My atmospheric physics research has been directed towards obtaining a better understanding of electrical discharges within the atmosphere. I have developed both hardware and low-level software to collect data and have written several high-level software programs to analyze that data. What follows is a brief chronological summary of some key results I have obtained. More detailed information can be found in some of my publications and on other web pages.

On the last day of operations, a 47-stroke triggered lightning discharge was observed with our interferometer system (Note: a normal cloud-to-ground (CG) discharge averages between 3-4 strokes!) The upward-going positive leader which was launched from the triggering rocket was not visible at radio-frequencies (RF). However, recoil-like events were observed propagating back towards the ground along channels through which the positive leader had passed. These events are similar to recoil events seen in intracloud as well as some CG discharges. An azimuth/elevation plot of the discharge is shown at left, with the trigger site at the lower right (the lightning propagated away towards a distant storm at low elevation angle in the lower left).

	In collaboration with Chip Scott, a fellow graduate student at New Mexico Tech, lightning images obtained by our interferometer during the summer of 1994 at Langmuir Lab were overlayed onto the RHI (vertical) scan planes of a dual-polarization radar situated at New Mexico Tech. This study revealed an apparent close correlation of the discharges with high reflectivity regions throughout the storm (left figure) as well as with the vertically-aligned ice crystals in an inferred strong electric field region in the upper portion of the storm (right figure).

On September 2, 1994, 3-dimensional lightning images obtained by the Kennedy Space Center Lightning Detection and Ranging (KSC LDAR) system, operated then by Carl Lennon and Launa Maier, were correlated with PPI (horizontal) NEXRAD radar scans through a thunderstorm developing over KSC. There is a close correlation between the lightning channels and the regions of highest reflectivity. This correlation supports the precipitation-based theory for thunderstorm charging. As the storm developed, the lightning became more complex. At the level of the negative charge layer (about 8 km altitude), lightning would often propagate from newer cells into older ones. One such discharge is shown in the neighboring figure. This particular discharge was initiated near the core of the cell just below center in the figure, produced a channel to ground, and launched horizontal channels into an older cell in the upper-right of the figure.

The occurrence of downbursts during the storm development on Sept. 2, 1994 was analyzed. The picture to the left shows lightning activity preceeding a downdraft of air at the ground. Note the close correlation between the plan position of the lightning (white) and the center of the subsequent downdraft (located halfway between the maximum velocity toward (purple) and away (yellow) from the radar). Lightning activity typically lead the occurence of downbursts by 6-10 minutes and was well correlated in plan position with them.

1. In late 1994, a new VME-based analog-to-digital (A/D) converter board was designed at New Mexico Tech. In early 1995, four A/D boards were assembled, their EEPROM and GAL chips were programmed, and the completed boards were thoroughly tested.

A VME-based Sonitech Dual-DSP (TI 320C40) board was used to process data output from the A/D boards. A real-time data acquisition program was written almost entirely in assembly language for the DSPs. A program was written in C on a Sun Sparc-station to interface with the DSPs, acquire the data, and display it real time. The whole system acquired its first data in late August.

3. A vertical baseline was added to the 2 orthogonal horizontal baselines used by the old system. Once operational in early September, the vertical baseline expanded the range of the interferometer to beyond 50 kilometers (from a previous maximum range of about 20 km).

Less than a week after our first sprite observations, spectacular images of close sprites were obtained on July 25, 1996. The animation to the left shows one of several examples obtained that night of fine dendritic streamer structure in sprites. The carrot sprite (so named because of a dramatic upward carrot shape) is followed by a slow upward-moving event within the body of the sprite. The tendrils, which ordinarily are seen branching downwards from carrot sprites, were obscured by a foreground cloud at the bottom of these images.

In 1996, a sferic-recording mode was implemented on our data acquisition system. One significant discovery was that energetic sprites were often accompanied by a slow electric field change, suggesting that the sprites themselves were conducting significant amounts of current. The plot to the right shows the red light produced by a sprite (plotted in red) compared with the electric field (plotted in green). The strong correlation between the curves clearly shows that the sprite was responsible for the field change (Note: the photometer trace was provided by Walt Lyons and Tom Nelson of FMA Research.)

The plot to the left view shows a sprite-producing discharge viewed from above with contours marking increasing levels of radar reflectivity and the discharge progression color-coded from blue to red with respect to time. The discharge originated in a convective core on the right side of the image and propagated towards the upper left into the stratiform region. The discharge was almost fully horizontally developed when it produced a +CG which in turn produced sprites above the discharge. The grid squares in the image are 25 km on a side. It is clear from this scale that sprite-producing discharges are much larger than ordinary discharges.

The animation to the right shows a portion of a 1000 frames/second high speed video sequence of a carrot sprite and some fainter column sprites. These were the first sprites to be successfully captured with a high speed video camera. This and numerous other high speed video sequences which followed clearly showed that sprites were initiated at high altitude (about 75 kilometers) and were composed of high-velocity streamers. Results from the high speed video study were highlighted in three different publications and in Science News.

On August 14, 1998, three energetic daytime sprite events were detected via their ELF fingerprint. The sprite events were associated with exceptionally energetic positive cloud-to-ground (+CG) discharges in southern Texas at >1000 km range from the detector. The electric field waveform of the first event is shown at left. A fast positive pulse, the +CG return stroke, is immediately followed by a large slow positive field change produced by a very large continuing current to ground. The second slow field change is not immediately preceeded by any fast pulses, giving away its true nature as a sprite field change. The daytime sprite events were published (and featured) in GRL (see publications) and were also highlighted in American Scientist.

In collaboration with Stanford University, the first sprites associated with -CGs were detected on August 29, 1998. The electric field waveform of the first sprite-producing -CG is shown at right. Unfortunately, all but the upper portion of the sprites (not shown) were blocked by clouds, so no one yet knows how -CG sprites compare visually to more common +CG ones.

On August 2, 1999, the Los Alamos Sferic Array (LASA) located a positive bipolar event only a few kilometers to the south of my electric field instrument, which was stationed at my house in Lemitar, NM. The electric field instrument detected the onset of a slow field change a full 30 milliseconds prior to the first detected VHF radiation from the discharge. Though slow electric field deflections without VHF are a common occurrence in rocket-triggered lightning (see 1994 results above), this is the first time that such behavior has been seen in more ordinary lightning. To make matters more interesting, the first VHF radiation produced was many orders of magnitude more powerful than that typically emitted by lightning and this radiation was associated with an energetic positive bipolar event.

The interferometer/sferic system acquired data at a location near the center of the STEPS campaign research area between May 25 and August 13, 2000. A very large amount of data was collected and placed on 222 CDROMs (about 130 GB). In addition, video data was collected of storm development and lightning discharges. The interferometer, sferic, and video data of select lightning discharges will be analyzed in detail and will also be compared to data obtained by other systems such as the 3-dimensional NMT Lightning Mapping Array.

The interferometer/sferic system was set up in Dominguito, Puerto Rico, just a few kilometers north of Arecibo Observatory. The Puerto Rico campaign was a joint effort between Penn State, New Mexico Tech, and Stanford. The purpose of the Puerto Rico campaign was two-fold: to search for ionospheric effects due to lightning and to compare the VHF (interferometer) development of lightning with that observed by radar (Arecibo 430 MHz UHF). An unexpected result from the campaign which occurred on the last night of observations was the video documentation of a blue jet event which reached up from a thundercloud top all the way to the ionosphere. This event was highlighted on the cover of Nature. The blue jet was captured with the assistance of a blue-sensitive ITT Night Vision GEN-III intensifier and was also documented via electric field data obtained by the sferic system. A false-colored image of the event is shown at left.

In an attempt to model the complex gain and phase pattern of the log-periodic FORTE antennas, I ran Numerical Electromagnetics Code (NEC) simulations in parallel across a large number of computers. The modeled response was then used on a large database of events to obtain an estimate of the power spectral behavior of lightning at VHF and to quantify how it varied between different types of discharge processes.

A collaborative effort was initiated by Walt Lyons in which 3D Lightning Mapping Array (LMA) data was analyzed along with electric or magnetic field change data of sprite-producing discharges over the STEPS 2000 region. I used 3D LMA data along with close electric field data collected in 2000 to obtain the charge moment change estimates of these discharges. These estimates were then compared with remote magnetic field measurements acquired by Steve Cummer. There was general agreement between our estimates, albeit with a systematic percentage offset which is still not completely understood.

I used a database of tower locations and heights to explore the possibility that large charge moment change lightning discharges were systematically launching upward leaders from tall towers. Using a few years of lightning data in Florida, it was found that towers of 400 meter height or greater have a high risk of launching an upward leader in the presence of a large charge moment change discharge.

The 2nd generation LASA system was developed using the 20MHz @12-bit Measurement Computing PCI-DAS4020 board. I wrote the data acquisition code for the board. The features of the system were as follows:

• Sub-microsecond time resolution using 1 PPS on channel 0
• Versatile triggering scheme as independent or slaved to another channel
• Configurable pre/post trigger lengths on up to 4 channels
• Capable of very high trigger and data rates
• No dead-time between triggers
• Resolution of up to 16-bits via software summation of raw 12-bit data
• Remotely operated and very stable

In addition, I wrote software which automated the task of retrieving the massive amounts of data collected by each station. Well over a terabyte of data had been retrieved over the internet in less than a year after initial deployment in April, 2004. I also wrote analysis software in IDL for the new data set.

Since many of these activities are classified, I can not discuss them further. However, I can say that they were very time consuming.

I helped deploy the new Great Plains Network addition to LASA in April, 2005. I also made some minor modifications to the acquisition code in order to accomodate some new hardware configurations for the new array, including an experimental photometer setup at one of the sites. Within a month after initial deployment, I had dropped the bipolar trigger thresholds to an unprecedented sensitivity of <0.2 V/m.

Using LASA data from 2004 and 2005, I found several RHESSI TGFs which were sufficiently close (<1100 km) to one or more stations that I was able to resolve whether or not the coincident discharge was an intracloud or cloud-to-ground. In all cases, they were intraclouds. Furthermore, a couple of events were sufficiently close and impulsive that I was able to obtain source altitudes using ionosphere reflection pairs. These and other analysis results were recently published in a GRL article (see my publications).