The stars were quite clear, and the air was a bit thinner. As I watched the sky, a meteor came almost straight down at me and burst.
I heard it. Simultaneously with the burst.
No, I did not have long frizzy hair, but there was a metal wire fence nearby.
Yes, you can hear them sometimes. With a track along the sky, some people hear crackling and popping. The topic is getting more serious attention, but the effect has been observed for years: China 817AD, England 1719AD.
According to this paper the meteor track can be a meter wide. That's potentially a lot of ionized gas. If an oxygen atom boils off the meteor and is ionized, it is initally moving at about 1.1 to 7.2 E7 meters/second. The Earth's magnetic field isn't very strong--2.5-6.5 microTesla--but with such a large speed the V cross B isn't negligable. The initial force is of order 2 to 75 E-19 Newtons, which given that the oxygen atom is only about 2.7 E-26kg, gives quite the acceleration.
The full MHD solution is way harder than I can solve on the back of an envelope, but that initial acceleration sounds pretty promising.
Suppose you have a cylinder of ionized gas moving at meteor speeds. It will slow down very rapidly, but in the meantime MHD forces will push the positive ions one way and the electrons the other. When they slow down enough, they will pull back together to recombine. If this second timescale is long enough, you should get a column of postive and a column of negative charges moving toward each other, which should create a low frequency radio pulse.
The only problem with this model is that it doesn't work: radio wave generation is rare!
You can detect meteor tracks from the way they reflect radio waves, but they usually don't seem to make any themselves. For starters, this naive model assumes no turbulance, which isn't even close to reality where gas gets mixed quickly. And it doesn't deal well with a bang at the end. But it gives an idea of what some of the forces are.
Now that I think of it, I wonder how much lightning you get along meteor tracks. They're far higher than thunderclouds, but the sprites are quite high too, and there have to be interesting electric fields with the sprites. Whether the same sorts of fields exist away from above thunderclouds I don't know.