BCGS Technical Talk – March 10, 2016
Speaker: Jovan Silic, Jovan Silic and Associates Pty. Ltd., Victoria, Australia
Title: Fast Inversion of AEM data using 2.5D models with 3D geologic controls: Comparisons with 1D and 3D models
Date/Time: Thursday, March 10, 2015 @ 4:30 pm
Location: 4th Floor Conference Room, Room 451, 409 Granville St. (UK Building at Granville and Hastings), Vancouver
Fast inversion of AEM data using 2.5D models with 3D geological controls: Comparisons with 1D and 3D models
Jovan Silic, Jovan Silic and Associates Pty. Ltd., Victoria, Australia.
The advantages of 2.5D (2D geology, 3D source) airborne electromagnetic inversion in 3D geological mapping applications compared to the more commonly used CDI transforms or 1D inversions are demonstrated using examples from different geological settings.2.5 D inversion sections show:
(a) greater depth resolution,
(b) complex/real body geometries, not just layers and plates ( e.g. synclines ),
(c) tend to minimize the causative body dimensions compared to other methods, due to better constraints,
(d) data from whole survey lines can be inverted within a single inversion attempt.
The technology is realized using the numerical solutions afforded by the 2D finite-element method. This enables the accurate simulation of 3D source excitation inclusive of topography, non-conforming boundaries and very high resistivity contrasts. Solution is accurate for geo-electrical cross-section which is relatively constant along a strike length that exceeds the AEM system footprint.
Completely new inversion solver with adaptive regularization algorithm allows the incorporation of a misfit to the reference model and the model smoothness function.
We allow the use of a starting or reference geology/resistivity model to influence the inversion and the software is or can be incorporated within a 3D geological modelling software with an intelligent graphical user interface.
For speed, the software has been parallelized using Intel MPI and can be used on standard computing hardware or computing clusters. Data from survey lines with lengths exceeding 30 kilometres can be inverted on laptop computers.
Numbers of examples are used to demonstrate the viability of using 2.5 D models to invert the data at a survey scale.