Creating finite-difference time-domain models of commercial ground-penetrating radar antennas using Taguchi's optimization method

Craig Warren, Antonios Giannopoulos

    Research output: Contribution to journalArticlepeer-review

    125 Citations (Scopus)
    30 Downloads (Pure)

    Abstract

    Very few researchers have developed numerical models of ground-penetrating radar (GPR) that include realistic descriptions of both the antennas and the subsurface. This is essential to be able to accurately predict responses from near-surface, near-field targets. We have developed a detailed 3D finite-difference time-domain models of two commercial GPR antennas — a Geophysical Survey Systems, Inc. (GSSI) 1.5-GHz antenna and a MALÅ Geoscience 1.2-GHz antenna — using simple analyses of the geometries and the main components of the antennas. Values for unknown parameters in the antenna models (due to commercial sensitivity) were estimated by using Taguchi’s optimization method, resulting in a good match between the real and modeled crosstalk responses in free space. Validation using a series of oil-in-water emulsions to simulate the electrical properties of real materials demonstrated that it was essential to accurately model the permittivity and dispersive conductivity. When accurate descriptions of the emulsions were combined with the antenna models, the simulated responses showed very good agreement with real data. This provides confidence for use of the antenna models in more advanced studies.
    Original languageEnglish
    Pages (from-to)G37-G47
    JournalGeophysics
    Volume76
    Issue number2
    DOIs
    Publication statusPublished - 17 Mar 2011

    Keywords

    • finite difference methods
    • geophysical equipment
    • geophysical techniques
    • ground penetrating radar
    • radar antennas
    • remote sensing by radar
    • time-domain analysis

    Fingerprint

    Dive into the research topics of 'Creating finite-difference time-domain models of commercial ground-penetrating radar antennas using Taguchi's optimization method'. Together they form a unique fingerprint.

    Cite this