An Engineer in Wonderland – A most unusual satellite.

GOCE is an unusual satellite – the one that was launched recently to measure gravity accurately.


Most unusually, it is an aerodynamic satellite because it flies low at 250km altitude where there are still a few air molecules.

 PICTURE – GOCE is 5m long, weighs 1,100kg, and its chassis is carbon fibre

Its solar cells are stuck on the outside like old-style satellites rather that being on steerable wings, both because drag is less this way, and it can’t afford to be vibrated by steering motors.

In fact, it can’t afford to be vibrated by anything.

Allegedly, its on-board gradiometer is so sensitive that mounted on a supertanker it could measure the acceleration of the ship due to a snow flake landing on its deck.

And any vibration in the chassis might be confused with gravity variations, so we are told non-gravitational forces need to be rejected to 1 part in 100,000.

Hence there are no unnecessary moving parts. There may actually be none.  

Despite its dart-like profile, there is still drag and this is compensated for by an ion engine developed by Qinetiq in the UK. 

40kg of xenon – 30 months of ion fodder – is carried on-board. This is ionised by electrons from an internal cathode that sustains a plasma inside the thruster.

Generally space thrusters are full-on, or off, however GOCE’s is again unusual in keeping its on continuously, with its thrust throttled magnetically to precisely match residual drag.

Gravity is measured by two systems whose data is merged.

Course information comes from GPS (these satellites fly at 20,000km), with GOCE plotting its own position as it bobs up and down on Earth’s mass field.

Fine information comes from the gradiometer within which six masses are constrained – two 50cm apart in each of three orthogonal axis. 

Electrostatic forces suspend the masses, measure their position inside their cages, and servo them to the centre of each cage. 

The servo signal is actually the output signal. 

Common-mode acceleration of the mass pairs is proportional to the externally induced acceleration. 

Errors are reduced with heaters that constrain the assembly to within 10mK (over 200s).

You can imagine the gradiometer would be horrendously tricky to calibrate – and it is.
If you are curious, there is a wonderful idiot’s guide called ‘The Very Basic Principles of the GOCE Gradiometer In-Flight Calibration’.

Combining the two gravity measurement systems allows a geoid model to be derived with 100km spatial resolution and 1-2cm accuracy. 

And a model for gravity can, it is claimed, be determined to within 0.000001g (where g is 9.81ish m/s2 rather than one gramme – how did that potential confusion leak into the metric system?).

Just in case all this is not enough, there are also some reflectors on-board to allow ground-based laser range finding.



Respond below, or to



  1. Is there a convention for avoiding confusion?
    Perhaps we should write 0.000001gravity.

  2. “how did that potential confusion leak into the metric system?). ”
    Because “g” for gravity is not part of the metric system?
    A bit like the other one – how on Earth did sqrt (-1) ever get confused with current?

  3. “Launched on 17 March 2009,” says the European Space Agency. Not the Queen of Hearts 🙂

  4. No-one I’ve spoken to has heard anything at all about this.
    Well, I know it is April the first, but someone has evidently gone to a lot of trouble!
    Cheshire puss, perhaps?

Leave a Reply

Your email address will not be published. Required fields are marked *