Iceberg calving is an efficient ablation process which introduces mechanical instability to glacier systems and can cause non-linear climatic response. This thesis uses glaciological and limnological data to examine the relative contributions of calving and melting to mass loss at glacier termini, and the interplay between glaciological and limnological processes. Calving dynamics are investigated at two lake-terminating glaciers; Glaciar Leon in Chile and Fjallsjokull in Iceland. Glaciar Leon, a temperate, grounded outlet of the North Patagonian Icefield, terminates at an active but stable calving margin in Lago Leones. The calving rate of 880 m a-1 in a mean water depth of 65 m is high for lake-calving glaciers. Detailed survey of the physical limnology of Lago Leones, important for considering heat transfer to the subaqueous ice face, revealed thermocline development towards the terminus between spring and summer. Melting at the waterline along the glacier terminus facilitates calving by undercutting the subaerial calving cliff, and accounts for around a quarter of mass loss at the terminus. Waterline melting is also an important rate-controlling process for calving at Fjallsjokull. Precise quantification of melt rates (subaerial, waterline and subaqueous) at the termini of calving glaciers is difficult and hazardous, but this study has demonstrated the value of two techniques: (1) detailed survey of melt notch growth, and (2) use of a radiocontrolled boat to record water temperatures at the ice-water interface. Continuous automated monitoring showed that lake-level fluctuations are integral to calving behaviour, influencing calving event timing and size over diurnal and hourly timescales. Fjallsjokull is sensitive to climatic forcing whereas Glaciar Leon, which exhibits larger seasonal than annual fluctuations, is less sensitive. Additional controls on calving at both sites are (1) buoyancy, (2) longitudinal stretching, and (3) the force balance at the ice-water interface. Calving operates along a continuum defined by the relative importance of interacting calving mechanisms, to which the climatic response of calving glaciers is sensitive.

This work examines the use of pedo-properties to identify dune soil system responses to environmental change on the Sefton coast, based on the development of conceptual pedogenic models. Previous environmental change and shoreline dynamics are determined through O.S. maps and aerial photographs, while present day processes are investigated through a dune-toe photographic survey and seasonal monitoring by fixed point photography. Topsoil (0-5 cm) physico-chemical characteristics are presented in a series of baseline GIS maps, displaying spatial pedo-property variation across the dune landscape. Combined with vegetation data, topsoil analysis identifies 10 distinct pedo-environments. Physico-chemical characteristics of associated National Soil Resources Institute (NSRI) soil profile classifications and an exposed stratigraphic section are presented graphically in a proposed sequence of development. Topsoil and soil profile samples are analysed for soil pH, soil organic matter (SOM) content, particle size, geochemical composition and mineral magnetism. Significant differences (p <0.05) are apparent for the suite of topsoil characteristics collated, indicating discrete dune environments are influenced by specific soil properties. Distinct down-profile variations in soil characteristics are also apparent between dune environments, highlighting pedological dynamism. Multivariate Factor analysis groups bare sand and mobile dune communities into ‘frontal dunes’ and fixed dune community, pasture, scrub, deciduous woodland and coniferous plantations into ‘hind dunes’, separating these topsoil environments from heath and slack communities. Factor analysis also identifies linkages between pedo-characteristics within soil profile horizons, suggesting pedogenesis on the Sefton dunes initiates as raw sand, progressing to sand-pararendzinas through leaching of nutrients. Desalinization and decalcification processes lead to brown earth development, followed by increased acidicification, subsequently, resulting in micro-podzol formation. Groundwater gley soils are associated with dune slacks, where drainage is inhibited and anaerobic conditions prevail. Analysis of buried soils suggests such pedo-environment formations are cyclic, responding to phases of shoreline regression/transgression, dune activity and stabilization. Conceptual models are designed to graphically demonstrate pedogenesis under both erosion and deposition regimes on the Sefton coast. Regression equations and correlation coefficients between pedo-properties and distance from mean high water are used as a proxy for soil age, which represent lateral soil maturity from the unstable frontal dunes to the stable hind dunes inland. The models simulate formation and process of the full array of soil properties, accounting for geomorphological impacts and anthropogenic influences. This has great implications for dune managers by raising awareness of pedogenesis as an integral part of nature and associated habitats, which could be incorporated in future shoreline management plans (SMPs).

Located beyond the southern limit of glaciation in Britain, the upland granitic terrain of Dartmoor, south-west England, has been exposed to long intervals of intense periglacial activity during the Pleistocene. This region has been significant in debates about appropriate models of long-term landscape change, most notably two-phase versus single-phase models of landform evolution, and the development of tors. However, given the previous lack of quantitative techniques capable of constraining denudation and specific process rates, and thereby testing developmental models for these features, there remains much uncertainty in the interpretation of the classic landforms of the region. This study measures concentrations of the cosmogenic nuclide 10Be produced in-situ in quartz within the upper few metres of the Earth surface. These reflect the history of near-surface exposure to cosmic radiation of sampled material, and allow for the interpretation of exposure age and/or erosion rates of the land surface. This research utilises these cosmogenic nuclide values to evaluate geomorphological processes and investigate key aspects of landform development. These include the formation of tors in non-glaciated regions, the development of regolith and boulderfields under periglacial conditions, and the derivation of catchment-averaged denudation rates. This study provides the first quantitative measurement of erosion on tor surfaces in Dartmoor, with typical rates of 14-45 mm ka-1. These are relatively high and comparable to other components of the landscape. In addition, there is no clear relationship of cosmogenic nuclide concentration to tor dimensions. It is shown that the tors are dynamic landforms and simple, two-stage development is an inappropriate model. Catchment-averaged denudation rates are derived and these long-term rates of 20-94 mm ka-1 are significantly higher than modern, short-term values. Finally, downslope transport in a palaeo-periglacial blockslope is investigated using 10Be concentrations. This variety of landforms and scale of investigation facilitates an integrated approach to the understanding of catchment-scale erosional dynamics. In addition, the complex nature of landform development that is evident in the area provides challenges to the application of in-situ cosmogenic nuclides and highlights both the potential and limitations of the technique.

The Geropotamos aquifer on the north‐central coast of Crete, Greece, is invaded in some places by salt water from the Aegean Sea, with impact on freshwater supplies for domestic and business uses, including agriculture. The geological setting of the study area is considered complex, as Miocene biogenic limestones, marls, clays and conglomerates crop out in the central and the western part and clastic limestones and dolomites of the Tripolis and Plattenkalk nappe (the bedrock) in the eastern part of the study area. The phyllitequartzite nappe (which forms the oldest rock of the study area) lays on the northern part of Geropotamos basin. The local tectonic regime of the study area is characterized by faults of NW‐SE and NE‐SW directions. Investigation of the aquifer using Transient ElectroMagnetic method (TEM) and Vertical Electrical Resistivity (VES) measurement technique has resulted in 1D models and 2D/3D imaging of geoelectric structures, depicting the zones of salination of groundwater in the aquifer. 1179 TEM soundings in 372 sites have been carried out in a detailed survey grid (about 200m in X and Y dimension) and 3 VES soundings were acquired in three different sites (different geological conditions). For the 2 of them, multidirectional measurements were also acquired since the structure is more complex than a 1D model that VES technique is able to model. Moreover, 3 water samplings carried out. At each sampling, samples from 22 boreholes and 2 springs were analysed and 16 chemical parameters were determined. Detailed geochemical analysis, including Piper, Durov, Ternary, Stiff, Wilcox, Dispersion diagrams and Factors controlling the groundwater quality, was accomplished showing very good results and the relationship with the geophysical methods. All data were inserted in GIS environment and Groundwater Quality Maps were produced. Furthermore, Remote Sensing application, geological mapping and hydro‐lithological data showed that the physical characteristics of geomorphology and geology are in great relationship with the chemical and geophysical properties as well. Suggestions that Miocene evaporites led to groundwater salination are unconfirmed, and seawater intrusion is the most probable cause, supported by the results of this research. It is indicated that saline intrusion is likely to occur along fractures in a fault zone through otherwise low‐permeability phyllite‐quartzite bedrock, and it is emphasized the critical role of fracture pathways in salination problems of coastal aquifers.

This thesis considers the use of Interferometric Synthetic Aperture Radar (InSAR) for surveying temporal fluctuations in the velocity of glaciers in the Arctic region. The aim of this thesis is to gain a broader understanding of the manner in which the flow of both land- and marine-terminating glaciers varies over time, and to asses the ability of InSAR to resolve flow changes over timescales which provide useful information about the physical processes that control them. InSAR makes use of the electromagnetic phase difference between successive SAR images to produce interference patterns (interferograms) which contain information on the topography and motion of the Earth's surface in the direction of the radar line-of-sight. We apply established InSAR techniques (Goldstein et al., 1993) to (i) the 925 km2 LangjÖkull Ice Cap (LIC) in Iceland, which terminates on land (ii) the 8 500 km2 Flade Isblink Icecap (FIIC) in Northeast Greenland which has both land- and marine-terminating glaciers and (iii) to a 7 000 km2 land-terminating sector of the Western Greenland Ice Sheet (GrIS). It is found that these three regions exhibit velocity variations over contrasting timescales. At the LIC, we use an existing ice surface elevation model and dual-look SAR data acquired by the European Remote Sensing (ERS) satellite to estimate ice velocity (Joughin et al., 1998) during late-February in 1994. A comparison with direct velocity measurements determined by global positioning system (GPS) sensors during the summer of 2001 shows agreement (r2 = 0.86), suggesting that the LIC exhibits moderate seasonal and inter-annual variations in ice flow. At the FIIC, we difference pairs of interferograms (Kwok and Fahnestock, 1996) formed using ERS SAR data acquired between 15th August 1995 and 3rd February 1996 to estimate ice velocity on four separate days. We observe that the flow of 5 of the 8 outlet glaciers varies in latesummer compared with winter, although flow speeds vary by up to 20 % over a 10 day period in August 1995. At the GrIS, we use InSAR (Joughin et al., 1996) and ERS SAR data to reveal a detailed pattern of seasonal velocity variations, with ice speeds in latesummer up to three times greater than wintertime rates. We show that the degree of seasonal speedup is spatially variable and correlated with modeled runoff, suggesting that seasonal velocity changes are controlled by the routing of water melted at the ice sheet surface. The overall conclusion of this work is that the technique of InSAR can provide useful information on fluctuations in ice speed across a range of timescales. Although some ice masses exhibit little or no temporal flow variability, others show marked inter-annual, seasonal and even daily variations in speed. We observe variations in seasonality in ice flow over distances of ~ 10 km and over time periods of ~10 days during late-summer. With the aid of ancillary meteorological data, we are able to establish that rates of flow in western Greenland are strongly moderated by the degree of surface melting, which varies seasonally and secularly. Although the sampling of our data is insufficiently frequent and spans too brief a period for us to derive a general relationship between climate and seasonality of flow, we show that production of meltwater at the ice surface and its delivery to the ice bed play an important role in the modulation of horizontal flow speeds. We suggest that a similarly detailed investigation of other ice masses is required to reduce the uncertainty in predictions of the future Arctic land-ice contribution to sea level in a warming world.

Turkey is widely accepted to have formed from a collage of microcontinents that rifted from the northern margin of Gondwana and assembled from the Mesozoic to Mid Cenozoic in response to the closure, collision and suturing of numerous oceanic strands in the Eastern Mediterranean. Sedimentary-tectonic basins, which formed during ocean basin closure, can yield important information about the evolution, timing and processes related to the closure of these oceanic strands. The Darende Basin and the adjacent Hekimhan Basin are two sedimentary-tectonic basins which developed during the collision and suturing of the Neotethys Ocean in the Eastern Mediterranean. The Darende and Hekimhan Basins developed as part of the northern margin of the Tauride microcontinent during the collision and suturing of Neotethys. Both basins exhibit a Jurassic to Cretaceous regional carbonate platform 'basement' overlain by a dismembered ophiolite, which was emplaced southwards during the Late Cretaceous. The basins then developed in two main phases: In the Darende Basin the first phase is characterised by non-marine clastic sediments, overlain by transgressive shallow-marine rocks. In the Hekimhan Basin, hemi-pelagic facies are deposited synchronously with the eruption of within plate-type alkaline basaltic-trachytic lavas and associated volcaniclastic sediments (later intruded by a syenitic pluton) under an extensional tectonic regime. A Paleocene-aged unconformity followed. A second phase of basin evolution during the Eocene is characterised in both basins by the deposition of variable sedimentary facies including conglomerate, sandstone, marl, shallow-marine nummulitic limestone and evaporites (and localised basaltic eruptions). These record successive deepening, shallowing and finally emergence of both basins during the Late Eocene. The Oligocene is represented by continental fluvial deposits that are only exposed in the Hekimhan Basin. The deposition of faunally diverse, shallow-marine, Miocene limestones, Pliocene subaerial basalts and Pliocene-Recent continental deposits in both basins completes the sequence. The following tectonically and eustatically controlled stages of basin development are inferred: 1) Late Cretaceous extension initiated basin development (after ophiolite emplacement), possibly related to immediate isostatic compensation and on-going slabpull during northward subduction of the remaining Neotethyan oceanic crust. The eruption of within-plate lavas and the intrusion of alkaline syenite bodies in the Hekimhan Basin reflect this extensional setting; 2) Emergence of the Darende and Hekimhan Basins in the latest Cretaceous was possibly controlled by regional flexural uplift as the down-going plate approached the subduction zone to the north (and was possibly also influenced by eustatic sea-level change); 3) Early Eocene flexural subsidence related to ‘soft collision’ of the Tauride microcontinent with Eurasia, coupled with a significant eustatic sea level rise, allowed sedimentation to resume; 4) Mid-Late Eocene ‘hard collision’ resulted in regional uplift, progressive isolation and subaerial exposure of the basins; 5) Suture tightening and compression, during the Late Eocene- Miocene, resulted in reactivation of pre-existing extensional faults and terminated marine sedimentation. Both basins were affected by predominantly sinistral strike-slip faulting during the Plio-Quaternary westward tectonic escape of Anatolia.

Urbanisation is increasingly recognised as a major ecological pressure at the coast. By 2035, the Department for Environment, Food and Rural Affairs will have to spend £1 billion each year on flood defence and erosion control infrastructure if current levels of protection are to be sustained in England and Wales; this represents a substantial commitment to building new hard structures. Ecological research has shown that structures like seawalls, breakwaters, and harbour and port infrastructure are poor surrogates for undisturbed rocky shores. This, alongside substantial international policy drivers, has led to an interest in the ways in which structures might be enhanced for ecological gain. Virtually all of this research has been undertaken by ecologists, while the contribution of geomorphological understanding has not been fully recognised. This thesis presents an assessment of the two-way interactions between colonising organisms and the materials used to build hard coastal structures under a framework of biogeomorphology. The influence of material type and small-scale surface texture on early colonisation is assessed alongside detailed observations of the ways in which biota are involved in the alteration of substratum properties and behaviours through weathering and erosion in the intertidal zone. The research demonstrates that biotic (organisms) and abiotic (material substrata) components of coastal structures are inherently linked at various spatial and temporal scales through complex biogeomorphic interactions and feedbacks. Importantly, these interactions have consequences for the subsequent operation of ecological and geomorphological processes that are of relevance to urban marine ecology, weathering and rock coast geomorphology, and engineering. This thesis demonstrates the considerable potential to manipulate substratum-biota interactions on artificial structures for ecological gain, both directly and indirectly. More broadly, the explicitly interdisciplinary methodological approach adopted shows the value and necessity of integrated research for achieving useful, applied outcomes.

This thesis describes volcanic records from the Andean southern volcanic zone, based on the collection of field data between Calbuco and Puyuhuapi volcanic centres, with a particular focus on the Hualaihue peninsula, combined with existing records from the region as a whole. These data, extending the understanding of the volcanic history of southern Chile, are examined for evidence of spatial or temporal variability, which may be used to explore underlying controls on volcanic processes. All three volcanoes on the Hualaihue peninsula have been active in the Holocene. A large mafic scoria unit from Apagado is unusually primitive, providing a potential window into primary magma generation in the arc. Dynamically similar eruptions occurred at Hornopirén and widely along the regional scale Liquiñe-Ofqui fault zone (LOFZ). Although the Hualaihue centres are closely related, petrological evidence indicates a complex magmatic storage system. Effusive activity is predominant at Yate and Hornopirén, and the tephrostratigraphy of the Hualaihue area is dominated by units from Calbuco volcano, to the north. The 2008 eruption of Chaitén provided an analogue for past large explosive eruptions in the region, with tephra deposition reflecting variable eruption intensity in a changing wind field. The regional tectonic setting and the LOFZ influence dyke ascent, volcano morphology and, as demonstrated at Yate, edifice stability, determining the orientation of collapse. Explosive eruption records over the post-glacial period also indicate a limited response of volcanism to deglaciation, suggesting a control on magma storage arising from changing crustal stress regimes, both at the arc front and along the LOFZ. On short timescales, large earthquakes are shown to influence eruption rate across the arc, implying a triggering role for dynamic seismic stresses. This work demonstrates the existence of a range of external forces affecting Chilean arc volcanism, but the degree to which these are quantifiable is strongly constrained by the quality of the available data.

The Lewisian complex of northwest Scotland contains the oldest known rocks of the British Isles and is commonly cited as a classical example of a high-grade gneiss terrane. The grey gneisses consist of a suite of tonalite, trondhjemite and granodiorite (TTG) rocks. The mainland Lewisian is divided into three - the northern, central and southern regions. The central region is known to have experienced a high-grade metamorphic event at 2490 Ma but may also have been affected by earlier events. The northern and southern regions only attained amphibolite grade. LILE depletion suggests that the Lewisian gneisses have partially melted.Additionally, published geothermometric and geobarometric estimates far exceed conditions required for anatexis of metatonalites, even in the absence of a fluid phase. There is also a suite of 'early' granite, trondhjemite and tonalite sheets (some of which contain garnet) in higher-grade parts of the complex that have been proposed as the products of this anatexis. A further suite of granitic rocks (late granite sheets) commonly occurs in the northern region. A series of fluid-absent, partial-melting experiments was carried out using three amphibolite-facies tonalitic and dioritic starting materials at P-T conditions thought to have been equivalent to the granulite-facies metamorphism (0.8-1.2 GPa and 800-1000 °C) in order to investigate the origins of these felsic sheets and the compositions of such partial melts. A series of H[sub]2 O-saturated experiments was carried out at temperatures around the wet solidus and 0.6 GPa. A further two experiments were carried out using a NaCI-H[sub]2 O fluid at 0.6 GPa to determine the affects that such a fluid would have at high grades. These experiments have been carried out in conjunction with petrographic studies into retrogression in the central region gneisses and a geochemical investigation into the compositions many rock types typical of the Lewisian. The experiments have produced thermometric estimates for the major granulite-facies and partial-melt-forming event of 950 ± 50 °c. Garnet stability in the restite leads to an estimate of maximum P of 1.1 ± 0.1 GPa. If the origin of the garnet in some of the early sheets can be shown to be the solid product of biotite breakdown then 0.9 < P < 1.1 GPa. For the Gruinard Bay area P < 0.9 GPa. The experiments and geochemistry show that it is unlikely that the early sheets are the products of partial melting of tonalitic protoliths. The limited extent of pristine granulites and the very high temperatures shown to have been attained suggest that the granulite-facies metamorphism was not a truly regional event and that temperature was locally higher where granulite-facies assemblages are preserved, probably in response to the introduction of mafic material. The late sheets are shown not to be the products of H[sub]2 O-saturated partial melting of tonalitic material similar to that exposed at the surface and it is likely that a layer containing significant quantities of K-feldspar lies beneath the northern region.

This thesis assesses seasonal hydrological prediction in Great Britain. Firstly, the study evaluates river flow prediction using climate model output to drive a rainfall-runoff model in the Dyfi basin, Wales. Results show that climate model precipitation can not skilfully simulate Dyfi discharge. When a downscaling process is employed to generate precipitation time series, river flow forecast skill improves, but historical river flows still provide superior forecasts. Secondly, large-scale climatic control on British precipitation/discharge and European precipitation is investigated by correlation analysis. Results show spatiotemporal hydroclimatological variation, with western regions generally having stronger empirical relationships. River flow has weaker associations because of basin controls and evapotranspiration. The dynamic nature of precipitation/discharge generating mechanisms is not captured by the North Atlantic Oscillation Index. Thirdly, seasonal climate model forecast skill is evaluated. Limited skill exists over land and over all extratropical regions for forecasts beyond month-1; precipitation has lower skill than 2-metre air temperature and mean sea level pressure. Seasonal climate models exhibit higher idealised predictive skill indicating potential for future increases in actual predictive skill. In conclusion, seasonal hydrological prediction using a climate-to-river modelling chain could be improved through consideration of the uncovered spatiotemporal hydroclimatological variability and through seasonal climate modelling improvements.