Detailed storm descriptions and time-height diagrams are given as an appendix. Below in Table 1 is a summary of all 14 cases labeled with an arbitrary letter. The date and time of each storm is given. The maximum reflectivity the storm reached is also shown. Below the maximum reflectivity are the maximum heights of each of the other reflectivities. The next column shows the first lightning as well as the time it occurred after the first 25 dBZ radar echo. The reflectivity that the lightning originated at is also given. Below the reflectivity is the altitude of each of the other reflectivities during the initial strike. The lightning that the storm went on to produce is also shown. If the storm produced CG lightning, the polarity and the time it occurred is listed.

Storms without lightning activity vs. storms with lightning
Two storms without lightning and 12 storms with lightning were analyzed. The two that did not produce any lightning were storms A and B. Storm A intensified to 50 dBZ, however, the heights of the strongest reflectivities remained under 7 km. The reflectivity for storm B reached only 35 dBZ. In comparing the storms with no lightning activity to those that did, it seems that storms typically have the height of the 40 dBZ contour above an altitude of 7 km in order to produce lightning. For the field area, the 0 degree Celsius isotherm was typically at an altitude of 4-5 km, the -10 degree Celsius isotherm usually occurred at 5-6 km and the -15 degree Celsius isotherm occurred at 6-7 km. The cloud base was usually around 10-15 degrees, which roughly corresponded to 3 km.T

The storms studied during the field project uphold Dye's theory as stated in the introduction that ice collisions contribute to electrification. In each of the 12 storms that had lightning activity, there was no particular time in the storm evolution that initial lightning occurred. Some happened right before the storm reached maximum height, some right at the maximum height, and others right after.

Previous research showed that in Florida, the first lightning occurred when the storm reached an intensity of 35 dBZ (Dye et. al., 1989). They did not however state the height that this reflectivity must surpass. For one storm in Montana, the reflectivity of 45 dBZ exceeded an altitude of 6 km. Although more than one storm in those areas should be studied to determine if this is the case for all storms, the area of northeastern Colorado and western Kansas is different. It seems that here a storm should typically reach greater intensity and form at higher altitudes. If this is the case, then Dye et. al.'s hypothesis that the onset of electrification will occur at different radar reflectivities in different geographic regions due to the various concentrations and distributions of ice particle sizes and concentrations within clouds seems correct (Dye et. al., 1986).

Thunderstorms with only IC lightning vs. thunderstorms with CG strikes
Seven storms with only IC activity and 5 storms with both IC and CG lightning were analyzed. Interstorm comparisons were made by looking at the altitude of the initial strike, the height of the 40 dBZ contour, and the reflectivity at which lightning originated.

The altitude that the initial IC strike originated at, and the height of the 40 dBZ radar contour at the time of this initial strike, differed for the two types of thunderstorms. Those that only produced IC lightning had the initial strike originate at altitudes of 7-10 km. The IC lightning in storm G originated at 9.755 km at a time where the height of the 50 dBZ reflectivity was greater than 7 km. The range of 7-10 km was also the height of the 40 dBZ radar contour. For thunderstorms that went on to produce CG strikes, the altitude range was broader and included lower altitudes. These were between 6-10 km. The height of the 40 dBZ radar contour, however, was at a higher altitude. This was around 8.5-10 km.

The differences between reflectivity of origin, whether it originated in low or high reflectivity, also seems to characterize if a thunderstorm will go on to produce CG lightning. Thunderstorms that produced CG lightning did not originate in reflectivities of less than 30 dBZ. Storms J and N both originated in a reflectivity of greater than 30 dBZ. In both of these storms the height of the 25 dBZ radar echo was greater than 12 km. The height of the 35 dBZ contour was also greater than 9.5 km. Storm N did not reach 45 dBZ. Both storms did however have the initial IC lightning originate at an altitude between 8-9 km. Lightning that originated in high reflectivity, greater than 40 dBZ, occurred in storms K, L, and M. In these storms, the height of the 25 dBZ was less than 12 km during the initial IC strike. For storms L and K, the 45 dBZ contour was greater than 8 km.

For thunderstorms with only IC lightning, the heights of reflectivities during the first initial strike were also different depending on whether the lightning originated in high or lower reflectivities. Storms C and I had lightning that originated in 25 dBZ. For lightning that originated in 25-30 dBZ, the height of the 35 dBZ radar contour was less than 9.5 km. The height for the 40 dBZ radar contour was also less than 8 km. Storm D, F, G, and H had lightning that originated in high reflectivity, greater than the 40 dBZ contour, the height of the 50 dBZ contour was less than 8 km.

Thunderstorms with +CG lightning vs. -CG lightning
A total of 3 -CG lightning thunderstorms (K, L, N) and 2 +CG lightning thunderstorms (J, M) were studied. Case K switched polarity from negative to positive. Thunderstorms that produced -CG lightning had less time between the first 25 dBZ radar echo and the first CG strike. This time was around 15-30 minutes. The time between the initial IC strike and the first CG strike was also less. This time was between 5-20 minutes. The maximum intensity during the time of the first IC strike was 45 dBZ. Rison et. al. 1996 studied one storm in Florida where they found that a CG strike occurred five minutes after the initial IC strike. It did originate in high reflectivity. Although this was only one storm, it is consistent with the three cases observed here. The reflectivity of 50 dBZ was present at the time of the first IC strike for thunderstorms K and L. Storm N did not reach 50 dBZ during the initial IC lightning.

Storms with +CG lightning took longer for both the IC and the first CG lightning to initiate where both times were greater than one hour. Positive CG lightning is known to occur with severe weather and both cases here support this observation. At the time of the initial IC strike, the reflectivity present was lower than for a -CG lightning storm. The case that switched polarity was different than regular -CG cases. In case K, the time interval from the first IC strike to the first CG strike to be less than 5 minutes. The other two - CG lightning cases were greater than 5 minutes. Case K was also different in that all the radar reflectivities below 45 dBZ occurred at lower altitudes. The 50 dBZ reflectivity was lower in the storm than the other -CG lightning storms. More cases of storms that switch in polarity will have to be studied to see if this is representative.