electrification of that storm. A time-height diagram was produced once a storm was chosen for analysis.

A time-height diagram has height as the vertical axis and time as the horizontal axis and has a contour plot of radar reflectivity drawn in this height/time space. The construction of a time-height diagram can be seen below in Figure 2. Radar scans are performed at several elevation angles, with the lowest angle corresponding to the lowest altitude and the highest angle with the greater altitude. A radar sequence involves a complete horizontal rotation of the radar for each of the vertical elevation angles. To produce a time-height diagram, the lowest and highest altitudes of each reflectivity are plotted. This is done for each radar sequence until the storm dissipates. Once all the data points are plotted, the points are connected to form a contour map of radar reflectivity in height/time space. The appendix shows all the time-height diagrams for the 14 storms studied. Once the initial lightning activity was extracted from the LMA, it could be plotted on the time-height diagram and relationships between radar reflectivity and initial lightning activity could be shown.

Figure 2: Producing time-height diagrams