<p>Airborne electromagnetic (AEM) data can be acquired cost-effectively, safely and efficiently over large swathes of land. Inversion of these data for subsurface electrical conductivity provides a regional geological framework for water resources management and minerals exploration down to depths of ~200 m, depending on the geology. However, for legacy reasons, it is not uncommon for multiple deterministic inversion models to exist, with differing details in the subsurface conductivity structure. This multiplicity presents a non-trivial problem for interpreters who wish to make geological sense of these models. In this article, we outline a Bayesian approach, in which various spatial locations were inverted in a probabilistic manner. The resulting probability of conductivity with depth was examined in conjunction with multiple existing deterministic inversion results. The deterministic inversion result that most closely followed the high-credibility regions of the Bayesian posterior probability was then selected for interpretation. Examining credibility with depth also allowed interpreters to examine the ability of the AEM data to resolve the subsurface conductivity structure and base geological interpretation on this knowledge of uncertainty. <p> <b>Citation:</b> Ray, A., Symington, N., Ley-Cooper, Y. and Brodie, R.C., 2020. A quantitative Bayesian approach for selecting a deterministic inversion model. In: Czarnota, K., Roach, I., Abbott, S., Haynes, M., Kositcin, N., Ray, A. and Slatter, E. (eds.) Exploring for the Future: Extended Abstracts, Geoscience Australia, Canberra, 1–4.

The Girls in STEM statement addresses Strategy 2028 impact area of ‘enabling an informed Australia’ by increasing earth science literacy and engagement while addressing issues of diversity and inclusion. The Statement articulates Geoscience Australia’s efforts to engage girls in STEM, particularly as it relates to our education program.

To understand earth processes, geoscientists infer subsurface earth properties such as electromagnetic resistivity or seismic velocity from surface observations of electromagnetic or seismic data. These properties are used to populate an earth model vector, and the spatial variation of properties across this vector sheds light on the underlying earth structure or physical phenomenon of interest, from groundwater aquifers to plate tectonics. However, to infer these properties the spatial characteristics of these properties need to be known in advance. Typically, assumptions are made about the length scales of earth properties, which are encoded a priori in a Bayesian probabilistic setting. In an optimisation setting, appeals are made to promote model simplicity together with constraints which keep models close to a preferred model. All of these approaches are valid, though they can lead to unintended features in the resulting inferred geophysical models owing to inappropriate prior assumptions, constraints or even the nature of the solution basis functions. In this work it will be shown that in order to make accurate inferences about earth properties, inferences can first be made about the underlying length scales of these properties in a very general solution basis. From a mathematical point of view, these spatial characteristics of earth properties can be conveniently thought of as “properties” of the earth properties. Thus, the same machinery used to infer earth properties can be used to infer their length scales. This can be thought of as an “infer to infer” paradigm analogous to the “learning to learn” paradigm which is now commonplace in the machine learning literature. However, it must be noted that (geophysical) inference is not the same as (machine) learning, though there are many common elements which allow for cross-pollination of useful ideas from one field to the other, as is shown here. A non-stationary trans-dimensional Gaussian Process (TDGP) is used to parameterise earth properties, and a multi-channel stationary TDGP is used to parameterise the length scales associated with the earth property in question. Using non-stationary kernels, i.e., kernels with spatially variable length scales, models with sharp discontinuities can be represented within this framework. As GPs are multi-dimensional interpolators, the same theory and computer code can be used to solve geophysical problems in 1D, 2D and 3D. This is demonstrated through a combination of 1D and 2D non-linear regression examples and a controlled source electromagnetic (CSEM) field example. The key difference between this and previous work using TDGP is generalised nested inference and the marginalisation of prior length scales for better posterior subsurface property characterisation. <b>Citation:</b> Anandaroop Ray, Bayesian inversion using nested trans-dimensional Gaussian processes, <i>Geophysical Journal International</i>, Volume 226, Issue 1, July 2021, Pages 302–326, <a href="https://doi.org/10.1093/gji/ggab114">https://doi.org/10.1093/gji/ggab114</a>

HiQGA is a general purpose software package for spatial statistical inference, geophysical forward modeling, Bayesian inference and inversion (both deterministic and probabilistic). It includes readily usable geophysical forward operators for airborne electromagnetics (AEM), controlled-source electromagnetics (CSEM) and magnetotellurics (MT). Physics-independent inversion frameworks are provided for probabilistic reversible-jump Markov chain Monte Carlo (rj-MCMC) inversions, with models parametrised by Gaussian processes (Ray and Myer, 2019), as well as deterministic inversions with an "Occam inversion" framework (Constable et al., 1987). In development software for EFTF since 2020