Integrated seismic and electromagnetic model building applied to improve subbasalt depth imaging in the Faroe-Shetland Basin

Abstract

Subbasalt imaging has gained significant interest in the last two decades, driven by the urge to better understand the geologic structures beneath volcanic layers, which can be up to several kilometers thick. This understanding is crucial for the development and risking of hydrocarbon play models in these areas. However, imaging based on the reflection seismic data alone suffers from severe amplitude transmission losses and interbed multiples in the volcanic sequence, as well as from poor definition of the subbasalt velocity structure. We have considered a sequential imaging workflow, in which the resistivity model from joint controlled-source electromagnetic and magnetotelluric data inversion was used to update the velocity model and to improve the structural definition in the migrated seismic image. The quantitative link between resistivity and velocity was derived from well data. The workflow used standard procedures for seismic velocity analysis, electromagnetic data inversion, and well analysis, and thereby allowed detail control and input based on additional geophysical knowledge and experience in each domain. Using real data sets from the Faroe-Shetland Basin, we can demonstrate that the integration of seismic and electromagnetic data significantly improved the imaging of geologic structures covered by up to several-kilometer-thick extended volcanic sequences. The improved results might alter the interpretation compared with the imaging results from seismic data alone.

Integrating Electromagnetic Data with Other Geophysical Observations for Enhanced Imaging of the Earth: A Tutorial and Review

“…The overall geometry and thickness of the basalt layer are similar in both models, but there are differences in some of the details. In particular, the model of Panzner et al (2016) predicts a significantly stronger change in basalt thickness in the vicinity of the borehole compared to the model of Heincke et al (2017) and shows deeper high-velocity structures potentially associated with prerifted basement that do not appear in the model of Heincke et al (2017). Considering the sensitivities of the different data sets (Fig.…”

“…For P-wave velocities of less than 6000 m/s, the differences are relatively subtle, while for higher velocities the relationship by Panzner et al (2016) predicts significantly higher resistivities than the relationship of Heincke et al (2017). Even though at first sight the difference might appear dramatic, an order of magnitude or more difference in predicted resistivity, for velocities greater than 6200 m/s, in practice the impact on the inversion results is not that strong.…”

“…Both studies use a combination of seismic travel times and electromagnetic data for their inversion approach. However, while Panzner et al (2016) use a combination of MT and marine controlled-source electromagnetic data (CSEM), Heincke et al (2017) only utilize magnetotelluric measurements, but also include gravity data in the inversion. Furthermore, Panzner et al (2016) use a sequential analysis of the different data sets, where inversions are performed individually, but information is exchanged between the separate inversions.…”

“…However, while Panzner et al (2016) use a combination of MT and marine controlled-source electromagnetic data (CSEM), Heincke et al (2017) only utilize magnetotelluric measurements, but also include gravity data in the inversion. Furthermore, Panzner et al (2016) use a sequential analysis of the different data sets, where inversions are performed individually, but information is exchanged between the separate inversions. These kinds of integrated workflows are often used in hydrocarbon exploration instead of full joint inversions, either because a full joint inversion is considered intractable (Um et al 2014), or because the main goal of the inversion is to generate improved models for migrating seismic reflection data (Colombo and Stefano 2007;Colombo et al 2008;De Stefano et al 2011;Colombo et al 2014;Cui et al 2015;Takam Takougang et al 2015) or full waveform inversion (Zerilli et al 2016).…”

“…Considering the sensitivities of the different data sets (Fig. 10 in Panzner et al 2016), the addition of CSEM data helps to resolve these structures. In comparison, the gravity data included by (Heincke et al 2017) do not seem to contribute much to resolving structures at depth.…”