Search

Fluvial systems are regarded as a critical component of earth surface processes because they link the sediment source and the basin. The characteristics of luvial sediments are influenced by tectonic and climatic processes. Thus, sediment grains, as the basic element of fluvial sediments, are often used to reconstruct changing tectonic and climatic conditions. One common observation of fluvial sediments is a tendency for the bedload to become finer downstream. In many studies, large pebbles located far away from the sources have been interpreted to be caused by tectonic or climatic influences. In the Great Plains, USA, large pebbles can be found more than 1400 km away from the sources. In this area, it has been demonstrated that slow subsidence resulted in multiple cycles of fluvial incision and aggradation. A hypothesis can then be proposed that large pebbles downstream came from the recycling of the paleosediments. To test this hypothesis, three main geological tools were applied: grain sizes analysis, cosmogenic nuclides analysis and numerical modelling of cosmogenic 21Ne. In the grain sizes analysis, downstream fining curves were generated, and similarity statistics were done. Based on the result, three locations of probable lateral sediment input from recycled paleo-channel deposits were recognised. For each of these locations, grain size distributions with and without lateral sediment input were generated by combining the grain size fitting curves and the similarity distributions. From the comparison between the grain size distributions with and without lateral sediment input, it can be concluded that almost all the pebbles collected from these three locations are from recycling of palesediments. Cosmogenic nuclides accumulated when the sediments were exposed, so that they can provide information about the exposure and burial history of fluvial sediments. In this thesis, we analysed eighty-two samples (in total) for cosmogenic 21Ne and five samples for 10Be. A “steady-state” model and a “simplified migration” model were built to estimate the maximum concentration of cosmogenic 21Ne accumulated from the source to the target location (Keystone) with a constant migration rate. From the comparison between this maximum concentration of cosmogenic 21Ne and the concentrations of cosmogenic 21Ne measured from the pebbles, it can be concluded that most pebbles collected from Keystone have experienced a long time of storage. This provides support for the existence of recycling in the Great Plains. Next, based on the Ne/Be ratio, the oldest age of recycled sediments in the Great Plains was defined as at least Miocene in age. The result of cosmogenic nuclides analysis introduced an interesting apparent conflict that cosmogenic 21Ne concentrations contained within the samples collected from upstream are higher than those from downstream. To resolve this conflict, numerical models were built to test the controls of cosmogenic 21Ne accumulated during recycling. Two geological parameters, elevation and age of paleosediments, were tested in these models. As suggested from the results, elevation plays a dominant role in the accumulation of cosmogenic 21Ne during recycling. And, the paleosediments of Pliocene/Miocene are the main sources of the recycled pebbles in the Great Plains, but the portion of them plays a minor role in the distributions of cosmogenic 21Ne contained within the pebbles collected along the North Platte River. For the influence of grain size, the data show no relationship between the grain size and the concentration of cosmogenic 21Ne. Lastly, a previous study which used grain size to reconstruct paleo-river profile was tested to evaluate the impact of introducing recycling on the previous interpretations. As the result shows, regional grain size fining rate cannot be used directly to reconstruct the gradient of the whole channels, when recycling is dominant. To conclude, I have shown that in the Great Plains, recycling is the main reason for the existence of pebbles 1400 km away from the sources, and recycling happened in modern time can affect the paleosediments as old as Miocene in age. Besides, it was concluded that elevation is the main factor that influences the concentration of 21Ne accumulated during recycling, and the paleosediments of Miocene/Pliocene are the main sources of recycling. In the end, it was recommended that recycling should be taken into consideration in geological studies.

The measurement of terrestrial cosmogenic nuclides (TCN) in river sediments is commonly used to determine basin-averaged erosion rates over centennial-to-millennial timescales, and for comparing erosional processes between catchments. Despite the prevalence of the use of this method in geomorphological studies, uncertainties remain and continued efforts are needed to further refine techniques, identify limitations, and understand biases in these measurements and their conversion to erosion rates. For example, the role of event-triggered mass-wasting on transient perturbations of TCN signals is an area of ongoing community interest, and the impact of grain attrition during sediment transport on TCN concentrations in different grain size fractions has only recently been explored in detail. This thesis begins with a modeling study that explores the relationship between landslides, erosion, perturbations, and TCN measurements from a catchment in central Nepal. While it is widely accepted that basin-scale perturbations should temporarily dilute TCN concentrations as landsliding mobilizes new, low concentration material, the impact of the catastrophic release of hillslope sediment caused by a single event on TCN signatures has not yet been systematically investigated. The second chapter of this work looks at the role of the spatial and temporal distribution of landslides in the measurement and potential variability of TCN signals using a modeling approach. We then use the 25 April 2015 Mw 7.8 Gorkha earthquake, Nepal, as a case study to investigate how perturbations are transferred to and ultimately archived in the sedimentary record. Modeling results show that changes in detrital TCN concentrations are not necessarily an indication of an upstream perturbation and that TCN dilution in fluvial sediments should not be considered a universal characteristic following high-magnitude landslide-triggering events. Following the modeling approach, we analyze time-series measurements of 10-beryllium (10Be) concentrations in fluvial sands from catchments in central Nepal affected by coseismic landsliding during the Gorkha earthquake. The third chapter focuses on interpreting this data, which in contrast to the typically observed TCN dilution, exhibits an increase in concentrations in nearly all sampling locations in the years following the event. To explain this data, we consider landscape processes, landslide characteristics, and sediment export scenarios. We find that the earthquake likely mobilized material in more highly charged areas of the hillslopes that do not fail as frequently during the annual monsoonal and which are also farther from the river network. Building off of Chapter two, our results provide further evidence that extreme events do not systematically dilute TCN concentrations in fluvial sediments and are not likely a reliable marker in the sedimentary record. In the final chapter, we explore inherent sampling biases in TCN-derived basin averaged erosion rates. From these same study catchments in central Nepal, we also measure 10Be concentrations in the pebble fraction and find that pebbles systematically record higher erosion rates than sand. A simple model of abrasion processes shows that the offsets can be attributed to the downstream abrasion of coarser material, released by mass wasting, into the sand fraction. This suggests that grain-size biases in TCN measurements should decrease with increasing catchment size, and likely be minimal ~100 km downstream of landslide deposits. Overall, this thesis offers insight into the evolution of TCN signals and how they are measured and recorded in a dynamic landscape.

Central to examining orogenic belt formation and evolution are efforts to define the relative roles of tectonics and climate in landscape denudation. Studies in high altitude, high relief mountain settings such as the Himalayan-Tibetan orogen show that denudation can scale with tectonics, precipitation and/or topography. Steep north-south trending gradients in elevation, relief, rock uplift and precipitation make the NW Himalaya an excellent location to examine these controls of landscape denudation. To elucidate the erosional processes involved in sediment mobilization from primary sources at catchment headwaters, rates of lateral rockwall slope erosion are defined for 12 catchments between the Baltistan region of Pakistan and Garhwal, northern India. These rates are derived from 10Be cosmogenic nuclide concentrations in medial moraine sediment, and affirm that between ~0.02 and 7 m of lateral slope erosion through periglacial processes can be achieved across a single millennium in this setting, and >2 km when extrapolated for the whole Quaternary. This erosion is sufficient to influence topographic change at the catchment head, and emphasizes the importance of localized erosion when evaluating broader landscape change. Statistically significant relationships are apparent between slope erosion and apatite fission track cooling ages and inferred rates of exhumation. Rockwall slope erosion in the NW Himalaya is found to be primarily governed by tectonically driven rock uplift, which is the result of the Indo-Eurasian convergence and the geometry and shortening of the Main Himalayan Thrust. Precipitation occupies a secondary control. To examine the nature and timing of sediment transfer from primary sources to transient sinks, 18 new and 29 recalculated 10Be fan surface exposure ages define the timing of surface abandonment for a suite of 10 fans across the Greater Himalaya ranges in northern India. The fans are found to evolve during interglacials and periods of deglaciation, which are associated with a higher intensity monsoon. As glacial conditions or periods of relative cooling and increasing aridity become reestablished, aggradation ceases or becomes more restricted and the fan surfaces stabilize and/or are abandoned. A regional temporal framework of landform abandonment/aggradation cessation and incision events for the monsoon-influenced western Himalaya ranges (MWHR) and the semi-arid western Himalaya ranges (SWHR) are defined for the NW Himalaya. Regional landform abandonment events are consistent with the fan record, occuring during periods of relative cooling and increasing aridity, while regional incision events occur across various climatic conditions. Climate and climate-driven processes largely moderate sediment transfer in the NW Himalaya; however, the distribution of precipitation and temperature are themselves influenced by the structural evolution of the orogen. Although the primary controls of erosion and sediment transfer in this study are shown to be rock uplift and climate respectively, this research overall argues that the various components of Himalayan mountain range evolution cannot each be governed by one mechanism alone. The specific nature and rates of landscape change on the catchment scale are therefore influenced by longstanding catchment-specific feedbacks between topography, geology, surface processes, climate and tectonics, in addition to regional forcings on the mountain range and/or orogen scale.

Anthropogenic climate change is transforming Earth’s Polar Regions. Oceanic and atmospheric warming threaten to melt portions of the Antarctic and Greenland ice sheets and raise sea level dramatically by the end of this century. As a guide for the future, we can examine the geologic record to understand the surface processes that govern ice sheet behavior and establish chronologies of past climate change. In this dissertation, I focus on the Pleistocene glacial history of two polar field sites: the presently ice-free region of McMurdo Sound, Antarctica and formerly glaciated continental shelf of Melville Bugt, Greenland. In my research, I use a combination of detailed geologic mapping, geochemical and isotopic analyses of glacial sediments, and several Quaternary geochronologic techniques. I show that several marine-based ice sheets inundated McMurdo Sound and deposited a series of glacial sediments that record former ice sheet configuration and Antarctic surface processes. Mapping results and a robust radiocarbon chronology indicate that grounded marine-based ice in the Ross Sea overflowed into McMurdo Sound and attained maximum thickness after the global Last Glacial Maximum; ice receded in response to rising sea level and changing ocean circulation. Cosmogenic nuclide exposure ages of boulders from these glacial deposits do not record a simple history of ice extent, but instead the integrated effects of multiple surface processes operating below, along, and above the Antarctic ice sheets. I build upon the insights gained from Antarctica to develop a novel approach to analyzing glacial marine diamict, a mixture of grain sizes and lithologies, deposited on the continental shelf by an expanded Greenland Ice Sheet. Using cosmogenic nuclides, low temperature thermochronometry, and lipid biomarkers, I show that the Greenland Ice Sheet during the Early Pleistocene was similar to today. Extremely low cosmogenic nuclide concentrations indicate the ice sheet already eroded pre-glacial soil cover, but thermochronometry data suggest deep fjords had not yet formed. Abundant lipid biomarkers recovered in this diamict indicate that ice-free areas supported vegetation, similar to the modern ice sheet. This dataset demonstrates how the novel application of proven techniques to glacial sediments can unlock new information about ice sheet history and process.

An extensive alluvial apron of coalescing gravel fans blankets the western flank of the Chajnantor Plateau in the Atacama Desert of northern Chile. Remnant alluvial surfaces, terraces, and intermittent debris flow deposits preserved in this bajada indicate multiple intervals of aggradation, incision and terrace abandonment, and deposition. The high preservation potential and sensitivity to climate shifts of the region provides a unique opportunity to elucidate the sedimentary response to climate variations at an extreme of Earth’s climate. Cosmogenic 36Cl exposure dating, aided by mapping, is used to establish a detailed chronology of the depositional history of the Chajnantor alluvial apron. Alluvial surfaces and gravel deposits yield cosmogenic exposure ages ranging from 20.7 ± 1.4 ka to 419.2 ± 39.6 ka. Debris flow boulders confined to modern and ancient channels yield cosmogenic exposure ages ranging from 9.3 ± 1.1 ka to 202.5 ± 19.6 ka. One localized (Qcf1) and two extensive (Qcf2 and Qcf3) abandoned alluvial fan surfaces yield cosmogenic exposure ages of ~33 ka, ~55 ka, and ~145 ka respectively. These abandonment ages coincide with periods of moraine stabilization and deglaciation on the adjacent Chajnantor Plateau. Protracted periods of aggradation appear to coincide with periods of increased precipitation and glacial occupation of the Chajnantor Plateau, and are punctuated by phases of incision and surface abandonment during interglacial periods. Cosmogenic exposure ages reveal a transition from aggradation to incision during marine oxygen isotope stage II (MIS II). These results suggest that climate, specifically precipitation, is the dominant control on alluvial fan formation along the western margin of the Chajnantor Plateau.