I have a GPS tracker/transmitter I use in rockets. I use it to locate and recover a rocket. I realize that Altitude is not the strong suit in GPS, and don't expect it to be really accurate. Neither do I expect the lock on a typical GPS unit to be maintained during the rapid acceleration of a rocket.

However, at apogee, and under parachute, I find that the re-locked-on GPS altitude varies enormously from the verifiable height of the rocket trajectory. A rocket descending from 5000 feet might show between 800 and 1000 feet in height, calculated as an offset from the launch pad altitude (so no, this is not local terrain elevation causing a difference). The GPS-reported altitude will rise, but slowly compared to its ability to track lateral (relative to ground) movement.

I've been told that this might be an intentional limitation of capability, to help prevent malicious use of GPS chips, but I can't find documentation of this, and think instead the chip might just be designed for terrestrial uses of GPS, and not need to track altitude change rapidly (or think that rapid altitude change in a land or sea vehicle must be in error).

Does anyone know what limits altitude rate-of-change?

  • 2
    That would depend a lot on the unit itself and what GPS satellites they can use. I would expect a cheap unit to be unable to determine a good Z value even at ground level; GPS units only check every few seconds and will often interpolate a location based on the previous trajectory and speed. If you want any kind of accuracy in height I would recommend using an inertial navigation system tied to a base station at your launch pad, though you'd have to find a very sturdy system for the multi-G environment of rocket launch. Nov 19 '19 at 5:04
  • Thanks - I think timing and interpolation might combine to create the effect I'm seeing: if the system expects to interpolate to a height it's reached rapidly, even though it's not moving rapidly now, there will be quite a lag. I'll be supplementing with a barometric altimeter, vs. inertial nav.
    – jddj
    Nov 20 '19 at 13:33

There is no difference in the rate of change of altitudes with respect to planimetric coordinates. Satellite navigators calculate their position in a geocentric Cartesian system. At the equator, at longitude 0º, the altitudes can correspond to the X axis, at longitude 90º they can correspond to the Y axis, at the poles with the Z axis. And for all coordinates the navigator has the same refresh rate . The problem is the inaccuracy with which determine its altitude.

The main limitation of GNSS units in the determination of altitude is the result of the trilaterations being carried out at stations (satellites) located in the same direction (the sky).

This defines the shape of its error ellipsoid, which is disproportionately vertically oriented.

If we could measure trilaterations to stations well distributed in space (this would involve measuring the distance to satellites below ground) the error ellipsoid would have a spherical shape. The same imprecision / inaccuracy that is obtained in planimetric coordinates would be obtained in altimetric ones.

In its origins, the GPS network had a system of selective availability, in which an intentional degradation of the signal was introduced to induce a minimum error in the positioning.

But the satellite navigation system advanced a lot in civil society, many networks are fully operational in addition to GPS, forming the GNSS system among all. Since then the GPS network must have set aside selective availability.

Currently, the errors produced by simple frequency receivers without differential postprocessing are the product of the defective spatial distribution of the trilatations and the absence of ephemeris correction of the position of the satellites.

The inclusion of an additional band in the GNSS message is being implemented, in order to provide autonomous navigation vehicles with absolute positions of greater accuracy. (I regret not having sources to support this information.)

  • Thanks. This is a wonderful comment on the situation, which has enlightened me a bunch on GPS altitude, though I don't think it gets at what I'm seeing. Michael Stimson's comment above on interpolation and timing seems like a direct cause.
    – jddj
    Nov 20 '19 at 13:38
  • Honestly, I don't think satellite navigators will extrapolate trajectory and speed to estimate their position. It would be easier for me to believe that it takes a long time to compensate for the indeterminations in altitude by iterative methods to explain the phenomenon that you are observing, rather than to believe that they obey a mixed system of satellite and pseudo inertial positioning by default. Nov 20 '19 at 23:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.