A&A 367, 1056-1060 (2001)
DOI: 10.1051/0004-6361:20000436
M. Beech1 - L. Foschini2
1 - Campion College and Department of Physics, The University
of Regina, Regina, Saskatchewan,
Canada S4S 0A2
2 - Institute TeSRE - CNR, Via Gobetti 101,
40129 Bologna, Italy
Received 4 October 2000 / Accepted 27 November 2000
Abstract
We investigate the conditions under which Leonid meteoroids might
generate short duration (burster) electrophonic sounds. A "first order''
theory is employed to estimate the approximate electron number density in the
meteoroid ablation column as a function of time. Using the threshold
conditions discussed in an earlier communication (Beech & Foschini 1999) we
find that Leonid meteoroids more massive than about 0.1 kg can potentially
generate short duration electrophonic bursters.
Key words: meteors, meteoroids
In this paper, we are specifically concerned with the Leonid meteoroid stream. In an earlier study, Beech (1998) found that enduring electrophonic sounds might be produced from (of order of) metre-sized Leonid meteoroids. Clearly, such large objects are not going to be abundant in the Leonid stream at any one time, but it has been suggested by Beech & Nikolova (2000) that they might be deposited into the stream during mantle ejection events associated with the aging of comet 55P/Tempel-Tuttle. Reports gathered by Olmsted during the 1833 Leonid storm indicate that enduring electrophonic sounds, as well as electrophonic bursters, were heard and this is indicative of the presence of at least a few large Leonid meteoroids. Of more recent note, however, electrophonic bursters have reportedly been heard during both the 1998 (Darren Talbot, personal communication) and 1999 (Drummond et al. 2000; Beatty 2000) Leonid outbursts.
Beech & Foschini (1999) have suggested that the burster phenomenon can be explained in terms of a shock propagating within a meteor's plasma column. Within the framework of this model, it is argued, that the rapid movement of the electrons, with respect to the much more massive and slower moving ions, generates a sizable space charge (see e.g., Zel'dovich & Raizer 1967). A transient electrical pulse is generated in response to the development of the space charge, and provided the resultant electrical field strength variations are large enough it is suggested, following Keay (1980), that they might trigger the generation of audible sounds through an observer localized transduction process. The shock wave is produced, Beech & Foschini (1999) suggest, during the catastrophic break-up of the parent meteoroid.
Under ideal conditions the electric field generated would be symmetric, propagating equally in all directions. However, as is more likely, the presence of fluid dynamic instabilities will lead to preferential propagation in certain directions. Specifically, since the meteor plasma can be thought of as a fluid with a higher density than the surrounding atmosphere, both Rayleigh-Taylor and Kelvin-Helmholtz instabilities might develop. In the possible presence of these instabilities, any shock front will soon become distorted. One may therefore envision the situation in which the electric field is either focused or defocused in random directions by the growing filaments and small scale perturbations in the shock. Gull (1975) has described the complex instability modification of a shock front within the context of a stellar supernovae model. The same processes are likely to occur in airburst explosions, with obvious changes in energy scale.
In addition to the presence or not of appropriate transduction material, this instability feature may explain why electrophonic sounds often appear to be highly localized.
Experiments conducted by Keay & Ostwald (1991) suggest that an
electric field strength of at least 160 V/m is required to generate
electrophonic sounds. Beech & Foschini (1999) find that such a
threshold electric field can be generated provided the electron number
density within the plasma exceeds
m-3.
Ionization Theory | Composition | Z [deg.] |
![]() |
![]() |
![]() |
Jones (1997) | type IIIA | 0 | 0.028 | 80.35 | -8.6 |
20 | 0.031 | 80.60 | -8.7 | ||
45 | 0.051 | 81.30 | -8.9 | ||
60 | 0.095 | 82.23 | -9.2 | ||
type IIIB | 0 | 0.12 | 85.80 | -10.4 | |
20 | 0.14 | 85.90 | -10.4 | ||
45 | 0.22 | 86.73 | -10.7 | ||
60 | 0.41 | 87.60 | -11.0 | ||
Bronshten (1983) | type IIIA | 0 | 0.005 | 84.10 | -6.8 |
20 | 0.006 | 84.25 | -6.8 | ||
45 | 0.009 | 84.93 | -7.0 | ||
60 | 0.017 | 85.85 | -7.3 | ||
type IIIB | 0 | 0.021 | 89.30 | -8.5 | |
20 | 0.023 | 89.43 | -8.6 | ||
45 | 0.039 | 90.20 | -8.8 | ||
60 | 0.070 | 91.08 | -9.1 |
where V is the velocity,
is the ionization
coefficient,
is the mean atomic mass of meteoroid atoms, and
is the mass ablation rate. The mass loss rate can be
derived by solving the standard single-body equations of meteoroid
ablation (see, e.g. Öpik 1958; McKinley 1961).
To first order, the electron line density may be converted to an
electron number density
via the relation:
where
is the initial train radius.
Equation (2) is the limiting initial case of the diffusion
equation when
(see e.g., Mckinley 1961). Provided one
can establish analytically reasonable expressions for the ionization
coefficient,
,
and the initial train radius,
,
then
Eqs. (1) and (2) may be solved for numerically.
In this manner, the characteristics of the Leonid meteoroid just
capable of producing an electron number density of, say,
1019 m-3 (this places
above the burster threshold
condition derived by Beech & Foschini 1999) can be determined.
The core of the problem is the determination of the ionization
coefficient .
It is, by definition, the ratio of the number of
free electrons produced to the number of meteoroid atoms vaporized.
The ionization coefficient is strongly dependent upon the relative
speed with which the meteoroid's ablated atoms and atmosphere molecules
collide. Typically the ionization coefficient is expressed as a power law
in the velocity
where
and n are coefficients to be calculated.
Massey & Sida (1955) and later on Sida (1969) studied the collision
processes in meteor trails and argued that
can be expressed as the
ratio of the ionizing and momentum-loss cross sections. It is the sensitivity
of the ionization cross section to the relative velocity that introduces the
velocity dependency in
.
Massey and Sida proposed several different
ways of weighting the cross sections, taking into account different factors
such as the angle of scattering (see Jones 1997, for a discussion of the
inherent problems with the Massey and Sida model). Indeed, even though the
definition of
is quite straightforward, the complexity of particle
dynamics in the meteor plasma complicates the situation greatly, and the
correct weighting factors are still a topic of debate (for a recent review on
particle dynamics in meteor plasma's see Dressler & Murad 2000).
Bronshten (1983) presents a detailed analysis of the empirical and
theoretical calculations of
and finds that:
As with the ionization coefficient, the initial train radius is also a
difficult term to quantify. Generally speaking it will be of order the
atmospheric mean free collision length. Radar measurements of meteor trains
have been used to determine the initial train radius as a function of
atmospheric height. Baggaley (1970) found, for example, that the initial train
radius was some 3 m at an altitude of 115 km, and some 0.5 m at 90 km
altitude. These values of the initial train radius are some 2 to 20 times the
mean free path lengths at the respective atmospheric heights. Jones (1995)
has derived a relationship for the initial train radius as a function of the
number density
of air molecules. He finds:
The critical mass,
,
that we are looking for determines the
mass of the Leonid meteoroid that will just generate a maximum electron
number density of 1019 m-3 somewhere along its ablation track.
Meteoroids more massive than
will produce electron number
densities higher than the critical value, and subsequently should such
meteoroids catastrophically disintegrate (when
m-3) then the burster mechanism described by
Beech & Foschini (1999) may come into play.
The results of our calculations are presented in Table 1. The calculations begin at an atmospheric height of 190 km and various zenith angles (Z) of meteoroid entry have been assumed. We have used a constant luminous efficiency of 1% to determine magnitudes.
The data gathered together in Table 1 suggest that the minimum mass for a Leonid meteoroid to produce an electrophonic burster is somewhere in the range of 5 to 400 g depending upon zenith angle, composition and the velocity sensitivity of the ionization coefficient. The predicted brightness of the burster meteors is in the magnitude range of -7 to -11.
For a given zenith angle the variation in composition (types IIIA or IIIB) results in about a factor of 4 variation in the critical mass estimate. For a given composition and zenith angle, the uncertainty in the ionization parameters (expressed through the use of Bronsthen's formula and that of Jones) results in a variation of about 6 in the critical mass.
It is our basic belief that none of the observational studies published to date can claim to have clearly established a casual linkage between the generation of a characteristic VLF transient signal and the passage of a fireball. We offer this statement in the sense that none of the published papers present instrumental data that unambiguously shows the coincident detection of a meteor and VLF or electrophonic sound transient.
While it is not our intention to be overly critical of honest effort,
we would also like to point out that it is not obviously clear that the
correct line of experimental attack has always been employed. Firstly,
we should ask ourselves what it is that we are actually trying to measure.
Electric and magnetic fields behave differently depending on
the distance R between the source and the point of measurement (see, for
example, Jackson 1975, Chapter 9). Indeed, we can distinguish three
characteristic zones: the near field, the far field, and the intermediate
zone. Comparing the distance R to the source with the wavelength
,
we have
For more massive meteoroids that penetrate deep into the Earth's atmosphere, we are essentially in the intermediate and near field regions. Only in the far field zone would we actually have a plane transverse electromagnetic wave (and therefore a radiation field). In the burster model discussed above, it is most probable that the intermediate field condition applies and hence the electric and magnetic fields will behave differently. Within this context it may be no coincidence that the VLF transient events reported by Beech et al. (1995) and Garaj et al. (1999) relate to meteors seen close to the horizon (but see below). Indeed, if the source is close to the horizon the distance from the observer can be many hundreds of kilometres. In the long-range case, we have a radiation field and consequently the transmission of energy via the Poynting vector. Before embarking on an experiment to study electrophonic transients a clear decision has to be made with respect to whether it is the electric or the magnetic field component that is to be measured. This decision in turn dictates the choice of antenna to be used in the experiment. The Keay and Bronshten electrophonic model is based upon the relaxation of the geomagnetic field, approximated as a magnetic dipole source, and to study this component one must employ a loop antenna valid for frequencies in the range 50 Hz-50 kHz. The mechanism outlined by Beech and Foschini for burster electrophonics may be thought of as generating an electric dipole source, and consequently to study this component a vertical wire antenna should be used. If one wishes to test the transduction process directly a well-calibrated detector and microphone system will have to be developed.
In addition to the appropriate selection of an antenna, care should also be directed towards the choice of signal analyzer. The best choice of detector is probably an Electro Magnetic Interference (EMI) analyzer. Such devices are typically sensitive over a wide dynamical range of frequencies. And indeed, we note, that since one is trying to measure implusive or rapidly changing broadband signals an instrument with a small dynamical range will seriously compromise the quality of measurements being made.
The model for electrophonic sound generation described by Keay (1980, 1993) makes a clear prediction. That is, the VLF signal generated by the meteoroid interaction with the geomagnetic field will be sustained, possibly observable over an extended period of time, and it will be distinct from background atmospheric sources. While the exact VLF signal characteristics can not be predicted at this stage, it is unlikely that it constitutes a series of very-short duration transients. In this respect, we do not at this stage fully accept that the very short duration VLF transient signals presented by Garaj et al. (1999) in their survey paper represent anything other than background, atmospheric events. They may represent a real meteor related signal, but the evidence it not compelling. The VLF transient signal presented by Beech et al. (1995), on the other hand, has characteristics (signal duration and distinctness from natural atmospheric sources) that make it more believable as an actual meteor related event. Our point here is that unless the meteor generated VLF signal has characteristics that clearly distinguishes it from background sources, then one cannot simply claim that because short-lived VLF transient are observed at the same time that a meteor is seen that the two observations are causally related. It is not clear to us as yet whether the electrophonic bursters are likely to generate a significant VLF signal.
We re-iterate, however, that we do not expect every Leonid fireball of magnitude -7 (and brighter) to generate electrophonic burster sounds. The key point is that the meteoroid has to develop a shock wave at the same time that the electron number density in the plasma column exceeds 1019 m-3. In addition, to generate a measurable signal in the VLF range the meteor must be placed at a distance greater than some 75 km from the observer.
Many of the Leonid fireballs recorded in 1998 were found to began rapid ablation at heights in excess of 120 km. In addition, the average end height of Leonid fireballs was found to be about 85 km (Spurny et al. 2000). These results certainly indicate that Leonid meteoroids are made of very friable and easily ablated material. Indeed, Spurny et al. (2000) classified all but one of the Leonid fireballs that they recorded as type IIIB, which is typical of the weakest interplanetary bodies. The model outlined above suggests that a meteoroid has to penetrate to an altitude of about 80 to 90 km before it undergoes catastrophic disruption. On this basis we see no inherent reason why some Leonid fireballs might not produce electrophonic bursters.
One of the key aspects of the enduring electrophonic sound model developed by Keay and Bronshten was the production of very low frequency (VLF) radiation. This radiation is generated through an interaction of the highly ionized and turbulent meteor plasma train with the Earth's magnetic field. It is the transduction of the VLF radiation by objects close to the observer that results in the generation of audible sounds. In contrast, the electrophonic burster model outlined by Beech and Foschini does not specifically predict the generation of any VLF signal. Rather, it predicts the generation of a short-lived transient pulse (or pulses) in electric field strength. The observational campaigns conducted to date have mostly focused upon the detection of VLF radiation transients through monitoring magnetic field variations. Electrophonic bursters, we argue, are more likely to be observed through the generation of electric field transients.
Acknowledgements
We extend our appreciation to the referee, L. Bellot Rubio, for his comments and suggestions. This work has been partially supported by a grant from the Natural Sciences and Engineering Research Council of Canada and partially by MURST Cofinanziamento 2000. This research has made use of NASA's Astrophysics Data System Abstract Service.