The development and implementation of a totally predictive model for the annual thermal structure of a freshwater lake or reservoir is described in detail. Water quality depends to a great degree on dissolved oxygen profiles and nutrient concentrations and is thus governed by the temperature profile in the lake. The numerical model described here discusses fully this latter problem and also indicates strongly the direction for further research into the introduction of comprehensive biochemical cycles to improve this calculation. All possible variations in external and internal parameters are included. These values can be observed for a specific lake and future behaviour predicted on the basis of them. For an unbuilt reservoir, mean values can be taken for the parameters that cannot otherwise be determined from climatological, meteorological or geophysical sources. The transport mechanisms responsible for the development of the thermal profile during the year are determined completely by those parameters. The numerical representation is based on the assumption of horizontal homogeneity so that the one-dimensional heat transfer equation can be solved using a finite difference grid and a forward time step of one day. Decreasing the time step and modifying daily mean values to allow for diurnal variation in solar elevation permits the model to be used over a shorter time scale. The surface energy budget and wind speed (which it is found must be modified for lakes of small surface area) are the main forcing functions for the vertical mixing. It is shown that the average annual temperature structure of the lake is stable over a period of many years irrespective of the initial conditions imposed on the temperature profile. The problems of validating this model 'climate' against the observations of a single year (termed the lake 'weather' ) are evident.

An examination of the hydrological functioning of a wet grassland in southeast England is undertaken within the context of scale issues. The importance of wetland hydrological functioning is demonstrated, alongside highlighting a gap in the literature of how scale issues affect contemporary wetland hydrological science. The subsequent assessment of hydrological functioning at a field site at Otmoor, Oxfordshire, is thus undertaken at both the field and catchment scales. At the field scale, an intensive field campaign over 18 months establishes the dominant hydrological processes as being precipitation and evaporation, the latter with losses of up to 5 mmday1 driving an unexpected diurnal fluctuation in soil water levels. The dominant hydrological function was surface water storage, showing potential conflict with current land management practice of raising water levels for wetland restoration purposes. The impact of scale issues on numerical models is assessed through utilisation of a multi-scale model. At the catchment scale, the wetland's impact was assessed through increasingly complex numerical models, ultimately an event-based hydrodynamic model, and is shown to be significant to flood management downstream at Oxford. Decreasing peak flood flows through flood storage was the dominant function, as dictated by surface topography, whereas other online floodplain areas within the catchment increase time to flood peaks by attenuating flow through surface roughness, confirming the importance of wetlands to river flow. Surface roughness was therefore shown to be critical for wetland behaviour at different scales for different wetland types, indicating the importance of scale to wetland hydrological processes. The significance of initial model conceptualisation was demonstrated, and several recommendations were made for modelling procedures in order that scale issues be incorporated and prevented from causing complications in future modelling work. These include taking an iterative approach to increasing understanding through modelling, and linking models of different scales.

The extensive glacial landforms in the Baltic States and neighbouring countries have been used to infer the dynamic behaviour of the Scandinavian ice sheet. While pre-Weichselian phases of ice flow were identified, it was concluded that the majority of lineations within the Eastern Baltic formed during the Late Weichselian. Long (up to 21 km), well-defined lineations were found to have formed during the Late Weichselian maximum when the ice velocities were greatest. These form flow sets with a north-south trend. Lineations from the final deglaciation are shorter in length and form flow sets orientated at 170°. During the final deglaciation ice streams developed. Interstream areas generally coincide with regions of elevated bed rock. The interpretations resulting from these observations were combined with similar data from Finland to create a data set covering the area from the ice divide to beyond the limit of Weichselian ice. The spatial distribution of lineation size was examined using this data set and compared to output from a glaciological model. It was concluded that flow-parallel lineations were most likely formed by a single mechanism. The most likely mechanism was concluded to be subglacial deformation with the most active zone of lineation formation occurring within 100 km of the ice sheet margin. Variations in the frequency distribution of lineation length between Finland and the Eastern Baltic regions point to differences in the controls on lineation formation. Differences in the character of deglaciation between Finland and the Eastern Baltic were identified. Ice streams in Finland appear to have been more stable in location and about twice the width of those observed in the Eastern Baltic region. These differences may be explained by the interaction between the ice sheet and its substratum.

Reduced Impact Logging (RIL) introduced in some places has been reported to improve harvesting efficiency and environmental impact e.g. better design of skidtrails, less waste wood, reduces forest damage and cost. In this study, the effect of RIL on hydrological aspects was examined to assess its impact on direct runoff, streamflow, suspended sediment and unit hydrograph of the catchment, comparing RIL with conventional logging (CL) and an unlogged (UL) control. Compared to the control, the ratio of water yield to rainfall increased by about 53% with RIL and 98% with CL treatment after a period of 3 to 4 months following logging. Direct runoff with RIL remained similar to the control but increased by about 140% with CL treatment. Compared to CL treatment, the mean maximum suspended sediment concentration (Cmax.) levels can be reduced by more than six times and the daily suspended sediment load (S) by almost half if RIL treatment is implemented. The ratio of Cmax and S under UL, RIL and CL treatments are about 1 : 5 : 32 and 1 : 6 : 11 respectively. The main cause of erosion and sedimentation was soil damage due to compaction by harvesting machinery and soil cover reduction. RIL treatment is capable of reducing the runoff coefficient about half than that of CL: mean runoff coefficients for UL, RIL, and CL treatments are 0.027, 0.045 and 0.095 respectively. Unit hydrograph models for the two replications of each treatment showed shape consistency (e.g. time to peak is shorter with UL than RIL and CL treatments) and could be used to predict direct runoff from the related catchments. Finally, the study showed that supervision of harvesting - including planning and post logging treatment has an important role in reducing logging impact on hydrological behaviour.

The localisation of erosional denudation represents the principal mechanism through which compressional mountains systems are exhumed and is a fundamental parameter in their mechanical and geomorphological development. Exhumation gradients within the central Pyrenean interior have been identified by synchronous periods of exhumation and sedimentation. There is no structural or metamorphic evidence to indicate that tectonic denudation in this region has occurred, hence, the distribution of exhumation was principally controlled through spatially heterogeneous surface erosion. The controlling parameters on the spatial variability of erosion in the Pyrenees and the implications for the developing mountain belt are, to date, poorly understood. In this study, spatial and temporal gradients in erosion-driven exhumation within the central Pyrenees are quantified and the mechanisms through which they were generated are investigated. To this aim, a multidisciplinary approach is adopted. Low temperature thermochronology (apatite (U-Th)/He, apatite fission track and zircon fission track) is applied to constrain time-temperature histories and to characterise the regional distribution of exhumation during active orogenesis and during the post-orogenic period. Geological field data from the orogenic interior and form adjacent sedimentary basins is used to provide constraints on the tectonic evolution of the area. Results indicate the history of erosional denudation during and after orogenesis is considerably more complex than previous models envisaged and is governed by a number of discrete physical mechanisms. During active orogenesis, exhumation was controlled by the strongly spatially variable distribution of tectonically-forced rock uplift and the associated control on erosion. During the post-orogenic period, spatial variability in exhumation was maintained, but was principally governed by externally-forced shifts in the evolution and denudation of the Pyrenean landscape.

This study focused on the associations of U and Pu with humic substances and the implications for migration in deciduous broadleaf and coniferous forest soils. Humic substances were isolated from soil samples by (i) solvent extraction (0.045 M Tris-borate and 0.1 N NaOH) and (ii) novel gel electrophoretic extraction (pH 8.5; 0.045 M Tris-borate as running buffer). Gel electrophoresis was also used to extract humic substances from soil that had been 'pre-equilibrated' with a solution containing ²³⁶U or ²⁴²Pu (pH 4.2). The relative concentrations of humic substances in the solvent and electrophoretic extracts were obtained by UV-visible spectrophotometry (A₂₅₄ _nm) whilst the ²³⁸U concentrations in the extracts were determined by inductively coupled plasma-mass spectrometry (ICP-MS). In addition, the molecular size and functional group content of the humic molecules in selected fractions of the gel electrophoretic extracts were investigated using gel filtration chromatography and FTIR spectroscopy, respectively. The methodology developed in this study enabled identification of several operationally defined soil actinide pools: (i) labile (< 0.2 mm, at pH 4.2), (ii) hydrophilic-mobile (< 0.15 mm, negatively-charged at pH 8.5) and (iii) immobile (> 0.15 mm and/or uncharged at pH 8.5). The adopted procedures avoid the extremes of pH used in the traditional methods of extraction (pH 13) and fractionation (pH < 2) which may alter humic substances. It was found that the ²³⁸U content of the hydrophilic-mobile pool (i.e. the gel electrophoretic extract) accounted for 14-41 % and 15-34 % of the total ²³⁸U in the deciduous broadleaf and coniferous forest soils, respectively. In general, these values were grater than those obtained by 0.045 M Tris-borate solvent extraction but less than those obtained using 0.1 N NaOH as the solvent. A further advantage of the new methodology was that the distribution of ²³⁸U within the humic extract could be investigated.

No description available.

Avalanche release is a brittle process meaning that fracture toughness rather than shear strength may be the fundamental parameter controlling slope stability. The results of a series of experiments to measure the fracture toughness of soft slab are reported. Snow was found to be toughest in mode I. The mode II and mode III values are approximately equal, and about 70% of the mode I. The mode II and mode III values are approximately equal, and about 70% of the mode I value. The fracture toughness of a weak layer was found to be about 1 kPa m^1/2. In addition to shear strength and/or fracture toughness, the coefficient of basal friction between the slab and the substrate is of obvious importance to slab avalanche release. Various sources suggest that the coefficient of friction for dry slab is quite variable, but a value of about 0.6 seems quite typical. Incorporating this and other typical snow parameters into shear band models gives a critical crack size of the order of ten metres. This is much too large to be produced by a skier. We discuss two alternative release mechanisms for skier triggered avalanches. The first of these is the interaction between local skier damage and the complex pattern of internal shear cracks assumed to lead to natural failure. The other mechanism is the whumpf of flexural wave, which is caused by compressive collapse and subsequent loss of shear strength of the weak layer. The relative importance of these two processes combined with highly variable basal friction can explain the wide variety of different behaviour associated with skier triggered avalanches. Examples of such behaviour are: a skier triggering a well tracked slope, slopes that whumpf (and sometimes can be felt to move and crack) but do not avalanche, and remotely triggered avalanches.

This thesis studies geotechnical properties of former subglacial sediments to establish the conditions found at the base of an ice sheet. Whilst many advances have been made in the theoretical analysis of ice sheet dynamics, little quantitative information exists about the physical processes that occur beneath an ice sheet. These processes depend on the type of surface over which the ice flows and on whether the base of the ice is melting. Until recently, most models of ice sheet flow have assumed that ice moves over a rigid surface as a result of deformation within the ice, or by sliding at its base. It has subsequently become clear that where ice overlies unlithified sediments they can deform, thus contributing to the forward motion of the ice. The basal boundary is difficult to observe directly, however information can be obtained directly by studying the sediments left behind when the ice has retreated. An advancing ice sheet loads the sediments over which it flows. Sediments consolidate in response to this additional load, providing that water can drain from the sediment pores. Sediments underlying a basally melting ice sheet must still allow melt water to drain through them and thus will not consolidate to the same extent as a non-glacial sediment under the same load. The sediments expand on retreat of the ice however they retain a record of the maximum pressure to which they were subjected, known as the preconsolidation pressure. The preconsolidation pressure is assumed to have occurred at the glacial maximum when the ice sheet was in a steady state, however the existence of a meltwater flux means that the preconsolidation pressure preserved is not a simple consequence of the overburden of ice and sediments. A one dimensional drainage model is adopted to investigate subglacial groundwater flow. Melt water is assumed to drain vertically through an upper layer of low permeability sediments into an aquifer. The theoretical distribution of pressure in the one dimensional model is derived and a relationship between the effective pressure gradient, potential pressure gradient and gravitational gradient established.

This thesis is concerned with accelerated soil erosion in the Lammermuir Hills in South-east Scotland. It reviews the literature on erosion in upland Britain and presents the controversy surrounding this issue in southern Scotland. Remote-sensing procedures are used to examine the extent of erosion phenomena in the study area. Possible associations between erosion occurrence and certain soil types and land uses are analysed. Historical documents, maps and sequential aerial photographs are used to investigate possible changes in the size and distribution of erosion features over time. The results of further work on erosion rates obtained by field monitoring in various parts of the Lammermuira are also presented and analysed. These studies show that erosion is much less of a problem than the literature suggests. Only 1.4% of the area studied was found to be occupied by erosion forms and only 0.8% of the area appeared to be actively eroding. The rates of erosion measured in the field were found to be low and in keeping with findings obtained in other parts of upland Britain. The results also suggest associations between erosion occurrence and certain land-uses and soil-series. The possibility of using range management techniques to produce more definite information about land-use/erosion relationships is explored. The results suggest that while they are not eroding very much at present, some parts of the Lammermuirs are overburned and over-occupied by sheep and could well suffer from erosion problems in the future. The study confirms the usefulness of the standard methods of investigating erosion in upland areas, such as aerial photography mapping and erosion pin monitoring. It also suggests that the hitherto little-used range management approach has considerable promise for appraising erosion in those areas of upland Britain where rapid results are required.