15

There's no way for any probe that is sent to orbit/land on a distant body to get there only on precalculated flight parameters alone - minuscule errors in the beginning of flight would convert to huge differences in speed and trajectory near its end. Of course these errors must be corrected during the flight as they become apparent (and before they escalate to uncorrectable levels), but first - they need to become apparent. And while in the initial phase the probes would be able to measure their speed and location quite precisely using the GPS satellites, this is not viable, say, halfway to Saturn, and good few AU away from any near bodies.

So, what methods do the probes use to measure their own flight parameters respective to planned ones?

SF.
  • 54,970
  • 12
  • 174
  • 343
  • 3
    Probes don't locate themselves. We do. See http://space.stackexchange.com/a/2362/265 – Mark Adler Nov 08 '14 at 22:43
  • 2
    What about star trackers like the NAVCAM used on ESA's Rosetta probe? NASA JPL Description of Rosetta NAVCAM – jCoder Nov 09 '14 at 21:54
  • Also see: http://space.stackexchange.com/questions/942/how-can-spacecraft-navigate-without-contact-from-earth?rq=1 on how probes navigate without contact from earth. – ForgeMonkey Nov 17 '14 at 12:31
  • A star tracker is used to determine a probe's attitude, not its location. Using the parallax of nearby stars you may be able to determine whether you're near Earth or near Pluto, but that's not enough for accurate navigation. – Hobbes Jul 21 '15 at 07:30
  • @Hobbes: The star tracker will "see" most planets as well. It won't be entirely accurate but definitely more than parallax between stars. – SF. Jul 21 '15 at 09:07
  • It would see the planets, but (beyond earth orbit) it'd have no way to identify them. E.g. when the probe is orbiting Jupiter, Mars would appear on the opposite side of where we see it from Earth. So you already have to know roughly where you are in order to find Mars in the sky. Planet positions might be useful if you update your position regularly so you have a small delta between one measurement an the next. This is a much more complicated problem than what star trackers are built for, AFAIK. Might be good as a standalone question. – Hobbes Jul 21 '15 at 10:45
  • @Hobbes: It'd find it very surprising if the probe didn't roughly know its position - just using the clock and programmed trajectory. Then it can expect given planets in specific areas of the sky, and fine-tune own location using their precise positions. – SF. Jul 21 '15 at 11:28
  • 1
    @SF: it seems to be possible (http://adsabs.harvard.edu/abs/1992gnc..conf.1287V), but the star trackers I've found so far (incl the ones used on New Horizons) don't mention any positioning or planet-finding capabilities (http://www.aiad.it/aiad_res/cms/documents/SELEXGALILEOStarTrackers.pdf) – Hobbes Jul 22 '15 at 09:26

1 Answers1

11

Interplanetary probes generally don't have enough equipment on board to calculate their own position. So the location of a probe is measured by ground control, using data from the probe.

  • Distance and speed along the Earth-spacecraft axis is done by sending signals to the probe, having the probe transmit them back immediately and measuring the time elapsed plus Doppler shift. See Mark Adler's answer to a related question.
  • The remaining parameters are derived from the above with some calculation, plus data from the spacecraft. Navigation images can be used, for example: these show the stars and planets from the perspective of the spacecraft, this can be compared to the view from Earth.

For New Horizons' Pluto flyby, this process was complicated by the fact that Pluto's orbit wasn't precisely known. So the NH team used several methods (*) to measure Pluto's position accurately so they could refine NH's trajectory in turn.

*: the NH team went through Tombaugh's 1930s photographic plates and digitized them to get data from as far back as possible.

There are projects to design a method for space probes to measure their position independently, but none of those are anywhere near 'ready to use'.

Hobbes
  • 127,529
  • 4
  • 396
  • 565