KSC LDAR

The Lightning Detection and Ranging (LDAR) system was developed at the Kennedy Space Center (KSC) in order to provide real time lightning safety information (Maier et al., 1995a). The LDAR is able to accurately locate lightning radiation sources in 3 spatial dimensions and time by measuring the arrival time of impulsive VHF events at seven stations.

The operating frequency of the LDAR is centered at 66 MHz with a 3 dB bandwidth of 6 MHz. The 66 MHz VHF signal at each station is logarithmically detected and transmitted to the central site via video microwave links. Each of the analog signals is sampled at the central site with a 100 MHz rate and with 8-bit resolution. The time and amplitude of the peak signal within an 81.96 $\mu$s window is determined for each channel every 100 $\mu$s. This data is output to a data computation system which determines 3D source locations for those points which are observed at the proper retardation times by at least 4 stations.

The stations are positioned such that they are approximately evenly-spaced from each other on the periphery of a circle $\simeq\,$10 km in radius, with one station at the center (see Figures 3.1-3.2). The median location error within the LDAR network is only $\simeq\,$50 m (Maier et al., 1995a). Outside the network, the 3D location errors are due primarily to radial errors which increase with range at a rate faster than the linearly increasing transverse errors due to a fixed bearing error of $\simeq\,$0.5$^{\circ}$.

The individual flashes analyzed in this dissertation were reported by LDAR to be between 40-130 km range from the LDAR central station. The median radial error at 40 km was directly measured by Maier et al. (1995a) to be $\simeq\,$1 km for regular VHF pulses emitted from an aircraft which was equipped with a differential GPS navigational system. Assuming an approximate $r^2$ dependency of radial errors on range (Boccippio et al., 1999), the median radial error at 130 km range would be $\simeq\,$10 km.

The detection efficiency of the LDAR instrument decreasees with increasing source range. Climatologically-averaged areal source density declines exponentially with ground range at an e-folding length of 39-50 km (Boccippio et al., 1999). Thus, the detection efficiency at 40-130 km range would be a factor of $\sim$3-30 less than within the LDAR network.

One effect of the rapid drop in detection efficiency is that sources which are closer than their calculated range will be overrepresented relative to sources which are further away, resulting in a net outward bias in radial locations. This will also result in an upward bias in heights. Climatologically-averaged source density plots of LDAR heights with respect to range indicate that heights are ``pushed up'' at increasing range with an approximate $r^2$ dependency, though this effect is somewhat minor with only a $\simeq$1 km increase in height at 100 km range (Boccippio et al., 1999). Since most of the data points analyzed will be within 100 km range, the systematic increase in height will not be particularly important.