Introducing Astrape — Volume 1/3
In the first installment of our three-part series, we delve into the scientific rigor that underpins our software, Astrape, the world’s fastest and highest-resolution AEM inversion tool. This article highlights how Vorticity successfully applied this cutting-edge technology to a well-established benchmark dataset. Next Wednesday, we will explore the performance acceleration achieved through GPU-based systems.
The Caber Deposit: A well-studied benchmark
The Caber volcanic massive sulfide (VMS) deposit has long served as a benchmark site for airborne electromagnetic (AEM) surveys due to its known geology and complex structure. The deposit itself is thin and steeply dipping southwest beneath a conductive overburden. With traditional technology, inverting such a thin mineralized zone in 3D poses a significant challenge.
This vertically dipping conductor generates a characteristic double dip in the EM response, complicating 1D modeling and often resulting in a double feature known as “pant-legs.” These artifacts, common in 1D inversions, can lead to inaccurate borehole placement and drive up the costs of exploration. The limitations of 1D assumptions for capturing 3D structures are well documented, emphasizing the advantages of higher-dimensional inversions.
With Astrape, you have the flexibility to perform EM inversions using both 2.5D and 3D approaches. In the following sections, we present the results of applying the 3D inversion technique to model the Caber deposit. Both inversion methods are grounded in proven scientific principles, including the ‘moving-footprint’ technique for domain decomposition and the finite volume method for simulation.
3D Inversion
This method models subsurface variations across all three dimensions (x, y, and z), iteratively adjusting the model to minimize discrepancies between predicted and observed data in all directions. In a 3D inversion, source-receiver pairs from many flight lines are considered together to better identify the correct shape and density of the conductive material. The results of applying the 3D inversion to the Caber deposit, as shown below, confirm our ability to identify both the conductive overburden and the deeper copper-zinc deposit.
The deeper conductive feature exhibits higher conductivity than the overburden. One key advantage of full 3D inversion is its accuracy in modeling conductive features that align parallel to survey lines, a challenge for 2.5D inversion due to the projection transformation from 2D to 3D.
While 1D inversion offers a straightforward and computationally efficient approach, it is often inadequate for accurately modeling complex geological structures. Its assumption of horizontal layering can introduce misleading artifacts, such as the “pant-leg” effect. On the other hand, 3D inversion provides a comprehensive and precise representation of the subsurface by considering variations in all three dimensions. This method not only eliminates common 1D artifacts but also ensures that critical subsurface features are accurately resolved, leading to more reliable interpretations and better decision-making in resource exploration. For projects where geological complexity and precision are paramount, 3D inversion is crucial in offering a more accurate picture of the subsurface that leads to better outcomes.
Astrape seamlessly integrates into existing geophysics workflows, allowing subsurface inversion models to be imported into widely used geophysical visualization tools.
2.5D Inversion
Astrape also supports 2.5D inversions, where independent inversions are performed on each survey line. Unlike 1D inversion, the 2.5D approach can model topography and irregular subsurface structures, assuming a consistent structure along the strike direction over a distance greater than the AEM 3D source footprint. The independent inversions can be stacked together through interpolation or kriging to produce a full 3D model. By considering all source-receiver pairs along the flight line together, this method better constrains the shape and location of conductive resources.
Check back in next Wednesday for the second installment, where we focus on demonstrating the processing capabilities of Astrape.
More about Astrape:
References
- McMillan, M., Haber, E., & Marchant, D. (2018). Large scale 3D airborne electromagnetic inversion: Recent technical improvements. ASEG Extended Abstracts, 2018, 1. https://doi.org/10.1071/ASEG2018abT6_1F.
- Haber, E., & Schwarzbach, C. (2014). Parallel inversion of large-scale airborne time-domain electromagnetic data with multiple OcTree meshes. Inverse Problems, 30(5), 055011. https://doi.org/10.1088/0266-5611/30/5/055011.
- Cox, L. H., Wilson, G. A., & Zhdanov, M. S. (2010). 3D inversion of airborne electromagnetic data using a moving footprint. Exploration Geophysics, 41(4), 250–259. https://doi.org/10.1071/EG10003.
