3D inversion & visualization of DC data acquired at dense borehole locations

TY - ABST

T1 - 3D inversion & visualization of DC data acquired at dense borehole locations

AU - Domenzain, Diego

AU - Liu, Lichao

AU - Kühl, Anders Kristian

AU - Christiansen, Anders Vest

AU - Yelamos Vela, Ivan

PY - 2022/8/15

Y1 - 2022/8/15

N2 - We present a 3D inversion algorithm that solves for subsurface DC conductivity over a large computational domain, and using large DC voltage datasets. Our algorithm is efficient in computation time and memory, which enables our method to be used for monitoring sites with densely sampled electrode-borehole locations. We achieve efficiency by exploiting matrix-vector multiplication of sparse matrices. Hence, we never explicitly store or approximate the Jacobian of the data. Moreover, our algorithm is independent of the discretization method used in the forward model. In order to visualize apparent resistivities of borehole DC data, we present a physics-based method for obtaining pseudo-locations in 3D space. We present field data results within the scope of the project Guided Injection Remediation Monitoring (GIRem), acquired with an in-house DCIP measuring instrument called Adapt.

AB - We present a 3D inversion algorithm that solves for subsurface DC conductivity over a large computational domain, and using large DC voltage datasets. Our algorithm is efficient in computation time and memory, which enables our method to be used for monitoring sites with densely sampled electrode-borehole locations. We achieve efficiency by exploiting matrix-vector multiplication of sparse matrices. Hence, we never explicitly store or approximate the Jacobian of the data. Moreover, our algorithm is independent of the discretization method used in the forward model. In order to visualize apparent resistivities of borehole DC data, we present a physics-based method for obtaining pseudo-locations in 3D space. We present field data results within the scope of the project Guided Injection Remediation Monitoring (GIRem), acquired with an in-house DCIP measuring instrument called Adapt.

UR - http://www.scopus.com/inward/record.url?scp=85146670309&partnerID=8YFLogxK

U2 - 10.1190/image2022-3742200.1

DO - 10.1190/image2022-3742200.1

M3 - Conference abstract for conference

SP - 2025

EP - 2029

ER -