Meteor detecting

I gave this a quick and dirty try last night with an indoor whip and the only thing I received was spikes and noise from my laptop and switched mode PSU. Failed again, miserably!

Could the trace anomaly be caused by Graves altering the angle of its transmit lobe as it scans the sky? The blurb suggests the radar sweeps from horizon to horizon. Therefore the reflection angle could be affected by movement of both meteor and the ground angle of the transmission: could that explain the result?

Another influence might be alterations in the polarisation of the signal. The yagi is horizontally polarised when what bounces off a meteor trail is likely to be all over the place.

Dave

Edited By SillyOldDuffer on 20/06/2018 10:28:44

The 'blob' is not necessarily a point of maximum signal, but the meteor exploding part way though its final descent. Yes, polarisation of received signals happens all the time, but is most noticeable in faint signals where it means the signal strength may reduce to below the noise.

To counter this, a 'turnstile' or crossed dipoles aerial may be made, which will pick up any polarisation whatever it is.

The doppler shift still applies, but the meteor isn't heading for 'you', the trail begins where the meteor hits atmosphere dense enough to be resistant, and stops when the meteor burns out.

I was listening this morning and didn't hear any either!

Geoff

Dave (SoD) got me looking at the spec for the Graves radar and the azimuth beam width is 7.5deg with 4 beams operating simultaneously and each beam doing 6 steps to provide 180deg coverage. Each step is 3.2 seconds. Looking at Neil’s plot the event is active for just under 3sec. Indicating that the duration of the event is a function of the time beam is illuminating the target, not necessarily the time the target is active.

Pete

Just a quick observation,if I've read this correctly. If the object is falling throught the atmosphere albeit at an oblique angle when it is at high altitude there is a large velocity componant bringing it closer to the receiver and the transmitter. As it gets lower into the plane of the receiver transmitter the lateral motion will have an increasing effect. To put it simply down dominates initially and then away or towards becomes dominant at lower altitudes.

Also if the transit is crossing the line between the transmitter and receiver there is another velocity vector which moves towards and then away.

You could probablyreasonably assume that the atmosphere is 300miles thick and that most meterites burn up above about 50 miles.

Could that explain your traces?

regards Martin

Edited By Martin Kyte on 20/06/2018 11:22:38

Good idea; I'll build one and put it out in the garden. But that raises another question, how high off the ground should it be?

I'm thinking a turnstile might work better at 2 or 3 metres than higher because ground reflections might reinforce high angle signals. Or is it better high up because increasing the horizon distance allows the aerial to see more sky. Except I want to focus on sky between here and Dijon. I'm beginning to think Neil's yagi, or a crossed yagi is a better answer. Why is nothing ever easy?

This is a terrible distraction from another slow-running project, which itself is already keeping me out of my workshop. I'm doomed.

Dave

Height doesn't really matter because you want it to 'look' upwards, any other signals are unwanted. Some mount it in a foil-lined trough, but the jury is out on its effectiveness. Take care with the feed arrangements, one dipole is fed 90 degrees out of phase with the other, it is easy to do. Google 'turnstile antennas'.

Geoff

Agreed, it gives the line of sight delta-v rather than actual velocity.

ISS see the AF window at bottom right where it is visible as a VERY faint line dropping from 1200 to 700Hz as it recedes (over the middle east which is why it's so faint!):

The 'blobs' are what is most often see - they are known to represent the persistent ionisation trail left by the meteor and can be several seconds long (I've seen one that lasted a minute or more that was photographed by several people, including me). Two issues are (1) can you detect the 'meteor head' and (2) is it possible to detect it moving both towards and away from the observer?

I've been given a copy of a discussion paper (sadly without cover so we don't know where it's from) that states:

Does the RCS of a moving plate (plasma front) scatter enough signal to be observed? What
would a CW radar return look like from the situation described in this model?
A great deal will depend on the geometry of the observation ie the location of the radar
source, receiver and the direction of the trail. Without a mathematical description of the
geometry we cannot answer this question in detail. However it should be possible to make
clear some points.
Given sufficient illuminating power from a radar source at some general angle from the
meteor vector – and given sufficient plasma front RCS in the direction of the receiver, one
would expect to see a return signal with a changing Doppler shift with time as the LOS
resolved component changes with the geometry. If the meteor is slowing down, this
resolved velocity component would further decrease with time. (However many of the
plasma generation models in the literature assume a meteor particle has a near constant
velocity until its mass is exhausted).
Would there be a geometrical arrangement of radar source, receiver and meteor track
where the LOS velocity would decrease with time – as is sometimes observed?
If we take it that the initial return signal from the meteor would show some type of Doppler
shift depending on geometry, what might we see at later times? The moving plasma front
will at some point cease to develop (due to meteor mass being exhausted being too low in
the atmosphere) and a Doppler signal will disappear. However the stationary plasma trail
may still exist (lasting for 1 second or more) and a strong signal may be expected from this
large conducting object with a high RCS. It will though have little or no Doppler shift. High
altitude winds may impose some bulk movement on the trail, or it may break up due to
shear winds – and in such cases some small Doppler shift will be imposed on the signal.
The foregoing suggests that returned signals may sometime have two clear components
(ignoring wind effects for the moment) – a high Doppler weak signal component and a
stronger signal with little or no Doppler shift. In measurements made by the author of
meteor reflections from the Graves Radar, returns with these characteristics have
commonly been seen.

The emphasis is mine. The paper give this trace as an example of a meteor that approaches and recedes from the receiver:

People using 3 to 7 element Yagis consistently report the best results. Aim it about SSE.

Both horizontal and vertical polarisation are used, but it's not important as returns from meteors have random polarisation.

Some people point it up by 10 degrees. This bewilders me as I thought most astronomers accepted that the Earth is round and a horizontal antenna will be pointing high up in the atmosphere over Southern France.

Yagis need pointing, but at what? The radiant won't help because that's where the meteors 'seem' to originate, but are actually seen at widely separated angles. The gain provided by a yagi is not really necessary as the signals are generally strong enough. A turnstile needs no directivity.

Pointing the yagi upwards by 10 or so degrees minimises signals from terrestrial sources in that direction. You need a vertical polar diagram to show this, rather than the usual horizontal polar diagram, which shows the gain pictorially using the aerial location at the centre.

I forgot to add, simple meteor detection works better at lower frequencies, below 150 MHz.

Geoff

The Graves radar in southeast France!

It's a 180-degree wide array they use to plot the orbits of satellites, it's beamed up at the sky which is why it works well.

Greater gain gives a different profile for detections - more smaller ones from a narrower area. As the number increases exponentially as they get smaller you get more hits by looking at a smaller area with greater gain.

You (or someone) asked about the ISS, I got a much better return of it today. Next challenge will be to get ISS slowscan TV the next time they do a demo. It's transmitted at 145.8MHz using a hobbyist 25 watt Kenwood transciever!

Pointing Yagis - The Graves Radar has an elevation beam width of 25deg pitched up at 15deg. The latitude of Dijon is 47degN and by way of example Northampton is 53degN

If the Rx Yagi also had a beam width of 25deg then the pitch up should be around 13deg so 10deg would be OK as the elevation beam width would certainly be greater than 25deg as it needs to be 180deg in azimuth. There is something to be said for a better shape to the RX antenna polar diagram. Now if you were to have a 4 element array that stepped in sync with the Graves azimuth sweep....

ISS - The ISS plot is interesting as it has a greater doppler shift than the meteor returns. With such a strong return from a target that isn't breaking up on entry is interesting as it has similar characteristics to the meteor return that is allegedly breaking up.

The ISS return also has those faint spectral lines how are they explained? There are also the lines at 143,046,800 143,048,900 and 143,052,900 something in the receiver?

Like Dave this is becoming an interesting distraction, particularly as I have no chance whatsover of reciveing these signals from New Zealand.

Pete

Edited By Doubletop on 20/06/2018 21:40:56

An interesting document

https://britastro.org/radio/projects/Detection_of_meteors_by_RADAR.pdf

Pete

Edited By Doubletop on 21/06/2018 06:34:02

Edited By Doubletop on 21/06/2018 06:34:40

Edited By Doubletop on 21/06/2018 06:35:00

After nothing all morning I heard an echo lasting a couple of seconds at 9:36 GMT. Apart from that very little. I watched the background noise come up with the sun (proving it must be daylight!), and saw my main computer generate some spurs as it booted. Once it's running they mostly disappear. Listening to white noise is torture.

Thanks to doubletop for the link to Dr David Morgan's pdf - very informative.

For 'mislaid parts' reasons my turnstile is a dipole about 6 foot off the ground. Needs more work!

Dave

The meteor doesn't 'break up' it 'ablates' generating a trail of ionised plasma.

Extended meteor trails show diffraction patterns as they are extended targets.

The spectral colours just show changing intensity. The pulsing of the ISS trail is due to the phased scan of the Graves Radar. Note the timescale is probably different from the meteor plot.

Clearly some meteors may be missed by falling between graves pulses.

The vertical lines are probably radar transmitters, that at 143,048,900 is probably the back signal from Graves itself, allowing for a small error in receiver calibration, the others being ATC?

Yep. I've read that three times in the last week!

Thanks Neil

Somewhere in the thread somebody had suggested the meteors would break up, which I wasn't too sure about but if it had been the case then ISS and meteor signals would be markedly different.

The thing I have been trying to reconcile is the faint diagonal lines in your plots that are changing frequency over time. The examples in Dr David Morgan doc don’t appear to have the same characteristics. He has the frequency shift (velocity) over a relatively short timeframe without the powerful ‘burst’ in the middle*. However, having spent years in a past life, peering at spectrum analyser displays of different types of radar signals I do know that the settings of the equipment can present a display that doesn't necessarily represent what you are looking for. Your setup and Dr Morgan’s could be looking at the same event and have different presentations. (* I’ve just realised Dr Morgans velocity display effectively ‘folds’ the inbound and outbound frequency shifts and combines them into one value of velocity)

When you say "pulses" I think you mean dwells. As you know Graves is a CW radar and the scan dwells for 3.2secs at a particular azimuth angle. A 45deg scan taking 19.2secs to complete. You could well miss a lot of events while it was dwelling on another sector.

As I say this is all academic for me as I can’t see the signals down here to be able to get some kit and join in the fun.

The faint lines are probably receiver artefacts. SDR produces them if something in the receive chain is slightly out of balance. Even if the SDR module is properly adjusted not all computer sound-cards are wonderful. Also, being broadband, SDR can be overloaded - the lines may be the effect of a strong out-of-band signal.

Or they might be interference from nearby domestic, comms or industrial equipment. Even more likely because it's so close, the computer. If it matters enough to check, the strength of an actual signal varies when the aerial is swung, artefacts stay constant. Probably not radar or Air Traffic Control - no modulation.

Left my receiver running all day and noticed 4 or 5 echos. All small, the first event in the morning being by far the biggest. Microsoft doing a major update lost me a couple of hours. Bet that never happens to Jodrell Bank!

Dave

Any FM Broadcast transmitters at a suitable distance from you? Before Graves helpfully provided a clean signal in a quiet part of the spectrum, people did similar detections with distant FM transmissions. The hard part was finding one that wasn't blotted out by local broadcasters using the same frequency.