Spatially Constrained Inversion of Transient Electromagnetic Data for Near Surface Characterization
D. Rovetta1, D. Colombo1, G. McNeice1, and T. Yousuf1
Complementary geophysical measurements and their joint interpretation and inversion into a coherent velocity model can provide effective solutions for seismic exploration. Electromagnetic methods can improve the characterization of the complex subsurface by enhancing resolution in poor signal-to-noise seismic areas and by adding sensitivity to low velocity zones and hidden layers. For land/airborne acquisition, the loop-based magnetic source transient EM technique represents a robust methodology for obtaining high resolution images of the near surface. The high sensitivity of the method to conductors allows resolving shallow velocity inversions otherwise invisible to refracted seismic energy. Acquisition of transient EM data can be performed by ground-based magnetic loops or more efficiently by using helicopters and suspended loops, which enable the acquisition of large areas with spatial sampling comparable with that of seismic. Fast and robust inversion of such data is needed to impact the seismic processing chain during the evaluation of the near surface corrections. To achieve this goal, a new algorithm for spatially constrained inversion of EM data was developed, where the 1D inversion of the observations is regularized in a 3D sense by means of spatial constraints introduced by the model covariance matrix. The inversion of each model cell is correlated to the inversion of the cells of nearby vertical profiles occurring within a radius of influence. Both the spatial extension of the influence and the percentage correlations to be applied to the model cells are explicitly defined by the user. This scheme is effective in terms of computing time and quality of results. The developed algorithm was successfully demonstrated on synthetic and real data. The application of the spatially constrained inversion to transient EM soundings acquired over a subsurface deltaic structure in Saudi Arabia provides unprecedented resolution in the imaging of the shallow near surface. The developed inversion scheme can be optimally embedded in a simultaneous joint inversion procedure with seismic data to enhance near surface velocity models.
AAPG Search and Discovery Article #90188 ©GEO-2014, 11th Middle East Geosciences Conference and Exhibition, 10-12 March 2014, Manama, Bahrain