Below is a paper I wrote for HET602, the introductory class for Swinburne University of Technology astronomy master's students. It was my first paper and written for an intro class, as such I make absolutely no assurances that anything in this document is correct. I urge you to visit the sites in the Sources section for more accurate info. Back to astro part of the site
In the mid to late twentieth century the detection of meteors via radio became a scientifically useful method for professional astronomers to analyze and observe meteor showers. Around the beginning of the twenty- first century, the availability of technology and education allowed amateurs to join this in this pursuit. I have endeavored to apply this new approach to create an in depth and scientific observing program for the Eta Aquarid meteor shower in April and May of 2000. My results show that even though the tools and procedures may be well understood, there is no substitute for experience. However, with that proper experience amateurs can indeed make useful contributions to the study of meteor showers.
The Eta Aquarid meteor shower is a relatively new shower, having been discovered and confirmed in the nineteenth century. However, recorded observations of the shower date back to around 400 AD. It was not until 1870 when the shower was discovered by G.L Tupman. On April 30 of that year he plotted 15 meteors with the same radiant. On May 2-3 he plotted 13 more with almost an identical radiant. At the same time, a group of observers with the Italian Meteoric Association plotted 45 meteors from April 29 to May 5 of that year with almost the same radiant. A year later on April 29, now alerted to this possible new shower, Tupman plotted 8 more meteors with the same radiant and thus confirmed the Eta Aquarid meteor shower. In the following two decades research conducted by William F. Denning of the Italian Meteoric Association and Professor Alexander Herschel associated the new shower with the orbit of the famous comet Halley.
The shower was not very well defined until the 1930's because of a characteristic favoring of the southern hemisphere by the radiant. There were not many southern observers at the time to keep track of the Eta Aquarids. In 1935 a published investigation by Ronald A. McIntosh of Auckland, New Zealand, revealed a broad activity curve lasting from April 28 to May 16, with a peak from May 3 to 6. The hourly rate at peak was estimated to be around 10. However, in 1947 radio observations were just beginning and this shower was one of the first to be observed. Radio observations increased the peak 20% to 12 per hour based on the 1947 shower alone. Slowly but surely our understanding of this shower was beginning to grow.
During the 1950's and 1960's radio observations of this shower really came into its own. Thanks mainly to the use of radar, fantastic amounts of meteors could be detected eventually placing the Zenith Hourly Rate (ZHR) between 300 and 600 during its peak. However, the complex structure of such a broad shower also began to show. Some observations displayed multiple apexes while others showed a single peak which moved year to year. The only thing that everyone agreed upon was that the stream of debris encountered by the Earth was complex. By the early 1980's, McIntosh published a paper with another well known Eta Aquarid researcher, A. Hadjuk. Together they proposed a complex stream of material left over by Halley's comet centuries earlier. It was more like a shell of matter than a simple stream. That theory, though understood to be speculative, is still respected to this day.
The 2000 Eta Aquarid meteor shower was expected to peak at 17UT on May 5th. This is 1pm local time so my area of the planet will be facing away from the meteor stream. Rates for my latitude were expected to be around 10 per hour with rates increasing to 60 per hour in the southern hemisphere. The best time to observe meteor showers are in the hours before dawn because that is when the observer's area of the planet is perpendicular and heading at full speed into the path of the meteor stream. Eta Aquarids are known for being speedy (around 60 kilometers per second) and thus being bright with long lasting trails. These trails make them ideal for radio observation.
Observing this meteor shower for this project posed quite a bit of a challenge. I can thank geography for the first challenge. The Eta Aquarids strongly favor the southern hemisphere since the radiant just barely rises above the horizon before dawn for those in the northern hemisphere. The second challenge involved the technique of observation. Radio observation has many advantages over visual observation. First, it can be recorded easier for later analysis. Second, it can be used during poor weather and bright moon phases. Finally, it can detect meteors on average 2-3 magnitudes fainter than the human eye. However, it also presents new challenges. Some of these challenges include eliminating local radio interference from artificial noise sources, calibrating receiving equipment, finding the right frequency, and finally, reducing the data. The most important challenge unfortunately cannot be overcome, and that is the inability to differentiate Eta Aquarids from the sporadic background. It is because of this that a good baseline of observations must be made along with an attempt at basic statistical reduction.
Radio observation of meteor showers uses an indirect detection method called forward scatter. That is, instead of hearing the meteor itself (we will get back to that issue later) we are listening for reflections of other radio sources off the meteor, or more precisely, its tail. Forward scatter has its origins in World War Two when radio operators learned they could briefly pick up long distance stations off of large concentration of high altitude flak bursts. Today forward scatter is a popular hobby of HAM enthusiasts across the world. When a meteor enters the Earth's atmosphere, its speed is so great (averaging around 50 kilometers per second) that it leaves a trail of ionized plasma particles behind it. This trail glows in the visible spectrum and is what we see as a shooting star. In the radio spectrum, however, it is reflective to VHF frequencies and thus echos electromagnetic energy. These reflections occur best in the 50-80MHz range, but can be heard all the way up to 200MHz. The higher in the frequency you go, the more faint the signal is. In order to hear these echos you need to find a clear VHF frequency (Not an easy thing for an area as electromagnetically crowded as Boston, Massachusetts, where I live!). The most powerful transmitters in this frequency range are television and FM radio stations. TV stations are lower in the frequency, thus better choices. I found that channel 3 was not used in my area. Television has three different signals. One carries the audio, another the video, and yet another for the color signal. The audio signal is not popular because sometimes it is quiet, and it modulates (using frequency modulation) so you can miss some short meteors during moments when the transmitter is silent. The color signal is weak and was interfered with in my area. The video signal, however, is a good choice because it sends a constant strong carrier signal. The video signal for channel 3 in North America is 61.25MHz. I listened to 61.24 because 61.25 was a little noisy in my area and also because stations frequently broadcast at 10Hz offshoots (so as not to interfere with other close channel 3 stations).
Meteor observations are manifested as reflections of the distant transmitter on the otherwise quiet frequency. These reflections can sometimes be quite long and clear. Its not unheard of for people listening to FM radio transmitters to hear short bursts of music or radio commercials. However, for the most part the reflections are distorted by the movement of the ion trail by upper level winds. When listening to television video signals on continuous wave (CW) you listen for a number of sounds, but they all basically sound like sonar pings. Sometimes it is a very short and high pitch ping, sometimes it is a deep pong, other times it may be blowing over an empty bottle to create a whistle, sometimes a combination of any of these.
The sounds you hear tell you something about the meteor. First, the length of the reflection tells you the relative size of the meteor (similar to the magnitude and trail length for visual observers). Secondly, the type of sound itself tells you a bit. There are three main echoes. The first, and most common, is an underdense echo. These are meteors where the tail's free electron density is below 10^14 electrons per meter. Below this critical density, radio waves are scattered independently thus creating a unique sound. Underdense events typical correspond to meteors with a visual magnitude of 5-8 as seen by the naked eye. The signal strength of these echoes usually start quickly and then fade out exponentially. A second type of meteor echo is the opposite of this sound and, predictably, is called an overdense event. This event occurs when the ion trail is thick enough (over the critical density) for the electrons to cooperate when they reflect the radio waves. The sound of these events usually starts slowly, is deeper, and fades slowly. These meteors are typically seen visually as greater than 5th magnitude by the naked eye. A relative of this event is the oscillating overdense echo where the thick ion trail is twisted and broken apart by winds, thus creating numerous identical signals. There is also a hybrid of the overdense and underdense events called a transition echo. These are actually overdense events that sound like an underdense event (with a sharp rise in pitch at the beginning) because of a low reflection angle with the transmitter. Finally, you can sometimes pick out the sound of the meteor head itself. This is a rapidly descending tonal pitch that sounds like a bomb falling from a plane. This is caused by the doppler shift of the meteor descending through the atmosphere and is usually followed by an underdense or transition event when the meteor reaches a point where forward scatter begins called the fresnel zone.
For my observations I used a Winradio 1500i radio card. It is a radio receiving card that hooks into your computer. Despite being designed to be used under Windows, I used an open-source program under Linux to control the card. The card was wired through my apartment (and roommate's bedroom window!) and hooked up to a discone antenna on my porch. I am situated on a tall hill about 15 miles west of downtown Boston. Within 10 kilometers of me are transmitters for almost every radio and television station in town. Needless to say, the interference was horrendous. However, after many days of frustration and with the help of a $5 FM trap hooked to my antenna, I was able to block out the interference on the desired 61.24 frequency. My antenna, being a discone, is an omni- directional receiver. Because of that, I could pick up video signals from stations almost all around me. I tried horizontally aligning the antenna to hopefully cancel out a few (you only need one to do radio observing), but found that the standard vertical orientation was best. So with that in mind, I research what transmitters I will be listening to. As luck would have it, there was only one channel 3 transmitter in my area. It was WSFB TV out of Hartford, Connecticut, broadcasting at 100 kilowatts to the southwest of my location. This means that most of the echoes I hear will be from this transmitter. However, there are two more transmitters within 1000 miles of my location which are much fainter, but will be responsible for some of the fainter echoes. With all my research and preparation done, I was able to finally get down to observing.
My observing schedule was broken up into two main projects. First, I wanted to setup a long term baseline of activity. I wanted to do this because the Eta Aquarids are a shower with a broad peak. I also wanted to get a better plot for what kind of daily sporadic activity to expect. This plan involved observing for one hour centered on midnight every evening ten days before and after the expected peak. The second project involved more detailed observing during the peak itself. For this project I spent four and a half hours per night, from 11:30pm to 4:00am locally, for three nights centered on the peak. This would hopefully give me the data I need to find a true peak in the shower, possibly along with some secondary peaks. With the data that these two programs would provide I hoped to be able to accurate map out the 2000 Eta Aquarid meteor shower.
Below is a table of my observations.
First Project: Setting Sporadic Baseline (23:30-00:30)
| Day | 27 | 28 | 29 | 30 | 01 | 02 | 03 | 04 | 05 | 06 | 07 | 08 | 09 | 10 | 11 |
| Meteors | 50 | 55 | 51 | 62 | 74 | NA | 61 | 137 | 111 | 59 | 70 | 62 | 66 | 59 | 65 |
Note: On May 2nd equipment trouble prevented adequate observations during observing window.
Second Project: Observing The Peak(s)
May 4 May 5 May 6 00:00-00:59 57 130 97 01:00-:01:59 75 128 140 02:00-02:59 58 84 64 03:00-03:59 73 103 76 Total 263 445 377
Note: All times are in EDT (-4h UT)
The results overall show a trend consistent with the predictions, however there are a few unexplained discrepancies in the data. First, the peak did fall on May 5th, as predicted. However, the peak was supposed to be at 17UT (13 local) so I expected the second most active period to be the evening of May 4th. Either the peak was later than expected, a secondary peak appeared on the 6th, or observations were inaccurate for the task. My theory is that a secondary peak was encountered the following day due to the fact that the sporadic baseline also found the secondary peak, plus the Eta Aquarids are suspected of having many small secondary peaks. Another worry I encountered was the 2am drop off encountered on each day. I have no good explanation for this. Observer fatigue cannot be the answer since numbers rise the following hour. My suspicion is that possibly one of the television station video carriers goes off the air at 2am and comes back at 3am, or that sporadic rates increase enough at 3am to make up for the loss of the carrier. Since we are listening to multiple carriers, I noted which observations came within a second of each other. I considered these observations "double observations" and possibly being the result of multiple transmitters being bounced by the same meteor. Below is a table showing only double observations.
Double Observations
May 4 May 5 May 6 00:00-00:59 2 5 4 01:00-:01:59 4 16 17 02:00-02:59 2 2 4 03:00-03:59 6 4 4
The results show that double observations drop by a larger margin at 2am than the general observations. This supports the theory that a particular transmitter may be going out of service at that time. However, the results are purely circumstantial and require much more observation to determine the true cause.
Overall I am very happy with my results. They exceeded my expectations. At the beginning, I had quite a heavy learning curve ahead of myself and I was worried that I would not have everything understood well enough to make careful observations. With the help of many people on the Internet, especially members of the Meteorobs e-mail mailing list, I was able to overcome these obstacles and gain confidence in my observing skills. My next step will be to setup a better antenna system to cancel out all but a single transmitter and to perform frequency analysis of the observations to distinguish the types of meteors I am hearing. This was a very enjoyable project and I look forward to continuing this new hobby.
Appendix
Example #1
This example is 15 seconds long and features a "bong" sound followed by a long trail that slowly fades away. I believe this to be an overdense event with a low reflection angle to the transmitter. I recorded this meteor at 04:29UT on May 5th.
Example #2
This is an example of an oscilating event. The oscillation is caused by the twisting of the trail by upper level winds. It began as an underdense event. I recorded this meteor at 04:12UT on May 6th.
External Sources
Below is a sample visual spectrum analysis of a meteor observation by radio. I do not have the software to make visual analysis so I have obtained express written permission from James Richardson (richardson@digitalexp.com) of the American Meteor Society to use this one, made by himself. The page it is from is http://www.amsmeteors.org/audio/index.html.
This is a rare meteor head echo made when a meteor comes in at an extremely low angle to the transmitter. Three underdense events are at the beginning and one at the end.
Sources
"5/5/2000: The Meteor Shower" by Dr. Tony Phillips for NASA Science News. http://science.nasa.gov/headlines/y2000/ast02may_1.htm
"A Radiometer Audio Gallery" by James Richardson for The American Meteor Society. http://www.amsmeteors.org/audio/index.html
"Eta Aquarids" by Gary W. Kronk for The American Meteor Society. http://comets.amsmeteors.org/meteors/showers/eta_aquarids.html
"Scattering of Radio Waves Off Meteor Trails" based on "High resolution meteor profile interpretation", Proceedings of the IMC 1993. http://www.imo.net/radio/reflection.html
"IMO Meteor Shower Calendar 2000" by Alastair McBeath and Rainer Arlt for the International Meteor Organization. http://www.imo.net/calendar/cal00.html