Bayesian geological and geophysical data fusion for the construction and uncertainty quantification of 3D geological models

Hugo K.H. Olierook, Richard Scalzo, David Kohn, Rohitash Chandra, Ehsan Farahbakhsh, Chris Clark, Steven M. Reddy, R. Dietmar Müller
2020 Geoscience Frontiers  
Traditional approaches to develop 3D geological models employ a mix of quantitative and 14 qualitative scientific techniques, which do not fully provide quantification of uncertainty in the 15 constructed models and fail to optimally weight geological field observations against constraints from 16 geophysical data. Here, using the Bayesian Obsidian software package, we develop a methodology to 17 fuse lithostratigraphic field observations with aeromagnetic and gravity data to build a 3D model
more » ... build a 3D model in a 18 small (13.5 km × 13.5 km) region of the Gascoyne Province, Western Australia. Our approach is 19 validated by comparing 3D model results to independently-constrained geological maps and cross-20 sections produced by the Geological Survey of Western Australia. By fusing geological field data 21 with aeromagnetic and gravity surveys, we show that 89% of the modelled region has >95% certainty 22 for a particular geological unit for the given model and data. The boundaries between geological units 23 are characterized by narrow regions with <95% certainty, which are typically 400-1000 m wide at the 24 Earth's surface and 500-2000 m wide at depth. Beyond ~4 km depth, the model requires geophysical 25 survey data with longer wavelengths (e.g., active seismic) to constrain the deeper subsurface. 26 Although Obsidian was originally built for sedimentary basin application, there is reasonable 27 applicability to deformed terranes such as the Gascoyne Province. Ultimately, modification of the 28 Bayesian engine to incorporate structural data will aid in developing more robust 3D models. 29 Nevertheless, our results show that surface geological observations fused with geophysical survey 30 data can yield reasonable 3D geological models with narrow uncertainty regions at the surface and 31 shallow subsurface, which will be especially valuable for mineral exploration and the development of 32 3D geological models under cover. 33 34 3 65 framework for doing this by using Markov chain Monte Carlo (MCMC) sampling methods for 66 estimation and uncertainty quantification of free parameters. Previous studies using Bayesian 67
doi:10.1016/j.gsf.2020.04.015 fatcat:ekwqodeyxvgp3jyggiqf2bv7zu