| dc.description.abstract | The Fourier transform generates a time-averaged amplitude spectrum of time series; seismic
data however is non-stationary, i.e., the frequency content of seismic data changes with time and depth. Spectral decomposition is an essential tool in seismic exploration for analyzing seismic data. Modern spectral decomposition methods such as the Short-Time Fourier Transform, the Continuous Wavelet Transform, and the S-transform address the non-stationary nature of seismic data. The above-mentioned common spectral decomposition methods are however inadequate for certain high-resolution seismic interpretation purposes. In addition, the above methods are performed on individual traces and do not take into account the continuous nature of geological structures or significant events of interest.
The first part of this dissertation presents and describes a new method called the Auxiliary S-transform. The Auxiliary S-transform is an invertible spectral decomposition method designed to significantly improve on the resolution of the S-transform by making use of the multi-dimensional nature of seismic data as well as the separation of seismic events in slowness coordinates. Multi-trace information is utilized with seismic processing techniques such as the linear Radon transform and the parabolic Radon transform. The linear Radon transform and the parabolic Radon transform are ideal for separating seismic events and extracting coherency information because they intrinsically take into account the velocity and curvature of seismic events. The workflow transforms data into the time-frequency-slowness or time-frequency-curvature domain where seismic events are better separated and the coherency attribute is more easily accessible. A filter can be applied in this domain to remove unwanted noise and further enhance the separation between events, to improve the temporal resolution of the method.
The Auxiliary S-transform is applied to synthetic data and its performance is compared to that of the S-transform. It is also applied to real seismic data for a shallow hydrocarbon environment. The results demonstrate that the Auxiliary S-transform has superior temporal resolution at all frequencies compared to the S-transform. The results also demonstrate that the Auxiliary S-transform is suitable for imaging the lateral continuity of seismic events and geological structures compared to the S-transform.
The second part of this dissertation presents a new quantitative approach for estimating the critical moment in a petroleum system. The petroleum system concept spans the spatial and temporal extent of all elements and processes required for the generation and preservation of petroleum. The critical moment of a petroleum system is the moment with the highest probability for the generation–migration–accumulation of hydrocarbons. It is an important concept in petroleum exploration risk assessment because the stratigraphic and geographic extents of a petroleum system are determined at the critical moment. In petroleum systems, thermal history data, burial history data, and vitrinite reflectance data may be unavailable, unreliable, or incomplete; this introduces significant uncertainty in the choice of the critical moment. This study presents a quantitative probabilistic framework for estimating the critical moment and quantifying the associated uncertainty in such cases. The quantitative probabilistic framework defines a probabilistic early bound and late bound for the critical moment (which, combined together, is termed the critical range) and then estimates the moment with the highest numerical probability of generation–migration–accumulation. It defines the uncertainty associated with the critical moment as half the absolute value of the critical range. In cases with little ambiguity or duplicity in the timing of petroleum system elements and processes, the critical range converges to one point, which is also the critical moment. The quantitative probabilistic framework introduces consistency to the critical moment estimation problem and quantifies the level of uncertainty in the estimation. This significantly reduces the risk involved in petroleum exploration assessment. | |
