The evolution of the Brazos and Colorado fluvial/deltaic systems over the last 190,000 years is examined. The data set includes high-resolution seismic data, lithostratigraphic data from offshore platform borings, cores and gamma-ray logs, biostratigraphic data, oxygen isotope data, and radiocarbon dates. Major periods of fluvial/deltaic deposition for the Brazos and Colorado fluvial/deltaic systems have been mapped through time, and across 20,000 square kilometers of the Texas continental shelf and upper slope. Chronologic control has allowed an independent assessment of the timing and nature of fluvial/deltaic deposition relative to the SPECMAP oxygen isotope curve. The Brazos and Colorado fluvial/deltaic systems have responded in phase with eustasy and facies distribution is related to fourth-order and fifth-order eustatic cycles. The Brazos fluvial/deltaic system developed deltas with highstand geometries during each of the oxygen isotope stages 5e, 5c, 5a, and 3. Fluvial incision and sediment bypass characterize the lowest sea-level stands. The Stage 2 to Stage 1 transgression is marked by a planar, time-transgressive ravinement surface. During the transgression, the Brazos deltas backstepped across the shelf. An oxygen isotope record, generated from benthic foraminifera, records the Mississippi meltwater pulse during the Stage 2 to Stage 1 transgression. Identification of isotope Stage 5 and Stage 3 is made possible by correlating the isotope data with paleoenvironmental data and regional downlap surfaces from seismic data. The interpretation of sea-level history and its role in fluvial/deltaic sedimentation has, therefore, been arrived at using three branches of stratigraphy (i.e., oxygen-isotope stratigraphy, biostratigraphy, and seismic stratigraphy). Lithologic data provide the ground truth for the seismic facies analyses. An additional aspect of this study was the acquisition of a low-cost, high-resolution 3-dimensional seismic survey. The experimental 3-D survey successfully imaged an incised valley and its sedimentary fill. The survey covered a 750 m x 2000 m area over the Stage 2 Brazos incised valley.

A portion of new seismic reflection/refraction data collected over the Iberia margin were processed and interpreted, resulting in the recognition of two previously unknown normal faults in the deep crust. These faults, named Q and R, formed during the Triassic-Cretaceous rifting of Pangea, and formation of the Atlantic basin. Previous studies have shown a third surface, the low-angle S reflector, in the area. S has previously been interpreted as a detachment surface or as a velocity interface. A model is proposed where extension was accommodated along Q, R and S during the rifting of Pangea. R and S acted as detachment surfaces as the hangingwall block moved westward. Late in the rifting process, the geometry of R and S was altered by motion of underlying serpentinized mantle material. This upwelling of the mantle controlled the shape of Sin three dimensions. Total extension across the margin is estimated at 1.67.

Gas hydrates were discovered in a mud diapir in the leased block Mississippi Canyon 798 - Gulf of Mexico, through piston coring. Subsequently, a seismic experiment was set up to investigate the dynamics behind the hydrate formation. Wide aperture seismic traveltime data obtained from the experiment have been inverted to estimate 2D P-wave velocity models of the five shot lines. The results from modeling indicate the presence of free gas in regions that show up as zones of high reflectivity on the reflection profiles. The topography of the study area suggests presence of active salt bodies, which in turn, makes it plausible for the gas in the Mississippi Canyon 798 to have deeper sources.

This study documents the evolution of the interdeltaic central Texas continental shelf over the last 350,000 years. The dataset consists of high-resolution seismic data, platform boring descriptions, oxygen isotope analyses, and radiocarbon dates. A strong correlation exists between seismic facies and lithofacies enabling seven stages of evolution to be mapped. The distribution of these facies is primarily controlled by fourth-order glacio-eustatic cycles. Seismic stratigraphy, oxygen isotope analyses, and radiocarbon dating were integrated to develop an independent sea-level curve for the area. Longshore and surface currents transport large volumes of sand into the area which is deposited as widespread, thick $(>$10 meters) barrier-bar highstand sand bodies on the inner shelf. Sediments are deposited and preserved in a repetitive manner during each glacio-eustatic cycle. This implies a predictable pattern of deposition, therefore, allowing for the development of depositional models that can be applied to ancient deposits in the exploration for hydrocarbons.

The evolution of the Matagorda/Lavaca Bay complex demonstrates that although eustasy was an important control on incised valley flooding during the Holocene, antecedent topography, climate change, and barrier migration significantly impacted the resulting stratigraphic architecture. With the acquisition of 110 km of high resolution seismic data and 100 m of sediment cores, lithofacies and seismic facies of both bays were mapped in detail revealing 4 abrupt changes in depositional environments. These changes occur across major flooding surfaces at 9600, 7700, 7000, and 5000 yr and include the backstepping of deltaic and estuarine deposits as well as the onset and termination of spit accretion. Comparison of the evolution of Matagorda Bay with other Gulf Coast bays reveals that, with the exception of the 9600 yr event, flooding events were diachronous across the Gulf Coast, implying that climatic and autocyclic forcing mechanisms played a larger role in coastal evolution than previously recognized.

First arrival traveltimes and associated amplitudes are essential components in Kirchhoff migration and modeling. Traditionally they have been determined by ray tracing. However ray tracing does not give traveltimes on a regular grid, and is not guaranteed to produce the minimum traveltime. Seismic traveltimes in three dimensions can be computed efficiently and accurately on a regular grid using an essentially nonoscillatory ("ENO") Hamilton-Jacobi (HJ) second order scheme. The scheme can be implemented in fully vectorizable form. Several examples illustrate the effectiveness of this approach to traveltime computation. A similar accurate scheme is required to solve the transport equation for the amplitudes associated with the first arrival traveltimes. A second-order Runge-Kutta upwind finite difference scheme is constructed for this purpose. The transport equation involves the traveltime gradient and Laplacian which must be evaluated using the output of the eikonal scheme $\tau.$ The error in $\tau$ is second order accurate, hence the approximation to the traveltime Laplacian is zeroth order accurate, and there is no reason to expect the traveltime Laplacian, hence the amplitude, to converge. One remedy to ensure that the traveltime Laplacian is sufficiently accurate to guarantee convergence, is to use a higher order scheme, say third order ENO upwind scheme to solve the eikonal equation. Preliminary numerical results are presented to demonstrate the third-order accuracy of the HJ-ENO numerical flux in spatial directions (x and y) and of TVD (total variation diminishing) Runge-Kutta method in z-direction.

Continental rifting is a fundamental component of the plate tectonic cycle. The West Iberia passive margin is a classic example of a nonvolcanic rifted margin. The West Iberia margin contains an enigmatic north-south ridge of serpentinized peridotite located within the ocean-continent transition. Interpretation of multichannel seismic data and tectonic subsidence analyses suggests that the ridge is located within a broad zone of exhumed mantle that has been serpentinized. This implies that seafloor spreading does not immediately follow continental breakup. Where the peridotite ridge is well-developed, it parallels a deeply-penetrating, west-dipping normal fault. Hydrodynamic circulation drove seawater down this fault close to the beginning of seafloor spreading and caused a concentration of serpentinization at its base. This water-driven process of formation accounts for the variability of the ridge along strike. Prestack depth migration of a 340 km long seismic reflection profile across the margin served as the basis for stratigraphic interpretation. The proximal margin displays horsts and grabens, with 1--2 km thick synrift deposits from Early Cretaceous (Valanginian) rifting. The deep water margin displays rotated blocks with distinct internal seismic patterns. These blocks formed during the final Cretaceous rifting event that led to continental breakup, and contain earlier Valanginian synrift deposits as part of the rotated blocks, thus providing evidence of two separate rifting stages along this profile. A large mantle-penetrating fault located seaward of the last rotated block of continental crust provides evidence for a third and final Cretaceous rifting event that may have been coeval with early seafloor spreading. Two independent estimates of crustal thickness along this line indicate stretching factors of 50% on the proximal margin (corresponding to a continental crustal thickness of ∼16 km), increasing to 100% in the deep water. Plate tectonics is one of the most important concepts that Earth Scientists can convey to the public. While this concept is often presented at the middle school level in Texas, the appropriateness of teaching plate tectonics to middle school students has been questioned. We present data that suggest that middle school students grasp the abstract concepts of plate tectonics as well as college students.

Pedro Channel is the deepest, central-most seaway on the northern Nicaragua Rise. A revised bathymetric map illustrates a seafloor dissected by numerous canyons and gullies as well as portions of a drowned carbonate bank with keep-up bank morphology. The 3.5 kHz echogram data interpretation in Pedro Channel indicates the banks provide both a line and point source of sediment to the channel floor. Concentric facies belts along the bank margins illustrate the line source concept. Localized zones of coarser-grained deposits at the bases of canyons along the bank margins depict the point source concept. The synthetic seismograms for ODP Site 1000 have low correlation coefficients (0.321). Visual correlation of the synthetic seismogram with SCS data indicates a good correlation of seismic facies with various lithologic intervals. An interpretation of SCCS and MCS data reveals that periplatform sedimentation has dominated Pedro Channel from the early Miocene to recent. Dredge haul analysis suggests that a neritic carbonate bank drowned partially in the middle Oligocene and finally in the early Miocene. Erosion affects the edges of this drowned bank in the middle Miocene. Faults in Pedro Channel illustrate characteristics typical of sinistral strike-slip faults including vertical to sub-vertical faults, faults that splay upward, forced folds, and linear fault traces in map view. ODP Site 998 was selected over a small portion of the Cayman Rise on a small topographic high. Synthetic seismograms for ODP Site 998 have low correlation coefficients (0.168). Despite the low correlation coefficients, turbidite-rich core intervals correlate with SCS intervals having low continuity seismic reflectors. The predominantly pelagic sedimentation at this Site results in a generally continuous seismic reflection character. ODP Site 1001 is located just north of the Hess Escarpment. Synthetic seismograms have low correlation coefficients (0.335), but do permit correlation of SCS A$\sp{\prime\prime}$ and B$\sp{\prime\prime}$ horizons with an Eocene-Miocene unconformity and basaltic basement respectively. Hydrosweep data indicates that the Hess Escarpment is dissected by numerous canyons incising along faults. The northwest trending normal faults are suggestive of dextral strike-slip motion.

This thesis presents the results of traveltime inversion of seismic wide-angle reflection/refraction data for strongly varying media. The two-dimensional velocity structure along two lines across the central Chilean margin (near Valparaiso) has been obtained and assessed in terms of resolution, uncertainties, and non-uniqueness. The traveltime inversion method was used to model the data and assess the model reliability. A tomographic approach was used to assess the objectivity of the structures in the final models. The final models include slope sediments, the Valparaiso forearc basin, subducting sediments, an accretionary wedge, upper and lower continental crust, a two-layer oceanic crust, and uppermost mantle. The thesis results were compared with the results of forward modeling of the same dataset and with results of analogous seismic surveys across convergent margins worldwide.

As the results of seismic reflection and ground penetrating radar (GPR) studies become more prevalent as input for quantitative groundwater and engineering studies, it is important to evaluate traditional approaches to data processing and analysis. Where conventional methods fail it is necessary to investigate and/or develop non-traditional approaches to data analysis. I present five stand-alone studies that are focused on characterizing shallow aquifers using seismic reflection and GPR data processing and analysis. Each of the projects involves a new approach to data analysis either through alternative processing strategies that are not widely applied in environmental studies or development of new processing methods and/or algorithms. The first two studies are focused on seismic reflection imaging problems that arise in the shallow environment. I first present a detailed discussion of the errors that can result from conventional normal-moveout (NMO) processing, and the application of pre-stack depth migration (PSDM) to improve image accuracy. Second, I apply dip-moveout (DMO) processing to a data set from the Puget Sound. DMO is rarely applied in environmental studies, but can improve velocity analysis and image quality where there are dipping layers or scattering events. The final three studies are focused on attribute analysis of GPR and seismic reflection data in direct detection studies. Over the past 10--15 years, direct detection has been used successfully in the energy industry to identify oil and gas reserves from exploration scale seismic reflection data, but is a new approach to the analysis of GPR and shallow seismic reflection data. Direct detection studies use reflected wave attributes such as amplitude, frequency content, and phase to estimate material properties. I first present a detailed discussion of GPR amplitude vs. offset (AVO) analysis for direct detection of free phase non-aqueous phase liquid contaminants (NAPLs). The analysis is not straightforward and requires careful consideration of electromagnetic dynamic properties. Second, I present a shallow seismic case study where a predictable AVO response was detected. Finally, I present a new wavelet decomposition and time-frequency representation, and illustrate applications in GPR attenuation analysis for NAPL detection and lithology characterization.