Item

Abstract

For GEOMAR’s time-domain CSEM data excited with two perpendicular horizontal dipole polarizations (Figure 1, Hölz et al., 2015), the use of rotational invariants is advantageous because it can provide representations of measured horizontal field components which are independent of the orientations of both the transmitters (TXs) and receivers (RXs, Figure 2). The use of rotational invariants was successfully applied for the 1D interpretation of timedomain CSEM data in Hölz et al. (2015) and Swidinsky et al. (2015). For performing 2D CSEM inversion in frequency domain, the GEOMAR’s time-domain data is first transformed into frequency-domain. It is worth mentioning that the acquired data were first processed to yield transient E fields for each RX-TX pair. The CSEM data selected in a varying frequency range can be used for interpretation, which includes both the low and high frequency data. In this study, we have developed a frequency-domain 2D inversion code using rotational invariants, which could be used for inverting the GEOMAR‘s CSEM data. The forward solver for 2D inversion is based on a staggered finite-difference code for simulating frequencydomain CSEM responses using the total field approach (Li et al., 2017). The singular source point is distributed by a pseudo delta function for numerical accuracy. The rotational invariants will also be computed in the frequency domain. For inversion, the Gauss-Newton (GN) optimization is used for fast convergence and accutate resistivity image reconstruction. Synthetic tests indicate its validity. Future work will focus on the interpretation of CSEM data collected in the Black Sea area for gas hydrate exploration.