Presentation: Estimative of gravity-gradient tensor components via fast iterative equivalent-layer technique

We present an iterative cost-effective approach for computing the gravity-gradient tensor components from the gravity data (g_z-component) using a fast equivalent-layer technique. Instead of solving linear systems and matrix multiplications, the method estimates the mass distribution on the equivalent layer through an iterative inversion algorithm. The 2D mass distribution is represented by a finite discrete set of point masses located directly below each observed data. The initial approximation is iteratively updated by adding a mass correction that is proportional to the residuals. Finally, the gravity-gradient tensor components are computed by multiplying the corresponding transformation matrix by the estimated mass distribution. Advantages of this method relative to the Fourier approach are it requires neither a regular grid nor an even observation surface. Tests on synthetic and real gravity data from the Vinton salt dome, USA, show that we must remove the regional gravity data first to calculate gravity-gradient tensor components with the equivalent-layer technique.