The Sounds of St.Michel



                       Joachim Köppen,

                       Observatoire de Strasbourg,

                       11, Rue de l’Université,

                       F-67000 Strasbourg





During the week of 21 to 26 Sept. 2009 we were near the small village St.Michel l’Observatoire in the Provence for a workshop of the APLF, the French Association of Planetariums. This was the first time that I was in this region. For my lectures I had also brought my apparatus to listen to natural raedio emissions on very low frequencies, but I had not bothered to bring my tape recorder.


A brief test during the first evening showed me that this plateau next to the village is a location with very little man-made electric noise. Surprise ! … fortunately I did bring my audio cable which enabled me to connect the receiver to my notebook computer. Therefore I could add to my lecture data that I had caputered "on site"!

The Tweeks                           


After sunset the lowest layer of the ionosphere (the D-Layer) vanishes due to the absence of solar radiation. Since this layer is responsible for the absorption of radio waves during daytime, one hears at night the signals of lightnings at larger distances (more than about 3000 km). These signals are characterized by a sound which is somewhat musical, as if one plucks a guitar string. In the spectrum one notes that the sharp impulse of a spheric is followed by a short tail, at a frequency of about 1.6 kHz, which is determined by the height of the ionosphere.

During one evening I recorded a number of nice tweeks with rather long tails, which also exhibit tails at the harmonic frequencies : The screenshot below shows the presence of tails at up to 5 times the fundamental frequency. This feature is not due to the overload of the receiver, because one can also find even stronger tweeks which have no harmonics. Thus, it is a non-linear effect of the propagation of these radio waves in the ionosphere.

The other very nice result is the observation of tweeks with different shapes: The further away the tweek is produced, the longer is the tail. In the following image - obtained during our night of observations at OHP – one sees several 'short' tweeks with short tails also visible in the first harmonic. But one also sees one tweek without harmonic in whose tail the frequency drops more slowly. A simple modeling permits the determination of the distance as well as the height of the ionosphere: The 'short' tweeks have their origin in lightnings at about 800 ... 1000 km distance (e.g. Spain, Italy, Rumania, ...), the long one comes from a distance of about 5000 km, which could be Senegal, Ghana, Nigeria, Ethiopia, Iran, Kazakhstan, etc.)

The Whistlers

Sept. 24th

On my first day without a lecture, I grabbed the chance to spend all the afternoon outside, seeking shadow under the trees and listening when in the spaces between the trees. Nothing of interset … at last, at 17:00 I heard a very faint whistling sound and I started recording – just to let the chance not escape. 165 seconds after the start of the sound file, there is something: the analysis shows a faint track between 3 and 2 kHz

At 235 seconds, there is a second signal, somewhat stronger:

Sept. 25th

I left the lecture room several times, about once every hour. All afternoon nothing to talk about, and nothing ar 17:00, … but the first whistle showed up around 18:00. The recording started at 18:09. Almost immediately, one finds a weak, but nice whistler:

It is worth being shown in a zoom, which covers an interval of one second. It is plainly evident that this is a pure tone whistler whose frequency takes about 0.2 second to pass from 6 kHz to 4 kHz.

Comparison with observations in Kiel (2006)

Where do the whistlers come from? In August 2006 I had recorded after a strong spheric a two-hop whistler as well as his four-hop fellow, followed by a rapid whistler, that is one of the one-hop variety!

If we look at a zoom of the one-hop whistler, we see that it takes about 0.2 seconds for the frequency to drop from 6 kHz to 4 kHz, the same value I observed in St.Michel … these whistlers are evident coming from Southern Africa!

The 2-hop whistler is about twice as slow: it takes about 0.4 seconds for the change from 6 to 4 kHz. This difference leaves no doubt about our identification!

Half a minute later, there is another nice one:

Another ten seconds later, one can hear another whistler, preceded by an even weaker sound:


The zoomed view shows clearly two whistlers between 6 and 3 kHz, with the same speed - the slope of the tracks in the spectrogram - and both are pure tone whistlers.

There are two more whistlers, very faint, at 91 and 353 seconds after the file's start, but they are not of sufficient interest to be shown here.

Half an hour later – at 18:27 – I recorded a second audio file which contains nice examples which I shall show already zoomed:


At that afternoon there had been a jolly thunderstorm cloud visible in direction of OHP. It produced very strong spherics which lasted a good fraction of a second. No lightning flash was seen, and no thunder heard. The spectrum below shows such a strong local discharge, followed by the faint whistler!

All observed whistlers are rather faint, but they are very nice and proper! Thanks to this environment of absent man-made electric and electronic noise it was possible to record them with my Ricola box receiver and a whip antenna of 40cm!


The audio files can be downloaded from

All the Audio files have a length of 5 seconds; the filename gives the data and the time of the master file, and the seconds of the offset between the start of each file and that of the master file.