Detecting and Modeling the Manganese ore body through micro-gravity data

In this paper, linear optimization (linear programming or LP) is implemented for 3D modeling of synthetic gravity data and real microgravity data situation in Safo mine using MATLAB programming. And all the results and pictures are given in paper.

This method is used to derive sharp images of subsurface, which show the geometry of the anomalous body in the region of interest. An L1-norm cost-function is applied in optimization to avoid obtaining smooth image through inversion. The density distribution of the subsurface is modeled with a uniform grid of cells. The model parameters, vector m, are the density contrasts of each cell that they can be obtained by minimizing the L1-norm objective function using linear programming while satisfying some inequality constraints such as an upper limit for density contrast.

The tests show that this inversion can give us a well shaped model even in the presence of substantial noise. This method can be used in exploration situation with sharp boundaries (density contrast) such as salt flank, ore bodies, dikes, etc.