Castagna, John P.2018-07-102018-07-10May 20162016-05May 2016http://hdl.handle.net/10657/3184Bandwidth-extension of a 3D seismic dataset from the Paradox Basin in Colorado showed a two-fold improvement in vertical resolution that allows for the interpretation of very subtle lateral discontinuities using seismic attributes. Common seismic discontinuity detecting attributes like coherence, chaos and curvature showed a noticeable increase in fault detection when applied on bandwidth extended data as compared to the original seismic data. A composite attribute was calculated by running these conventional attributes on various frequency band outputs from spectral decomposition and combining them using principal component analysis. This composite attribute can much more reliably delineate and locate very subtle faults compared to the individual input attributes. Validating the different conventional discontinuity attributes and the composite attribute against the seismic amplitude data showed that the subtle faults picked up by the composite attribute correspond to geologically-realistic faint discontinuities that are below the noise level of any of the input attributes alone. The orientations of seismically derived faults were validated by comparing them to faults seen at nearby wells. The reduction of noise and increase in spatial resolution allows for a reliable calculation of a fault density attribute. This attribute was compared against rock physics inclusion models through elastic inversion for quantitative interpretation of fractures in the seismic data showing a relationship between the presence of faults, the concentration of fractures, and the saturating fluid.application/pdfengThe author of this work is the copyright owner. UH Libraries and the Texas Digital Library have their permission to store and provide access to this work. Further transmission, reproduction, or presentation of this work is prohibited except with permission of the author(s).SeismicAttributesSeismic faultsBandwidth extensionInverseRock physicsPrincipal component analysisSpectral decompositionBrittlenessAzimuthAVOImpedanceInversionReflectivitySparse2018-07-10Thesisborn digital