none 10.1190/geo2012-0246.1 Society of Exploration Geophysicists Society of Exploration Geophysicists 186 62220263 5981 2025121114475213200 10.1190 2025-12-11T19:48:01Z 2013-05-14T05:17:52Z 72 Geophysics 1942-2156 0016-8033 05 1 2013 05 01 2013 78 3 Hormoz Jahandari Memorial University of Newfoundland 1 , Department of Earth Sciences, St. John’s, Newfoundland, . E-mail: h.jahandari@mun.ca ; cgfarquh@mun.ca . Canada Colin G. Farquharson Memorial University of Newfoundland 1 , Department of Earth Sciences, St. John’s, Newfoundland, . E-mail: h.jahandari@mun.ca ; cgfarquh@mun.ca . Canada ABSTRACT Minimum-structure inversion is one of the most effective tools for the inversion of gravity data. However, the standard Gauss-Newton algorithms that are commonly used for the minimization procedure and that employ forward solvers based on analytic formulas require large memory storage for the formation and inversion of the involved matrices. An alternative to the analytical solvers are numerical ones that result in sparse matrices. This sparsity suits gradient-based minimization methods that avoid the explicit formation of the inversion matrices and that solve the system of equations using memory-efficient iterative techniques. We have developed several numerical schemes for the forward modeling of gravity data using the finite-element and finite-volume methods on unstructured grids. In the finite-volume method, a Delaunay tetrahedral grid and its dual Voronoï grid are used to find the primary solution (i.e., gravitational potential) at the centers and vertices of the tetrahedra, respectively (cell-centered and vertex-centered schemes). In the finite-element method, Delaunay tetrahedral grids are used to develop linear and quadratic finite-element schemes. Different techniques are used to recover the vertical component of gravitational acceleration from the gravitational potential. In the finite-volume scheme, a differencing method is used; in the finite-element method, basis functions are used. The capabilities of the finite-volume and finite-element schemes were tested on simple and realistic synthetic examples. The results showed that the quadratic finite-element scheme is the most accurate but also the most computationally demanding scheme. The best trade-offs between accuracy and computational resource requirement were achieved by the linear finite-element and vertex-centered finite-volume schemes. 05 1 2013 05 01 2013 05 14 2013 G69 G80 10.1190/geo2012-0246.1 https://pubs.geoscienceworld.org/geophysics/article/78/3/G69/300958/Forward-modeling-of-gravity-data-using-finite https://pubs.geoscienceworld.org/seg/geophysics/article-pdf/78/3/G69/3230820/geo2012-0246.pdf https://pubs.geoscienceworld.org/seg/geophysics/article-pdf/78/3/G69/3230820/geo2012-0246.pdf http://geophysics.geoscienceworld.org/cgi/doi/10.1190/geo2012-0246.1 0016-8033 Geophysics Barbosa 62 1997 10.1190/1.1444275 Gravity inversion of basement relief using approximate equality constraints on depths Potential theory in gravity and magnetic applications Blakely 1996 0956-540X Geophysical Journal International Cai 162 2005 10.1111/j.1365-246X.2005.02711.x Fast finite-element calculation of gravity anomaly in complex geological regions 0926-9851 Journal of Applied Geophysics Chasseriau 52 2003 10.1016/S0926-9851(02)00240-9 3D gravity inversion using a model of parameter covariance Handbook of numerical analysis Eymard 2000 Finite volume methods 0016-8033 Geophysics Farquharson 73 1 2008 10.1190/1.2816650 Constructing piecewise-constant models in multidimensional minimum-structure inversions 0926-9851 Journal of Applied Geophysics Farquharson 68 2009 10.1016/j.jappgeo.2009.03.007 Three-dimensional modelling of gravity data using finite differences 0949-7714 Journal of Geodesy Fašková 84 2010 10.1007/s00190-009-0349-7 Finite element method for solving geodetic boundary value problems Numerical computation of internal and external flows: The fundamentals of computational fluid dynamics Hirsch 2nd ed. 2007 The finite element method in electromagnetics Jin 2nd ed. 2002 Foundations of potential theory Kellogg 1967 10.1007/978-3-642-86748-4 0016-8033 Geophysics Li 61 1996 10.1190/1.1443968 3-D inversion of magnetic data 0016-8033 Geophysics Li 63 1998 10.1190/1.1444302 3-D inversion of gravity data 0956-540X Geophysical Journal International Li 152 2003 10.1046/j.1365-246X.2003.01766.x Fast inversion of large-scale magnetic data using wavelet transforms and a logarithmic barrier method 0956-540X Geophysical Journal International Mackie 115 1993 10.1111/j.1365-246X.1993.tb05600.x Three-dimensional magnetotelluric inversion using conjugate gradients 0956-540X Geophysical Journal International May 187 2011 10.1111/j.1365-246X.2011.05167.x Optimal, scalable forward models for computing gravity anomalies 0016-8033 Geophysics Okabe 44 1979 10.1190/1.1440973 Analytical expressions for gravity anomalies due to homogeneous polyhedral bodies and translations into magnetic anomalies 0016-8033 Geophysics Pilkington 74 1 2009 10.1190/1.3026538 3D magnetic data-space inversion with sparseness constraints 0016-8033 Geophysics Rodi 66 2001 10.1190/1.1444893 Nonlinear conjugate gradients algorithm for 2-D magnetotelluric inversion Saad 1990 SPARSKIT: A basic tool kit for sparse matrix computations Si 2004 TetGen, a quality tetrahedral mesh generator and three-dimensional Delaunay triangulator, vol. 3 0044-2275 Journal of Applied Mathematics and Physics Waldvogel 27 1976 10.1007/BF01595137 The Newtonian potential of a homogeneous cube 0016-8033 Geophysics Zhang 69 2004 10.1190/1.1778235 Three-dimensional crustal structure of central Taiwan from gravity inversion with a parallel genetic algorithm The finite element method, vol. 1: The basis Zienkiewicz 5th ed. 2000