Frequency-dependent analysis of seismic data is gaining more and more attention in the seismic industry and academia, simply because it opens potential opportunities not only for indicating hydrocarbon anomalies, but also for estimating fluid mobility properties of underground reservoirs. While there has been significant advancement in the interpretation of seismic amplitude-versus-offset (AVO) anomalies, there is a lack of theory to guide the interpretation of frequency-dependent analysis. The area of frequency-dependent seismic attributes to predict the fluid-mobility properties of underground reservoirs is in the early stage of development.

In the first part of my dissertation, I give a brief review of Biot's poroelastic wave theory. In the second part, a crossplot technique has been developed, which integrates AVO and spectral analysis and thus is called Spectra Crossplot. It provides a quick and useful tool to map hydrocarbon anomalies in Class 2 AVO environments.

In the third part, I examined the characteristics of the reflection-coefficient variations as a function of frequency at an interface between a non-dispersive medium and a dispersive medium. Similar to AVO classification, the Amplitude versus Frequency (AVF) curves are generally divided into three classes. This classification provides a guide to frequency-dependent interpretations. These three classes of AVF responses have been confirmed by forward modeling based on Biot's wave-propagation theory. A field example is used to illustrate one of the AVF responses that I propose.

In the fourth part, I study the permeability effects on seismic amplitudes based on analytical and forward-modeling analyses. It is shown from the analyses that if only permeability changes, the permeability difference may cause significant amplitude changes in the transition frequency range, which is confirmed by numerical modeling results. I also discussed the reason why the relation between amplitudes and the fluid mobility of a reservoir is sensitive at low frequencies. A field example is presented in the end to show the correlation between the low-frequency amplitudes and reservoir permeability. This result is very encouraging for future research in estimating permeability values from seismic.