Seismic waves are used to study the Earth, exploit its hydrocarbon resources, and understand its hazards. Extracting information from seismic waves about the Earth’s subsurface, however, is becoming more challenging as our questions become more complex and our demands for higher resolution increase. This dissertation introduces two new methods that use scattered waves for improving the resolution of subsurface images: natural migration of passive seismic data and convergent full-waveform inversion.
In the first part of this dissertation, I describe a method where the recorded seismic data are used to image subsurface heterogeneities like fault planes. This method, denoted as natural migration of backscattered surface waves, provides higher resolution images for near-surface faults that is complementary to surface-wave tomography images. Our proposed method differ from contemporary methods in that it does not (1) require a velocity model of the earth, (2) assumes weak scattering, or (3) have a high computational cost. This method is applied to ambient noise recorded by the US-Array to map regional faults across the American continent. Natural migration can be formulated as a least-squares inversion to furtherer enhance the resolution and the quality of the fault images. This inversion is applied to ambient noise recorded in Long Beach, California to reveal a matrix of shallow subsurface faults.
The second part of this dissertation describes a convergent full waveform inversion method for controlled source data. A controlled source excites waves that scatter from subsurface reflectors. The scattered waves are recorded by a large array of geophones. These recorded waves can be inverted for a high-resolution image of the
subsurface by FWI, which is typically convergent for transmitted arrivals but often does not converge for deep reflected events. I propose a preconditioning approach that extends the ability of FWI to image deep parts of the velocity model, which significantly improves the chances for finding hydrocarbon deposits.
|Date of Award||Dec 2015|
|Original language||English (US)|
- Physical Sciences and Engineering
|Supervisor||Gerard Schuster (Supervisor)|