A fast algorithm for the large-scale joint inversion of gravity and magnetic data is developed. The algorithm uses a non-linear Gramian constraint to impose correlation between the density and susceptibility of the reconstructed models. The global objective function is formulated in the space of the weighted parameters, but the Gramian constraint is implemented in the original space, and the non-linear constraint is imposed using two separate Lagrange parameters, one for each model domain. It is significant that this combined approach, using the two spaces provides more similarity between the reconstructed models. Moreover, it is shown theoretically that the gradient for the use of the unweighted space is not a scalar multiple of that used for the weighted space, and hence cannot be accounted for by adjusting the Lagrange parameters. It is assumed that the measured data are obtained on a uniform grid and that a consistent regular discretization of the volume domain is imposed. Then, the sensitivity matrices exhibit a block-Toeplitz-Toeplitz-block structure for each depth layer of the model domain, and both forward and transpose operations with the matrices can be implemented efficiently using two dimensional fast Fourier transforms. This makes it feasible to solve for large scale problems with respect to both computational costs and memory demands, and to solve the non-linear problem by applying iterative methods that rely only on matrix–vector multiplications. As such, the use of the regularized reweighted conjugate gradient algorithm, in conjunction with the structure of the sensitivity matrices, leads to a fast methodology for large-scale joint inversion of geophysical data sets. Numerical simulations demonstrate that it is possible to apply a non-linear joint inversion algorithm, with Lp-norm stabilisers, for the reconstruction of large model domains on a standard laptop computer. It is demonstrated, that while the p = 1 choice provides sparse reconstructed solutions with sharp boundaries, it is also possible to use p = 2 in order to provide smooth and blurred models. The methodology is used for inverting gravity and magnetic data obtained over an area in northwest of Mesoproterozoic St Francois Terrane, southeast of Missouri, USA.
- Award ID(s):
- 1720472
- NSF-PAR ID:
- 10309730
- Date Published:
- Journal Name:
- GEOPHYSICS
- Volume:
- 86
- Issue:
- 3
- ISSN:
- 0016-8033
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
SUMMARY -
We exploit the different but complementary data sensitivities of ground-penetrating radar (GPR) and electrical resistivity (ER) by applying a multiphysics, multiparameter, simultaneous 2.5D joint inversion without invoking petrophysical relationships. Our method joins full-waveform inversion (FWI) GPR with adjoint derived ER sensitivities on the same computational domain. We incorporate a stable source estimation routine into the FWI-GPR. We apply our method in a controlled alluvial aquifer using only surface-acquired data. The site exhibits a shallow groundwater boundary and unconsolidated heterogeneous alluvial deposits. We compare our recovered parameters to individual FWI-GPR and ER results, and we compare them to log measurements of capacitive conductivity and neutron-derived porosity. Our joint inversion provides a more representative depiction of subsurface structures because it incorporates multiple intrinsic parameters, and it is therefore superior to an interpretation based on log data, FWI-GPR, or ER alone.more » « less
-
We have developed an algorithm for joint inversion of full-waveform ground-penetrating radar (GPR) and electrical resistivity (ER) data. The GPR data are sensitive to electrical permittivity through reflectivity and velocity, and electrical conductivity through reflectivity and attenuation. The ER data are directly sensitive to the electrical conductivity. The two types of data are inherently linked through Maxwell’s equations, and we jointly invert them. Our results show that the two types of data work cooperatively to effectively regularize each other while honoring the physics of the geophysical methods. We first compute sensitivity updates separately for the GPR and ER data using the adjoint method, and then we sum these updates to account for both types of sensitivities. The sensitivities are added with the paradigm of letting both data types always contribute to our inversion in proportion to how well their respective objective functions are being resolved in each iteration. Our algorithm makes no assumptions of the subsurface geometry nor the structural similarities between the parameters with the caveat of needing a good initial model. We find that our joint inversion outperforms the GPR and ER separate inversions, and we determine that GPR effectively supports ER in regions of low conductivity, whereas ER supports GPR in regions with strong attenuation.more » « less
-
more » « less
Abstract A portion of water not consumed by crops during flood irrigation can flow back across the surface or through the subsurface to adjacent surface water bodies and streams as return flow. Few studies have directly addressed subsurface processes governing return flow and the importance of structural complexity on hydrologic process representation. It is challenging to measure and model these subsurface flow paths using traditional hydrologic observations. In this study, we assess the impact of subsurface structural complexity on vadose zone flow representation in a two‐dimensional transport model by varying structural complexity derived from background geophysical data. We assessed four model structures each with three soil types of homogeneous hydrologic properties, two of which were evaluated with and without an anisotropy factor. Wetting front arrival times, derived from time‐lapse electrical resistivity measurements during flood irrigation field experiments, were used to evaluate the different representations of soil profile structures. These data indicated both vertical and lateral preferential flow in the subsurface during flood irrigation. Inclusion of anisotropy in the saturated hydraulic conductivity field improved the ability to model subsurface hydrologic behavior when flow processes shifted from uniform to heterogeneous flow, as occurs with lateral subsurface return flow under flood irrigation driven by a large pressure gradient. This reduced the need for detailed spatial discretization to represent these observed subsurface flow processes. The resulting simple three‐layer model structure was better able to model both the vertical and lateral flow processes than a more complex geospatial structure, suggesting that overinterpretation of smoothed inverted profiles could lead to misrepresentation of the subsurface structure.
-
First-arrival traveltime tomography is an essential method for obtaining near-surface velocity models. The adjoint-state first-arrival traveltime tomography is appealing due to its straightforward implementation, low computational cost, and low memory consumption. Because solving the point-source isotropic eikonal equation by either ray tracers or eikonal solvers intrinsically corresponds to emanating discrete rays from the source point, the resulting traveltime gradient is singular at the source point, and we denote such a singular pattern the imprint of ray-illumination. Because the adjoint-state equation propagates traveltime residuals back to the source point according to the negative traveltime gradient, the resulting adjoint state will inherit such an imprint of ray-illumination, leading to singular gradient-descent directions when updating the velocity model in the adjoint-state traveltime tomography. To mitigate this imprint, we solve the adjoint-state equation twice but with different boundary conditions: one being taken to be regular data residuals and the other taken to be ones uniformly, so that we are able to use the latter adjoint state to normalize the regular adjoint state and we further use the normalized quantity to serve as the gradient direction to update the velocity model; we call this process ray-illumination compensation. To overcome the issue of limited aperture, we have developed a spatially varying regularization method to stabilize the new gradient direction. A synthetic example demonstrates that our method is able to mitigate the imprint of ray-illumination, remove the footprint effect near source points, and provide uniform velocity updates along raypaths. A complex example extracted from the Marmousi2 model and a migration example illustrate that the new method accurately recovers the velocity model and that an offset-dependent inversion strategy can further improve the quality of recovered velocity models.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Journal Article:
- https://doi.org/10.1190/geo2020-0373.1
