Figure 1. Distant FM radio or TV signal reflected from the ion trail of a meteor.

 

Here is a description of the apparatus and software used to detect "radio meteors". For over-the-horizon beacons, one can use distant FM stations, but a better signal-to-noise ratio and interesting audio sounds are heard using a TV video carrier. Channel 2 in Boston (55.25 MHz) is a good source in the early morning when it is not transmitting digital TV signals. It is 160 miles away. Any of the lower VHF channels 2 through 6 is fine, but not a local station.

Figure 2. Block diagram of apparatus used to detect trails of ions caused by meteors.

 

The VHF radio I use is a Watkins-Johnson type 906A-6 which was made around 1960, but is still useful for this purpose. Any VHF receiver that will tune from 50 to 150 MHz and has an IF output should work fine. Commercial products like the Icom R-7000 or a similar VHF reciever is ideal for the purpose. My receiver has a 21.4 MHz IF output maximum of about 2 V p-p. The bandwidth is switchable, and 20 KHz seems to work best. This signal is fed to a simple demodulator and low pass filter, shown below in figure 3. In my apparatus, the output of the filter is then sent to a Metrabyte DAS-16 data acquisition card in a PC, and processed for display and data storage using a homebrew QBASIC program. The data sample rate is 10 per second. The microphone input of a sound card can be used for analog to digital conversion if sufficient documentation makes it possible, but as these cards sample at around 44 KHz, the resulting files will be large.

Figure 3. Demodulator and lo-pass circuit

 

When this system was first connected with the QBASIC code running and a dipole antenna outside, a distant FM station was tuned, and almost immediately generated presentable data. Some results are shown in figures 4 through 6 below. The signal to noise ratio is good, and even better with the TV carrier than with the FM stations. A thousand signal strength units is about a quarter of a volt of signal out of the demodulator and low-pass filter.

Figure 4. First light! The noise floor is mostly from FM station modulation.

 

On the web are many sites describing radio meteors and their detection. (see References)  Some of these mention "overdense" and "underdense" events, referring to the number of ions produced. Bigger meteors produce overdense events. The ion trails last longer. During some overdense events like the one shown in Figure 5. below, the audio output of the receiver produces a rapid chirping sound, like little birds. It is possible that the TV sidebands are being Doppler shifted into the video carrier frequency, where an audible hetrodyne note is heard. The mechanism for this may be the ion trail decelerating, resulting in a changing Doppler shift .

Figure 5. Long duration overdense event with prominent Doppler birdies in the audio.

 

When underdense events occur, sometimes a tone is heard in the audio range, lasting as long as the spike lasts. This may also be a Doppler effect. The pitch of the tone is most often unchanging, but occasionally it will vary. It is not like the fast chirping birdies in an overdense event.

Figure 6. Overdense and underdense events in the same 30 second time frame. Note the improved noise floor using the TV carrier.

 

Though radio meteors can be detected at any time of day, midnight to noon is the peak period for meteors, and airplane traffic becomes noticeable after 6 AM because of multipath interference effects. See Fig. 7. The atmosphere is most stable before dawn, so the best times for observing are in the early morning. For those who do not like to be awake in those hours, computer controlled automation is an option.

Figure 7. Multipath interference from an airplane.

 

Another interesting feature of these traces is the exponential tail associated with overdense events, for example, from 9-30 seconds in figure 5 and 12-14 seconds in figure 6. This tail is usually noisy. Many sources on the web say that this is due to multipath interference effects. Ok, maybe, but the spike at 25 seconds on figure 5 is probably a seperate underdense event.

One of my friends saw these traces and asked if I was sure these events were due to meteors. this is a really good question and one to be answered carefully. I have not answered her question yet. I have three ways of thinking about this. First, everybody else thinks they are meteors, but that's really not much of an answer. Secondly, compare the kinetic energy of incoming sand grains with the energy of a typical cosmic ray. No contest, the sand grain, at 40 km/s, has much more energy. Third, I have been keeping track of hourly rates for different dates of observation, so as to correlate hourly rates with known meteor showers. So far, the results are inconclusive, as I have unfortunately slept through the shower peaks! I'm pretty sure they are meteors, but not certain. In any event, it is fun to try to interpret what is seen.

References: Google on "radio meteors", and these sites will appear, along with many more!

http://www.spaceweather.com/glossary/nasameteorradar.html

http://www.imo.net/radio/

http://www.amsmeteors.org/radmet.html

http://www.net4you.co.at/user/kuneth/gallery.html