GO_3D_OBS: A benchmark crustal-scale geomodel

for seismic imaging methods assessment and next generation 3D surveys design

The GO_3D_OBS geomodel has been designed by Andrzej Górszczyk (Institute of Geophysics, Polish Academy of Sciences Department of the Geophysical Imaging) to assess seismic imaging methods for crustal-scale exploration mainly from sparse ocean-bottom seismometer (OBS) survey. This highly-realistic geomodel is inspired by the geology of the eastern-Nankai through subduction zone (Tokai area), offshore Japan, where the first OBS survey for FWI applications was carried out in 2001 by JAMSTEC. The most recent FWI results obtained with this dataset are presented in

• A. Gorszczyk, S. Operto, and M. Malinowski. Towards a robust workflow for deep crustal imaging by FWI of OBS data: the eastern Nankai trough revisited. Journal of Geophysical Research, https://doi.org/10.1002/2016JB013891, 122(6), pages 4601--4630, 2017.
• A. Gorszczyk, S. Operto, L. Schenini and Y. Yamada, Crustal-scale depth imaging via joint FWI of OBS data and PSDM of MCS data: a case study from the eastern Nankai Trough, Solid Earth, https://doi.org/10.5194/se-2019-33, 10, 765--784, 2019.

The GO_3D_OBS geomodel is distributed on a Cartesian regular grid with a 25m grid interval. The physical dimensions of the model are 175 km x 130 km x 30 km. It is parametrized by Vp, Vs, rho, Qp and Qs, hence it is an isotropic visco-elastic model. The Vs and rho models have been mainly inferred from the Vp model using the empirical relationships of Brocher (2005).

More details can be found in

• Górszczyk, A. and Operto, S.: GO_3D_OBS – The Nankai Trough-inspired benchmark geomodel for seismic imaging methods assessment and next generation 3D surveys design (version 1.0), Geosci. Model Dev, https://gmd.copernicus.org/articles/14/1773/2021/, 2021.

The geomodel can be downloaded from the official site here

For Geoazur users, download the model here

For assessment of 2D methods, we suggest to use the section 1001 as a reference one (y=25km).

• TARGET OF THE ISOTROPIC MODEL USED IN THE FOLLOWING PAPERS AS WELL AS THE FINITE-DIFFERENCE MATRIX AND THE 130 SPARSE RIGHT-HAND SIDES.

1. P.-H. Tournier, P. Jolivet, V. Dolean, H. Aghamiry, S. Operto and S. Riffo, Three-dimensional finite-difference \& finite-element frequency-domain wave simulation with multi-level optimized additive Schwarz domain-decomposition preconditioner: A tool for FWI of sparse node datasets, Geophysics, Submitted.
2. H. Aghamiry, A. Gholami, L. Combe and S. Operto, Accurate 3D frequency-domain seismic wave modeling with the wavelength-adaptive 27-point finite-difference stencil: a tool for full waveform inversion, Geophysicsl, Submitted.

VELOCITY GRID (WITHOUT PMLs) (233Mb) (.bin)

Size of the grid: n1=284, n2=1021, n3=201; Grid interval: d1=d2=d3=100 m.

In Fortran,

REAL(KIND=4) :: v(n1,n2,n3)

The first (fast) dimension is depth, the second dimension is inline direction, the third dimension is crossline direction.

The FDFD solution is verified against the solution computed with the Convergent Series Method (CBS) .

Reference: @article{Osnabrugge_2016_CBS,
title={A convergent Born series for solving the inhomogeneous Helmholtz equation in arbitrarily large media},
author={Osnabrugge, Gerwin and Leedumrongwatthanakun, Saroch and Vellekoop, Ivo M},
journal={Journal of computational physics},
volume={322},
pages={113--124},
year={2016},
publisher={Elsevier}
}

The CBS wavefield (real and imaginary part) can be downloaded here.

CBS wavefield: real part (bin) imaginary part (.bin)

IMPEDANCE MATRIX IN COO FORMAT

mata.bin, irna.bin, jcna.bin

file mata.bin: Complex-valued coefficients of the matrix.

files irna.bin, jcna.bin: row and column indexes of the non zero coefficients of the impedance matrix.

COMPLEX(KIND=SP):: mata(NNZ) (15Gb) (.bin)

INTEGER(KIND=4) :: (7.5Gb each) irna(NNZ) (.bin) , jcna(NNZ) (.bin)

NNZ=1811931418

Note that the matrix was built with constant density = 1 and constant Q=1000

SPARSE RIGHT-HAND SIDES (seabed node acquisition)

rhs_sparse.bin (.bin) , irhs_sparse.bin (.bin) , irhs_ptr.bin (.bin)

COMPLEX(KIND=SP):: rhs_sparse(NNZ_RHS)

INTEGER(KIND=4) :: irhs_sparse(NNZ_rhs), irhs_ptr(NRHS+1)

File rhs_sparse.bin: non zero coefficients of the 130 RHS vectors.

File irhs_sparse.bin: row index of the 130 RHS vectors.

File irhs_ptr.bin: index in a vector of size NRHS+1 of the first entry of each RHS

NNZ_RHS=94770

NRHS=130 Number of RHS. Roughly on the sea bed.

Contact:

Dr Andrzej Górszczyk, Assistant Professor
Institute of Geophysics, Polish Academy of Sciences
Department of the Geophysical Imaging
Ks. Janusza 64, 01-452 Warsaw
phone: +48 22 6915 838, mobile: (+48) 517 586 314