FD_CSEM

Software · Mon May 09 00:00:00 EDT 2005

OSTI ID:1246938

Weiss, Chester J.

FD_CSEM is a staggered-grid finite-difference forward solver for predictive modeling of low-frequency electromagnetic (EM) response of sub-seabed geology to a time-harmonic electric dipole transmitter located at some arbitrary location and orientation in the water column. The software is distributed as FORTRAN-90 source code only and should be considered vetted, albeit, research grade software intended in its present form for exploration of the physics surrounding the controlled-source electromagnetic (CSEM) experiment. Such knowledge, ultimately, is intended to assist in improvements in experiment design for resolving sub-seabed geology of commercial (e.g. hydrocarbons) or academic (e.g. magma bodies) interest. Documentation on code usage is contained within source code comments, which compose roughly half of the total lines of code. Input and output data formats of the code are documented within and the technically savvy user is encouraged to modify them as needed for a particular application. Multiple source terms (dipole antennae) can be evaluated simultaneously (through source superposition) in a single forward evaluation. Redistribution of the modified code is prohibited. However, users are encouraged to submit comments, improvements, bug reports, etc. to [e-mail], some of which may be included in future distributions. Software designed to predict the electromagnetic (EM) response of sub-seabed geology to a towed, time-harmonic electric-dipole transmitter. Military applications are not considered here. Software may be used in conjuction with imaging/inversion algorithms (not yet developed) in which the sub-sea bed geology is inferred from observations of EM fields. Staggered-grid finite-difference methodology originally described by Yee (1965) and applied to land-based geophysical exploration by Newman and Alumbaugh (1996) is used here as the algorithmic foundation of the present software. To summarize, the time-harmonic Mwrwell equations are transformed into a large, complex-linear system of equations, which is solved using the iterative, two-term coupled recurrence QMR method of Freund and Nachtigal (1996). To economize storage requirements, the matrix elements of the linear system are computed on the fly at each step of QMR iterative sequence. Improvements in solution accuracy are achieved through Gauss-quadrature weighting in the construction of system's right- hand-side. Dynamic memory allocation is employed to further economize resources during code execution. Multiple source terms (dipole antennae) can be evaluated simultaneously (through source superposition) in a single forward evaluation.

• Short Name / Acronym: FD_CSEM; 001869GENWS00
• Site Accession Number: NGOTP project 2836, LDRD project 53586
• Version: 00
• Programming Language(s): Medium: X; OS: os-x/windows/unix; Compatibility: Workstation
• Research Organization: Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: AC04-94AL85000
• OSTI ID: 1246938
• Country of Origin: United States