This airborne electromagnetic (AEM) dataset provides regional scale probabilistic inversion products from 141,000 line km of airborne electromagnetic (AEM) data from the AusAEM program. The two main benefits of a probabilistic inversion over a deterministic one are: - Loss of signal sensitivity at depth does not “fade to blue” by returning to the resistive deterministic reference model. - Ambiguous subsurface features become clearer when examining multiple probability percentiles. Further details are provided in the accompanying technical note. Conductivity products from the following surveys are available: - Frome 2011 - AusAEM 1 NT (2017) - AusAEM 1 QLD (2017) - AusAEM 2 Tranche 1 part (2019) - AusAEM 3 Eastern Resources Corridor (all 3 phases) - AusAEM 3 Western Resources Corridor (Kimberley, Central, Musgraves and South) The 10th, 50th and 90th and mean percentiles of log10 conductivity are provided in a variety of formats: - VTK structured grids - ASCII point clouds - ASEG-GDF2 files - GOCAD S-grids All products have been provided in the coordinate reference system (CRS) the original AEM data were provided in. The VTK unstructured grids are also provided in GDA94 geodetic longitude, latitude coordinates for ease of display in the same CRS.

<div> The High Quality Geophysical Analysis (HiQGA) package is a framework for geophysical forward modelling, Bayesian inference, and deterministic imaging. A primary focus of the code is production inversion of airborne electromagnetic (AEM) data from a variety of acquisition systems. Adding custom AEM systems is simple using a modern computational idea known as multiple dispatch. For probabilistic spatial inference from geophysical data, only a misfit function needs to be supplied to the inference engine. For deterministic inversion, a linearisation of the forward operator (i.e., Jacobian) is also required. For fixed wing geometry nuisances, probabilistic inversion is carried out using Hierarchical Bayesian inference, and deterministic inversion for these nuisances is done using BFGS optimisation. The code is natively parallel, and inversions from a full day of production AEM acquisition can be inverted on thousands of CPUs within a few hours. This allows for quick assessment of the quality of the acquisition, and provides geological interpreters preliminary subsurface images together with associated uncertainties. These images are then used to create subsurface models for a range of applications from natural resource exploration to its management and conservation.</div><div> </div> This abstract was submitted to/presented at the 8th International Airborne Electromagnetics Workshop (AEM 2023) (https://www.aseg.org.au/news/aem-2023).

<div>Much of Australia has been surveyed with low-flying airborne electromagnetic (AEM) instrumentation under Geoscience Australia’s AusAEM program. Acquired AEM data allow for imaging the earth's buried geology down to depths of 300-500 m. Such imaging is crucial for managing Australia’s subsurface minerals, energy and groundwater resources, by allowing geoscientists to build a 3D framework of the shallow geological architecture. However, individual AEM lines can be up to 500 km long, data are acquired every 10-12 m, and conventional electromagnetic conductivity imaging methods based on optimisation are unable to accurately characterise the subsurface imaging resolution. Bayesian probabilistic methods can do so, but at significant computational cost if naively used. Efficient Markov chain sampling strategies with parameter dimension reduction, which leverage the high-performance distributed computing capabilities inherent in the Julia language, have now made large scale Bayesian AEM imaging possible. In this work we show the results of imaging using the Julia-based, open-source, High Quality Geophysical Analysis (HiQGA) package, on continent-wide data using Bayesian probabilistic methods. We are unaware of any similar analysis at this scale, routinely using 41,600 cpu-cores for up to three hours in semi-embarrassingly parallel fashion on the National Computational Infrastructure’s Gadi cluster at the Australian National University. Consequently, deeper geology can be mapped, and subsurface 3D geology can be rapidly demarcated using posterior percentiles of conductivity, when contrasted with deterministic methods. Compared to the cost of AEM acquisition, extraction of subsurface information with computation at scale greatly increases the economic and social return on public AEM data acquisition. Abstract presented at the 2024 Supercomputing Asia Conference, Sydney NSW (SAC2024)