Seismic inversion using low‐frequency seismic impedance trend computed from CSEM data

Abstract

Summary Seismic inversion is highly dependent on the initial P-wave impedance model. We generally use filtered well log data combined with moveout velocities to constraint this model, but generally at shallow depths no well log data is available. Also, conventional seismic reflection data has an information gap in the ~2-10 Hz frequency band, and does not contain information about the background impedance trend. On the other hand, controlled-source electromagnetic (CSEM) data have low resolution and contain information in this gap about the earth’s resistivity structure. Rockphysics relations between resistivity and porosity, and acoustic impedance and porosity can be used to extract information from the CSEM data to provide constraints on the low-frequency and the shallow seismic impedance section. Porosity is the main link that allows us to develop and use these cross-property relations. We used Archie’s equation and soft-sand model to estimate the low-frequency seismic impedance trend from CSEM data. We combined this trend with well log data information to estimate an initial model for seismic inversion. Results show that appropriate cross-property relations can capture the background low-frequency impedance trend, particularly at shallow depths where no well information is available, and can provide useful constraints on prior models for inversion. Validation test showed that residuals are lower, and the fit is better with the background model contrained by the impedance trend obtained from CSEM data.