fagu98.bib

@ARTICLE{papadopoulos_fagu98,
  AUTHOR = {K. Papadopoulos and G. M. Milikh and P. N Guzdar and
		  A. S. Sharma},
  TITLE = {Gamma Rays Generation by Lightning},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F126},
  MONTH = {November},
  ANNOTE = {A22A-03 poster},
  ABSTRACT = {A new and novel mechanism for generation of gamma
		  rays observed at equatorial latitude by the Gamma
		  Ray Observatory is proposed.  Several new
		  interacting physics elements are involved.

                  -The first element is the generation of runaway
		  electrons by thunderstorm electric field at
		  altitudes 16 - 20 km slightly below the region where
		  the runaway electrons become magnetized ($z \sim 20)
		  km. 

                  -Upward leakage of the runaways above 20 km allows
		  for the presence of helicon mode with frequency $f <
		  10 \sqrt{n_r}$ kHz where $n_r$ is the local density
		  in $cm^{-3}$ of the leaking electrons.

                  -ELF and VLF waves induced by lightning, reaching
		  this region undergo self-focusing creating channels
		  of enhanced electric field structure in the regions
		  of runaway leakage.

                  -The self-focused electric fields interact
		  resonantly with runaways inducing a new runaway
		  discharge at altitudes between 20 - 40 km

                  -Gamma rays are due to beams from the runaway
		  electrons forming at those altitudes.

                  A qualitative analysis will be presented.  The new
		  process can account for significant enhancement of
		  energy deposited by lightning in the middle
		  atmosphere with significant consequences to global
		  warming.}
}
@ARTICLE{smith_fagu98,
  AUTHOR = {D. A. Smith and D. N. Holden and X. Shao and P. R. Krehbiel},
  TITLE = {Physical Characteristics of Compact Intracloud
		  Discharges (CIDs)},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F127},
  MONTH = {November},
  ANNOTE = {A22A-11 poster},
  ABSTRACT = {Compact Intracloud Discharges (CIDs) are distinct,
		  isolated electrical discharges that occur within
		  intense regions of thunderstorms.  Unlike other
		  intracloud and cloud-to-ground lightning events that
		  occur as sequential breakdown processes and often
		  last for a large fraction of a second, CIDs are
		  singular impulsive events that last for only a few
		  tens of microseconds of less.

                  In recent years, CIDs and the thunderstorms in which
		  the events have occurred have been studied using a
		  variety of resources including ground-based electric
		  field change and broadband HF arraysk, satellite
		  instrumentation including the Blackbeard receiver
		  and FORTE satellite, NEXTRAD weather surveillance
		  radars, and the National Lightning Detection
		  Network.  Based on the data from these resources,
		  CIDs have the following physical characteristics:

                  CIDs begin with an initial breakdown stage as
		  indicated by RF radiation that lasts for a few
		  microseconds.  Ionization proceeds at a velocity on
		  the order of one third of the speed of light and
		  forms a channel that is between a couple hundred and
		  many hundred meters in spatial extent.  The channel
		  is predominately  GET abstract from AGU site..}
}
@ARTICLE{krehbiel_fagu98,
  AUTHOR = {P. R. Krehbiel and R. Thomas and W. Rison and
		  T. Hamlin and J. Harlin and M. Davis},
  TITLE = {Lightning Mapping Observations During {MEaPRS} in
		  Central {O}klahoma},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F127},
  MONTH = {November},
  ANNOTE = {A22A-12 poster},
  ABSTRACT = {A 3-dimensional lightning mapping system was operated
        over a county-wide area northwest of Oklahoma City in
        conjunction with the MEaPRS project during June of 1998.
        Observations were obtained of the total lightning activity in
        several supercell storms, a tornadic storm, and storms having
        a large fraction of positive cloud-to-ground lightning.  A
        large, fast-moving storm system that propagated over the
        network produced normal cloud-to-ground (CG) and intracloud
        (IC) lightning discharges in a number of coexisting cells and
        occasional discharges of large horizontal extent.  One hybrid
        IC/CG flash had an overall extent of 75 km and exhibited a
        bilevel structure in the main negative and upper positive
        charge regions of the storm.  The channels in the upper
        positive charge region decreased to the altitude of the main
        negative charge as the discharge progressed into an apparent
        stratiform region away from the core of the storm, indicating
        that the positive charge region similarly decreased in
        altitude.  A third, slightly lower level of activity was also
        observed that may be associated with lower positive charge.
        
        Lightning in the supercell storms was essentially continuous
        and had a more amorphous structure, with discharges occurring
        in rapid succession at different locations that filled the
        horizontal area of the storm every minute.  Preliminary
        results for one predominantly positive CG storm showed that
        +CG discharges were associated with normal polarity IC
        discharges early in the storm, but that later in the storm's
        evolution the IC flashes appeared to be of inverted polarity,
        that is, between an upper negative and main positive charge
        region.  The latter result remains to be confirmed by
        electrostatic field change observations.  Lightning in the
        tornadic storms remains to be analyzed.  Overall, the great
        plains storms are observed to be extremely active
        electrically.}
}
@ARTICLE{rison_fagu98,
  AUTHOR = {W. Rison and P. R. Krehbiel and M. Davis and
		  T. Hamlin and J. Harlin and T. Barber and M. Jones},
  TITLE = {A Deployable 3-Dimensional Lightning Mapping System},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F127},
  MONTH = {November},
  ANNOTE = {A22A-13 poster},
  ABSTRACT = {A lightning mapping system based on Kennedy Space
        Center's real-time LDAR (Lightning Detection and Ranging)
        system has been developed that utilizes GPS timing at a number
        of widely spaced locations to image lightning discharges in
        three spatial dimensions and time.  The system was
        successfully deployed in two field programs in 1998 -- in
        central Oklahoma during May and June and in central New Mexico
        during August and September.  Lightning-produced radiation
        signals are received in an unused television channel (Channel
        3, 60--66 MHz) and the time of arrival of impulsive events is
        independently measured at each station using the GPS timing
        signals.  Overall timing accuracies of 40--50 ns rms are
        achieved on actual lightning events, corresponding typically
        to 50--150 m location errors.  In the 1998 field programs, the
        peak VHF radiation signal was timed in successive 100~$\mu$s
        windows at 10 locations over a 50-km diameter area.  This
        provided excellent location information over a 100-km diameter
        area and useful information out to several hundred km (with
        increasing errors at larger distances).  An automatic,
        non-linear least squares solution procedure is used to locate
        the radiation sources and provides reliable locations when an
        event is observed at 6 or more stations.  The system typically
        locates 500--1000 sources per second of lightning activity.
        3600 sources were located during one horizontally extensive
        flash that lasted 2.5 seconds.  The sources clearly show the
        temporal development and channel structure of lightning
        flashes and sometimes the dendritic structure as well.  The
        system has been designed to be relatively low cost and is
        PC-based, using a custom programmable-logic PC card to
        digitize and time the signals.  A wireless communications
        network (with 115~kbaud data rate for each remote station) is
        currently being used to control the network and will be used
        to automatically transmit data to a central site for real time
        processing and display of the locations.  Electric field
        change and field mill sensors will also be added at each site
        to further characterize the lightning and the storm
        electrification.}
}
@ARTICLE{symbalisty_fagu98,
  AUTHOR = {E. M. D. Symbalisty and R. A. Roussel-Dupr{\'{e}} and
		  V. A. Yukhimuk},
  TITLE = {The Transition From Red Sprite to Columniform Sprite
		  in the Context of Runaway Air Breakdown Theory},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F129},
  MONTH = {November},
  ANNOTE = {A22A-26 poster},
  ABSTRACT = {The results of numerical simulations of columniform
        sprites, in the context of the runaway air breakdown mechanism
        (A. V.  Gurevich, G. M. Milikh, and R. Roussel-Dupré, {\it
        Phys. Lett. A}, {\bf 165}, 463, 1992), are presented. The
        results include intensity and spectra of optical emissions,
        and the time dependent secondary (slow) and primary
        (relativistic) electron concentrations. We recall that the
        simulations model the transient electrical environment above a
        large mesoscale thunderstorm complex due to a normal lightning
        stroke. The transition from a normal sprite discharge to a
        columniform sprite is proposed to be a function of the
        configuration of the driving electric field. In our 2-D,
        axi-symmetric simulations, with simple models for the driving
        fields, we can produce the transition by varying the strength,
        location, and spatial size of the initiating positive cloud to
        ground lightning stroke. For example, a point discharge
        (spatial size less than 1 km) is more likely to produce a
        columniform sprite than a spatially extended lightning
        discharge, with all other parameters being equal. We are using
        the recently recalculated runaway avalanche rates
        (E. M. D. Symbalisty,R. Roussel-Dupré, and V. Yukhimuk, {\it
        IEEE Transactions on Plasma Science}, scheduled to appear in
        October 1998 issue). We also examine the impact of different
        ambient electrical conductivity profiles and compare our
        results to observations (E. M. Wescott, D. D. Sentman,
        M. J. Heavner, D. L. Hampton, and W. A. Lyons, {\it University
        of Alaska Preprint}, 1997).}
}
@ARTICLE{lyons_fagu98,
  AUTHOR = {W. A. Lyons and T. E. Nelson and E. R. Williams and
		  J. A. Cramer and T. R. Turner},
  TITLE = {Changes in the Electrical Structure of
		  {U}. {S}. Thunderstorms Ingesting Smoke from the 1998
		  {M}exican Fires},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F129},
  MONTH = {November},
  ANNOTE = {A22B-02 poster},
  ABSTRACT = {As well documented by both satellite imagery and surface 
        aerosol concentration measurements, smoke from the vast 
        1998 El-Nino drought-related forest fires in Mexico and 
        Central America frequently advected into the United States. 
        Specifically,  between about 7 April to 8 June, most air 
        masses over the southern plains of the U.S. were influenced 
        by the smoke. A strong frontal zone crossing the region 
        between 14 - 18 May resulted in smoke being advected as far 
        north as Ontario and eastward to New England. As documented 
        by the National Lighting Detection Network (NLDN), 
        thunderstorms developing within these contaminated air 
        masses had extraordinary changes in their electrical 
        characteristics. Specifically, large thunderstorm systems 
        exhibited very high percentages of cloud-to-ground (CG) 
        lightning flashes with positive polarity (+CGs). While +CGs 
        are typically less than 10% of the total, many affected 
        storms, such as those with the mid-May frontal system, 
        sustained +CG percentages of 50 to 90\% during their lifetime. 
        Moreover, the peak currents within these +CGs were much 
        higher than  normal. Throughout a two month period, storms 
        in the southern plains averaged three times the normal 
        frequency of +CGs and twice the average peak current. Prior 
        research suggests +CGs have far greater potential for 
        starting fires, disrupting electrical systems, and possibly 
        producing larger amounts of NOx. It has been confirmed that 
        lightning from the 1998 smoke-influenced storms produced 
        unprecedentedly large numbers of optical sprites in the 
        mesosphere. Similar though less pronounced +CG enhancements 
        may have been observed in Florida sea breeze thunderstorms 
        during that state's siege of summer wildfires. Some possible 
        mechanisms which might explain these unexpected changes will 
        be presented.}
}
@ARTICLE{nelson_fagu98,
  AUTHOR = {T. E. Nelson and W. A. Lyons and R. A. Armstrong and
		  E. R. Williams and D. M. Suszcynsky and R. Strabley
		  and M. Taylor and L. Gardner},
  TITLE = {Some Initial Results from {SPRITES}'98},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F135},
  MONTH = {November},
  ANNOTE = {A31A-01 poster},
  ABSTRACT = {SPRITES'98 was conducted at the Yucca Ridge Field
        Station near Fort Collins, CO from mid-May through August
        1998. Using multiple remote sensing technologies, coordinated
        measurements of individual sprite and elve events were
        obtained. Both red and blue low-light imagers (LLTVs) and
        multi-color broad- and narrow-band photometers monitored
        optical emissions. Evidence of ionization within sprites
        obtained in previous programs was confirmed from simultaneous
        4278 and 4709 nm emissions. High speed (1000 fps) images of
        sprites reveal new details of the temporal sequence of
        events. In addition, numerous high speed videos of the entire
        CG lightning event for both positive and negative polarity
        strokes provide new insights on the continuing currents and
        horizontal dendrite discharges. These videos were coordinated
        with optical measurements made with a photodiode array plus
        VLF and ELF transients (Q-bursts) made at MIT's Rhode Island
        Schumann resonance observatory. Storms ingesting smoke from
        Mexican fires produced unprecedentedly high percentages of
        +CGs, which in turn had twice the normal peak currents. A
        record number of sprites (nearly 400 in 208 minutes) were
        imaged above one such smoke-influenced storm.  Sprites again
        were typically associated with storms with radar echoes larger
        than 10,000 km2, but some exceptions were noted, especially
        bursts of sprites in the dying phases of some supercell class
        thunderstorms.  An unusual interaction between a meteor and a
        sprite was recorded.  Monitoring was also coordinated with
        scientists from Tohoku University, New Mexico Tech, Stanford,
        and the University of Alaska.}
}
@ARTICLE{strabley_fag98,
  AUTHOR = {R. Strabley and D. M. Suszcynsky and
		  R. Roussel-Dupre and E. M. D. Symbalisty and
		  R. A. Armstrong and W. A. Lyons and T. A. Nelson},
  TITLE = {Video and Photometric Observations of a Possible
		  Metor-Triggered Sprite/Jet Event},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F135},
  MONTH = {November},
  ANNOTE = {A31A-02 poster},
  ABSTRACT = {This paper presents video and photometry data of a possible 
        meteor-triggered sprite/jet event.  The data was recorded 
        with both white-light and blue-light Xybion video cameras 
        with millisecond time-stamping, a gps time-tagged broadband 
        red photometer with 50 microsecond time resolution, and a VLF
        receiver.  The event was collected on August 1, 1998 during 
        the SPRITES '98 campaign at Yucca Ridge Field Station 
        operated by FMA Research in Ft. Collins, CO. The event 
        consisted of three stages, (1) the observation of a 
        moderately bright meteor, (2) the development of a sprite in 
        the immediate vicinity of the meteor as the meteor reached 
        the 60 - 70 km altitude range, and (3) a slower forming ``jet''
        of luminosity that appeared during the late stages of the 
        sprite and propagated back up the ionization trail of the 
        meteor.  The event is analyzed in terms of its geometry, its 
        association with the meteor, and the implications to existing 
        theories for sprite and jet formation.}
}
@ARTICLE{rairden_fagu98,
  AUTHOR = {Rick Rairden and Stephen Mende},
  TITLE = {Intensity-Calibrated Sprite Observations from New Mexico},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F135},
  MONTH = {November},
  ANNOTE = {A31A-03 poster},
  ABSTRACT = {Image intensified video cameras fielded at the Langmuir
        Laboratory near Socorro NM recorded numerous Sprite events
        during the summer 1998 campaign.  The Lockheed Martin/UC
        Berkeley effort involved field testing of a new CID camera,
        and ground-based trials of a NASA camera, veteran of the Space
        Shuttle Tethered Satellite missions.  Calibrated intensity
        results will be presented for white light and red-filtered
        imagery obtained July 19-22.}
}
@ARTICLE{hardman_fagu98,
  AUTHOR = {Simon F. Hardman and Richard L. Dowden and James
		  B. Brundell and John L. Bahr and Zenichiro Kawasaki
		  and Craig J. Rodger},
  TITLE = {Sprites in {A}ustralia's {N}orthern {T}erritory},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F135},
  MONTH = {November},
  ANNOTE = {A31A-04 poster},
  ABSTRACT = {Sprites have been observed near Darwin in Australia's
        Northern Territory (NT).  These are the first confirmed ground
        observations of sprites outside America.  The spatial and
        temporal properties of sprites seen in the NT are similar to
        those measured in the United States.  A number of 'dancing
        sprites' were observed, where a succession of sprite columns
        form and decay, appearing to move across the sky.  On one
        night almost all the NT dancing sprites progressed from right
        to left, or towards a more southerly bearing, suggesting that
        the direction of dancing sprite movement is determined by
        large scale cloud structure.}
}
@ARTICLE{fullekrug_fagu98,
  AUTHOR = {Martin Fullekr\"ug},
  TITLE = {{ULF/ELF} Magnetic Fields of Sprite-Associated
		  Lightning Flashes},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F135},
  MONTH = {November},
  ANNOTE = {A31A-05 poster},
  ABSTRACT = {During the sprite campaign 1998, the Institut f\"{u}r
        Meteorologie und Geophysik at the Universit\"{a}t
        Frankfurt/Germany deployed a network of 3
        instruments to measure horizontal magnetic
        field variations of sprite-associated lightning flashes
        at ULF/ELF frequencies ranging from 0.01 Hz to 1 kHz.
        The measurement instruments are
        located in Santa Cruz/California, Socorro/New Mexico and
        Saskatoon/Canada.
        The network is
        GPS time synchronized with an accuracy of 20 usec
        between stations and enables the
        triangulation of lightning flashes with continuing current
        by use of time of arrival difference techniques.
        The slow tails of sprite-associated lightning flashes
        are often associated
        with slowly varying magnetic fields on the order
        of 100-200 ms which are extremely efficient in the
        excitation of globally observable Earth-ionosphere
        cavity resonances. The measurments at three different
        stations are interpreted as quasi-static magnetic fields
        of sprite-associated currents and
        physical properties of these currents will be derived.
        Some of the sprite-associated lightning flashes
        exhibit additional
        occurrences of ultra-slow tails in the Pc1 frequency range
        \mbox{(0.2-5.0 Hz)}.
        The measurements at three different stations are used to
        quantify the directional dependence
        of ultra-slow tails along and across the
        magnetic field lines and
        an interpretion in terms of discrete
        excitations of ionospheric Alfven resonances
        will be discussed.}
}
@ARTICLE{huang_fag98,
  AUTHOR = {E. W. Huang and E. R. Williams R. A. Boldi and
		  S. J. Heckman and W. A. Lyons and T. E. Nelson and
		  M. J. Taylor and C. T. Wong},
  TITLE = {Criteria for Sprites and Elves based on {S}chumann
		  {R}esonance Observations},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F135},
  MONTH = {November},
  ANNOTE = {A31A-06 poster},
  ABSTRACT = {Ground flashes with positive polarity associated with
        both sprites and elves excite the Earth's Schumann resonances
        to amplitudes several times greater than the background
        resonances.  Theoretical predictions for dielectric breakdown
        in the mesosphere are tested using ELF methods to evaluate
        vertical charge moments of positive ground flashes.
        Comparisons of the measured time constants for lightning
        charge transfer with the electrostatic relaxation time at
        altitudes of nighttime sprite initiation (50--70~km) generally
        validate the electrostatic assumption of predictions made
        initially by C.T.R. Wilson.  The measured charge moments
        (200--2000~C-km) are large in comparison with ordinary
        negative lightning, but are insufficient to account for
        conventional air breakdown at sprite altitudes.  The measured
        charge moments however are sufficient to account for electron
        runaway breakdown, and the long avalanche length in this
        mechanism also accounts for the exclusive association of
        sprites with ground flashes of positive polarity.  The
        association of elves with large peak currents (50--200~kA)
        measured by the National Lightning Detection Network in a
        bandpass beyond the Schumann resonance range is consistent
        with an EMP mechanism for these events.}
}
@ARTICLE{satori_fag98,
  AUTHOR = {G. Satori},
  TITLE = {El Nino Related Variations of Global Lightning
		  Activity as Shown by Schumann Resonances},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-07 poster},
  ABSTRACT = {Schumann resonance (SR) frequencies and amplitudes
        have been recorded for the vertical electric
        component at Nagycenk Observatory in Hungary since
        May of 1993. The excitation source of SR is the 
        global lightning activity which is responsive to
        surface air temperature. The SR frequencies are
        indicative to the areal extension of global
        lightning activity  and the variations of source-
        receiver geometry, as well as to the speed of 
        north-south annual migration of thunderstorm 
        regions. All these parameters exhibit El Nino
        related variations as shown by SR observations. 
        The north-south asymmetry of the land/ocean 
        ratio plays an important role in the El Nino
        related redistribution of the temperature
        dependent global lightning activity.   }
}
@ARTICLE{schlegel_fagu98,
  AUTHOR = {K. Schlegel and M. Full\"ekrug},
  TITLE = {Changes of {S}chumann-{R}esonance Parameters During
		  High energy Solar PArticle Events},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-08 poster},
  ABSTRACT = {Regular Mail}
}
@ARTICLE{woodard_fagu98,
  AUTHOR = {R. Woodard and M. J. Heavner and D. R. Moudry and
		  D. D. Sentman and E. M. Wescott and J. S. Morrill
		  and C. Siefring and E. J. Bucsela and D. L. Osborne
		  and J. T. Desroschers and H. Nielsen and J. Winick
		  and J. Kristl and T. Hudson and L. M. Peticolas and
		  V. Besser},
  TITLE = {Spatial Variation of Ion and Neutral Emissions in Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-09 poster},
  ABSTRACT = {The EXL98 aircraft observations included both
		  broadband cameras and narrowly filtered cameras.
		  The paper will present a comparison between a camera
		  with response such that it primarily measures
		  spatial molecular nitrogen first positive (neutral)
		  emissions and a camera filtered for 427.8 nm ionized
		  molecular nitrogen first negative emissions.  The
		  comparison of the spatial region of neutral
		  emissions to the spatial region of ionized emissions
		  is a crucial issue for estimations of the total
		  energy of sprites.  A wide range of variability
		  between the two cameras has been observed.  An
		  example of a sprite which looks almost identical in
		  the two systems will be presented, and an example of
		  a sprite which appears in the broadband camera but
		  not in the 427.8 nm camera will also be presented.
		  In addition to several examples of sprites observed
		  in both camera systems, a detailed analysis of the
		  responses of the two systems will be presented.}
}
@ARTICLE{moudry_fagu98,
  AUTHOR = {D. R. Moudry and M. J. Heavner and D. D. Sentman and
		  E. M. Wescott and J. S. Morrill and C. Siefring},
  TITLE = {Morphology of Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-10 poster},
  ABSTRACT = {The currently documented forms of sprites span the range from
         rather simple columniform sprites (c-sprites) in the form 
        of vertical lines to much more complex forms. Some are 
        popularly referred to as ``carrot-shaped'' sprites, as 
        ``wishbone'' sprites, or as ``dancing sprites'', where one 
        sprite is followed several ms later by another, then another.
        In addition, within the sprite structures, more intense 
        balls and branches of illumination are sometimes visible. 
        This study looks at the data gathered by University of 
        Alaska in the summers of 1995, 1996 and 1998, and attempts 
        to categorize sprite structures by their shape from the 
        simplest to the most complex forms. This work builds on the 
        preliminary morphology study done by Desrochers, EOS 
        Supplement, 76, 46, November, 1995.}
}
@ARTICLE{osborne_fagu98,
  AUTHOR = {D. L. Osborne and J. Tobolski and D. Sentman and
		  E. Wescott and J. Winick and C. Siefring},
  TITLE = {The {EXL}98 Aircraft},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-11 poster},
  ABSTRACT = {Airborne measurements of sprites, jets, and other middle
        and upper atmospheric electrical effects of lightning permit
        diagnoses of spectral regions not accessible from the ground
        because of atmospheric absorption and scattering, as well as
        giving direct visual access to the associated thundercloud
        tops in certain flight profiles. The EXL98 project utilized a
        Gulfstream 2 business class jet aircraft to study sprites from
        flight altitudes of 12 to 13.5 km. The aircraft carried a
        suite of 8 separate co-aligned low light level television
        imaging systems with spectral responses covering a wide range
        of wavelengths 340 nm to 4.3 microns. In this paper we
        describe these instruments and the corresponding optical
        requirements for the aircraft viewing ports. The aircraft
        structural and power engineering and safety issues involved in
        creating the EXL98 aircraft are discussed, along with the
        challenges they presented for performing instrument
        integration, in a medium altitude aircraft environment, which
        must be compliant with Federal Aviation Administration (FAA)
        regulations. We describe the Global Positioning Satellite
        (GPS) technology that was utilized to provide both in-flight
        synchronization of the aircraft television systems, and
        synchronization of the aircraft instruments with University of
        Alaska ground imaging systems on Mt. Evans, CO and Jelm
        Mountain, WY. }
}
@ARTICLE{benesch_fagu98,
  AUTHOR = {W. M. Benesch and J. S. Morrill and C. Siefring and
		  E. J. Bucsela and J. H. Bowels and M. J. Heavner
		  and D. R. Moudry and D. D. Sentman and E. M. Wescott
		  and D. L. Osborne and J. T. Desrochers and
		  H. Nielsen and L. M. Peticolas and J. Winick and
		  J. Kristl and T. Hudson},
  TITLE = {Overview of {NUV} Observations During {EXL} 98},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-12 poster},
  ABSTRACT = {During the recent EXL 98 mission to observe sprites and other
        related phenomena from aircraft, some of the video cameras 
        were configured to make narrow band and spectral observations
        of emissions in the near ultra-violet (NUV) and blue spectral
        region(~3200 - 5000$\AA$).  The spectral features in this region 
        are primarily due to emissions from neutral and ionized 
        states of N$_{2}$, namely the 2PG and 1NG band systems.  The need
        for aircraft observations arises from the effect of 
        atmospheric attenuation in this spectral region when 
        observations are made from lower altitudes and longer slant 
        paths.  These images and spectra will be used to estimate 
        electron energies as well as examine sprite morphology.  In 
        this presentation, we will discuss NUV/Blue measurement 
        philosophy and motivation, the details of the instrumental 
        techniques and present a variety of observations made during 
        the EXL 98 mission.}
}
@ARTICLE{green_fagu98p,
  AUTHOR = {B. D. Green and W. T. Rawlins and M. E. Fraser},
  TITLE = {Kinetics of Excitation of Infrared Fluorescence by Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-13 poster},
  ABSTRACT = {Regular Mail}
}
@ARTICLE{siefring_fagu98,
  AUTHOR = {C. L. Siefring and P. A. Bernhardt and J. S. Morrill
		  and D. D. Sentman and E. M. Wescott and
		  M. J. Heavner and D. L. Osborne and E. J. Bucsela},
  TITLE = {Correlation Between Ambient Near InfraRed ({NIR})
		  Airglow and Sprite Structures},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-14 poster (moved to talk?)},
  ABSTRACT = {An Indium-Gallium Arsinide (INGAAS) NIR camera (spectral
        range 900 and 1700 nm) was flown on the EXL98 missions to
        monitor hydroxyl airglow which gives indications of density
        structures in the neutral atmosphere, i.e., gravity waves.
        The sensitivity of the NIR camera combined with the brightness
        of the airglow in the NIR, and the fact that the camera was
        operated from an airborne platform at low elevation angles,
        makes these observations quite unique.  These airglow
        observations span a much greater altitude range than is
        typical for ground based visible or NIR measurements.  The NIR
        experiment was extremely successful and made the first
        measurements of NIR emissions from Sprites.  Also of
        importance are cases where a clear correlation was seen
        between structures in the neutral density and the
        shape-and-form of Sprite emissions (both in the visible and
        the NIR).  In a number of cases the terminal altitude and
        bright and dark bands in the Sprites where aligned with the
        ambient airglow.  Also the characteristic tilt of the Sprites
        was often aligned perpendicular to gravity wave fronts.  We
        will discuss these observations and their implications, e.g.,
        these observations would appear to support 'classical'
        breakdown mechanisms as opposed to a runaway breakdown
        mechanism.
        
        *The work at NRL was sponsored by NASA and ONR.}
}
@ARTICLE{deehr_fagu98,
  AUTHOR = {C. S. Deehr and E. M. Wescott and D. D. Sentman and
		  H. C. Stenbaek-Nielsen and M. J. Heavner and
		  D. R. Moudry and C. L. Siefring and J. S. Morrill
		  and E. J. Bucsela},
  TITLE = {New Evidence for Ionization of Blue Starters and
		  Blue Jets},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F136},
  MONTH = {November},
  ANNOTE = {A31A-15 poster},
  ABSTRACT = {Blue jets and blue starters were first documented by
        1994 aircraft observations of a severe storm near Texarkana,
        Arkansas. They appear to propagate upward out of the top of
        thunderstorms at speeds of order 100 km/sec.  Blue jets
        propagate up to about 40 km but blue starters, which resemble
        blue jets, terminate abruptly after only a few km of upward
        travel.  Single blue jets or starters have been recorded from
        aircraft near Kansas and Central America.  During the EXL98
        aircraft observations of July 1998, numerous blue starters
        were recorded on two evenings over severe storms in the upper
        Midwest.  We have previously presented evidence from color TV
        images of blue jets and starters that suggested that the blue
        light must have an ionized N2 component. The 1998 observations
        included both white light, 427.8 nm (N2+1N) and 340 nm N2 2PG
        band filtered images.  The 427.8 nm filter was carefully
        designed to exclude the N2 2PG band emissions at 426.8 nm.
        The 340 nm images show that the blue starters are not upward
        lightning flashes.  The 427.8 nm data clearly demonstrate that
        the blue starters are ionized, and by association that blue
        jets are also ionized beams. The knowledge of the ionization
        allows a better estimate of the energy involved and effects on
        the upper atmosphere by these phenomena. We compare the
        lightning activity and hail in the 1998 storms with the
        previous observations and conclusions.}
}
@ARTICLE{gerken_fagu98,
  AUTHOR = {E. A. Gerken and U. S. Inan and
		  C. P. Barrington-Leigh and M. Stanley},
  TITLE = {Results from a New Telescopic Imager: A Survey of
		  Sprite Structures},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-16},
  ABSTRACT = {Two intensified CCD video cameras were coupled with a 40
        cm aperture Newtonian telescope with a Dobsonian mount to
        create a new, high-resolution sprite imager, called the
        Dobsonian Sprite Experiment (DSE).  The two cameras of the DSE
        have fields of view of $0.92^\circ\times 0.7^\circ$ and
        $12^\circ\times9^\circ$ to allow both a broad view of each
        sprite and detailed imagery of one region.  The experiment was
        fielded during July and August, 1998 at the Langmuir
        Laboratory for Atmospheric Research near Socorro, New
        Mexico. Simultaneous Very Low Frequency and Extremely Low
        Frequency magnetic field recordings were made at Langmuir and
        at Stanford, California.  During this period many sprites were
        observed at ranges of up to 900 km.  This paper reports on the
        morphology of observed sprites, their association with storm
        activity and sferic waveforms, and the variety of fine
        structure observed in the telescopic imager.}
}
@ARTICLE{remick_fagu98,
  AUTHOR = {K. J. Remick and D. D. Sentman and E. M. Wescott and
		  M. J. Heavner and D. R. Moudry},
  TITLE = {Small Scale Strcture in Sprite Tendrils},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-17 poster (withdrawn)},
  ABSTRACT = {High resolution images of sprites reveal that they often
         possess a tendril structure attached to a much brighter main
         body at higher altitude.  When they are present the tendrils
         originate at altitudes of about 60-65 km, extend downward to
         45-50 km and exhibit a variety of structural detail and scale
         sizes, from wispy and barely discernible forms to narrow,
         straight or curved channels, or forked lightning-like fractal
         patterns.  To date the University of Alaska video imagers
         have resolved this structure to a few tens of meters.  In
         this talk we roughly categorize the various structures using
         a topological metric applied to images obtained during
         various University of Alaska ground and aircraft field
         campaigns, taking into account the image resolution and
         effective point spread function of the camera systems derived
         from stellar images.  Results of this analysis are used to
         assign an effective cross section to the tendrils at given
         altitudes, thereby permitting improved estimates to be made
         of the optical emissivity of the tendrils compared to
         previous estimates based on the overall dimensions of
         sprites.}
}
@ARTICLE{sukhorukov_fagu98,
  AUTHOR = {A. I. Sukhorukov and P. Stubbe},
  TITLE = {{QT} whistler excitation by strong lightning},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-18 poster},
  ABSTRACT = {We propose that, in addition to the conventional 
        quasi--longitudinal (QL) whistlers, the sprite/elves 
        triggered lightning can produce the quasi--transversal 
        (QT) whistlers, a mode which does not appear in the 
        linear Earth--ionosphere transmission. The source of 
        the excitation is the small--scale nonlinear transient 
        current in the D-E regions, associated with fine 
        structure of sprite/elves--discharges. An analytical 
        excitation model is presented. The QT mode has a number 
        of peculiar signatures, such as a large $E_\parallel$ 
        component accelerating electrons along the geomagnetic 
        field, strong link to the lower hybrid resonance, much 
        smaller dispersion at frequencies $\omega \gg \omega_{LH}$ 
        and larger group velocities than those of the QL whistler. }
}
@ARTICLE{levtov_fagu98,
  AUTHOR = {S. J. Lev-Tov and U. S. Inan and T. F. Bell},
  TITLE = {Possible Mechanisms of ``Early/fast'' {VLF} events
		  caused by lightning},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-19 poster},
  ABSTRACT = {Subionospheric very low frequency (VLF) waves provide a 
        powerful remote sensing tool for detection of changes in the 
        lower nighttime ionosphere.  For example, changes in D-region 
        electron density or temperature are manifested in changes in 
        VLF wave amplitude and phase.  One class of events in which 
        the wave amplitude and phase change in a very unique manner 
        is the so-called ``early/fast'' VLF events.  Early/fast events 
        were first recognized fifteen years ago and are clearly 
        associated with lightning discharges.  They are ``early'' in 
        that they occur within 20ms of the causative lightning 
        discharges; and they are ``fast'' in that the VLF signal change 
        takes place over about 20ms.  In recent years, some of these 
        events have been found to be associated with sprites.  
        Although several physical mechanisms of VLF early/fast events 
        have been put forth, none of these has reproduced the 
        magnitude of some of the larger signal changes observed in 
        early/fast VLF events.  In this paper we analyze new data 
        sets of ``early/fast'' events with and without sprites.  We 
        also explore possible mechanisms for the production of the 
        early/fast event through  conductivity changes in the 
        ionospheric region overhead the causative lightning (and, if 
        present, sprites).  Sophisticated three-dimensional 
        subionospheric VLF propagation and scattering routines are 
        used to quantitatively interpret effects of  possible 
        ionospheric profiles causing early/fast events.}
}
@ARTICLE{dowden_fagu98,
  AUTHOR = {R. D. Dowden and J. B. Brundell and M. A. Cliverd
		  and C. J. Rodger},
  TITLE = {Decay Rates of the Perturbation Phasors of Low and
		  High Latitude Trimpis},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-20 poster},
  ABSTRACT = {Sprite-associated amplitude and phase perturbations of
        subionospheric VLF transmissions (``Trimpis'') logged as a
        function of time can be transformed to scattered amplitude and
        phase versus time by simple phasor subtraction of the
        unperturbed phasor. The form of the scattered amplitude decay
        with time is related to the form of the plasma decay rate with
        respect to altitude. Over the range of at least 50 km to 80
        km, attachment is the dominant loss term so at any altitude
        the plasma decays exponentially with time. Neglecting a small
        temperature dependence, the attachment coefficient is
        proportional to the square of the neutral number density and
        so the decay rate decreases exponentially with a scale height
        of 3.5 km. The reflection coefficient of sprite plasma at VLF
        is a strong function of plasma density only near a
        ``threshold'' value. Thus contributions to the total scattered
        VLF signal from plasma at altitudes below where the plasma
        density is instantaneously passing through this threshold
        value are negligible, while contributions from higher
        altitudes are still as large as they have been since plasma
        formation. This transition in reflection coefficient from
        essentially unity to essentially zero occurs over only a few
        km.
        
        Here we examine very strong Trimpis of high signal/noise ratio
        (SNR) of two types.  The Type 1 Trimpis we examined were
        ``fast'', occurred on entirely tropical paths, from NWC (21
        deg, 48' S, 114 deg, 9' E) to Darwin area (12 deg, 26' S, 130
        deg, 59' E) during the lightning season and were associated
        with sprites.  We assume that these were produced by sprite
        plasma.  The Type 2 Trimpis were observed at Faraday and
        Rothera, Antarctica, on both NAA (44 deg, 39' N, 67 deg, 17'
        W) and NPM (21 deg, 25' N, 158 deg, 9' W).  From this we
        deduced that the scatter (or perturbing) source for these Type
        2 Trimpis was near Antarctica in a region devoid of lightning
        (and presumably of sprites) and so produced by electron
        precipitation.
        
        For both types, the rate of exponential decay decreased with
        time.  From the form of this decay we deduced that the plasma
        extends down from the base of the ionosphere to at least 60 km
        in the case of Type 1 Trimpis and to about 75 km in the case
        of Type 2 Trimpis. }
}
@ARTICLE{miyamura_fagu98,
  AUTHOR = {K. Miyamura and I. Nagano and S. Yagitani and
		  B. Barnum and R. H. Holzworth and M. C. Kelly},
  TITLE = {Wave forms of lightning-generated {VLF} waves:
		  {C}omparison between a full-wave analysis and rocket
		  measurements},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-21 poster},
  ABSTRACT = {A new full-wave technique has been developed to
        rigorously compute space and time evolution of an ELF/VLF
        electromagnetic wave in the ionosphere radiated from a current
        source located above the ground.  We can examine detailed wave
        forms as well as dynamic spectra of the linear electromagnetic
        wave fields at an arbitrary location in a horizontally
        stratified ionosphere.  This technique has been applied to
        study propagation of VLF waves generated by various kinds of
        lightning strokes modeled with transmission line currents,
        especially to examine their role in generation of luminous
        emissions above lightning such as Sprites and Elves.
        
        By using the full-wave technique in this study we investigate
        in detail space and time evolution of lightning-generated VLF
        electromagnetic fields in the ionosphere, to discuss their
        linear propagation characteristics.  The computed VLF wave
        forms are directly compared with those actually measured over
        thunderclouds by the Thunderstorm-III rocket launched from
        Wallops Island.  This enables us to estimate possible
        magnitude and shape of lightning strokes which generated the
        VLF waves measured onboard the rocket.  Quantitative
        examination of amplitudes and phases of those wave forms would
        even clarify nonlinear response of the ionosphere to an
        extremely intense VLF wave generated by a strong lightning
        stroke.}
}
@ARTICLE{veronis_fagu98,
  AUTHOR = {G. Veronis and V. P. Pasko and U. S. Inan},
  TITLE = {Characteristics of Mesospheric Optical Flashes
		  (Elves) Produced by Lightning Discharges},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-22 poster},
  ABSTRACT = {A new two-dimensional cylindrically symmetric
        electromagnetic model of the lightning-ionosphere
        interaction includes effects of both the lightning radiated
        electromagnetic pulses (EMP) and the quasi-electrostatic
        (QE) fields, thus allowing effective studies of
        lightning-ionosphere interactions on time scales ranging
        from several $\mu$s to tens of ms. The temporal and 
        spatial evolution of
        the electric field, lower ionospheric electron density and
        optical emissions calculated with the new model are used to
        theoretically investigate the effects of the lightning
        return stroke current waveform (i.e., the current rise and
        fall time scales) and of the observational geometry on the
        optical signals observed with a photometer. For typical
        lightning discharges of $\sim$100$\mu$s duration the
        ionospheric response is dominated by the EMP induced heating
        leading to the highly transient and laterally expanding
        optical flashes  known as elves. The optical signal
        characteristics are found to be highly sensitive to both the
        observational geometry and the current waveform. The onset
        delay with respect to the lightning discharge, the duration
        and  the peak magnitude of optical emissions are highly
        dependent on the elevation and azimuth angles of field of
        view of individual photometric pixels. The shape of the
        optical signal clearly reflects the source current waveform.
        For a waveshape with relatively slow rise time
        ($\sim$50$\mu$s) a double pulse shape of the photometric
        signal is observed. 

}
}
@ARTICLE{rousseldupre_fagu98,
  AUTHOR = {R. A. Roussel-Dupre and E. M. D. Symbalisty and V. Yukhimuk},
  TITLE = {Initiation of Intra-Cloud Discharges by Runaway Air
		  Breakdown},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F137},
  MONTH = {November},
  ANNOTE = {A31A-23},
  ABSTRACT = {Simulations of intra-cloud discharges initiated by
        runaway air breakdown are presented.  Results of new kinetic
        calculations (E.M.D. Symbalisty, R. Roussel-Dupré, and
        V. Yukhimuk, IEEE Transactions on Plasma Science, scheduled to
        appear in October, 1998 issue) are used to characterize the
        relativistic electron distribution function.  Both
        conventional and runaway breakdown are included in the
        simulations; however, the discharge is initiated by a runaway
        streamer that develops when the electric field strength
        exceeds one-third the threshold for conventional breakdown.
        Optical, radio-frequency, and gamma emissions are computed and
        compared with existing data.  Other details of the discharge
        including typical temporal and spatial scales and the
        development of steady-state saturation electric fields will be
        discussed in the context of observations.}
}
@ARTICLE{grard_fagu98,
  AUTHOR = {R. J. L. Grard},
  TITLE = {The Role of Turbulence and Convection in the
		  Electric Charging of Thunderstorm Clouds},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-24 poster},
  ABSTRACT = {The electric current which circulates downwards in the Earth 
        atmosphere results from the motion of positive and negative
        ions drifting in opposite directions under the influence of
        a predominantly vertical potential gradient. It is observed
        experimentally that a body, such as a balloon or a gondola,
        moving upwards against the electric field collects an
        excess of positive ions; conversely, a falling body such
        as a water drop, acquires a negative charge. This mechanism
        just requires that the velocity of the body exceed that of 
        the drifting small ions, which is less than a few cm/s at
        altitudes under 30 km in a fair weather atmosphere. In a 
        similar way, parcels of hot and cold air ascending or
        descending in a cloud are selectively charged. Some ions
        are captured by aerosol particles and form large ions which,
        due to their extremely reduced mobility, are subsequently 
        transported with the draft against the force applied by the
        electric field. Charge separation increases the ambient
        electric field and current density, which in turn enhances
        the charging process. This phenomenon takes place in spite of
        the reduced conductivity which prevails within clouds, due to
        the recombination of small positive and negative ions on 
        aerosol particles. The proposed mechanism is particularly 
        efficient during the growth phase of a thundertorm cloud, 
        where the air motion results essentially from rising eddies.
        It is shown that, under certain conditions, the electric
        field may grow exponentially from a few 10 V/m to a few 100
        kV/m in a period of about 20 minutes and that a typical
        thunderstorm cloud can generate a current of the order of
        1 A.}
}
@ARTICLE{lehtinen_fagu98,
  AUTHOR = {N. G. Lehtinen and U. S. Inan and T. F. Bell},
  TITLE = {Effects of Upward Driven Runaway Electrons in the
		  Conjugate Hemisphere: {C}onjugate {S}prites?},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-25 poster},
  ABSTRACT = {Runaway acceleration of relativistic electrons driven by 
        mesospheric quasi-static electric fields following large
        positive cloud-to-ground lightning discharges has been put
        forth as a possible physical mechanism for sprites and 
        terrestrial gamma-ray flashes. The intense and upward 
        propagating runaway electron beams are ultimately trapped
        along the Earth's magnetic field lines and travel to the
        geomagnetically conjugate region. At that point, the runaway
        electron beam encounters the Earth's atmosphere and
        is backscattered, producing light, ionization and x-rays,
        very much like a beam of precipitating auroral electrons.
        We calculate the characteristics (energy and pitch angle
        distribution) of the runaway electron beam as a function
        of the intensity of the parent lightning and the geomagnetic
        latitude. We discuss the pitch-angle scattering of the
        electrons due to beam-plasma interaction during their
        propagation along the geomagnetic field line.
        The backscattering of electrons, as well as optical and
        x-ray emissions in the conjugate region is estimated using
        the Monte Carlo approach.}
}
@ARTICLE{yukhimuk_fagu98,
  AUTHOR = {V. Yukhimuk and R. Roussel-Dupre and E. Symbalisty},
  TITLE = {X-ray and Radio Pulses Produced by Red Sprites, simulation
		  results},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-26 poster},
  ABSTRACT = {Upward discharges driven by the runaway air breakdown
  mechanism (A. V. Gurevich, G. M.  Milikh, and R. Roussel-Dupré,
  Phys. Lett. A, 165, 463, 1992) can generate significant x-ray
  emissions as well as radio emissions. The results of numerical
  simulations of the x-ray, radio and optical emissions from Red
  Sprite discharge are presented. The sprite simulations are done
  using the recently recalculated runaway avalanche rates
  (E. M. D. Symbalisty, R.  Roussel-Dupré, and V. Yukhimuk, IEEE
  Transactions on Plasma Science, scheduled to appear in October 1998
  issue). The x-ray emissions are generated as a result of
  relativistic beam deceleration by air, their time evolution and
  angle distribution are calculated taking into account the
  attenuation of x-rays in the atmosphere. The radio emissions are
  produced as a result of the relativistic avalanche growth and may
  consist of one or two pulses. The first pulse precedes the onset of
  the high altitude optical emissions and is associated with the beam
  formation at lower altitude ~ 18-30 km, the source of the second
  pulse approximately coincides in space and time with high altitude
  optical transient. The time separation between two pulses is in
  range 150 - 800 microseconds, but can appear to be different for
  observers because of the spatial separation of their sources. The
  results are compared with observations.}
}
@ARTICLE{barnum_fagu98,
  AUTHOR = {B. H. Barnum and R. H. Holzworth and M. C. Kelley},
  TITLE = {Simultaneous Ground and Rocket Based Optical
		  Measurements of Lightning Flashes},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-27 poster},
  ABSTRACT = {During the Thunderstorm III rocket flight from Wallops
        Island VA, during the night of September 2, 1995, simultaneous
        rocket and ground based measurements were made of Cloud to
        Ground (CG) and Intracloud (IC) lightning flashes.  The
        results of this study show that the CG triggered IC events are
        much brighter at the rocket than the CG events. This is due to
        the source altitude and the incloud light scattering. The
        higher altitude of the IC lightning channels produce much
        brighter optical signals at the rocket.  The associated IC
        events usually occur within 20 milliseconds of the initial CG
        stroke, and initiate from near the top of the CG lightning
        channel, as shown by the Wallops Island LDAR VHF system.
        Because of the close timing between the CG and IC strokes,
        rocket or satellite measurements can easily mistake the
        brighter stroke for the actual CG lightning.
        
        The CG and IC strokes each initiate an upgoing whistler wave
        in the ionosphere, but the CG stroke typically has much
        more power at lower VLF and ELF frequencies. The VLF and ELF power
        spectra give a way to discern the nature of the lightning source for
        space based optical sensors.
        
           Other results of this study show that there is no
        correlation between the NLDN measured peak currents and CG
        stroke optical brightness for over 100 NLDN CG cases at
        Wallops Island.}
}
@ARTICLE{morss_fagu98,
  AUTHOR = {D. A. Morss},
  TITLE = {{SPARKE} ({S}pherical {P}ropagating {A}tmospheric
		  {R}adiative {K}inetic {E}mission): Fireball in the
		  Sky?},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-28 poster},
  ABSTRACT = {Sprite imaging using low light, image intensification
        video cameras in the Omaha, Nebraska area on 22 August 1996
        revealed a phenomena that was unexplained when reported at the
        Dec 1996 AGU Fall meeting in San Francisco. At that time, a
        number of probable causes had already been ruled out, e.g.,
        space debris re-entry, meteorites, aircraft, helicopters, and
        insects. Estimates of range and speed were speculated at that
        time based upon the probable distances from and direction to
        the nearest thunderstorms on that evening. Assuming a
        thunderstorm-related initiation, it was estimated that SPARKE
        seemed to have originated some 750 km away, and traveled with
        a speed of approximately 3000 km/sec.  Subsequent
        investigation of sources and data associated with atmospheric
        ``lights'' included the general area of ball lightning. While
        most of the reported sightings and probable mechanisms of
        origin did not seem applicable to SPARKE, the reports of
        microwave simulation warranted further investigation.
        Consequently, tests were conducted using a small microwave
        oven with a carbon source placed within the cavity.  The
        phenomena seen in the oven led to video capture of the
        activity using the multispectral image intensification video
        cameras.  It then became a task of examining videos of this
        observed phenomena at very slow speed to find similarities
        (albeit at much reduced power and speed) to SPARKE.  Given
        that up-scaling the power of the microwave and sizes of the
        carbon sources is not outside the bounds of reality, the
        phenomena morphology observed during the microwave experiments
        may similarly be possible in nature.  We thus conclude that it
        is entirely conceivable that the combination of events
        necessary to generate SPARKE can be explained with natural
        events associated with thunderstorm lightning.  Details of the
        experiments and images of the associations will be presented
        at the conference.}
}
@ARTICLE{valdivia_fagu98_b,
  AUTHOR = {Juan A. Valdivia and Gennady M. Milikh},
  TITLE = {Gamma Ray Spectra due to thunderstorms},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-29 poster},
  ABSTRACT = {Recent observations of high energy photons (x-rays,
        gamma-rays),of atmospheric origin, at different altitudes have
        revealed some interesting phenomena. Measurements include
        balloons [Eack et al., 1997] as well as satellite observations
        [Fishman et al., 1994]. These observations may provided us
        with the possibility of discovering the underlying physical
        process in the generation of the gamma rays. It is generally
        accepted that these gamma rays are related to runaway
        electrons produced by thunderstorms, but there are a number of
        possible mechanism capable of producing such runaways
        (Quasi-static fields, electromagnetic pulses, whislers, etc).
        
        The objective of this report is to analyze the spatial and
        temporal evolution of the spectrum of high energy photons
        produced at different heights and by different initial source
        distributions. This model, that include photon diffusion
        (Compton scattering) and photo ionization, can be used to find
        the location of the runaway source. The different models of
        the generation of the runaway electrons are used to compute
        the source spectrum of the high energy photons, which are then
        compared with observations using the photon propagation model
        described above.
        
        Fishman et al., Science, 264, 1313, 1994.}
}
@ARTICLE{eack_fagu98,
  AUTHOR = {K. B. Eack and D. Suszcynsky and W. Beasley and
		  W. D. Rust and R. Roussel-Dupre and E. Symbalisty},
  TITLE = {A High Time-resolution balloon-borned X-ray Detector},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F138},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31A-30 poster},
  ABSTRACT = {We have designed and flown a balloon-borne x-ray detector
        with high time resolution to look for the existence of 
        runaway breakdown processes in sprites.   A secondary goal
        is to obtain additional data on x-ray emissions previously
        observed from inside the thunderstorm, but with greater time
        resolution.  The x-ray spectrometer samples 3 energy bins 
        at a rate of 15kHz.  In addition to the x-ray detector, an
        electric field change instrument was also incorporated into
        the package and data acquisition system to examine the 
        hypothesis that an electrostatic field change above a 
        thunderstorm complex is associated with sprites.
          Although this instrument was designed with specific 
        sensors, the data acquisition and telemetry systems are 
        both much more generic allowing other sensors be used with 
        the existing payload design. }
}
@ARTICLE{kirkland_fagu98,
  AUTHOR = {M. W. Kirkland and D. M. Suszcynsky and S. O. Knox
		  and R. C. Franz and J. Guillen and J. Green and
		  R. E. Spalding},
  TITLE = {Observations of Lightning from space using the
		  {FORTE} photodiode Detector},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F142},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-01 talk},
  ABSTRACT = {We present the initial results obtained from
        observations of optical lightning activity as observed by the
        Photodiode Detector (PDD) onboard the 800 km-circular,
        low-earth orbiting Fast On-orbit Recording of Transient Events
        (FORTÉ) satellite. The PDD is a silicon photodiode sensitive
        to wavelengths from 0.4 to 1.1 microns, with a peak in the
        responsivity at 0.85 microns, and having an 80° field-of-view
        corresponding to a footprint diameter of approximately 1200 km
        on the Earth's surface. This footprint matches that of the
        Lightning Location Sensor and the 3-dB footprint of the
        primary antenna connected to the RF system. In the most
        frequently employed trigger mode, the PDD captures a
        1.9-millisecond time history of transient optical waveforms
        with 15-microsecond resolution. While optical reflections from
        spacecraft or ocean surfaces do result in the detection of
        false events, these false events are easily discriminated from
        optical events attributed to lightning activity. We examine
        the PDD data collected since launch and compare our
        observations to ground-based and aircraft-based observations
        of optical lightning emissions. We present examples of the
        data and infer source optical energies and additional
        propagation path lengths due to clouds.}
}
@ARTICLE{knox_fagu98,
  AUTHOR = {Stephen O. Knox and Abram R. Jacobson and Paul Argo
		  and Robert Franz},
  TITLE = {Correlation of {FORTE} satellite radio-frequency
		  lightning observations with {NLDN} stroke reports},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F142},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-02 talk},
  ABSTRACT = {The FORTÉ satellite has completed a six-month (April -
        September 1998) focused campaign observing lightning over
        North America, in order to study the correlation (if any)
        between radio-frequency emissions and very-low-frequency
        sferics from lightning. FORTÉ is in a 70* inclination,
        circular low-Earth orbit and makes several passes per day
        within radio sight of North America.
        
        FORTÉ records discrete records of very-high-frequency (VHF)
        lightning signatures. A record typically lasts 400 microsec
        and is triggered by a multichannel coincidence trigger.  FORTÉ
        carries a GPS clock and subsecond counter allowing 1-ms
        absolute timing of registered waveforms. On the order of
        ten-thousand such waveforms can be captured and downloaded per
        day.
        
        We are particularly interested in the time relationship
        between FORTÉ VHF records and National Lightning Detection
        Network (NLDN) stroke times. We correct the FORTÉ time-stamp
        back to the putative source (indicated by NLDN stroke report
        rendered with typically $<$50-microsec uncertainty), and
        construct a histogram of NLDN - corrected FORTÉ timestamps. We
        had expected this histogram to reveal a preponderance of FORTÉ
        triggers preceding (by milliseconds to tens of milliseconds) a
        cloud-ground stroke. Instead, we find that the dominant (and
        perhaps only) correlation is prompt. That is, to within our
        methodological uncertainty of ± 100 microsec, when a VHF burst
        seen by FORTÉ is detectably coincident with an associated
        NLDN-reported stroke, then the temporal coincidence is
        indistinguishable from being prompt.
        
        We will discuss these issues as well as what we mean by
        detectably coincident. We will also show the characteristics
        of the promptly-coincident VHF bursts and show to what extent
        they differ from VHF bursts which are not associated with
        sferics.}
}
@ARTICLE{argo_fagu98,
  AUTHOR = {Paul E. Argo and Abram Jacobson and Matt Kirkland
		  and Robert Massey},
  TITLE = {Initial Comparisons of Optical and Radio Frequency
		  Satellite Observations of Thunderstorms with Ground
		  Base Sferics},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F142},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-03 talk},
  ABSTRACT = {The FORTE satellite, launched 29 August 1997, carries 
        radio-frequency-receivers for the study of lightning, as 
        well as fast photodiode radiometer and CCD imaging 
        detectors.  In the first year of operations we have detected 
        more than 2 million radio-frequency events, and more than 
        one million optical events we believe to be associated with 
        lightning.  Over the continental United States we have used 
        both a Los Alamos based sferics array and the National 
        Lightning Detection Network data to provide ground truths 
        for our measurements.  The results are reminiscent of the 
        ancient story about the blind men feeling the elephant.  
        This is in part due to the fact that each of the detection 
        methods is sensitive to emissions created during different 
        parts of the entire lightning process.  We will discuss 
        this, and use the multi-view perspective  to help understand 
        more of the storm emission processes.}
}
@ARTICLE{wiens_fagu98,
  AUTHOR = {Kyle C. Wiens and Robert S. Massey and Xuan-Min Shao
		  and Marc H. Eberle and Kenneth B. Eack},
  TITLE = {The Los Alamos Electric-Field-Change Sensor Array},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F142},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-04 talk},
  ABSTRACT = {We have deployed an array of electric-field-change 
        sensors, located in New Mexico, to help 
        identify the lightning processes responsible 
        for the VHF RF signals detected by the FORTE 
        satellite. The sensors are located at Los Alamos, 
        Socorro, Roswell, and Tucumcari. Each station consists 
        of a flat-plate antenna followed by an integrating 
        amplifier and a line driver, whose output is recorded 
        by a 1 MS/s 12-bit digitizer in a PC. The time 
        constant of the integrator is 1 ms, and waveforms 
        are recorded for 8 or 16 ms, with 50% or 25% pre-trigger 
        samples, respectively. The trigger circuit is 
        bipolar, and the trigger levels can be controlled 
        by the PC. A GPS receiver and timing card within 
        the PC allow time-tagging of the trigger time to 
        two microsecond accuracy. When taking 8-ms records, 
        the system can trigger on and record events 
        separated by as little as 30 ms for periods of several 
        seconds. The hard disk has a capacity of 2 GB, 
        which is well over 100,000 events.
        
        The PCs run the Linux operating system, and 
        data and control links to Los Alamos are via 
        the internet. We currently download waveform 
        data for events that are reported (within a 
        1.1 ms coincidence time) by three or more stations. 
        Event locations are computed with a simple 
        time-of-arrival algorithm, and are found to 
        agree reasonably well with locations of events 
        observed by the National Lightning Detection 
        Network (NLDN). The array has been operational 
        since mid-May 1998, with very little down-time. 
        On active days, stations may record 30,000 or 
        more events, with well over 5000 events reported 
        by three or more stations.
        
        Data from the array will be reported in this 
        and other papers in this session.}
}
@ARTICLE{shao_fagu98,
  AUTHOR = {Xuan-Min Shao and David A. Smith and Kenneth B. Eack
		  and Robert S. Massey and Kyle C. Wiens},
  TITLE = {Observations of Compact Intracloud Discharges ({CID}s)},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F143},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-05 talk},
  ABSTRACT = {Data from a number of resources were used to study
        thunderstorms and thunderstorm radio emissions that occurred
        in the southwestern US between May and September of 1998.
        These resources included the Los Alamos sferic array,
        instrumentation on-board the FORT\'{E} satellite, the National
        Lightning Detection Network, and pertinent NEXRAD weather
        surveillance radars.  The two former resources are described
        in detail in other papers of this session.
        
        A primary goal of our campaign was to study compact intracloud
         discharges (CIDs), unique and isolated thunderstorm events
         that have been identified as sources of narrow positive and
         narrow negative bipolar electric field change pulses (NPBPs
         and NNBPs), subionospheric pulse pairs (SIPPs), and
         transionospheric pulse pairs (TIPPs).  CIDs are powerful,
         singular discharges that occur in regions of high electric
         field strength and high charge density.  Their physical
         characteristics are described in detail in another paper of
         this session.
        
        Many hundreds of CIDs were observed by the sferic array during
         the summer of 1998 and were located using methods of
         differential time of arrival.  Distances from the discharges
         to stations of the array ranged from just a few kilometers to
         many hundreds of kilometers.  Source heights were determined
         using relative delays between CID groundwave signals and
         reflections from the earth and ionosphere.  In some cases,
         FORT\'{E} data were also used to determine 3-D source
         locations.  CID locations have been compared to those of
         other thunderstorm electrical discharges and to images of
         radar reflectivity.  The events occur in localized regions of
         thunderstorms that also produce regular intracloud and
         cloud-to-ground lightning.  Radar reflectivity data show that
         these localized regions correspond to thunderstorm cores
         where reflectivities in excess of 40 dBZ exist.
        
        Further analysis of data from this summer should provide
        insight into the following questions: Do CIDs of opposite
        polarity occur at different altitudes within thunderstorms?
        Do CIDs occur only in regions of high reflectivity?  Do all
        regions of high reflectivity produce CIDs?  What are the
        temporal and spatial relationships between CIDs and other
        intracloud and cloud-to-ground thunderstorm electrical
        discharges?  Can phases of thunderstorm or different storm
        types be identified as being more or less likely to produce
        the events?}
}
@ARTICLE{krider_fagu98,
  AUTHOR = {E. Philip Krider and Launa M. Maier and Martin
		  J. Murphy and Donal W. Schibert},
  TITLE = {The Onset of Electrification in Florida Thunderstorms},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F143},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-06 talk},
  ABSTRACT = {The NASA Kennedy Space Center (KSC) and USAF Eastern Range
        (ER) currently operate a network of 31 electric field mills
        (the LPLWS system), 7 VHF pulse time-of-arrival receivers
        (the LDAR system), and a 5-station network of gated,
        wideband magnetic direction finders (the CGLSS system) to
        identify possible lightning hazards to launches and to a
        variety of ground operations. The field mill data can be
        used to detect the onset of cloud electrification and also
        to determine the locations of any lightning-caused changes
        in the thundercloud charge distribution. Previous reports
        have shown that intracloud (IC) flashes tend to produce two
        clusters of LDAR sources in convective storms, one typically
        at 7--8 km altitude and the other at 9--10 km. The field
        mill data show that the upper cluster corresponds to a
        region of positive charge and the lower cluster to a region
        of negative charge. The LDAR sources in cloud-to-ground
        (CG) flashes tend to begin at or just below the lower edge
        of the negative region and propagate downward toward a lower
        positive charge region that is near the 0$\deg$C level.
        
        During the onset of electrification, the lower positive
        charge center usually appears first in the electric field
        records that are close to isolated cells, i.e., the electric
        field is directed downward. Later, the field reverses
        polarity, and the effects of the negative charge at higher
        altitude dominate the field pattern under the storm.
        The implication of these results for thunderstorm
        electrification and lightning warnings will be discussed.}
}
@ARTICLE{macgorman_fagu98,
  AUTHOR = {D. R. Mac{G}orman and M. Stolzenburg and W. D. Rust
		  and T. Marshall and P. Krehbiel and R. Thomas and
		  W. Rison and W. Beasley and K. Eack and M. Lockwood},
  TITLE = {Electric Field Profiles and Lightning in Storms with
		  Frequent Positive Cloud-to-Ground Lightning},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F143},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-07 talk},
  ABSTRACT = {Eight balloon soundings of the electric field have been
        obtained in severe storms in which the majority of ground
        flashes lowered positive charge to ground, instead of the 
        more usual negative charge.  These storms occurred in 
        Oklahoma, Texas, and Kansas on four different days.  Two
        soundings were obtained in the strong convection of
        different isolated supercell storms on one day, and one was
        obtained completely through the anvil of yet another
        isolated storm on a second day.   Two soundings were 
        obtained in the convective line of each of two severe squall 
        lines that produced mostly positive ground flashes.  The 
        remaining sounding was in the transition region between the 
        convective line and stratiform region of one of these squall 
        lines.  Preliminary analyses of the soundings are compared 
        with published soundings from storms that produced primarily 
        negative ground flashes.  Cloud-to-ground lightning data 
        from the U.S. National Lightning Detection Network have been 
        analyzed for one of the isolated storms.  The evolution of 
        positive ground flash rates relative to both radar-derived 
        storm parameters and severe weather was similar to that 
        observed by previous studies of isolated severe storms 
        having frequent positive ground flashes.  Lightning also was 
        mapped in three dimensions for one of the isolated storms in 
        which a sounding was obtained and for several other severe 
        storms that occurred in central Oklahoma during field 
        operations.  Initial analysis of mapping data for some 
        positive ground flashes shows that they were bilevel hybrid 
        flashes, each flash having two layers of extensive 
        horizontal intracloud structure.  The layers are inferred to 
        correspond respectively to upper positive charge and lower 
        negative charge in the thunderstorm charge distribution.  
        Later in the storm, bilevel intracloud flashes appeared to 
        occur in an inverted charge distribution, with upper 
        channels propagating through negative charge and lower 
        channels propagating through positive charge.  Cloud flash 
        rates in Oklahoma storms typically were much larger than 
        ground flash rates, particularly during periods when the 
        storms produced severe weather.}
}
@ARTICLE{thomas_fagu98,
  AUTHOR = {Ronald J. Thomas and William Rison and Timothy
		  Hamlin and Jeremiah Harlin and Paul R. Krehbiel},
  TITLE = {3-Dimensional Lightning Observations in Central New Mexico},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F143},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-08 talk},
  ABSTRACT = {Following its initial operation in Oklahoma, Tech's
        3-dimensional lightning mapping system was set up in August of
        1998 over a 60-km diameter area around Langmuir Laboratory and
        the New Mexico Tech campus in central New Mexico.  The results
        show that lightning activity in New Mexico storms has a
        simpler structure and smaller extent than in the larger Great
        Plains storms.  Lightning discharges are also much less
        frequent in a storm, compared to the nearly continous activity
        observed in Oklahoma.  Intracloud lightning often has only a
        vertical channel but in more developed storms shows a clear
        bilevel structure, undoubtedly associated with the main
        negative and upper positive charge regions of the storm.
        These results agree with previous observations both in New
        Mexico and in Florida.
        
        A new finding is that cloud-to-ground discharges sometimes
        show considerable activity in a third region below the
        negative charge level, presumably associated with lower
        positive charge.  The overall lightning activity in the first
        storm to be looked at (the storm of August 18) exhibited a
        trilevel structure, consistent with the tripole model of storm
        electrification, and appears to map out the basic charge
        structure of the storm.  Similar structures have been observed
        in the more complex Oklahoma storms.  A complete analysis of
        the August 18 observations is expected to show the evolution
        of the lightning activity throughout the life cycle of the
        storm and will be compared to dual-polarization radar
        observations of storm structure.  The relation between the
        lightning channel levels and the storm charge regions will be
        investigated at Langmuir Laboratory in the summer of 1999
        using balloon-borne electrostatic field measurements by
        Marshall, Stolzenburg, and Rust.}
}
@ARTICLE{hayakawa_fagu98,
  AUTHOR = {Masashi Hayakawa and Alexander P. Nickolaenko and
		  Irina G. Kudintseva and Stanislav V. Myand and
		  Leonid M. Rabinowicz},
  TITLE = {Natural Sub-Ionospheric {ELF} pulses in Time Domain},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F143},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A31D-09 talk},
  ABSTRACT = {There are problems when constructing the time domain
        solution for the natural electromagnetic signals of the
        extremely low frequency band. Such pulses originate from
        the powerful lightning strokes and propagate in the
        spherical Earth-ionosphere guide for long distances. Since
        the wave attenuation in the lower part of ELF band is
        small, the pulses multiply circle the Earth's circumference
        giving rise to the Schumann resonance phenomenon. Recent
        studies indicate on possible link between the red sprites
        and the Q-burst pulses observed in the Schumann resonance
        band (from a few to some tens Hertz frequencies) or with
        the slow tail atmospherics. The standard way of obtaining
        such a solution is following.  The mode is used in the
        frequency domain. The zonal harmonic series representation
        is applied, which is an expansion of the fields over the
        Legendre polynomials. Then, the fast Fourier transform is
        applied  to obtain the time domain solution. There are two
        areas on the sphere where the solution in the frequency
        domain becomes divergent. These are the vicinities of the
        source and its antipode, where the radio wave focusing
        occurs. To solve the problem everywhere in the spherical
        Earth-ionosphere cavity, we construct the solution directly
        in the time domain. For this purpose we make an analytical
        Fourier transformation of each term in the zonal harmonic
        series representation and then compute the time domain
        fields by direct summing of the series obtained
        analytically.
        
        The main result of the report is following.
        
        (i) One fails to describe the field at the point where the
        source was regardless the time when computing the field in
        the frequency domain first and then applying a FFT
        algorithm. The time domain representation obtained allows
        computing the electromagnetic pulse everywhere in the
        Earth-ionosphere cavity for  t > 0, including those
        multiply crossed the Earth's circumference.}
}
@ARTICLE{sentman_fagu98_a,
  AUTHOR = {D. D. Sentman},
  TITLE = {The Effects of lightning on the middle and upper
		  atmospheres: Snapshot of a Rapidly Evolving Discipline},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F164},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-01 talk (invited)},
  ABSTRACT = {Interest in the electrical effects of tropospheric
        lightning on the middle and upper atmospheres took a sudden
        leap in 1989 with the serendipitous video capture of a
        cloud-to-stratosphere discharge by R. Franz and colleagues.
        Shortly thereafter similar signatures were found by W.  Boeck
        and colleagues in space shuttle video tapes.  Since these
        seminal reports appeared, activity in this new field has
        exploded as other researchers scrambled to determine the
        nature and extent of this new phenomenon. Beginning in 1993
        and continuing to the present, ground and aircraft based video
        studies have amassed thousands of images of these middle
        atmospheric apparitions, now generally referred to as
        sprites, whose luminous flash within the mesosphere, at
        altitudes of 50-90 km, coincides with a large underlying
        positive cloud-to-ground lightning discharge.  Spectral
        measurements have identified the emitting atmospheric species
        of red sprites (N2), and optical signatures of accompanying
        ionospheric heating have been detected (ELVES).  A new, even
        more elusive and apparently unrelated electrical phenomenon
        known as a blue jet was also discovered to spurt optically
        upward directly from the top of a thunderstorm to altitudes of
        about 40 km (130,000 ft). Measurements of electromagnetic
        emissions have identified several distinctive signatures
        associated with sprites: ULF impulses, Q-bursts in the ELF,
        and slow tails in the VLF frequency bands.  Ground-based VLF
        studies have shown that a certain class of gamma ray bursts
        observed from earth orbit originate in thunderstorms, and are
        possibly part of the same process that produces sprites and
        related phenomena.  The focus of current ground and aircraft
        experiments is to extend optical remote sensing into the
        ultraviolet and infrared regimes as a way to determine the
        energy budget of sprites/ELVES/jets, and to study their wide
        band electromagnetic emission properties.  Several satellite
        initiatives are underway to perform topside studies of these
        events unimpeded by the terrestrial atmosphere.  This talk
        will present an overview of past and present observations in
        the rapidly evolving discipline of the effects of lightning on
        the middle and upper atmospheres, and point out several areas
        of possible significance of these processes in the larger
        geophysical system.}
}
@ARTICLE{inan_fagu98,
  AUTHOR = {Umran S. Inan and Christopher P. Barrington-Leigh
		  and Elizabeth A. Gerken and Timothy F. Bell},
  TITLE = {Telescopic Imaging of Fine Structure in Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F164},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-02 talk},
  ABSTRACT = {Recent theoretical work [{\it Pasko et al.}, {\it GRL,
        25}, p. 2123, 1998] has indicated that the sprite optical
        luminosity as observed in regular video may be constituted by
        a superposition of a large number of highly structured
        streamers having fine structure with scale sizes of few
        meters.  To look for the presence of such fine structure, we
        use a 40-cm wide, $\sim$2-m tall Newtonian telescope with a
        field-of-view of $0.92^\circ\times 0.7^\circ$ ($\sim$10-m
        video resolution at a range of 500-km), boresighted with a
        wide field-of-view ($12^\circ\times9^\circ$) video camera.
        The rotatable and tiltable mounting base of the telescope
        provides easy elevation and azimuth control, while aiming of
        the telescope is facilitated by the known phenomenology of
        sprites, i.e., that they often tend to occur at approximately
        the same location for tens of minutes.  Both the wide- and
        narrow-fields-of-view are recorded with image intensified CCD
        video cameras, with simultaneous recording of the broadband
        very low frequency sferic signal from causative lightning
        flashes.  The first observations with the telescopic imaging
        system were conducted from the Langmuir Laboratory annex,
        located near the 11,000 ft high South Baldy peak, providing
        clear views of thunderstorms in Kansas, Colorado, Texas, and
        the north-west coast of Mexico.  Preliminary data from the
        telescope show the presence of a fascinating complex of fine
        structures within the body of the sprite, clearly indicating
        the presence of columnar channels of luminosity with lateral
        extent $<$10-m.  In one case, on July 13, 1998, observing a
        storm ata range of $\sim$580 km, many tens of
        multiply-oriented (but mostly vertical) streamer-like
        structures were simultaneously observed within the field of
        view, with lateral scales ranging from $<$10-m to up to
        100-m.}
}
@ARTICLE{fukunishi_fagu98,
  AUTHOR = {H. Fukunishi and Y. Takahashi and Y. Watanabe and
		  A. Uchida and M. Sato and W. A. Lyons},
  TITLE = {Frequent Occurrences of Elves Discovered by Array
		  Photometer Observations},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F164},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-03 talk},
  ABSTRACT = {Fukunishi et al.[1996] found that Elves occur just after
        the onset of sferics but preceding to sprites. However, during
        the SPRITES'98 campaign carried out at Yucca Ridge Field
        Station, Colorado in July and August 1998, we discovered that
        elves occur very frequently without sprites. For example, we
        identified 52 elves events without sprites in 6 hours during
        the night on August 1-2, 1998. Since the duration of elves is
        about 1 ms, it is difficult to identify them on video images
        except for unusually bright events. However, the
        identification of occurrences of elves is easy by using array
        photometer data since elves always show time delays with
        decreasing elevation angles, but such time delays are not
        observed for cloud flashes. We used two sets of 16-channel
        array photometers set up in vertical and horizontal,
        respectively. Each array photometer has individual
        fields-of-view of 0.7 x 11 degrees and the total field-of view
        of 11 x 11 degrees and the sampling time of each channel is 50
        microseconds. Several important features were found. First,
        elves are excited not only by positive cloud-to-ground
        lightning discharges (CGs) but also by negative CGs. Second,
        the durations of optical flashes measured by the vertical
        array photometer are 200-500 microseconds for individual
        channels and the onset of flashes delays with decreasing
        elevation (time delay from top to bottom is 300-600
        microseconds). Third, the durations of optical flashes
        measured by the horizontal array photometer are 300-900
        microseconds and the onset of optical flashes is simultaneous
        with the almost same luminosity variations at all
        channels. The observed features of elves are consistent with a
        model proposed by Inan et al. [1996] in which the optical
        emissions are produced as a result of heating by
        electromagnetic pulses (EMPs) from lightning discharges.  The
        frequent occurrences of elves suggest that lightning-induced
        EMPs are playing an important role for heating of the lower
        ionosphere.}
}
@ARTICLE{barringtonleigh_fagu98,
  AUTHOR = {Christopher P. Barrington-Leigh and Umran S. Inan
		  and Elizabeth A. Gerken},
  TITLE = {Temporal and Spatial Structuve of Sprites and Elves:
		  Photometric and {CCD} Observations},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F164},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-04 talk},
  ABSTRACT = {Data acquired with a high-speed photometric array (the
        Fly's Eye) combined with high-resolution CCD imaging and Very
        Low Frequency sferic recordings allow the precise measurement
        of the onset delay, development, and duration of red sprites
        and elves.  The two complimentary optical instruments provide
        the first simultaneous documentation of the spatial fine
        structure and submillisecond temporal characteristics of
        sprites and elves.  Elves are unambiguously identified by
        means of their rapid lateral expansion as observed with the
        Fly's Eye and are clearly produced by both positive and
        negative cloud-to-ground flashes. Bright sprites are found to
        have typical durations of 30-100 ms, and the altitude
        dependence of their observed structure and optical properties
        is discussed.  Elves and sprites are detected in both blue and
        red photometers, and preliminary analysis indicates that the
        ratio of blue and red emission intensities is consistent with
        that expected for air breakdown at high altitudes.}
}
@ARTICLE{sentman_fagu98_b,
  AUTHOR = {D. D. Sentman and E. M. Wescott and J. Winick and
		  C. Siefring and J. Morrill and D. Baker and P.
		  Bernhardt and M. Heavner and D. Moudry
		  and D. Osborne and J. Desrochers and L. Peticolas
		  and V. Besser},
  TITLE = {The {EXL}98 Sprites Campaign},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F164},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-05 talk},
  ABSTRACT = {The EXL98 (Energetics of Upper Atmospheric Excitation by
        Lightning, 1998) project was a joint effort by the University
        of Alaska, Air Force Research Laboratory, and Naval Research
        Laboratory to study the energy budget of sprites, jets and
        ELVES using jet aircraft and ground remote sensing methods.
        The EXL98 field campaign was conducted from Colorado and
        Wyoming during July, 1998, with aircraft flights targeted on
        nocturnal thunderstorm systems in the Middle and Upper Plains
        states.  Simultaneous sprite observations were obtained from
        the airborne imaging platform carrying a suite of eight low
        light level television (LLTV) camera systems, and at two
        ground locations equipped with LLTV, a spectrograph, and high
        speed photometers. Campaign and flight mission planning were
        carried out at the Laboratory for Atmospheric and Space
        Physics (LASP) at the University of Colorado.  The
        measurements focused on low light level imaging across a wide
        range of wavelengths, including the near ultraviolet (NUV),
        visible, near infrared (NIR), and medium infrared (MIR).  This
        talk will review the science rationale for the EXL98 project,
        discuss aircraft and ground resources utilized during the
        campaign, and present preliminary results.}
}
@ARTICLE{bowles_fagu98,
  AUTHOR = {Jeffrey H. Bowles and Carl L. Siefring and Jeff
		  S. Morrill and Paul A. Bernahrdt and Davis
		  D. Sentman and Dana R. Moudry and Eric J. Bucsela
		  and Daniel L. Osborne and Eugene M. Wescott and
		  Matthew J. Heavner},
  TITLE = {Do Sprites Sometimes Connect to the Cloud Tops?},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F164},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-06 talk (withdrawn)},
  ABSTRACT = {In this paper we present exciting results from the EXL98
         mission which indicate that Sprites may occasionally connect
         from the ionosphere all the way down to the cloud tops.
         During the EXL98 mission one of the major goals was to obtain
         Near-UltraViolet (NUV) and Near-InfraRed (NIR) measurements
         of Sprites.  Because of atmospheric absorption and
         scattering, to fulfill these goals it was necessary to fly
         the plane close to the storms (100-200km).  As a result we
         obtained many images which view from the horizon (cloud tops)
         into the tendril portion of the Sprites.  The NRL NUV camera,
         running without a filter, observed two cases where it
         appeared that a secondary breakdown started from near the
         horizon and propagated upward toward the remnants of a
         Sprite.  The UAF narrow field visible camera was often
         saturated by scattered light from the lightning flash in this
         area.  However, on a number of occasions this camera showed
         'fingers' extending from the horizon to the bottom of the
         Sprites as it came out of saturation. One of these cases
         corresponds to the observations made with the NUV camera of a
         possible secondary breakdown following the Sprite by several
         video frames.  The evidence for a 'direct' connection from a
         Sprite to the cloud tops is not conclusive, but the
         possibility is extremely important.  We hope these results
         will spur further experimental and theoretical investigations
         into this possibility.
        
        *The work at NRL was sponsored by NASA and ONR.

}
}
@ARTICLE{besser_fagu98,
  AUTHOR = {Veronika Besser and Dana R. Moudry and Matt
		  J. Heavner and Davis D. Sentman and Eugene
		  M. Wescott and Jeff S. Morrill and Carl Siefring and
		  Daniel L. Osborne and Don L. Hampton},
  TITLE = {Case Study of a Sprite},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-07 talk},
  ABSTRACT = {On July 22, 1998, at 04:57:43 UT, a sprite over central Iowa 
        was imaged from the EXL98 aircraft from a very close range. 
        The sprite was captured in six cameras: a wide field (approx.
         60$^\circ$ field of view (FOV)) monochrome camera, and five 
        narrow field cameras (approx. 16$^\circ$ FOV): a monochrome 
        camera, an intensified color camera, a camera filtered at 
        427.8 nm (FWHM 1.6 nm), a near-ultraviolet (340 nm) camera, 
        and a near-infrared (0.8-1.7 $\mu$m) camera. The entire 
        sprite was captured in the wide field camera, the other 
        cameras imaged only the tendrils. 
        \\In addition, approximately 100 ms after the onset of the 
        sprite, spike or finger-like features were seen propagating
         upward from the cloudtops towards the lower portions of the 
        tendrils of the sprite, stopping short of reaching them. 
         This is the first time this feature has been reported.}
}
@ARTICLE{pasko_fagu98,
  AUTHOR = {Victor P. Pasko },
  TITLE = {Theoretical Models of Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-08 talk (invited)},
  ABSTRACT = {Spectacular luminous optical emissions
        appearing in clear air above thunderstorms and known as
        sprites provide dramatic evidence of strong
        electrodynamic coupling between tropospheric lightning
        discharges and the mesosphere/lower ionosphere. In this
        talk we will provide a review of existing interaction
        and coupling mechanisms proposed for the explanation of
        experimentally observed features of sprites (i.e.,
        their spatial structure, optical spectra and time
        dynamics). We will discuss  the conventional as well as
        the relativistic runaway breakdown processes stimulated
        at mesospheric altitudes by large quasi-electroctatic
        field transients following intense positive  cloud to
        ground lightning discharges, and the role of these two
        breakdown mechanisms in sprite production. We will also
        discuss the properties of sprites on a variety of
        spatial scales, including their fine spatial structure
        ($\sim$1 m), as well as their contribution  to the
        large scale atmospheric electric circuit (scales
        $>\sim$100 km). }
}
@ARTICLE{fernsler_fagu98,
  AUTHOR = {Richard F. Fernsler and Harvey L. Rowland},
  TITLE = {Theory and Experiment for Sprites and Elves},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-09 talk},
  ABSTRACT = {Analytic models were previously given [1] for elves and
        red sprites produced in the upper atmosphere, and recent
        experiments provide support for the model predictions. The
        underlying criterion is gas breakdown from the electric fields
        induced by lightning discharges, and results were given for
        both vertical and horizontal discharges. The radiation field
        from a lightning return stroke was found to be capable of
        producing flat, thin, short-lived elves just below the
        ionosphere, provided the return-stroke current exceeds ~50
        kA. Similarly, the quasi-static field from a continuing
        current was found to be capable of producing red sprites
        provided the charge moment (charge times height of charge)
        exceeds ~150 Coulomb-km.  In addition, red sprites are
        predicted to elongate and filament as they propagate down in
        altitude via enhancement of the field at the tip.  Similar
        enhancement does not occur for the radiation fields producing
        elves.  In this talk the two models are briefly reviewed and
        then compared with experiment.
        
        Work supported by NASA and ONR.
        
        1. J. Geophys. Res. 101, 29653 (1996).}
}
@ARTICLE{armstrong_fagu98,
  AUTHOR = {Russell A. Armstrong and David M. Suszcynsky and
		  Robert Strabley and Walter A. Lyons and Thomas
		  Nelson},
  TITLE = {Simultaneous Multi-Color Photometric and Video
		  Recording of {S}PRITES and their parent lightning
		  mechanisms and energy deposition},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-10 talk},
  ABSTRACT = {We have obtained new multi-color photometry data
        simultaneously with blue, red and high-resolution broad-band
        video imagery of sprite events which indicates that more than
        one mechanism can cause sprites.  The time-resolved emissions
        from the nitrogen first-positive, second positive and first
        negative systems indicate different time histories.  We have
        found that the relative time histories are not unique from
        event-to-event but generally fall into two (or perhaps three)
        categories.  We find that the time histories can be
        interpreted to yield both the ionization level and the energy
        accommodation.  Data has been obtained which confirms
        ionization in sprites varying from very strongly ionized to
        relatively weakly ionized.  The simultaneous red and blue
        video images associated with the sprite events indicates that
        columnar sprites exhibit different energetic profiles than
        carrots.  We will present the systematic analysis of the
        time-histories of the photometric emission behavior of the
        sprite events, correlate them with the corresponding imagery
        and discuss implications for the electrodynamic modeling of
        the phenomena.}
}
@ARTICLE{taylor_fagu98,
  AUTHOR = {Mike J. Taylor and Larry C. Gardner},
  TITLE = {Simultaneous Red and Blue Imaging of Sprites During
		  the Yucca Ridge 1998 Campain},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-11 talk},
  ABSTRACT = {As part of the Sprite-98 campaign conducted at Yucca
        Ridge, CO two intensified Image Isocon cameras and a bare CCD
        imager were used to investigate the spectral morphology of
        Sprites.  The CCD and one Isocon camera were filtered to
        observe the nitrogen first positive band emission at 665 nm
        while the second Isocon camera was used to detect any blue
        emissions originating at ~400 nm.  A fourth, Xyibon, camera
        was unfiltered and used to provide reference image data.  All
        four cameras were co-aligned and mounted on a motorized
        alt-azimuth tripod providing accurate pointing information.
        Joint observations were conducted during a one month period
        from 8 July to 5 August, and despite the limited storm
        conditions Sprites were imaged successfully on several nights
        in the red emission but only occasionally were they detected
        simultaneously at blue wavelengths.  An initial comparative
        analysis of the red and blue Sprite signatures and their
        associated lightning discharges will be presented together
        with an unusual observation of light apparently propagating up
        from an isolated Sprite and re-illuminating a meteor trail.}
}
@ARTICLE{heavner_fagu98,
  AUTHOR = {M. J. Heavner and D. R. Moudry and D. D. Sentman and
		  E. M. Wescott and J. S. Morrill and C. Siefring and
		  E. J. Bucsela and D. L. Osborne and J. T. Desrochers
		  and H. Stenbaek-Nielsen and J. Winick and J. Kristl and
		  T. Hudson and L. M. Peticolas and V. Besser},
  TITLE = {Ionization in Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-12 talk},
  ABSTRACT = {The EXL98 sprite observing campaign of July 1998
        involved multiple cameras across the wavelength range 300 nm
        to 5 microns.  A filter at 427.8 nm was used on one camera to
        detect emissions from the first negative group of molecular
        nitrogen, which would be a direct observation of the amount of
        ionization.  The filter was carefully selected to reject 426.8
        nm light and any other second positive (non-ionized) molecular
        nitrogen emissions which have made previous blue filtered
        observations of sprites difficult to interpret with regards to
        the ionization question.  The paper will present observations
        made with the filtered camera and simultaneous EXL98 aircraft
        images as evidence for the observation of ionization.  The
        evidence for ionization will be discussed in the context of
        previous observational campaigns.}
}
@ARTICLE{groves_fagu98,
  AUTHOR = {Keith M. Groves and John M. Quinn and E. Russell
		  Shinn and Peter Ning and Matthew R. Cox},
  TITLE = {An Upper Limit on Ionization in Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-13 talk},
  ABSTRACT = {During the past few years, much progress has been made
        towards characterizing the morphology of red sprites and
        understanding the source mechanisms leading to their
        occurrence. A combination of optical measurements and
        theoretical calculations have been applied to determine the
        levels of ionization produced in sprite discharges, but no
        definitive results have yet been produced. During July-August
        1996 we collected radio frequency (RF) measurements of
        lightning-induced atmospheric effects in northeastern Colorado
        using a 28 MHz coherent backscater radar. Such a radar serves
        as a powerful diagnostic of electron density. A
        CCD-intensified video system was co-located with the radar to
        verify the occurrence of sprites within the radar beam. The
        combined data sets were compared with CG lightning events as
        recorded by the National Lightning Detection Network to
        correlate RF signatures with individual lightning strikes.
        
        More than 200 sprite events were seen simultaneously with the
        radar and the optical system; numerous more events were
        observed with the radar in the absence of good optical
        conditions.  No echoes associated with sprite occurrence were
        observed.  An analysis of the radar sensitivity and a
        consideration of various scattering models enable one to place
        an upper bound on the levels of ionization present in red
        sprites. The resulting ionization values are strongly
        dependent on the assumed structure and extent of the
        scattering region. High resolution optical measurements and
        theoretical analyses are needed to constrain the scattering
        calculations further. Additionally, a number of uncorrelated
        radar echoes were observed during the campaign, sometimes in
        conjunction with thunderstorm activity.  The significance of
        these echoes and their origin will be discussed.}
}
@ARTICLE{swenson_fagu98,
  AUTHOR = {Gary R. Swenson and Rick L. Rairden},
  TITLE = {What is the source of the sprite seed electrons?},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F165},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A41C-14 talk},
  ABSTRACT = {It is clear from spectral measurements of the N2 1PG 
        vibrational distribution originatin from Sprites, 
        that the energy distribution of 
        electrons responsible, had a mean of $~$1 eV.  This energy is
        orders of magnitude colder than auroral secondary electrons.
        This cold source cannot be explained by normally produced 
        secondaries from electron induced ionization of local
        molecules whose ionization thresholds are typically $>$15 eV.  
        A mesospheric source of low energy electrons can originate 
        from negative ions, where electron release energies are 
        typically .5-1.5 eV.  It will be shown that measured 
        brightnesses of Sprite cores are consistent with 
        climatological average negative ion densities available.  
        The remaining question is to explain the souce of 'seed' 
        electrons, which when exposed to the broad electric field, 
        accelerate to remove the electrons from the parent molecules.}
}
@ARTICLE{green_fagu98,
  AUTHOR = {Bryon David Green and Terry Rawlins and Russ Armstrong},
  TITLE = {Molecular Excitation as a Probe of Sprite Mechanisms},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F175},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-01 talk (invited)},
  ABSTRACT = {REGULAR MAIL}
}
@ARTICLE{winick_fagu98,
  AUTHOR = {Jeremy R. Winick and Joseph A. Kristl and Thomas
		  Hudson and Davis D. Sentman and Eugene M. Wescott
		  and Laila Jeong and Carl Siefring and Daniel Osborne},
  TITLE = {Do Sprites produce infrared emission? Preliminary
		  Results from {EXL}98},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F175},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-02 talk (withdrawn)},
  ABSTRACT = {We present the preliminary results of measurements taken
        by MWIR filtered video cameras from high altitude aircraft
        flights during the EXL98 campaign. AFRL and USU Stewart
        Radiance Laboratory provided video cameras filtered for the
        2.8 $\mu{}m$ and 4.3 $\mu{}m$ bands.  These bands are enhanced
        in aurora as a result of emission from chemiluminscently
        produced NO(v) ($\Delta{}$v=2 at 2.7-3.0 $\mu{}m$) and CO$_2$
        ($\Delta{}$v=1 at 4.3 $\mu{}m$) produced from near-resonanat
        vibrational energy transfer from N$_2$(v). These processes are
        initiated primarily by secondary electrons. In sprites we
        anticipate most electrons are heated by discharge electric
        field as opposed to being secondaries from ionization
        processes. Observations of copious N$_2$ first positive
        emission in sprites indicates that electrons of greater than 7
        eV are present. Since the cross section for production of
        N$_2$(v) is very large for electrons of 2-5 eV,enhancement of
        CO$_2$ 4.3$\mu{}m$ is likely. Formation of NO(v$\geq{}2$)
        depends upon higher energy electrons that dissociate and/or
        ionize N$_2$, and its formation in sprites is less
        certain. However, NO(v$\geq{}2$) emission is a useful
        indicator of NO formation and can tell us whether sprites have
        any photochemical impact in the region of the mesosphere where
        the NO density is near its minimum.
        
        In light of these observations we will examine the mplications
        for IR radiance for space-based limb viewing geometry in which
        case the background emission, atmospheric transmission, and
        field-of-view limitations are less severe. We will also
        discuss the similarities and differences between aurora
        mechanisms and likely sprite mechanisms.}
}
@ARTICLE{bernhardt_fagu98,
  AUTHOR = {Paul A. Bernhardt and Carl L. Siefring and Jeff
		  S. Morrill and Davis D. Sentman and Eugene
		  M. Wescott and Matthew J. Heavner and Daniel
		  L. Osborne and Eric J. Bucsela},
  TITLE = {Near-InfraRed ({NIR}) Measurements During the
		  {EXL}98 Campaign},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F175},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-03 talk},
  ABSTRACT = {A NIR digital/video camera operating in the spectral
         range between 900 and 1700 nm was flown on the EXL98
         missions.  This camera had two functions. First, it was used
         to monitor nighttime airglow which gives indications of
         density structures in the neutral atmosphere.  The camera
         uses a relatively new technology Indium-Gallium Arsinide
         (INGAAS) focal plane array which is sensitive in the NIR.  It
         offers many advantages over similar measurements in the
         visible because of its sensitivity and the relative
         brightness of the airglow in the NIR.  In fact, from the
         EXL98 airborne platform it was possible to monitor airglow
         modulations at the full 30 frame/sec video rate.  Second, was
         to search for excitation of N2 first positive emissions in
         the NIR by Sprites.  The NIR experiment was extremely
         successful -- observing both the first Sprite related NIR
         emissions and the first correlation between structures in the
         neutral density and visible and NIR Sprite emissions.  In
         addition, a number of cases were observed where features in
         the background airglow appeared to be related to structures
         in the thunderstorm clouds.  In this paper we will present an
         overview of these measurements.
        
        *The work at NRL was sponsored by NASA and ONR.}
}
@ARTICLE{bucsela_fagu98,
  AUTHOR = {E. J. Bucsela and J. S. Morrill and C. Siefring and
		  M. J. Heavner and D. R. Moudry and D. D. Sentman and
		  E. M. Wescott and D. L. Osborne and W. M. Benesch},
  TITLE = {Estimating Electron Energies in Sprites from
		  1{NG}/2{PG} Intensity Ratios},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F175},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-04 talk},
  ABSTRACT = {As part of the EXL 98 mission which observed sprites from an 
        aircraft, video images were taken with narrow passband 
        filters at 4278$\AA$  and 3407$\AA$ .  These passbands 
        correspond to the N$_2$$^+$ 1NG (1,0) and N$_2$ 2PG (0,0) 
        bands, respectively. These filters were designed 
        specifically to have very narrow passbands in order to 
        exclude nearby contaminating emissions.  This set of image 
        data will be used to generate intensity ratios of the above 
        spectral features, specifically 1NG(1,0)/2PG(0,0) which can, 
        in turn, be related to the population ratio of the 
        N$_2$$^+$(B) and N$_2$(C) states.  By making assumptions 
        about the electron distribution we 
        can estimate the electron energies implied by the observed 
        intensity ratios. In this presentation, we will discuss the 
        details of both the method of estimating energies and the 
        observing techniques.   Preliminary results from the EXL 98 
        mission will be discussed.}
}
@ARTICLE{takahashi_fagu98,
  AUTHOR = {Yukihiro Takahashi and Yoshiaki Watanabe and Akihiro
		  Uchida and Masaaki Sera and Mitsuteru Sato and
		  Hiroshi Fukunishi},
  TITLE = {Energy Distributions of Electrons Exciting Sprites
		  and Elves Inferred from the Fast Array Photometer
		  Observations},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F175},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-05 talk},
  ABSTRACT = {Recent spectral observations and numerical model
        calculations have suggested that the 1st and 2nd positive
        bands of N$_2$ are dominant emissions in sprites and elves,
        although it has been also reported that the sprite emission
        contains the 1st negative band of nitrogen molecule ions
        (N$_2$$^+$). The intensity ratio of 1st and 2nd positive bands
        of N$_2$ gives information on the energy distribution of
        electrons which excite nitrogen molecules through collision
        processes. In order to investigate the spectral
        characteristics of sprites and elves we employed two
        multi-anode array photometers (MAPs) in the SPRITES '98
        campaign. One of the MAPs with a shrap cut filter measures the
        1st positive band in the range 560-850 nm, while the other
        with a blue filter measures the 2nd positive band in the range
        350-450 nm. Each photometer has 16 channels with individual
        fields-of-view of 0.7\deg $\times$11\deg and samples data at a
        rate of 20 kHz. During the campaign, we observed 22 sprites
        and analyzed the ratio of 1st and 2nd positive bands in detail
        for 7 events of them. It was found that the 2nd positive band
        of N$_2$ was remarkably enhanced at the initial phase of
        columnar sprites. We will discuss the temporal and spatial
        changes of electron energy distributions quantitavely for both
        sprites and elves based on these MAP data.}
}
@ARTICLE{bering_fagu98,
  AUTHOR = {Edgar A. Bering and James R. Benbrook and Eugene
		  M. Wescott and Davis D. Sentman and Hans
		  C. Stenbaek-Nielsen and Walt A. Lyons},
  TITLE = {The {SPRITES} 99 Balloon Campaign},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-06 talk},
  ABSTRACT = {There are several competing models for the production of
         sprites, jets and elves. It has become clear it is not
         possible to select between these models using only
         ground-based data, owing to the fact that the ground shorts
         out the field signatures of interest. Consequently, a balloon
         campaign was proposed and funded. The campaign, which was
         originally scheduled for July, 1998, has been rescheduled for
         July and August of 1999. The 1999 Sprite Balloon Campaign
         will attempt to conduct six high altitude balloon flights
         from Ottumwa, Iowa, during the dark of the moon in the summer
         of 1999. Nominal campaign dates include July 6-21, and August
         4-19, with lower priority fall-backs of June 7-21 and
         Sept. 3-17. The balloons will be launched at sunset, float at
         a nominally constant altitude of 32 km and drift westward at
         a predicted speed of 40 knots. Cutdown will occur after
         sunrise in central Colorado.  Balloon tracking will be made
         available in real-time on redundant Web sites.  The balloon
         payloads will be instrumented with dual three axis electric
         field detectors, three axis fluxgate and induction
         magnetometers, X-ray scintillation counter, Geiger-Mueller
         tube, upward looking high-speed photometer, vertical current
         density ammeter, conductivity measurements, and an ambient
         temperature thermometer. A multiply redundant telemetry
         scheme will give us five orders of magnitude of dynamic range
         in field amplitude sensitivity. The use of event triggered
         on-board memory will allow us to sample 8 quantities at a
         digitization rate of at least 50 kHz per channel during
         sprite events. The efficiency of the on-board event trigger
         will be checked by using 4 broad-band analog waveform
         channels. Ground observations may include low light level TV
         observations from as many as three sites: tentatively, WIRO,
         on Jelm Mtn., Wyoming, Mt Evans, Colorado, and Yucca Ridge,
         Colorado. At least one of these sites will also have a fast
         photometer.  Other topics that will be covered include: why
         the program was postponed and the trigger criteria for the
         on-board fast memory.}
}
@ARTICLE{chern_fagu98,
  AUTHOR = {Jyh-Long Chern and Lou-Chuang Lee and Kuan-Ren Chen
		  and Rue-Ron Hsu and Han-Taong Su and Chin-Chun Tsai},
  TITLE = {{ISUAL} Project: Observations of Red Sprites on
		  {T}aiwan's {ROCSAT}-2},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-07 talk},
  ABSTRACT = {A satellite observation mission of red sprites has been
        planned on ROCSAT-2, which is to be launched in June 2002.
        ROCSAT-2, the second scientific satellite of Taiwan, is
        a low-earth orbit sun-synchronous satellite, with an
        altitude 891km and inclination angle 98.9$\deg$. 
        The scientific payload, ISUAL (Image of Sprite; Upper 
        Atmospheric Lightning), will contain a two-dimensional 
        CCD imager, 6 narrow-band photometers, and 2 photometer
        arrays.  The ISUAL imager is a limb viewing instrument
        pointing in a side direction, which is perpendicular to 
        the satellite moving direction and away from the sun. 
        The main objectives are (1) to obtain the global
        distribution of red sprites, (2) to determine the 
        spatio-temporal dynamics of red sprites, (3) to identify 
        the UV bands, (4) the degree of ionization in the sprite
        emission region; (5) to observe airglows and auroras.  
        International collaborations have been established and will 
        be further encouraged.}
}
@ARTICLE{marshall_fagu98,
  AUTHOR = {Lee H. Marshall},
  TITLE = {An Overview of Electromagnetic Measurements Relating
		  to Sprites, Elves and Jets},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-08 talk},
  ABSTRACT = {During the past several years, numerous measurements
         have been made of lightning- related middle and upper
         atmosphere (30-100km) optical phenomenon such as sprites,
         elves, and jets.  The relationship of these events to parent
         lightning strokes and storms suggests energy coupling
         processes upward from the lower atmosphere.  The large
         geometry and time-scale of sprites and particularly jets, and
         comparatively short time-scale of elves suggest energy
         coupling processes that could be quite different from each
         other and whose duration may vary over several orders of
         magnitude.  For these reasons, electromagnetic measurements
         have been made, in conjunction with optical observations,
         extending from the sub-hertz ULF range, through the widely
         studied ELF and VLF ranges, the LF and HF ranges and up into
         the VHF and UHF ranges.  This paper will present an overview
         of the electromagnetic measurements relating to sprites,
         elves, and jets.  Both electric and magnetic field
         measurements over a wide range of frequencies and distances
         will be discussed.  Also, apparent differences between the
         theoretical explanations and the empirical observations will
         be discussed, and the areas greatly in need of additional
         observation will be outlined. }
}
@ARTICLE{bell_fagu98,
  AUTHOR = {Timothy F. Bell and Chris Barrington-Leigh and
		  Elizabeth A. Gerken and Umran S. Inan},
  TITLE = {Continuing Currents, Charge Transfer, and Optical
		  Emissions Associated with Red Sprites Observed in
		  Thunderstorms in {M}exico},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-09 talk},
  ABSTRACT = {Red Sprites consist of large scale, transient, luminous
        , predominently red structures in the mesosphere which appear
        above thunderclouds following positive cloud-to-ground
        lightning discharges. Over the past few years Sprites have
        been extensively studied and documented in the midwestern US,
        where they are very common during the summer months. Although
        it is well known that Sprites occur in storms at other
        locations, such as Florida, Australia, the Amazon basin, and
        Africa, much less is known of the characteristics of these
        Sprites.  During the summer of 1998, Stanford University
        operated a number of instruments at the Langmuir Laboratory in
        New Mexico in order to observe Sprites that occurred within ~
        900 km of the laboratory. On a number of evenings vigorous
        Sprite activity was observed in thunderstorms in northern
        Mexico. In the present paper we present optical images of the
        Mexican Sprites in conjuction with calculations of the
        continuing current and positive charge transfer associated
        with each Sprite, and discuss the differences between Sprites
        in the midwestern US and northern Mexico.}
}
@ARTICLE{wescott_fagu98,
  AUTHOR = {E. M. Wescott and D. D. Sentman and
		  H. C. Senback-Nielsen and M. J. Heavner and
		  D. R. Moudry and U. S. Inan and T. F. Bell and
		  M. Stanley},
  TITLE = {`Columniform' Sprites: Their Optical
		  Characteristics, {ELF} and {VLF} Signatures, and
		  Relationship to Lightning},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {FAll AGU 1998, A42D-10 talk (withdrawn?)},
  ABSTRACT = {This paper summarizes observations since 1995 of a
        distinctive form of sprites associated with positive CG
        flashes in mesoscale thunderstorms, and discusses physical
        mechanisms for their features. We have examined the
        relationships between the visible forms, the causative
        lightning current, and the ELF and VLF waveforms.  The sprites
        are characterized by long vertical columns about 10 km long,
        less than 1 km in diameter, often with several black bands
        near the bottom. Some show faint tendrils extending above and
        below the column, with distinctive features.  Three
        dimensional triangulation of what we define as a columniform
        sprite (c-sprite) event on the evening of June 19, 1995 showed
        that the individual elements had an average terminal altitude
        of 86.7 km and an average bottom of 76.2 km. The sprite
        columns are nearly vertical in video imagery, however some
        show obvious curvature.  On some evenings c-sprites are the
        dominant form of sprite activity above thunderstorms: On the
        night of July 17, 1998 we observed 21 examples over a severe
        storm in New Mexico, some associated with ELVES.  Previous
        work involving comparison of c-sprite forms vs National
        Lightning Detection Network (NLDN) positive cloud-to-ground
        current shows a progression from simple, thin vertical forms
        to brighter and more complicated forms with increasing
        current. Investigators from New Mexico Tech using high speed
        TV systems (> 1000 fps) have found that all forms of sprites
        begin as c-sprites in the first ms, and evolve into the many
        other forms of sprites. The question that we address in this
        paper is: What physical processes start c-sprites, and what
        determines how far they will develop into other more
        complicated forms?}
}
@ARTICLE{reising_fagu98,
  AUTHOR = {Steven C. Reising and Umran S. Inan and Timothy
		  F. Bell and Yukhiro Takahashi and Maasaki Sera},
  TITLE = {Further Evidence of Electrical Current in Sprites
		  Using Measurements of {ELF} Radio Atmospherics with
		  Simultaneous High Time-resolution multi-anode array
		  photometer observations},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-11 talk},
  ABSTRACT = {The magnetic field signature of extremely low 
        frequency (ELF, here 
        defined as 15-1500 Hz) radio atmospherics
        (``sferics'' for short), originating in positive
        cloud-to-ground (CG) discharges in a Kansas and Texas 
        panhandle thunderstorm and measured $\sim$500 km 
        away at Yucca Ridge Field Station, Colorado, 
        consists predominantly of a positive-going pulse of 
        $\sim$1 ms width. 
        Of the ELF sferics which were launched by sprite-producing 
        positive CG flashes, roughly 20\% exhibit a second 
        ELF pulse, which has been shown to be simultaneous with 
        Sprite luminosity by comparison with high time
        resolution ($\sim$30 $\mu$s) photometer measurements
        [{\it Cummer et al., GRL, 25,} 1281, 1998].  The explanation
        of the second ELF pulse as radiation from electrical
        currents in sprites is consistent with recent results 
        of theory and simulation [{\it Pasko et al., GRL, in
        press}, 1998].
        
        We present new results of simultaneous ELF measurements
        of radio atmospherics and multi-anode array photometer
        (MAP) measurements which demonstrate that the second ELF 
        pulse and sprite luminosity occur simultaneously and that
        the delay between the causative positive CG and the sprite
        varies between 1 and 70 ms.  In addition, our data set 
        indicates a clear distinction between ``columnar'' sprites,
        with 
        a delay of $<$3 ms between CG and sprite, and ``carrot''
        sprites, with a delay of $>$4 ms between CG and sprite.
        Elves are often observed preceding the occurrence of
        column sprites.  The MAP data measure
        vertical structure in sprites with $\sim$10 km altitude 
        resolution and provide time series of optical intensity 
        with $\sim$50 $\mu$s resolution.  
        Comparison of MAP data with the second ELF pulse allows a
        preliminary assessment of the altitude distribution of 
        electrical currents in sprites.  }
}
@ARTICLE{hale_fagu98,
  AUTHOR = {L. C. Hale and L. H. Marshall and C. L. Croskey and
		  J. D. Mitchell},
  TITLE = {The Roles of Mesospheric Aerosol in Sprite-Related
		  Phenomena},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F176},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-12 talk},
  ABSTRACT = {}
}
@ARTICLE{cummer_fagu98,
  AUTHOR = {Steven A. Cummer and Mark Stanely and Paul Krehbiel
		  and Marx Brook},
  TITLE = {Simultaneous {ELF}-Radiating Currents and
		  Submillisecond Sprite Development},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F177},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-13 talk},
  ABSTRACT = {A critical parameter in current sprite theories is the
        charge transfer in the source lightning discharge.  This
        charge transfer is related to the mesospheric electric
        field thought to be the driving force in sprite
        production.  A direct comparison of theory and observation
        requires simultaneous observations of sprite optical
        emissions and lightning charge transfer.  Such a study has
        proved difficult because of fast time resolution required
        to image sprite dynamics.
        
        Studies of simultaneous ELF radio atmospherics and
        single-channel photometer data have shown that a peak of
        ELF radiation is often time correlated with bulk sprite
        brightness [Cummer et al., GRL, 25, 1281, 1998; Brook et
        al., EOS, 78, F82, 1997; Bell et al., EOS, 78, F70, 1997]
        which has been interpreted as the radiation from electric
        current in the sprite itself.  But it remains to be seen
        how this ELF-radiating sprite current relates to detailed
        sprite characteristics.
        
        To answer both of these questions, we use an ELF
        propagation model to extract radiating currents from
        simultaneous ELF radio atmospherics.  Combining this with
        submillisecond time resolution sprite images, we
        quantitatively determine for individual events the
        threshold of lightning charge transfer for sprite
        initiation and the relationship of charge transfer and
        sprite development. In the same manner, we investigate the
        relationship between optical sprite characteristics and
        sprite current flow.  These results will be interpreted in
        the context of existing theories of sprite production.}
}
@ARTICLE{stanley_fagu98,
  AUTHOR = {Mark A. Stanley and Marx Brook and Steve A. Cummer
		  and Chris P. Barrington-Leigh and Elizabeth A. Gerken},
  TITLE = {Broadband Detection and Characterization of Day-Time
		  Sprites and of Negative CGs Which Initiated Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {F177},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, A42D-14 talk},
  ABSTRACT = {The optical detection of sprites during the day-time has
        been impractical due to their low intrinsic brightness.
        However, the recent discovery of a unique ELF signature due to
        sprites [Brook et al., EOS, 78, F82, 1997; Bell et al., EOS,
        78, F70, 1997; Cummer et al., GRL, 25, 1281, 1998] has made it
        possible to search for their presence during the day.
        
        Several sprite field changes were detected during daylight
        hours on August 14, 1998 at Langmuir Laboratory, NM.  Each of
        the field changes was delayed from an energetic positive CG by
        anywhere from 11 to 15 milliseconds.  The positive CGs were
        located by the National Lightning Detection Network (NLDN)
        within a large storm system at least 2 hours prior to sunset.
        The current moment for the parent discharge and sprite will be
        extracted from the waveforms and will be compared with
        night-time positive CGs and sprites.  It will be shown that
        these sprites were more energetic than those which occur at
        night.
        
        On the night of August 29, 1998, small columniform sprites were 
        detected on video coincident with some energetic negative CG elves.
        Based on joint photometry and video measurements, the sprites were 
        initiated shortly after the negative CGs and persisted for several 
        milliseconds.  The source current moment for the negatives will be 
        determined and compared with that of sprite-producing positive CGs.}
}
@ARTICLE{valdivia_fagu98_a,
  AUTHOR = {J. Valdivia},
  TITLE = {Intracloud Lightning and Red Sprites},
  JOURNAL = {EOS Supplement},
  YEAR = 1998,
  VOLUME = 79,
  NUMBER = 45,
  PAGES = {FNNN},
  MONTH = {November},
  ANNOTE = {Fall AGU 1998, invited talk (Scarf Award)}
}

This file has been generated by bibtex2html 1.69