THE LIGHTNING AND RADIO EMISSION DETECTOR (LRD) INSTRUMENT L. J. LANZEROTTI^1,6, K. RINNERT^2, G. DEHMEL^3, F. O. GLIEM^4, E. P. KRIDER^5, M. A. UMAN^6, G. UMLAUFT^2 , and J. BACH^4, 7 Abstract. The Lightning and Radio Emission Detector (LRD) instrument will be carried by the Galileo Probe into Jupiter's atmosphere. The LRD will verify the existence of lightning in the atmosphere and will determine the details of many of its basic characteristics. The instrument, operated in its magnetospheric mode at distances of about 5, 4, 3, and 2 planetary radii from Jupiter's center, will also measure the radio frequency (RF) noise spectrum in Jupiter's magnetosphere. The LRD instrument is composed of a ferrite-core radio frequency antenna (~100 Hz to ~100 kHz) and two photodiodes mounted behind individual fisheye lenses. The output of the RF antenna is analyzed both separately and in coincidence with the optical signals from the photodiodes. The RF antenna provides data both in the frequency domain (with three narrow-band channels, primarily for deducing the physical properties of distant lightning) and in the time domain with a priority scheme (primarily for determining from individual RF waveforms the physical properties of closeby-lightning). 1. Introduction Prior to the Voyager 1 flyby of Jupiter, the presence of lightning in that planet's atmosphere had been speculatively suggested as an energy source both for the non-thermal radio emissions from the planet (which are easily detectable at Earth) and for the production of certain nonequilibrium chemical species. The suggestion regarding nonthermal radio emissions was soon effectively dismissed by the work of several researchers, in particular Zheleznyakov (1970). However, the second suggestion, related to possible non-equilibrium chemical processes, may well have some validity (Sagan et al., 1967; Woeller and Ponnamperuma, 1969; Chadha et al., 1971; Bar-Nun, 1975; Bar-Nun et al., 1984; Bar-Nun and Podolak, 1985; Prinn and Owen, 1976). In spite of the uncertainty prior to the Voyager 1 encounter regarding the existence of lightning on Jupiter, NASA believed the question of Jovian lightning to be of sufficient importance that a joint U. S.-German lightning experiment was selected for flight on the Galileo Probe (Lanzerotti et al., 1983). The Voyager 1 mission produced two night-side photographs showing strong optical evidence for lightning in Jupiter's atmosphere (Cook et al., 1979; Borucki et al., 1982). Data from Voyager's plasma wave experiment (PWS) show the existence of whistler waves, most probably generated by lightning discharges, propagating in Jupiter's magnetosphere (Gurnett et al., 1979; Kurth et al., 1985). However, these observations cannot provide information on several key parameters related to possible Jovian lightning (e.g., frequency of occurrence, intensity, distribution, global distribution, source height distribution, cloud charging, and discharging). Hence, measurements of lightning in situ in the Jovian atmosphere are crucial to understanding some of the electrical properties, as well as the chemistry and dynamics, of this giant planet's atmosphere. Because of the possible existence of lightning on Venus (e.g., Ksanfomality, 1979; Taylor et al., 1979) and Saturn (e.g., Kaiser et al., 1983; Burns et al., 1983; Zarka, 1985), as well as on Jupiter, there has emerged a new field of research that could be called planetary lightning. Several reviews of planetary lightning have already been written (Levin et al., 1983; Williams et al., 1983; Rinnert, 1982, 1985; Lanzerotti et al., 1988a, 1989a) and a book chapter devoted to the subject is contained in a recent monograph on the lightning discharge (Uman, 1987). At least one thesis already exists on Jovian lightning (Williams, 1986). The LRD instrument has been designed to take into account large uncertainties in the nature of possible Jovian lightning. For example, since Jupiter has no well-defined surface close to the cloud system, there will be no cloud-to-ground discharges, which are the best understood type of lightning on Earth. Lightning in general, and cloud discharges specifically, are very complex physical phenomena and can generate a large variety of RF pulse types and trains: unipolar pulses, bipolar pulses, asymmetric pulses, groups and bursts of pulses. The LRD instrument is designed as a compact and versatile instrument which allows a characterization of these signals with maximum sensitivity and maximum dynamic range. During the design phase, prototype instruments have been intensively tested with Earth lightning during several measuring campaigns. The final instrument characteristics have been set with acceptable margins for the unknown conditions to Jupiter. Modelling of the propagation of RF signals in the frequency range of the LRD instrument in Jupiter's atmosphere shows that direct propagation of signals will occur to distances of order 10^4 km (Rinnert et al., 1979). Hence, it is likely that Jovian atmospheric discharges with the energy of a typical cloud-to-ground discharge on Earth (order 10^8 J) will be detected at 10^4 km or more distance within the atmosphere with the LRD instrument. As noted below, the LRD instrument also includes a 'superbolt' channel, in order to count extremely large events. Hence, in light of all the above, the flown lightning detector instrument must be designed to be as sensitive as possible, limited only by spacecraft noise. The instrument must also cover as large a dynamic range as possible. 2. Cloud Structure and Lightning Most Earth lightning is produced by the charge generated and separated in cumulo nimbus clouds that develop in unstable air containing water vapor (see, for example, the reviews in Uman, 1987). The instability is most often produced by solar heating at the ground or by the interaction of air masses driven together by winds and convection. As the moist air rises and cools, condensation occurs and ultimately precipitation forms. Charging of particles in the air can occur by various processes, including a thermoelectric effect and collisions between water droplets, supercooled water, and ice in various forms. On Earth, it is thought that the positive charges are predominant on the smaller particles and are raised preferentially by vertical convective motion. In Earth's atmosphere, charges can be separated to distances of the order of kilometers. Most lightning discharges in the Earth's atmosphere occur as intracloud lightning and neutralize tens of Coulombs of charge. Cloud-to- ground lightning generally transfers negative charge from the lower portion of the cloud to the Earth's surface. Fig. 1. Temperature, altitude, and pressure in the Jovian cloud system as given by the model of Weidenschilling and Lewis (1973). The water ice cloud for a depleted water abundance (10^-3 solar abundance) at Jupiter is also shown (adapted from Atreya, 1986). At the top is a cartoon schematic of the global atmospheric convective motion. A sketch of the temperature-density-altitude-pressure profile of the Jovian atmosphere is shown in Figure 1, based upon the cloud model of Weldenschilling and Lewis (1973). The height region for a water ice cloud for an atmospheric mixing ratio 10^-3 that of the solar abundance is indicated by the hatched area (Atreya, 1986). At the top of Figure 1, the global convection motions in the Jovian system are indicated schematically. The cloud model in the lower portion of Figure 1 assumes stationary atmospheric conditions, solar composition, and chemistry in thermal equilibrium (i.e., no sources such as lightning to change energy from any existent electromagnetic fields to sound, heat, RF, etc.). More recent cloud models, deduced primarily from spacecraft optical measurements, contain either no, or only a thin, NH4SH cloud, and add aerosol layers to account for the haze seen above the ammonia clouds (Sato and Hansen, 1979; West and Tomasko, 1980; Marten et al., 1981; Orton et al., 1982; Owen and Terrile, 1982). Further, Bjoraker et al. (1986) deduced from Voyager infrared interferometer spectrometer (IRIS) measurements at 5 mm that water is depleted by a factor of ~50 relative to solar abundances in the 2-6 bar atmosphere region. With the type of atmosphere illustrated in Figure 1, a cartoon comparison can be made of Earth's and Jupiter's cloud structures, as is done in Figure 2. The dominant Fig. 2. Cartoon representation of the possible locations of separated charges in the atmospheres of Earth and Jupiter, showing comparisons of the cloud locations, altitudes, and temperatures. atmospheric chemical species in each planet's atmosphere are boxed. On Earth, the topography of the land surface can at times also determine cloud location for the separated charges. Obviously, such an orographical influence should not exist at Jupiter, and the lightning should primarily be of the intra-cloud, cloud-to-cloud, and cloud-to-'air' types. Especially important for establishing the conditions for localized, intense, upward convective motions in the Jovian atmosphere has been the recent modeling work of Stoker (1986) to understand the vertical structure of the equatorial plumes. She finds that, for appropriate parameters for cumulus cloud formation, moist convection can produce the equatorial plumes. For a solar water abundance in the atmosphere, peak vertical convective velocities of ~150 m s^-1 are found. The model results of Stoker (1986) suggest that the water abundance is closer to solar than to the depleted water conditions suggested by Bjoraker et al. (1986). Lunine and Hunten (1987) use the idea of moist convective plumes and the model plume of Stoker (1986) to conclude that such plumes can explain the apparent depletion of the IRIS detected water vapor on Jupiter without invoking a global depletion of water from solar abundance. It is interesting to note that neither Stoker (1986) nor Lunine and Hunten (1987) made mention of the possibility of lightning production in convective regions such as these plumes, even though atmospheric conditions in the models, such as aerosols and upward convective velocities, are present to produce electrical charge separations and discharges (e.g., Williams et al., 1983; Levin et al., 1983). Fig. 3. Schematic illustration of Probe descent into Jupiter's atmosphere and the locations in the magnetosphere of LRD data acquisition. 3. Overview of LRD Objectives As illustrated schematically in Figure 3, the Galileo Probe will be released from the Orbiter spacecraft 150 days prior to atmospheric entry. Radio frequency measurements are made in a reduced mode of operation at altitudes of ~5, 4, 3, 2 planetary radii from the center of Jupiter. These data are stored in the Probe memory and then read out during the atmospheric descent phase of the mission. During the atmospheric descent, the full complement of LRD data are acquired until the loss of the Probe signal by the over-flying Orbiter and/or the demise of the Probe due to atmospheric pressure and heat. The science objectives of the LRD instrument are shown in Figure 4 for both magnetosphere and the atmosphere measurements. Measurements will not be made in the Jovian ionosphere (schematically shown by the horizontal dashed line). The RF data obtained in the magnetosphere will be analyzed also jointly with the Probe Energetic Particle Instrument (EPI) data to gain understanding of magnetospheric particle dynamics. In the magnetosphere, statistics on the characteristics of individual waveforms measured during a sampling interval will be accumulated at the four different altitudes. In addition, noise levels at three different spectral frequencies (3, 5, 90 kHz) will be determined during the measurement intervals. Fig. 4. Science objectives of the LRD instrument. In the atmosphere mode, in addition to statistics on the waveforms and the spectral noise levels at the three narrow-banded frequencies, individual waveforms will be detected, saved, and transmitted to Earth. Such waveforms will provide powerful additional diagnostic capabilities for Jovian RF signals. The LRD instrument, as noted above, has been designed to be as sensitive as possible, limited only by the spacecraft noise, and to be as versatile as possible, limited only by the imposed limitations on power, bit rate, and reliability considerations. It is within these constraints that the scientific objectives (Figure 4) will attempt to be achieved. Extensive measurements with Earth lightning have been made (see below) and these will be continued in order to gain the maximum understanding of the operational characteristics of the instrument, and therefore the maximum science from the Probe descent through Jupiter's atmosphere. Fig. 5. Diagram of the sensors and the several different data outputs of the LRD instrument. In both the magnetosphere and atmosphere modes the component of the Jovian magnetic field perpendicular to the Probe spin axis will be determined. These data will be used for analyses of EPI data and for determining the spatial distribution of the sources of some of the detected lightning signals. Further, these data will give engineering data on the Probe spin rate. The output data from the LRD instrument is schematically shown with the two sensor types in Figure 5. The single ferrite core antenna ('magnetic sensor', MS) and two optical sensors (OS1; OS2) are shown. The optical sensors provide the optical (OPT) signals. The MS provides the input signal for the three principal RF data channels: the waveform analyzer (WFA) for statistics and waveform snapshot, the spectrum analyzer (SPA) with the three narrow-banded frequencies F1, F2, F3, and the magnetic field/spin rate determination B0E. Fig. 6. Picture of LRD instrumentation. 4. LRD Instrument Overview. A picture of the electronics box, assembled antenna, optical sensors, and the EPI detector housing is shown in Figure 6. The parameters of the instrument (including the EPI) are given in Table I. TABLE I LRD and EPI instrument parameters ------------------------------------------------------------------------------ Weight: 2.5 kg Size: Electronics plus four sensors RF antenna: 32 cm x 3.3 cm dia. Electronics box: 13 cm x 11.5 cm x 15 cm Optical sensors: 4 cm x 2.3 cm dia. each, two pieces EPI sensor box: 4 cm x 15.2 cm x 7.4 cm Power: EPI bias: 1 W for 1 hour Pre-entry: 3 W for 84 min (nominal) Entry: 3 W for duration Data rate: 10^4 bits stored; 544 bits/64 s at entry; stored and real time data transmitted at 8 bps. During descent, a measuring period lasts 256 s and provides one data frame of 256 bytes. ------------------------------------------------------------------------------ Radio wave sensor. The elements of the RF sensor are pictured in Figure 7. The RF signals in the Jovian atmosphere in the frequency range ~100 Hz to ~100 kHz are detected using the 32 cm long ferrite core antenna (Dehmel, 1989). An electric antenna is not used because of expected electrostatic noise during Probe descent through the Jovian atmosphere. The MS is mounted in a plane perpendicular to the spin axis of the Probe. As the Probe rotates with a nominal 10 rpm rate, a relatively high voltage is induced due to the strong Jupiter magnetic field (~4 G). This signal has to be suppressed and, in effect defines the lower band limit. The upper band limit was set to 100 kHz according to the sampling frequency of the signal (see below) and in order to suppress the detection of any high-frequency noise generated in the Jovian magnetosphere which can penetrate the Jovian ionosphere at frequencies higher than its maximum critical frequency. The antenna is mounted outside the Probe body and is thus subject to large ambient temperature changes: from about 0 degrees C during cruise, to -150 degrees C after heat shield release, to +150 degrees C during descent. The strong ferrite core support structure is filled with insulating ceramic material to smooth this temperature variation. The antenna is wrapped in a conducting fabric as a shield for high-frequency stray fields from the nearby telemetry antenna. The antenna also contains a low-noise preamplifier with a dynamic range of 80 dB. The LRD instrument performs its magnetospheric measurements when the Probe is encapsuled in the Probe heat shield. Hence, the after heat shield cover was designed with a special RF window over the antenna location. All of the thermal and structural Fig. 7. Picture of the elements of the LRD ferrite core antenna. Left to right: Piece of ferrite rod, coil, electronics, support structure, mounted antenna filled and enveloped with thermal isolation. (Not shown: outer-most cover of microwave absorbing material.) The dimensions of the assembled antenna shown on right are 32 cm long by 3.3 cm diameter. loads, as well as the electromagnetic responses with and without the heat shield, including operation with the Probe telemetry, have been extensively tested during the development and fabrication program. Optical sensors. For an unambiguous verification of lightning as the source of measured magnetic pulses, a coincidence with light pulses is important. For this purpose, two identical sensors are mounted to look in opposite directions about the perpendicular to the Probe spin axis. Each sensor consists of a photodiode with an amplifier behind a fisheye lens. Both sensors are sensitive to the visual spectrum and together cover nearly a full 4*pi field of view. Special care has been taken to thermally decouple the lenses (exposed to the ambient temperature) and the electronics (which are coupled to the Probe structure). The signals of both sensors are added and provide measurements of the overall atmospheric brightness and of the light pulses (Figure 4). Fig. 8. Overview block diagram of the LRD electronics. The numbers on the lines indicate the actual number of connections. Electronics. Figure 8 gives an overview of the analog and digital electronics and the signal paths of the sensor outputs. Four major sub-elements of the electronics are indicated by the dashed boxes and are described below. B0E: The voltage induced due to the rotation of the antenna within the Jovian magnetic field provides a measure of the magnitude of the magnetic field component perpendicular to the Probe spin axis and of the Probe spin period. This signal is also used to sector RF measurements with respect to the ambient magnetic field direction. SPA: The SPectral Analyzer consists of three narrow-band frequency channels centered at 3, 15, and 90 kHz. In each of the three channels the rectified and averaged signal is output as a measure of the narrow-band noise level, and a pulse height analyzer measures the amplitude distribution with 8 amplitude levels spaced by 10 dB. One of the narrow-band channels is sectored using the spin phase information of B0E. During the Pre-Entry (magnetosphere) mode of Operation (see below) the 3 kHz-channel is subdivided into two channels: parallel and perpendicular to the magnetic field. During the Descent mode of operation the 15 kHz channels is subdivided into 4 channels. This is the most 'sensitive' portion of the instrument. For example, an 'average' Earth lightning at a distance ~20 km could trigger a threshold ~60 dB above the present Probe noise level (see Rinnert et al. (1979) for discussions of propagation conditions in Jupiter's atmosphere). WFA: The Wave Form Analyzer uses the full antenna bandwidth. Since the output signal of the magnetic antenna is proportional to the time derivative of the magnetic wave field, the antenna output first is integrated to regain the correct waveform. This signal is then fed to a gain change amplifier and is continuously digitized every 4 ms to provide the input for the waveform snapshot (see below). The integrated signal is amplified, rectified and averaged to provide the wideband signal noise levels and to obtain statistical characteristics of the waveforms. There are four threshold levels spaced by 10 dB following a gain change amplifier (0, 10, 20, 30 dB). Signals exceeding the lower threshold are treated as events and are characterized by the highest threshold which is exceeded and by the duration between consecutive crossings of the lower threshold, which corresponds to either a 'pulse duration time' or a 'pulse gap time'. As noted, there are 4 amplitude levels; the duration times and the gap times are sorted into 8 time bins for each distribution: ------------------------------------------------------------------------------ Duration time distribution Gap time distribution ------------------------------------------------------------------------------ < 8 micros < 8 micros 8-24 micros 8-40 micros 24-40 micros 40-168 micros 40-72 micros 168-680 micros 72-138 micros 0.68-8.7 ms 138-266 micros 8.7-72.2 ms 266-522 micros 72.2-1100 ms > 522 micros > 1.1 s ------------------------------------------------------------------------------ The gain change amplifier adjusts the pulse height analyzer to the appropriate noise level as it is controlled by the number of events in order to guarantee that pulses are characterized which sufficiently exceed the background noise level. A bipolar pulse, for example, is characterized by two amplitudes, two duration times (one for each half period) and one gaptime between the threshold crossings of the two half periods. Another channel is devoted to extremely large events, so-called 'super bolts' (see Figure 5). The threshold of this channel is set 20 dB above the highest WFA threshold. OPT: The OPTical channel adds the output signals of both optical detectors. An averaged d.c. signal from the combined detectors provides a measure of the 'brightness' of the Probe's atmospheric environment. The pulses from the combined signals correspond to flashes. There is no information on the optical spectrum, the pulse amplitude or other pulse characteristics. Not shown in Figure 2 is the In Flight Testgenerator (IFT). This device inserts 128 well-defined electrical pulses into the electronics in place of the MS signals. These pulses are processed the same way as the antenna signals and give test data on the operation of the instrument. WF SNAPSHOT: The Wave Form Snapshot channel provides a window of 1 ms out of the continuously sampled signal (sample period 4 micros). A block diagram of the WF Snapshot electronics is shown in Figure 9. The peak amplitude of the measured pulse defines the trigger point. The 1 ms signal interval contains 64 pre-trigger samples covering 256 micros before the pulse peak, 64 post-trigger samples covering 256 micros after the pulse peak, and 61 samples (every second sample) covering 488 micros. The time resolution of the first half of the time interval is therefore 4 micros and the second half is 8 micros. Although Fig. 9. Block diagram of the waveform 'snapshot' analyzer of the LRD instrument. the antenna signal is continuously sampled, only one one-millisecond interval is saved for transmission during each measuring period of 256 s. This is because of the limited data rate available for the instrument. Two switched pipeline memories are used for this purpose. The signal interval that is transmitted during each measuring interval (see below) is defined by a rotating priority scheme (largest amplitude waveform with optical coincidence; largest amplitude waveform; and random). The instrument is controlled by an 1802 microprocessor which determines the modes of operation (including the EPI measurements), generates the statistical distributions, determines coincidences between optical and magnetic pulses and performs the data formatting (most of the science data are logarithmically compressed). One standard data frame consists of 256 bytes in total and contains the information gathered during four major frame periods of 64 s each. The first one quarter (64 bytes) of the LRD science data frame contains all of the LRD statistics information (spectral and waveform statistics), number of events, average levels, and gain information. The other three quarters (192 bytes) of the data frame contain the waveform samples as signed numbers (-128 to +127) and waveform status information. The second and third quarters of the LRD science data frame contain the values of the samples which are made each 4 micros, while the fourth quarter contains the values of those samples which are each 8 micros. In addition, the data frame contains engineering data such as voltages, currents, and temperatures. 5. Modes of Operation Magnetosphere mode. The LRD instrument will operate in the pre-entry phase at distances from the planet's center of about 5, 4, 3, and 2 Rj. The instrument is switched on by the Probe timer at these locations. In this 'magnetosphere mode' the EPI is also in operation. As the Probe is still encapsulated within the heat shield, the MS is less sensitive and the optical sensors are covered. The outputs of the LRD instrument are as noted in the previous section, but without the waveform snapshots. The magnetosphere mode data set at each of the four locations consists of a 64 byte data frame with statistics and the 3 kHz spectral channel subdivided into parallel and perpendicular (to the magnetic field) channels. The data are stored in the Probe memory for transmission during the atmospheric descent phase of the mission. Atmosphere mode. When the LRD instrument is switched on at descent the instrument begins with a test cycle (IFT) and the first data set contains the test pulse data. After that, the instrument runs continuously until the end of the mission and outputs a complete data set every four major frame periods, 256 s. These data sets contain spectral data (the 15 kHz channel being sectored), waveform statistics data, a 1 ms time interval with a selected waveform, optical data and miscellaneous data such as magnetic field component, spin period, and engineering data. The number of complete data sets achieved during descent depends upon the length of time that the Probe survives and/or the length of time that the Probe relay signal is successfully acquired by the over-flying Orbiter. For example, if the total atmosphere data time is ~48 min, then 10 data sets would be sent back (the 11th would be acquired but there would be no time for transmittal). The 10 data sets would contain one test data set and 9 science data sets. 6. Tests and Calibrations Because of severe constraints as to weight and power for Probe subsystems, the LRD instrument is very compact. Further, extensive on-board compression of the data is necessary because of the limited available data rate. All sensors and instrument characteristics, of course, have been extensively tested and calibrated. For example, radiation tests were carried out on the sensor electronics and pressure tests were made of the vented electronics box. These latter tests caused a stiffening piece to be added to the microprocessor chip. The calibrations could be verified over long periods because of the delays of the launch of the Galileo spacecraft. A further verification of the instrument parameters is provided by the on-board implemented test generator (ITG), as noted in Section 4. Fig. 10. Upper panels: Two cloud-to-ground discharges measured with the Galileo Probe lightning experiment engineering unit during a lightning storm near Lindau in August 1984. Lower panels: The power spectrum of each discharge. Most important for the definition, design verification, and test of the instrument have been the extensive campaigns devoted to measurements of Earth lightning. Some of the results of these campaigns and their relevance to both characterizing the instrument and to providing new data on Earth lightning have been reported in the scientific literature over the years (Rinnert et al., 1984, 1985, 1988, 1989; Lanzerotti et al., 1988b; 1989b). On Earth, cloud-to-ground discharges are quite different in their RF characteristics Fig. 11. Examples of types of intra-cloud discharge magnetic field waveforms measured in Lindau during three different lightning storms. than are intra-cloud discharges. These differences are being carefully studied by our team using the LRD engineering unit (which has identical electrical characteristics to the instrument flying on the Probe). These studies are being made in order to eventually be able to evaluate the atmospheric breakdown processes, the electrical currents, and the possible chemical processes associated with Jovian lightning. As an example, shown in the upper portions of both sides of Figure 10 are the magnetic field wave forms of two cloud-to-ground return strokes measured in Lindau with the Galileo LRD engineering instrument unit during a nearby lightning storm in 1984. The power spectrum of the RF signals for each of these discharges was calculated for us by D. J. Thomson (AT&T Bell Laboratories) after first mathematically removing the impulse response of each signal. The power spectra were then calculated using four prolate spheroidal data windows in the time domain and a fast Fourier transform algorithm (see Lanzerotti et al., 1989b). Each RF power spectrum clearly falls with increasing frequency and each spectrum has definite structure that corresponds to the 'secondary peaks' evident in the time domain traces following the initial impulse. The RF signals of intra-cloud Earth lightning discharges have been much less studied than those of cloud-to-ground discharges (see, e.g., Uman, 1987). On average, ground discharges are much more powerful than cloud discharges. The ratios of the amplitudes of return strokes to cloud discharges appear to range from about 20 to 1 at ~3 kHz, to ~10 to 1 at 10 kHz, to ~1 to 1 at 100 kHz (Brook and Ogawa, 1977). Shown in Figure 11 are examples of different types of magnetic field waveforms produced by cloud discharges and measured using the engineering unit (Lanzerotti et al., 1989b). From top Fig. 12. The gap time distribution provided by the LRD instrument during two lightning storm intervals. to bottom in the figure, the cloud discharges observed in 1 ms include a single bipolar pulse, a series of quasi-periodic bipolar pulses filling the entire measuring interval, and a sequence of (possibly approximately periodic) bipolar pulses apparently modulated by an overall signal envelope. LRD analysis of the intervals between pulses (gap times) obtained during two storm periods is shown in Figure 12 (Rinnert et al., 1989). The large numbers of gap times less than 8 micros are probably due to the gaps between successive portions of rectified bipolar pulses. There is a significant minimum in gap times between 56 and 88 micros, and then a maximum for gap times between ~10 and ~170 ms. Additional studies have been made on Earth lightning with the LRD engineering unit to study gap time and other distributions as a function of whether the originating RF signals were predominantly intra-cloud or cloud-to-ground (Rinnert et al., 1989). The field campaigns with the LRD instrument have yielded interesting new scientific results in the discipline of lightning research. Such studies will be continued in the years to come until the Probe descends through the Jovian atmosphere in December 1995. Acknowledgements This work has been supported in part by NASA contract NAS2-997 to the University of Florida and in part by the Bundesministerium fuer Forschung und Technologie, Germany by contract 01QJ0976. We thank the NASA Ames Galileo Project Office personnel over the many years of the program, especially Al Wilhelmi, Benny Chin, Ed Tischler, and Charles Sobec. References Atreya, S. K.: 1986, Atmospheres and Ionospheres of the Outer Planets and their Satellites, Springer-Verlag, Heidelberg. Bar-Nun, A.: 1975, Icarus 24, 86. Bar-Nun, A. and Podolak, M.: 1985, Icarus 64, 122. Bar-Nun, A., Noy, N., and Podolak, M.: 1984, Icarus 59, 162. Bjoraker, G. A., Larson, H. P., and Kunde, V. G.: 1986, Astrophys. J. 311, 1058. Borucki, W. J., Bar-Nun, A., Scarf, F. L., Cook, A. F., and Hunt, G. E.: 1982, Icarus 52, 492. Brook, M. and Ogawa, T.: 1977, in R. H. Golde (ed.), Lightning, Academic Press, New York, p. 191. Burns, J. M., Showalter, M. R., Cuzzi, J. N., and Durison, R. H.: 1983, Icarus 54, 280. Chadha, M., Flores, J., Lawless, J., and Ponnamperuma, C.: 1971, Icarus 15, 39. Cook, A. F., Duxbury, T. C., and Hunt, G. E.: 1979, Nature 280, 794. Dehmel, G.: 1989, in R. Ball and K. J. Overshott (eds.), Sensors, 5, VCH Publ., New York, p. 205. Gurnett, D. A., Shaw, R. R., Anderson, R. R., Kurth, W. S., and Scarf, F. L.: 1979, Geophys. Res. Letters 6, 511. Kaiser, M. L., Connerney, J. E. P., and Desch, M. D.: 1983, Nature 303, 50. Ksanfomality, L. V.: 1979, Kosm. Issled. 17, 747. Kurth, W. S., Strayer, B. D., Gurnett, D. A., and Scarf, F. L.: 1985, Icarus 61, 497. Lanzerotti, L. J., Rinnert, K., Krider, E. P., Uman, M. A., Dehmel, G., Gliem, F. O., and Axford, W. I.: 1983, in L. H. Ruhnke and J. Latham (eds.), Proc. Atmospheric Electr., Deepak Publ., Hampton. Lanzerotti, L. J., Rinnert, K., Krider, E. P., and Uman, M. A.: 1988a, Proc. VIII Int. Conf. Atmospheric Electr., Uppsala, p. 749. Lanzerotti, L. J., Thomson, D. J., Maclennan, C. G., Rinnert, K., Krider, E. P., and Uman, M. A.: 1988b, Proc. VIII Int. Conf. Atmospheric Electr., Uppsala, p. 480. Lanzerotti, L. J., Rinnert, K., Krider, E. P., and Uman, M. A.: 1989a, in M. J. S. Belton, R. A. West, and J. Rahe (eds.), Time-Variable Phenomena in the Jovian System, NASA SP-494, p. 374. Lanzerotti, L. J., Thomson, P. J., Maclennan, C. G., Rinnert, K., Krider, E. P., and Uman, M. A.: 1989b, J. Geophys. Res. 94, 13221. Levin, Z., Borucki, W. J., and Toon, D. B.: 1983, Icarus 56, 80. Lunine, J. I. and Hunten, D. M.: 1987, Icarus 69, 566. Marten, A., Rovan, D., Baluteau, J. P., Chedin, A., and Scott, N.: 1981, Icarus 46, 233. Orton, G. S., Appleby, J. F., and Martonchick, J. V.: 1982, Icarus 52, 94. Owen, T. and Terrile, R. J.: 1982, J. Geophys. Res. 87, 11193. Prinn, R. G. and Owen, T.: 1976, in T. Gehrels (ed.), Jupiter, University of Arizona, Tucson, p. 319. Rinnert, K.: 1982, in H. Volland (ed.), Handbook of Atmospherics, CRC Press, Boca Raton, FL, p. 99. Rinnert, K.: 1985, J. Geophys. Res. 90, 6225. Rinnert, K., Lanzerotti, L. J., Krider, E. P., Uman, M. A., Dehmel, F., Gliem, G. O., and Axford, W. I.: 1979, J. Geophys. Res. 84, 5181. Rinnert, K., Lanzerotti, L. J., Dehmel, G., Gliem, F. O., Krider, E. P., and Uman, M.: 1984, Proc. VII Int. Conf. Atmospheric Electr., p. 481. Rinnert, K., Lanzerotti, L. J., Dehmel, G., Gliem, F. O., Krider, E. P., and Uman, M. A.: 1985, J. Geophys. Res. 90, 6239. Rinnert, K., Lauerdale, R., Lanzerotti, L. J., Krider, E. P., and Uman, M. A.: 1988, Proc. VIII Int. Conf. Atmospheric Electr., Uppsala, p. 882. Rinnert, K., Lauerdale, R., Lanzerotti, L. J., Krider, E. P., and Uman, M. A.: 1989, J. Geophys. Res. 94, 13229. Sagan, C. E.. Lipincott, E. R., Dayhoff, M. D., and Eck, R. V.: 1967, Nature 213, 273. Sato, M. and Hanse, J. E.: 1979, J. Atmospheric Sci. 36, 1133. Stoker, C. R.: 1986, Icarus 67, 106. Taylor, W. W. L., Scarf, F. L., Russell, C. T., and Brace, L. H.: 1979, Nature 279, 614. Uman, M. A.: 1987, The Lightning Discharge, Academic Press, New York. Weidenschilling, S. J. and Lewis, J. S.: 1973, Icarus 20, 465. West, R. A. and Tomasko, M. G.: 1980, Icarus 41, 279. Williams, M. A.: 1986, 'An Analysis of the Voyager Images of Jovian Lightning', Ph.D. Thesis, University of Arizona, Tucson. Williams, M. A., Krider, E. P., and Hunten, D. M.: 1983, Rev. Geophys. 21, 892. Woeller, F. and Ponnamperuma, C.: 1969, Icarus 10, 386. Zarka, P.: 1985, Astron. Astrophys. 146, L15. Zheleznyakov, V. V.: 1970, Radio Emission of the Sun and Planets, Pergamon Press, Oxford.