# How do you calibrate the accelerometers of an IMU in space, outside the GPS range?

IMU measurements are corrupted by drifts and errors, which accumulate over time. In space, gyroscopes can be calibrated using e.g., a star tracker as an absolute reference. GPS can serve as an external reference for the accelerometers, but only reliably beneath the network of GPS satellites. So, how does one calibrate the accelerometers outside the GPS range, for instance in orbit around the Moon?

• These are not duplicates to your question, but you may fin the answers posted to each one interesting. How can a deep-space spacecraft determine in real time the direction of delta-v? as well as Accelerometer in space and also Autonomous Navigation for deep space missions
– uhoh
Commented Oct 11, 2018 at 8:19
• What should the accelerometer measure in orbit around the Moon? The spaceship is in zero gravity there as long as no thrusters are used. So drift errors may be calibrated, the signal should be zero, if not adjust compensation.
– Uwe
Commented Oct 11, 2018 at 11:44
• Indeed, the accelerometers should measure the impact of activating the thrusters. But can't they also be used to detect perturbations (3rd body, radiation pressure)? I gather that when the spacecraft is in circular orbit the accelerometers prompt zero, but when the spacecraft is in elliptic orbit and accelerates towards its periapsis, don't the accelerometers notice this? Commented Oct 11, 2018 at 12:12
• @Wouter:When the spacecraft is in elliptic orbit, is it in zero gravity like in a circular orbit? A circular orbit is just a special case of elliptic orbit. Eccentricity may be big, small, very small or zero. Why should zero gravity exist only for the very rare case of a perfect circle?
– Uwe
Commented Oct 11, 2018 at 12:44
• @Wouter, to further clarify, "gravity" as we experience on earth is thrust upwards from the ground resisting our body's acceleration downward. We are constantly trying to accelerate at 9.8m/s/s downward, so the ground has to push on us at 9.8m/s/s upwards. We are at 1G. In orbit, there's nothing pushing us from the "pure" effect of gravity alone, so even if we're in a (admittedly impossible) highly complex 42-body orbit with thousands of fly-bys, we'll still be in free fall and will always be in a null-G frame of reference. Commented Oct 11, 2018 at 17:37