Electrophonic Meteors

Meteors are what are commonly known as "shooting stars".   However, these are not stars at all.  Meteors are small grains of dust left behind by comets as they orbit the sun.  Comets are basically dirty snowballs that heat up as they approach the sun.  Some are periodic (like Halley and Temple-Tuttle) returning every so often to visit us.  As they approach the sun they evaporate or sublimate, with tiny dust particles ejected, remaining behind in space in the same orbit as the comet.  It may be hundreds or thousands of years before one of these dust particles collides with a passing object such as the Earth.  And when this happens, we see these meteors burning up as they enter the atmosphere.  Larger particles and small rocks from asteroids can result in huge fireballs or bolides as they enter the atmosphere, lighting up the whole sky [NASA movie].


For centuries people have claimed to hear sounds while observing the large meteors (bolides) burning up as they enter our atmosphere.  But even Edmund Halley (from comet Halley fame) in 1719 concluded that this was pure fantasy, since light travels much faster than sound, and therefore there is no way that anyone would be able to hear anything simultaneously as they see the meteor tens to hundreds of kilometers away.  But the anecdotes kept coming in.  Were these people really crazy, or imagig things?

An Australian scientist Colin Keay suggested in the 1980s that perhaps the meteors generate radio waves.  As with our home radio, we hear the news exactly at 8:00am no matter how far we are from the radio station.  The reason is that the radio waves carrying the information travel at the speed of light to all of our homes, and then our radio has a method of converting the radio waves (that we cannot hear since they are electromagnetic waves) to sound waves (pressure waves) that vibrate the speaker, and our ear drums.  So if the meteor sends out a radio wave together with the light we see, the radio waves and the light would both arrive at the same time at the observer.   However, the radio waves then needs to be converted to sound waves.  Well this can be done by simply vibrating any conduting object at the same frequencies of the radio waves.  This is how speakers produce sound in your home radio system.  Metal fences, glasses, hair, plants, etc.  can all be used as transducers to convert the radio waves to sound waves.  This conversion of electromagnetic waves to acoustic/sound waves is known as electrophonics.

Now if the radio waves do exist, they need to be in the frequency range that people can hear (20-20,000 Hz).  This range of radio waves is known as the extremely low frequency (ELF) and very low frequency (VLF) range, and is also the range used for lightning research.  Hence my interest in this topic.  If meteors do produce ELF/VLF radiation as they enter the atmosphere, then perhaps our sensitive lightning detectors in the Negev desert (below) would be able to pick up these signals. Below is also shown a dynamic spectrum of ourVLF data for a 15 seconds period (x-axis) showing the frequency (y-axis) of electromagnetic waves detected at our stations.  The vertical lines are from natural impulsive lightning sferics, while the horizontal lines are mainly from man-made VLF transmitters used for communications with submarines.


During the 1999 Leonid meteor shower (which occurs every year close to the 18/19 November), we were ready for the challenge.  We set up our instruments in the Negev desert, as part of a large scientific experiment to study the Leonid meteors called LeonidMAC.  The name of this meteor shower comes from the region in the night sky (Leo constellation) where the meteors enter the atmosphere.  The meteors during the Leonids originate from comet Temple-Tuttle that returns to visit Earth every 33 years.  However, the meteors in any particular year can be from visits of Temple-Tuttle hundreds of years ago since we do not pass through the same part of the debris stream every year.

During the 1999 Leonids we detected very brief (~10 msec) pulses of VLF radiation that did not look anything like the normal lightning pulses we observe every day.  The pulses were longer, less impulsive, and actually had a minimum in radiation in the 5kHz range, exactly where the lightning radiation peaks.  However, the pulses were very short relative to the optical meteors themselves.  Anyone who has seen a meteor entering the atmosphere knows they can last up to 1 second or more.  The VLF pulses we found were much shorter.  Were these really from the meteors?  So we decided to count them.   We knew that during the night of 18 November 1999 there was a huge peak in optical meteor activity around 0200 UT.  After counting these strange ELF/VLF pulses for the night before the meteor shower (16/17 Nov) and the night of the shower (17/18 Nov) we found that the number of these pulses tracked the optical counts very well (see below).  The night prior to the shower showed no significant changes through the night, while during the meteor shower (black curve below) the ELF/VLF counts reached a maximum, decaying together with the optical counts.  However, the ELF/VLF data showed another peak around 0045 UT not seen in the optical data.  Could there have been another weaker shower of meteors not bright enough to be seen, yet large enough to produce radio waves?  In addition, the ELF/VLF pulse counts reached 15,000 per hour! Compare that with the 400 per hour observed optically in the Negev.  Could we be detecting many small meteors that cannot be detected optically, or does each meteor produce a number of pulses?  This is still an open question.


But back to the electrophonics.  None of the observers out that night heard any sounds while observing the Leonid meteors.  There were also no big bolides.  However, we now think we have found the evidence that can explain the reports of people hearing sounds while observing bright shooting stars.

Another project we have worked on in recent years is to see whether the meteors themselves may impact the characteristics of the lower ionosphere (around 100km) on nights of meteor showers.  For this study we looked at the background VLF noise levels (Vd values) to see if they change on nights of meteor showers.  The noise is generally produced by lightning activity  in our region, or even thousands of kilometers away from Israel.  What we noticed was that on nights of meteor showers, the noise levels went down (Vd goes up).  Why does the noise decrease?  We think that the meteors change the conductivity of the upper atmosphere, resulting in more absorption of the distant lightning signals before they arrived in Israel.  The amount of lightning activity did not change, but the VLF pulses arriving in Israel did decrease due to the changing characteristics of the Earth-ionosphere wavegiuide. Below we can see the plot showing the  value of Vd  as a function of hour of day (x-axis) and day of year (y-axis) for two meteor showers (red dots and dashed line show peaks in optical observations of meteors).  The plot shows that the noise level changes are related to the meteor showers, while the impact is large during nighttime than during daytime.


    Price, C., and M. Blum.  ELF/VLF emissions detected from the Leonids ’99 meteorites, 2000: Earth, Moon and Planets, 82-83, 545-554.
    Reuveni, Y., C. Price, Y. Yair and R. Yaniv, 2010:  The connection between meteor showers and VLF atmospheric noise signals, J. Atmos. Electr., 31 (1), 23-36.
Related websites:

University of Kentucky
Australian site
ELF/VLF meteor signals