P1.12 CHARACTERISTICS OF CLOUD-TO-GROUND LIGHTNING IN WARM-SEASON THUNDERSTORMS S.A. Fleenor * , C. J. Biagi 1 , K. L. Cummins 2 , E.P. Krider, and K. Kehoe 3 Institute of Atmospheric Physics, University of Arizona, Tucson, AZ 1 Currently, Electrical and Computer Engineering, University of Florida, Gainesville, FL 2 Also, Vaisala, Inc., Thunderstorm Business Unit, Tucson, AZ 3 Currently, CIMMS, University of Oklahoma, Norman, OK 1. INTRODUCTION Five measurement campaigns have been conducted in southern Arizona (AZ), northern Texas and southern Oklahoma (TX-OK), and in the Great Plains (GP) of eastern Colorado, western Kansas and western Nebraska in order to evaluate the performance of the U.S. National Lightning Detection Network TM (NLDN) after an upgrade in 2002-2003 [Cramer et al., 2004; Kehoe and Krider, 2004; Cummins et al., 2006; Biagi et al., 2007]. This dataset has also been used to quantify how the characteristics of confirmed cloud-to- ground (CG) flashes vary with geographic region. In each campaign, lightning was recorded using digital video cameras that were synchronized to GPS time (with 16.7 msec resolution, see Parker and Krider (2003) and Biagi et al. (2007)), and the results were compared with NLDN reports that provided the time, polarity, location, and an estimate of the peak current (I p ) for each stroke [Cummins et al., 1998]. In the GP, radar imagery was combined with NLDN reports to show when and where in the storm development the positive and negative flashes occurred, and to determine if the flashes we recorded were biased by the sampling. In this paper, we will discuss the parameters of positive and negative CG flashes in all three regions and summarize the radar results in the GP. 2. RESULTS 2.1 Negative strokes The values of I p for negative strokes recorded on video varied considerably between different recording sessions in all geographic regions [Biagi et al., 2007]. Table 1 summarizes the mean and median values of I p in AZ in 2003 and 2004, TX-OK in 2003 and 2004, and the GP of eastern Colorado, western Kansas and western Nebraska in 2005. Results are listed separately for first strokes, subsequent strokes (SS) that created a new ground contact, and subsequent strokes * Corresponding author address: Stacy A. Fleenor, Univ. of Arizona, Dept. of Atmospheric Sciences, Tucson, AZ 85721-0081; e-mail: fleenor@atmo.arizona.edu . that remained in a pre-existing channel. Table 1 also shows the mean NLDN detection efficiency (DE) in each region for each type of stroke and the overall flash DE. Note that the flash DE is greater than 90% in all regions. The distributions of I p for negative first strokes are shown in Figure 1. Note in Table 1 that the median I p for first strokes in TX-OK is 18% less than the median in AZ, and the median I p in the GP is 13 % larger than in AZ. Given the large sample sizes, these differences are likely significant. The median I p for subsequent strokes that produced a new ground contact also vary between regions, while the median I p values for subsequent strokes that remain in a pre-existing channel are similar in all regions. The standard deviation of first strokes in TX-OK is much larger than in AZ and in the GP, primarily because TX-OK has greater fractions of both low and high values of |I p | (see Figure 1). 2.2 Negative multiplicity and number of ground contacts per flash Values of video multiplicity are listed in Table 2. Note that when the |I p | of the first stroke is ≤ 10 kA, the multiplicity tends to be smaller than when the |I p | > 10 kA in all regions. The largest multiplicity (3.6) in Table 2 was in AZ. Because the time-resolution of the video camera was limited to 16.7 ms, we expect that all multiplicities in Table 2 actually underestimate the true values by about 11% [Biagi et al., 2007]. The percentage of negative CG flashes that produce a given number of ground contacts (GC) are summarized in Table 3. It should be noted that the fractions of flashes that produce a single ground contact, and the average number of ground contacts per negative CG flash, are similar in all regions. 2.3 Positive Strokes An unexpectedly large number of positive CG flashes, relative to negative (204 positive flashes and 103 negative flashes) were recorded on video in the GP 2005 campaign, and the number and percentage of positive flashes varied considerably between recording