Research Of The Theory Of Joint Inversion Of Gravity And Magnetic Data Based On Cross-gradient Function

Potential field exploration is the first method that be applied to geophysicalprospecting. During the past half a century, this method has been developed greatlyand reached a mature state relatively. In these decades, potential field method hasbeen widely used in regional tectonic study, oil-gas exploration and solid mineralprospecting. During the whole process of potential field method, data inversion is thekey issue. So, the inversion technologies of potential field data developed greatly inthe past decades. And all of these technologies played an important role in differentareas.In this paper, we studied the inversion theory of potential field data in great detail.And we accomplished these algorithm with computer programs based on the study ofinversion theories. The forward modeling of gravity and magnetic data is adoptedbased on uniform grid subdivision of Cartesian coordinate system. The forwardprocessor is done by the calculation of discrete integral formulas. The inversionequations are solved using conjugate gradient method. Meanwhile, regularization isincorporated in the objective functions. Regularization skills can stable the solving ofobjective equations. Furthermore, different constraints are also incorporated such asdepth-weighting, bounds constraint, compress constrain and so on. Varies constraintscan help to improve inversion results.On the basis of studying the single inversion of gravity data and magnetic data,we utilized the structural coupling to build up the relationship between gravity andmagnetic data. In this way, we can accomplish the joint inversion of gravity data andmagnetic data. Inversion of the theoretical model data demonstrate that our inversionalgorithm with cross-gradient constraints can utilize different data to improve theinversion results. The results of joint inversion show much enhancement comparedwith single data inversion.The research mainly contained the following aspects:(1) The forward modelling of gravity data and magnetic data; (2) Single inversion of gravity data and magnetic data with regularizations;(3) How to incorporate different constrains in the inversion objective functions;(4) Set up the joint inversion of different data based on cross gradient function.