Crystallization temperatures (T) and oxygen fugacities (fO2) of kimberlite magma estimated from oxides included in olivine phenocrysts from eight kimberlite pipes in the central Slave Province, Canada, are compared to the degree and character of resorption observed in diamonds recovered from these kimberlites. The mechanism of diamond oxidation in kimberlite melts and the rate-controlling parameters for this reaction are explored in oxidation experiments. The T and maximum fO2 recorded by olivine - chromite pairs at an assumed pressure of 1 GPa are 970° -- 1070°C and 2.2 - 3.1 log units below the nickel - nickel oxide (NNO) buffer. This mineral assemblage crystallized from a magma with 1 1 to 28 mol% of liquid, 10 mol% of earlier-precipitated olivine phenocrysts and 62 to 79 molc7o of mantle xenocryst olivine. The T - fO2 values vary between kimberlites from Northwest and Southeast clusters within 150°C and one log unit, respectively, and form a trend of decreasing fO2 and increasing crystallization T in the southeast direction. This trend corresponds to substantial differences in the diamond populations. A detail description of morphological forms and surface resorption features for five diamond parcels (> 7000 stones) show an increase in diamond resorption with increase in kimberlite crystallization T and more extensive surface etching in more oxidized kimberlites. The surface etch features on diamonds are determined by the conditions in the kimberlite melt, whereas some of the volume resorption occurs in the mantle and its relationship with the melt conditions is obscure. The diamond grade is higher in kimberlites with lower fO2 confirming the effect of the melt conditions on diamond preservation. Diamond oxidation experiments at 1350°C to 1500°C and 1 GPa produced only surface graphitisation, and no diamond resorption in volatile undersaturated melts. In contrast, volatile oversaturated conditions produce resorption features seen in diamonds recovered from kimberlites, suggesting that the process of diamond resorption is its reaction with the fluid and not with the melt. Both CO2 and H2O oxidize diamonds at a similar rate, but produce very different surface features. Therefore, the surface features of natural diamonds may provide information on the H2O/CO2 ratio in the kimberlitic fluid. The morphologies of diamonds from this study imply high H2O/CO2. The scarcity of surface graphitisation and presence of highly resorbed diamonds in kimberlites suggest presence of free fluid phase in kimberlite magmas for the most of their history. The diamond oxidation is not affected by the physical properties of diamonds.

Bridge Glacier is a prominent eastward-flowing valley glacier located on the east side of the Pacific Ranges within the southern British Columbia Coast Mountains. The terminus of Bridge Glacier has retreated at rates ranging from 0 to 125 m/year over the last 50 years and currently calves into proglacial Bridge Lake. Field investigations of the recently deglaciated terrain at Bridge Glacier in 2002 and 2003 led to the discovery of detrital boles and glacially-sheared stumps. Dendroglaciological analyses of this subfossil wood allowed for the construction of five radiocarbon-controlled floating tree-ring chronologies. The relative age and stratigraphic location of these samples revealed that Bridge Glacier experienced at least four periods of significant advance during the late Holocene: a Tiedemann-aged advance at ca. 3000'4C years BP. an unattributed advance at ca. 1900 '4C years BP. a First Millennial Advance at ca. 1500 '4C years BP. and an early Little Ice Age advance at ca. 700 '4C years BP. Lichenometric investigations at eight terminal and lateral moraine complexes led to the recognition of early Little Ice Age moraine-building events during the late 13th to early 14th centuries. with subsequent Little Ice Age episodes in the mid 15th. early 16th. mid-late 17th, early 18th, mid-late 19th, and early 20th centuries. These interpretations provide an exceptional long-term perspective on the extent and character of a glacier within this region during the late Holocene.

This thesis documents hydrothermal alteration patterns exposed along major fault scarps at Pito Deep, a tectonic window into the upper ocean crust formed at the very fast-spreading (>140 mm/yr) East Pacific Rise. Two main study areas were examined, each covering at least five lateral kms across the sheeted dike complex, exposing > four km of relief. and revealing >70,000 years of spreading history. Hydrothermal alteration patterns reveal vertical and lateral variation in the degree of alteration, dominant secondary mineral assemblages, peak temperatures of alteration, and metal depletion on the scale of 10s to 100s of meters. Amphibole and chlorite are the most common secondary minerals and replace clinopyroxene, interstitial zones, and plagioclase. Mineral assemblages indicate alteration temperatures of 250-450°C were common throughout most of the sheeted dike complex. Geothermometry indicates a range in alteration temperatures from 902-505°C and 350-60°C for amphibole-plagioclase and chlorite thermometry respectively. The highest temperatures are recorded in both deformed and undeformed dikes indicating that deformation was not always synchronous with peak alteration temperatures.

The initial discovery and settlement of the islands of Oceania is an important issue in Pacific anthropology. I test. two methods by which new island groups might be discovered: drift voyages and downwind sailing. I focus on the region of the initial eastward expansion into Remote Oceania by the Lapita people. Simulations are driven by high resolution surface wind and current data from atmosphere and ocean models forced by real observations and which capture the high degree of seasonal and interannual variability in the region. Both drift and sailing voyages can account for the discovery of all the islands in the Lapita region based on initial starting points in the Bismarck and Solomon archipelagos. Eastward crossings are most probable in the Austral summer and fall when the probability of occurence of westerly winds is highest. Contact with islands in the arc from Santa Cruz to New Caledonia is viable in all years and is particularly probable in the Austral summer. Pathways further to the east as far as Tonga and Samoa are plausible when considering anomalous westerlies which occur in certain years. Other key crossings in Polynesia are also possible when considering this interannual variability, much of which is associated with El Nino events. Many of my findings differ from an important, earlier modelling study performed by Levison et al. (1973).

This thesis presents estimates of marine gas hydrate distribution and concentration obtained from various geophysical methods. The study area is located in the accretionary prism of the Northern Cascadia subduction zone, offshore Vancouver Island. Canada. The primary objective of this study was to assess the applicability of a suite of geophysical methods in estimating marine gas hydrate distribution and concentration. The measurements tested are downhole log electrical resistivity and seismic velocity, multi-channel seismic (MCS) velocity, and seismic amplitude vs. offset (AVO) of a gas hydrate-related bottom-simulating reflection (BSR). The downhole log data are from Integrated Ocean Drilling Program Expedition 311, along a transect of four wells, and the seismic data are from a conventional 2-D MCS line along the well transect. Gas hydrate distribution and concentration estimates along the well transect exhibit high spatial variability, both from site to site, and within any given site. On average. estimates from electrical resistivity measurements give 5-15% gas hydrate pore space saturation. whereas velocity-based estimates are 15-25%. Some intervals in both cases show concentrations over 40%. Nonlinear Bayesian inversion of seismic AVO data yields a gas hydrate concentration estimate of 0-23% of the pore space. These results lead to the conclusion that resistivity and velocity data are effective tools for estimating marine gas hydrate concentration. The main uncertainty in the resistivity analysis is the in situ pore fluid salinity, whereas the main uncertainty in the velocity study is the magnitude of the bulk sediment velocity increase associated with gas hydrate occurrence (related to how gas hydrate forms). It is shown here that AVO of a gas hydrate BSR is not a useful method to estimate marine gas hydrate concentration. The method lacks the shear-wave velocity resolution necessary to add useful constraints to what is already known from compressional-wave velocity information.

This thesis is aimed at investigating the possibilities of using vertical line array (VLA) data to image the gas hydrate stability zone (GHSZ) in the northern Gulf of Mexico. The presence of gas hydrate can be inferred from seismic evidence such as bottom simulating reflectors (BSRs) or changes in seismic velocity. The petroliferous northern Gulf of Mexico is noted for its obvious absence of BSRs, a characteristic it shares with other active passive margins with mobile salt and/or shale, which have high propagation velocities for seismic waves. This makes the imaging and the identification of the gas hydrates a challenging process with conventional seismic techniques. Therefore. new techniques in data acquisition. processing and analysis are sought to improve the imaging of complex areas. The new, unconventional seismic data acquisition technique used here is the VELA. This work defines a seismic processing flow that has been developed to extract velocity, travel-time and amplitude information from VLA data to predict the hydrate distribution over the surveyed area. Specialized amplitude versus offset analysis and inversion is applied to the VLA data using a Bayesian inversion approach to provide estimates and uncertainties of the viscoelastic physical parameters at an interface. This thesis will compare the inversion of the 2-D seismic reflection data collected in 1998 by the USGS and in 2002 by the Center for Marine Resources and Environmental Technology (CMRET) to the VLA data collected in 2002 and 2003 by the CMRET to assess the value of a VLA in monitoring changes in the near-surface sediments that can be associated with the presence of gas hydrate.

I examine uncertainty in climate change mitigation using the Jaina theory of reality, anekdntavada, and post-normal science, focussing on carbon dioxide capture and storage (CCS). Aneknntavada reveals all conceptual knowledge, not just science, as reductionist. Traditional science is ekanta, failing to acknowledge its reductionism, and this has led to incautious and contextually inappropriate application of scientific results. However, through Jaina principles and post-normal methodology, we can avoid scientific and technological ekdntavadism. For example, we must ensure "continuous iterative control" is possible in any CCS project. Accordingly, geological and mineral carbonate storage are preferable to oceanic storage. The IPCC report on CCS is not a post-normal document, since only scientific and technical experts prepared it. This is not necessarily a criticism: the report could be an expert input preceding post-normal dialogue. However, since values enter scientific undertakings early on, a broader community of stakeholders should perhaps have prepared the Special Report.

This dissertation work applies seismic tomographic inversion methods to two different datasets - one to address the earthquake hazard within the Strait of Georgia and the other to estimate hydrate concentration and distribution in the continental slope off Vancouver Island. In the first part of the study, seismic refraction/wide-angle reflection data from onshore-offshore experiments in 1998 and 2002 were inverted for a smooth three-dimensional (3D) velocity structure down to depths of 6-7 km beneath the Strait of Georgia, a seismically active region where an earthquake swarm (with magnitude up to 5) occurred in 1995-1997. The objectives were to map structures that contribute to seismic hazard evaluation in the Georgia Basin. The main structural features obtained from the inversion are: a northeast-southwest trending hinge line at the location of the earthquake swarm, where the basin deepens rapidly to the southeast; a northwest-southeast trending velocity discontinuity that correlates well with the surface expression of the shallow Outer Island fault; sediment thickening from north to south; and basement uplift at the San Juan Islands, possibly caused by a thrust fault. In the second part of the dissertation, seismic single channel and wide-angle reflection data collected in September 2005 were analyzed for a 2D profile of ocean bottom seis¬mometers (OBSs) on the continental slope region off Vancouver Island, near ODP Site 889 and IODP Site 1327. The objectives were to determine the shallow sediment velocity structure associated with marine gas hydrates and to estimate the hydrate concentration in the sediment pore space. Combined inversion of single channel and OBS data produced a P-wave velocity model down to the depth of the BSR at 230 m below seafloor. Strong attenuation of P-waves below the BSR indicates the presence of free gas. To investigate structures below the BSR, forward modelling of S-waves was carried out using the data. from the OBS horizontal components. Both the P- and S-wave models match very well with the sonic log data from ODP Site 889 and IODP Site 1327. The increase in P-wave velocity of the hydrate bearing sediments relative to the background no-hydrate velocity was utilized to estimate the hydrate concentration by using a simple porosity-reduction equation. An average concentration of 15% was estimated from the P-wave velocity model. Prestack depth migration was applied to the OBS data to image the structure along the 2D profile containing the OBSs. The primary and multiple arrivals were migrated separately. Conventional migration of the primary arrivals produced an image with a very narrow illumination and the shallow subsurface layers including the seabottom were not imaged. However, migration of the OBS multiples, using a mirror imaging technique, pro¬duced a continuous structural image of the subsurface including the shallowest layers. The lateral illumination is much wider with a quality comparable to that of vertical incidence reflection data.

This dissertation presents experiments with the UVic Earth System Climate Model that explore the sensitivity of global climate simulations to an increase of horizontal resolution into the ocean eddy-permitting range. Model versions, with resolutions ranging from 1.8' (latitude) x3.6' (longitude) to 0.2'x0,4'. are evaluated in control states and in response to climate perturbations. The effect of resolution on ocean volume transport, ocean heat transport, water mass formation, and sea ice distribu¬tion are investigated within the framework of three foci: 1) the Atlantic meridional overturning circulation and its response to surface freshwater forcing: 2) the South-ern Ocean overturning circulation and its response to poleward intensifying winds concomitant with increasing atmospheric C02; 3) the Southern Ocean temperature and Antarctic Circumpolar Current response to poleward intensifying winds alone. These sensitivity studies shed light on the ability to draw firm conclusions from coarse resolution modelling results. First, it is found that the simulation of western boundary currents. sea ice and meridional heat. transport. in the North Atlantic Ocean are improved with increasing resolution. A slowdown of the Atlantic meridional overturning circulation is robustly produced at both coarse and ocean eddy-permitting resolutions in response to fresh-water forcing applied evenly over the Labrador Sea and exclusively along its western boundary. An evaluation of the forcing impact on different regions of NADW for-mation with tracers reveals that increased Labrador Sea deep convection at higher resolution may mitigate the influence of better resolved boundary current transport. With increasing resolution. there is less cooling in the subpolar west Atlantic. more cooling in the subpolar east. Atlantic, and greater variability in the deep ocean re-sponse to the western boundary forcing. Second, it. is found that as resolution increases the interior ocean circulation be-comes more adiabatic, with deep water formed in the North Atlantic tending to upwell more in the Southern Ocean and less in the low-latitude oceans. For some density classes the transformation rate derived from surface buoyancy fluxes can provide a proxy for the net meridional transport in the upper Southern Ocean. The response of the Southern Ocean overturning to poleward intensifying southern hemisphere winds concomitant with increasing atmospheric CO2 suggest that the circulation associated with the formation of Antarctic Intermediate Water is likely to strengthen through the 21st century. Third, the zonal mean structure of the Southern Ocean temperature response to poleward intensifying winds through the 21st century, with warming between 40-55°S and cooling at higher and lower latitudes. remains robust as model resolution increases. However, the migration of distinct ocean fronts at finer resolutions can produce a strong local intensification of the temperature response. There is also less high latitude cooling and a greater loss of sea ice thickness at eddy-permitting resolutions, in conjunction with a significant increase in southward ocean eddy heat transport. Antarctic Circumpolar Current transport is found to increase in response to the forcing at both coarse and eddy-permitting resolutions.

The purpose of this research is to gain understanding of the occurrence and origin of streamlined forms in central British Columbia. More than 50,000 landforms, primarily drumlins and crag-and-tail ridges, were digitally mapped over an area covering five 1:250,000 NTS map sheets. Visual Basic programs were written to statistically analyze the streamlined forms database and to simulate site-scale, two-dimensional glacial erosion. Results show three principal ice and/or meltwater flow directions: southeast flows probably originating in the Skeena Mountains, northeast flows from the Coast Mountains and Quanchas Range, and west flows originating east of the Babine and Telkwa Ranges. Rat-tails and striae occur up to 1680 m elevation and record uphill flow to the west in these ranges. Streamlined forms were investigated at outcrop scale (e.g. rat-tails), landform scale (e.g. drumlins) and landscape scale (as defined by closed contours). On bedrock outcrops, cross-cutting striae are common and they both parallel and cross-cut rat-tails. Small rat-tails occur on, and parallel to, larger rat-tails but they do not cross-cut, suggesting a different origin than striae. Rat-tails are interpreted as being formed by subglacial meltwater flows, an interpretation supported by the glacial erosion model. Lowland streamlined forms (e.g. drumlins and crag-and-tails) are interpreted as either glacially-formed ridges subsequently shaped by meltwater floods or as being formed entirely by meltwater floods. This interpretation is largely based on the common occurrence of interconnecting hairpin furrows around these streamlined forms and on the demonstrated association of hairpin furrows with fluvial erosion. The results of topographic analysis indicate that an interconnecting system of valleys separates uplands that can be objectively defined by single (closed) contours. The aspect ratios of the uplands are highly correlated (L/W = 2.38, R2 = 0.89) with values that are similar to those reported for braid bars and erosional residuals thought to have been formed by glacial outburst floods. This upland/lowland landform assemblage may, in places, represent streamlined erosional residuals within braided channel networks formed, at least in part, by subglacial or glacial outburst floods.