Seismic imaging with traveltime and waveform inversion: Naga Thrust and Fold Belt, India
Zelt, Colin A.
Doctor of Philosophy
I demonstrate the utility of traveltime and waveform inversion in depth imaging of seismic data with the help of two 2-D multichannel seismic lines, Reg-07 and PO-03, shot perpendicular to the trend of the Naga Thrust and Fold Belt (NTFB), India. I use both lines to demonstrate that a) a velocity model from traveltime inversion is suitable for pre-stack depth migration (PSDM) and can effectively serve as a starting model for frequency-domain full-waveform inversion; b) both PSDM and waveform inversion yield interpretable depth images but a combination of the two best describes the geology; and c) there still exists untested and unexploited hydrocarbon potential in the NTFB. Using the Reg-07 data I establish a working methodology for combining traveltime inversion with waveform inversion and PSDM. The results are validated using a nearby well. Using the PO-03 data I first demonstrate the advantages of combining traveltime inversion with PSDM for depth imaging; the combined method is referred to as unified imaging. Unified Imaging simultaneously yields a velocity model and a depth image that are consistent with each other. The velocity model from unified imaging is validated using an in-line well. The depth image from unified imaging reveals the presence of a triangle zone along PO-03 that was previously unknown and could be promising for exploration. Waveform inversion using the velocity model from unified imaging reveals the presence of a complex conjugate fault system in the supra-thrust along PO-03 which was also previously unknown and could also be promising for exploration. Traveltime inversion estimates a velocity model that is representative of the large scale features of the subsurface. This makes traveltime inversion a necessary first step regardless of the final choice of imaging--PSDM or waveform inversion. While PSDM uses the traveltime model as a whole for estimating a reflectivity image, waveform inversion enhances the resolution of a part of the traveltime model to yield a detailed acoustic property map. Even though the data and model requirements for PSDM are less stringent than for waveform inversion, waveform inversion appears to be resolving structural features that are imaged inadequately by PSDM.