Practical 3D inversion of large airborne time domain electromagnetic data sets

Yang, Dikun; Oldenburg, Douglas W.

doi:10.1071/aseg2012ab173

Cited by 7 publications

(2 citation statements)

References 3 publications

(3 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When handling a large‐scale inversion, gravity data can be clipped into several base‐scale gravity data of the same size for the prediction stage, and then the inversion results can be combined (Phillips et al., 2010; Yang & Oldenburg, 2012). Therefore, in theory, even if the inversion task increases, the time cost of the DL inversion method barely increases.…”

Section: Inverse Analysismentioning

confidence: 99%

Deep Learning 3D Sparse Inversion of Gravity Data

Huang

Liu

et al. 2021

JGR Solid Earth

View full text Add to dashboard Cite

Gravity measurements have a wide range of applications in the detection of natural resources, mapping bedrock topography, and microgravity research (Nabighian et al., 2005). From a geophysical perspective, the main outcome when interpreting gravity data is the determination of the geometric and physical characteristics of rock bodies.It is generally assumed that the subsurface reconstruction that is obtained from observed data is composed of a mass of discrete rectangular blocks with contrasting densities (Boulanger & Chouteau, 2001). Further, due to noise in field data, the application of inverting gravity data is limited, and gravity sources are inherently non-unique. In recent years, several approaches have been proposed to obtain a reasonable inversion solution, most of which have been employed for sparse inversion (Liu et al., 2018). Indeed, researchers expect to develop a sparse solution, Here, most elements of density vectors become zero, and a few remain large.Machine learning (ML) is a branch of artificial intelligence that has produced good results when applied to prediction systems, fraud alerts, image recognition, and spam filters (Bobadilla et al., 2013;Ravisankar et al., 2011). The main reason for this success is that it allows computer systems to "learn" by driving a large number of data pairs for training. In the field of geophysics, the application of nonlinear techniques in ML can be traced back to the 1990s. Specifically, Guan et al. (1998) combined a neural network with the inverted theory of gravity and magnetic anomalies, thereby developing a pseudo back propagation (BP) neural network method that can invert gravity and magnetic data. Although this method does not have a sample set to update the network weights, it loses the ability of self-learning. Guo et al. ( 2012) investigated a 3D gravity inversion method with a large number of parameters based on a BP neural network. Compared with the BP neural network, the radial basis function (RBF) provides the additional advantages of nonlinear mapping under constraint conditions and good generalization capabilities. On this basis, Geng and Yang (2013)

show abstract

Section: Inverse Analysismentioning

confidence: 99%

Deep Learning 3D Sparse Inversion of Gravity Data

Huang

Liu

et al. 2021

JGR Solid Earth

View full text Add to dashboard Cite

show abstract

“…We again quote Macnae: "While the formal inversion process is practical for targets already identified as being of interest, it (3D inversion) is unlikely to be routine in the near future as a means of processing complete surveys." More recently, Yang and Oldenburg (2012) have suggested a workflow of randomly sampled data forming a grid of multiple subdomains ("tiles") that are independently inverted, and then postinversion stitching of those multiple "tiles" into a single 3D model. However, this approach still has an inherent problem of how to best stitch different "tiles" together.…”

Section: From Mega-cell To Giga-cell Inversionmentioning

confidence: 99%

Inverting airborne geophysical data for mega-cell and giga-cell 3D Earth models

et al. 2012

View full text Add to dashboard Cite

Today's mineral exploration is driven by the simple fact that discovery rates have not kept pace with the depletion of existing reserves. To improve discovery rates, there is an industry-wide consensus on the need to increase the “discovery space” by exploring under cover and to greater depths. This attracts increased risks which may be mitigated by improved targeting. To do this, mining geophysics needs to shift toward 3D geological models founded upon improved petrophysical understanding and geophysical inversion. Regardless of the inversion methodology used, all geological constraints manifest themselves in the user's prejudice of an a priori model, upper and lower bounds, and choice of regularization. However, the practice of geologically constrained inversion is not the major problem needing to be addressed. It is known (and accepted) that geology is inherently 3D, and is a result of complex, overlapping processes related to genesis, metamorphism, deformation, alteration and/or weathering. Yet, the mining geophysics community to date has not fully accepted that geophysics should also be 3D, and most often relies on qualitative analysis, 1D inversion, and deposit-scale 2D or 3D inversion. There are many reasons for this unfortunate deficiency, not the least of which has been the lack of capacity of existing 3D inversion algorithms. To date, these have not been able to invert entire surveys with sufficient resolution in sufficient time to practically affect exploration decisions.

show abstract

Massively parallel regularized 3D inversion of potential fields on CPUs and GPUs

C̆uma

Zhdanov

2014

Computers & Geosciences

View full text Add to dashboard Cite

Practical 3D inversion of large airborne time domain electromagnetic data sets

Cited by 7 publications

References 3 publications

Deep Learning 3D Sparse Inversion of Gravity Data

Deep Learning 3D Sparse Inversion of Gravity Data

Inverting airborne geophysical data for mega-cell and giga-cell 3D Earth models

Massively parallel regularized 3D inversion of potential fields on CPUs and GPUs

Contact Info

Product

Resources

About