Two-dimensional Kalman filter approach to airborne vector gravimetry

Abstract The paper presents a new approach to the airborne vector gravimetry problem. The idea of the approach is to take into account spatial correlation of the gravity field to improve observability of horizontal components of the gravity disturbance vector (GDV). We consider the GDV determination problem given airborne data at a set of parallel survey lines assuming that lines are flown in the same direction at a constant height above the reference ellipsoid. We use a 2-D random field model for the gravity field at the flight height. The random field is governed by two autoregressive equations (one in the direction along the lines, the other across the lines). Then we pose the estimation problem simultaneously for the GDV horizontal components and systematic errors of an inertial navigation system at all the lines simultaneously. The developed estimation algorithm is based on 2D Kalman filtering and smoothing techniques. Numerical results obtained from simulated data processing showed improved accuracy of the gravity horizontal component determination.