We describe an effective method for joining the benefits of inversion of different kinds of measurements. We show the simultaneous joint inversion objective function, which allows users to link different inversion domains, like seismic with gravity or seismic with magnetotellurics. This function can be extended to any number of domains and does not require that they be sampled on the same grid. Our parallel implementation allowed us to scale well with large volumes of data and a large number of unknowns, and it has already been included in production workflows. Furthermore, it is generic and constitutes a framework where new inversion techniques can be easily plugged in. We also present two different ways of linking various inversion domains for establishing relationships between different model domains and how they can be chosen and used to achieve the best result. Applications of the algorithm using synthetic and field data produce model features generally not achievable with single-domain inversions. Specifically, we applied our technique to real data from the Walker Ridge area in the Gulf of Mexico, and we used the results to reinterpret and remigrate seismic data. The final migrated section clearly found improved quality with respect to previous efforts. Our results document the fundamental importance of integrating nonseismic methods with seismic techniques to increase the image quality of geologically complex areas.
Cost-effective Seismic Data Reprocessing for Subsalt Imaging Enhancement