3. EM/GPR Imaging Technologies
SIP Parameter Estimation From the Multi-frequency Data.
Jeong-sul Son, Jung-Ho Kim and Myeong-Jong Yi
Korea Institute of Geoscience & Mineral Resources, Korea.
Abstract
The IP (induced polarization) method has ability of providing valuable information about the structural and hydraulic nature of porous media and fluids contained in the media. Until these days, most of IP interpretation techniques use the chargeability property only. In principally, IP method can give more valuable information about subsurface structures and medium. It provides the improved lithological characterization of subsurface if complex resistivity (conductivity) approach is used, since surface polarization effects can be separated from electrolytic and surface conduction effects. In this study, we develop IP inversion technique using complex resistivity and expand it to the multi-frequency data set, and then extract the SIP (spectral IP) parameter from them. To invert the multi-frequency data simultaneously, we propose the special object function which minimizes the difference of each updating model vector. With the proposed object function, we may enhance the noise characteristics of whole data set and get more accurate estimation of complex conductivity. When the multi-frequency inversion algorithm was applied to the synthetic data set with difference noise levels, it reduced the unwanted structure that came from the noise. To extract the hydraulic parameter from the multi-frequency inversion result, we use the global optimization technique such as SA (simulated annealing method). Our proposed method will give more reasonable estimation of SIP parameter.
Last modified: Fri May 12 19:35:36 2006