NavList:

Several navigation systmes use the Doppler shift as a satellite passes to get the closest approach and distance from the satellite's orbit. EPIRBs and COSPAS/SARSAT use this in reverse for the sat to find the EPERB. The Argos system used to track wild animals and ocean currents uses its own satellites and dopler location, as did the pre-GPS TRANSIT system. I know this is not exactly celestial navigation but I was wondering how accuratly you could get a fix by observing the Doppler shift as a satellite goes by. Certainly with the amateur communication satellites one can observe a significant shif in the UHF or VHF frequency, and it should be easy to automate measuring this with a cheap software defined radio dongle (like a TV dongle). I wonder how accurate the frequency is on the oscillators and how accurately we know teh emphemeris data? My guess is provided the orbit is fairly well known it should be as accurate as an EPIRB (a couple of nautical miles). 

Polar orbit sats go once around every 90 minutes so it suggests you only need 0.25s accuracy timing the inflection point of the frequency variation to get latitude to within 1'.  The rate of change of freqency gives latitude so I presume less accurate.

I am new to NavList so it could well be that this debate has been had exhaustively already (like the vulnerabilities of GNSS systems) but I would be keen to know if anyone has tried it. Like DSLR sun sights it is one thing where the technology (in this case SDR dongles) just made it so much easier.

Bill Lionheart