Estimating Gravity Acceleration from Static Atomic Gravimeter by Kalman Filtering

We present the construction of a two-state model of the atomic gravimeter and the associated Kalman recursion to estimate gravity acceleration from an atomic gravimeter. It is found that the Kalman estimator greatly improves the estimation precision in the short term by removing the white phase noise. The residual noise of the estimates follows 0.13 mu Gal/root s for more than 100 s and highlights a precision of 0.34 mu Gal at the measuring time of a single sample, even with no seismometer correction.