Surface wave inversion in complex systems

Dispersion curves estimated from field data can be inverted to supply near surface velocity models. Since the problem is hill posed, strongly non linear and mix-determined, the inverse problem suffers from strong solution non uniqueness particularly for complex velocity models. Global search methods explore the solution space and supply a “picture” of the solution non uniqueness enabling a proper model parameterization to be used for linearised inversion. A possibility for reducing the broadness of the possible equivalent solutions is the introduction of constraints to the solution. Constraints are traditionally obtained from other geophysical tests or boreholes, but significant improvement in the inversion results can be obtained also introducing the higher modes of propagation or inverting simultaneously the dispersion curves relative to different locations in the case of complex 2D/3D systems.