Biological phosphorous (P) -cycling in marine, terrestrial and atmospheric realm is key to evolutionary and climatic changes in Earth history. Oxygen isotope composition of phosphate reveals mechanisms of bond breaking and reforming during P-cycling as well as ambient water oxygen isotope compositions. I first review the state-of-art knowledge of phosphate P-O bond breaking and reforming induced by enzymatic processes (Chapter 1). Literatures in both the geochemistry and biochemistry have shown that on the cellular level, equilibrium exchange between phosphate and water happens within the cell and a non-equilibrium kinetic isotope effect occurs in the bond breaking and forming catalyzed by extracellular enzymes. Triple oxygen isotope could potentially separate the mixed signals between the two dominant processed which are otherwise hard to interpret. After we have concluded that P-O bond cannot be readily altered without extensive biological activities, we move to explore the history of land colonization by biota. I obtained a set of drill core samples from a paleoweathering profile of Middle Cambrian age (~500 Ma). The extracted phosphate oxygen isotope signature shows typical igneous ä18O and ∆17O values with no change between pristine and weathered igneous rocks. This is in contrary to the 13.2 change in the modern profile as demonstrated by Dustin Boyds thesis, suggesting a lack of any significant P-cycling in Middle Cambrian land surface (Chapter 2). The same triple oxygen isotope approach was applied to explore the weathering nature of weathering rinds, which shows a 0.8 excursion from pristine to weathered rinds, implying that biological activities are playing a role in the formation of the weathering rinds (Chapter 3). Finally, a set of phosphate samples extracted from a paleoweathering profile from the Late Permian Emeishan Large igneous provinces (ca. 260 Ma) reveals not only a modern-like land biological P-cycling but also local meteoric waters triple oxygen isotope composition at that time, opening up a promising venue for studying paleo-precipitation and paleoaltimitry (Chapter 4). Much of this study is explorative and has revealed many new applications of triple oxygen isotopes of phosphate. It paves the way for a more systematic study of geological materials with extensive sampling in space and time.

Structural, thermodynamic, and transport properties of several geologically important silicate liquids have been investigated at pressures relevant to deep Earth via first-principles molecular dynamics (FPMD) simulations. Results from CaO, CaSiO₃, model basalt (eutectic mixture of 36 wt. % anorthite and 64 wt. % diopside), hydrous model basalt (model basalt with 5 wt. % water), and mid-oceanic ridge basalts (with 9.9 wt. % FeO and 2.4 wt. % Na₂O) are presented. This work is a milestone in the density functional theory based calculations of compositionally complex systems including molten iron bearing silicates and a major step to understanding the natural magma. Simulation results show that liquid structure changes considerably on compression with all of the mean cation-oxygen coordination numbers increasing nearly linearly with compression. Most coordination changes occur at pressures below 30 GPa, which are accountable for rapid initial increase of melt density on compression. Melt compositions studied here along with other silicate melts remarkably show the same pressure evolution for Si-O coordination increasing from 4- to 6-fold over the entire mantle regime, but considerably differing in O-Si coordination. This change implies that the coordination polyhedra serve as building blocks of all silicate melts but how they act together to control the melt behavior might vary among different melts. CaO and CaSiO₃ liquids are much more compressible than their solid counterparts indicating the possibility of liquid-solid density crossovers at high pressure. In basaltic melts, the magnitudes of density changes due to the Fe and H₂O components are such that the melts including hydrous melt may be buoyantly stable at one or more depths in the mantle. Calculated self-diffusion and viscosity coefficients of the basaltic melts at zero pressure closely follow an Arrhenian law with activation energies ranging from 79 to 158 kJ/mol. However, the pressure variations of these coefficients requires a non-Arrhenian representation with variable activation volume. Predicted differences in viscosity of all basaltic melts with other silicate melts are subtle at shallow depths (up to ~20 GPa), suggesting the viscosity of major magma-forming silicate melts might not change much over that regime. Studies of hydrous model basalt reveals that the speciation of the H₂O component consists of mostly hydroxyls and molecular water at lower pressure, which change to more complex extended forms at higher pressure. Our calculation also shows that the volume of mixing in a melt-water system is nearly zero for most of the mantle pressure regime with negative enthalpy of mixing. Thus, the complete miscibility of H₂O with silicate melt favors the hypothesis that deep-mantle melts can be hydrogen-rich.

Research focused on three areas of karst hydrogeology and sediment transport that have been poorly studied in the past: the role of particulate inorganic carbon transport in calculating carbon sink rates in karst; rapid changes in surface vs. subsurface sediment mixing in karst conduits; and comparison of landscape denudation calculations using dissolved carbonate load vs. total dissolved/sediment load. Carbonate bedrock weathering is a significant component of the atmospheric carbon sink. Particulate inorganic carbon (PIC) in bed and suspended sediment load of karst waters is frequently dismissed as insignificant for calculating denudation and carbon transport/sink rates, but PIC flux has not been adequately studied. PIC fluxes were quantified in fluviokarst settings using RFID-tagged cobbles and gravel to track subsurface bed load flux, along with remote loggers and lab analysis of sediment and water chemistry for suspended and dissolved loads. PIC contributed about 10.3 percent annually to total inorganic carbon removal and additional carbon sequestration at Blowing Cave basin. Cosmogenic and fallout radionuclides 7Be and 137Cs were also utilized to develop a better understanding of rapid, short-term changes in sediment flux and transport through karst aquifers. The ratio of surface-derived sediment to stored and remobilized subsurface sediment in cave streams was estimated by tracking isotopic changes across storm runoff events. During a large storm event, approximately 34% of the total sediment flux during the 4-day runoff event originated from surface erosion. Landscape denudation rates in karst settings have been calculated using a number of different methods. The most widely applied basin-wide method is measurement of dissolved carbonate load at base-level springs, calculating denudation using the area of exposed carbonate outcrop, and assuming other outcrop types are in equilibrium. In mixed carbonate/noncarbonate lithologies (which includes most fluviokarst settings), much of the landscape is also eroded by transport of sediments through conduits. The basin average rate of landscape denudation (aggregating both carbonate and non-carbonate areas) was 48.5 mm/ka, but denudation rates for carbonate and non-carbonate outcrop areas were not in equilibrium.

Barataria Bay, one of the largest receiving basins for the Mississippi deltaic complex, is the location of a proposed river-sediment diversion for delta restoration. In order to determine how the sediment in the receiving-basin may respond to diversion flows, twenty-five sediment vibracores were collected from a 115 km2 study area located near Myrtle Grove and Bayou Dupont, southeast of New Orleans, LA. These cores were subject to multiple tests, including gamma bulk density scans, grain size analysis, and loss-on-ignition, in order to identify the lithology and stratigraphy. In addition, 137Cs and 14C dating techniques were employed in order to construct a geochronology. A subdelta lithofacies succession was identified and stratigraphically correlated across the basin, indicating more than one subdelta cycle in the sediment record. Geochronology suggests at least one St. Bernard subdelta entered dormancy within the range of 2130 to 2770 ± 30 14C years BP, a period that lasted a minimum of 860 ± 30 14C years, followed by Plaquemines-Belize subdelta progradation that ceased between 280 to 870 ± 30 14C years BP. The presence of channel sands and surviving St. Bernard age peats in the near-surface suggests resistance to compression and subsidence at depths greater than 2 m, providing a viable foundation for stable platform development from the mineral sediment nourishment of a large-scale diversion.

Newly acquired multibeam data from the Whales Deep Basin show lineations that extend to the shelf break, providing documentation that the West Antarctic Ice Sheet (WAIS) extended across the entire continental shelf during the Last Glacial Maximum (LGM). Following the LGM, the grounding line retreated approximately 60 km where it was pinned on a bathymetric high and began constructing a composite grounding zone wedge composed of seven individual wedges. Stabilization of the grounding line was thus caused by construction this composite wedge, elevating the grounding line by ~140 meters. Sedimentological data indicate that a small ice shelf with a calving front near the shelf edge formed over the outer shelf immediately after the grounding line began retreating from the shelf edge. The ice shelf broke up sometime after the fourth wedge was constructed. After construction of GZW7, the grounding line abruptly shifted ~200 km to the south where retreat slowed near the northern end of Roosevelt Island. Concomitant with grounding-line retreat, a second larger paleo Ross Ice Shelf reformed over the middle continental shelf. This large ice shelf then retreated abruptly and the present-day open water environment was established. These data indicate a complex deglaciation, which involved several discrete steps of the grounding line during the initial stages followed by a final major retreat of both the grounding line and calving front. The sedimentologic framework presented here is significant because it is a necessary step towards developing a detailed retreat chronology for this sector of the WAIS.

Seiche waves in Yellowstone Lake at ~78, ~51, and ~25 minute periods and heights <10 cm can cause measurable strain (< 40 ns) as observed on borehole strainmeters both near (<300 m) and far (~20 km) from the lake. Observations of the correlation between the seiche waves and the associated strain response can be used to constrain the rigidity of the upper crust, the depth to magma bodies in the subsurface, and the viscosity (related to melt percentage) of the magma bodies in place. Lake level was directly measured with campaign absolute pressure gauges deployed in the West Thumb basin, Breeze Channel, and central basin of Yellowstone Lake, and 13 large seiche events (¡Ý7 cm) were identified from these observations. Periodic strain transients were measured on borehole strainmeters in the caldera during the 13 large seiche events. The peak-to-trough amplitude and phase of the seiche wave and strain response were estimated by fitting a multi-frequency sinusoid to the time series over a 4-hour window during each seiche event. It was found that the relationship between an applied seiche load and the associated strain response is self-consistent and linearly proportional at each borehole strainmeter in the caldera. Over multiple seiche events the observed strain response is consistent with a modeled strain-field produced by a seiche load on a two-layered viscoelastic model defined by free parameters Young¡¯s modulus (E1), plate thickness (H), and shear modulus ratio (¦Ì2/¦Ì1). The two-layered viscoelastic model represents a solid upper crust overlying a partially molten body which may be small pockets of melt (<1 km thickness) or a larger magma reservoir. Results suggest crystallizing melt beneath Yellowstone caldera at depths (H) of ~4¨C8 km in the south-southeast and ~3-5 km in the north-northwest sections of the caldera. Temporal observations between strain meters, coupled with constraints of Young¡¯s and shear moduli suggest that melt in the shallow crust has a viscosity ¡Ü1013 Pa s.

Current point-bar complex models do not include subsurface unit bars as a normal feature. This study provides evidence for a potential buried unit bar amongst point-bar sediments of the large-scale, modern-day False River point-bar complex of the Mississippi River. We collect, process and interpret a two-dimensional, 150-m-long CMP seismic reflection profile that cuts perpendicularly across a major discontinuity surface in the False River point bar complex. The seismic source consists of a ground recoil device that fires a shotgun shell horizontally, producing shear waves. Multiple field experiments demonstrated which type of source and receiver provided the least amount of noise, with the most coherent incoming signal from reflections. LiDAR data allow the ridge-swale topography that exists above the point bar deposits to be readily mapped; this ridge-swale topography gives clues to the relative history of the meander bend. Seismic methods allow us to map the internal structure of the deposit, something that LiDAR cannot do. Gamma-ray and electrical conductivity data have previously been collected in a well located along our seismic line. These are correlated with the seismic data in order to assist with the interpretation. We find a seismic boundary dipping in the opposite direction that we anticipate in a point-bar complex. This may be a unit bar buried beneath many meters of point bar sediment, or may be the result of an erosive event. Unit bars add to the complexity of a point bar complex; they can lead to opposite-dipping boundaries than those caused by IHS layers. Two different models are created that could have resulted in the subsurface geometry seen in the seismic data. This study provides valuable insight into the evolution of fine-grained river systems, both modern and ancient.

The 3.29 Ga Pioneer Ultramafic Complex (WPC) is a 1 km thick east-northeast trending tilted (90° dip) section of layered ultramafic rock in the Weltevreden Formation in the west-northwest region of the Barberton Greenstone Belt (BGB). It was originally interpreted as an intrusive complex with slaty zones that have a linear horizontal fabric produced by shear. However, a recent study identifies these slaty zones as komatiitic tuffs and attributes their horizontal fabric to plane and cross bedding. The present study suggests the non-slaty zones are extrusive komatiitic flows with abundant pyroxene spinifex tops, normal volcanic textures and crystal sizes, and no indication of intrusive cross cutting. A geologic map is presented for WPC komatiites along with a stratigraphic section. The stratigraphic section includes 9 flows or flow sets (numbered WP 1 9). WP 1, WP 4, and the lower WP 7 are relatively undifferentiated massive olivine cumulate flows with minor interstitial pyroxenes. WP 2 and 3 are layered komatiitic flows with dunitic to peridotitic bases and increasing pyroxene content and changing pyroxene morphology toward pyroxene spinifex flow tops. Pyroxene spinifex komatiitic basalts are dominant in WP 6 8. Each flow type has substantially differing rock and mineral chemistry. Layered komatiitic flows are likely to represent more ponded lavas. WPC komatiites are mineralogically well preserved for Archean komatiites with abundant fresh olivine, orthopyroxene, pigeonite, augite, and chromite. The WPC komatiites are of the Al-undepleted geochemical type with rock and mineral Al2O3/TiO2 very similar to other units of the Weltevreden Formation. This geochemical signature is commonly attributed to the depth of partial melting in plumes, with Al-undepleted komatiites representing shallower partial melting (<450km), and also indicates that WPC komatiites may have a magmatic source related to that of the previously studied Weltevreden Formation komatiites. However, WPC komatiites differ from previously studied Weltevreden Formation komatiites in that they are somewhat less magnesian (maximum olivine Mg# 92.9 vs. 95.6), indicating a lower eruption temperature, but have much more complex and variable lithologic layering.

An outcrop in La Serre, France, was officially ratified by the ICS in 1989, and the IUGS in 1990, as the location of the Global Boundary Stratotype Section and Point (GSSP) for the FamennianTournaisian, and subsequently the DevonianCarboniferous (DC) boundary. GSSPs, like this one, are official outcrops that provide physical representations of geologic time boundaries, essentially geological standards that define geologic time, providing a vital framework to model a variety of interpretations of geological phenomena from paleoclimate to paleontological. It has been acknowledged that the GSSP in La Serre, France is in need of revision due to fossil reworking and general outcrop quality. The DC boundary has also traditionally been a challenging boundary to place with precision because of problems associated with Siphonodella sulcata, whose first occurrence is the current definition of the boundary. With the aim of improving DC boundary correlation, especially in the central United States, outcrops from the Woodford Shale in south-central Oklahoma and a New Albany Shale core from Johnson County, Indiana have been analyzed for cyclostratigraphy through measurement of mass-dependant, low-field magnetic susceptibility (χ), and also gamma radiation (GR). Gamma Ray Spectroscopy (GRS; field based) measurements were used for general correlation, along with previous conodont biostratigraphic work. Combined use of χ and GR measurements for 40K allows for a deeper layer of stratigraphic comparison. With the combined statistical techniques of the periodogram, multi-taper method (MTM), and wavelet analysis, a detailed timescale was pieced together for both the Oklahoma outcrops and Indiana core by comparing the periodic elements of the cyclostratigraphic signal represented by these geophysical proxies. Future work studying faunal assemblages can be compared with the cyclostratigraphic framework provided here to allow greater precision in interpretation.

One of the most pressing issues facing groundwater managers is saltwater intrusion. In coastal Louisiana this issue is especially prevalent. One location that is currently threated by saltwater intrusion is the industrial area of Lake Charles, Louisiana, where three high chloride areas have been detected within the underlying Chicot aquifer. Three sand units of the Chicot aquifer are present in Lake Charles: the 200-foot (200) sand, the 500 sand, and the 700 sand. Groundwater with elevated chloride concentrations was first noticed by industries in the early 1970s. An initial investigation determined that the northern and southern bodies had formed by upwelling of saline groundwater from the 700 sand. However, the origin of the salinity in the central body was not determined. The objective of this study was to determine the origin of the salinity for the central chloride body. Two sources of data were obtained from wells in the area: (1) spontaneous potential (SP) and resistivity logs from oil and gas and water wells (2) water quality data from United States Geological Survey (USGS) monitoring wells. The result of this study was the creation of a series of isoconcentration contour maps that help illustrate the movement of the saline groundwater in each aquifer layer over time. Results indicate that saline groundwater has been introduced into the aquifer from a variety of sources over time, including surficial contamination and upwelling of brine from the Lockport salt dome.