The aims of this work are to improve the understanding of the way in which a river system approaches equilibrium during and after sediment transporting events and by using experimental data and hypothetical models to gain an insight into mixed grain sized transport and its connection with local bed morphology. Large scale flume experiments were undertaken in a trapezoidal channel where a mixed grain sized river bed is subjected to a variety of imposed sediment loads. The river system was allowed to adjust naturally to the imposed conditions while flow and bed conditions were frequently monitored. Changes to both flow and bed occur at many scales and the superimposition of these changes adds to the complexity of the interpretation of hydraulic and bed data. Three experiments are reported in which the sediment influx differ. One in which no material is added. One in which a feed rate is established equal in magnitude to 50% of the initial transport rate for the equivalent non-fed case. In the third experiment the feed rate is doubled to 100% of the initial transport rate in the degradation experiment. The imposed load is continued for approximately 75% of the duration of the run or until a dynamic equilibrium transport rate is evident within the reach. The river channel is analysed at two distinct scales; grain scale and reach scale. Conclusions about the relative influence of grain and form effects are drawn on the basis of detailed surface texture information and reach scale measurements of the bed surface. A hypothetical model is proposed to show that the transport activity during a degradation experiment can be qualitatively linked to a wave like translation of grain dislodgement through the channel.

No description available.

Most documented investigations of the effects of land-use change on hydrological systems have considered the modification of forest areas. In this thesis, a headwater area in the North York Moors is used to examine the consequences of maintaining a land management regime which has received comparatively little observation in this context: controlled heather burning (muirburn). The effects of coniferous afforestation are also evaluated for selected variables. Particular attention is given to the responses of soil, moisture and evapotranspiration and the relationship between these two components.Simulated soil moisture deficits derived from empirical models are tested against measured values. Predictions based on Penman-Monteith evapotranspiration and 'layer' moisture deficits, along with an optimised soil-drying parameter, were found to simulate observed conditions most closely. A land-use change from open heather moorland to burnt ground promoted reductions both in evapotranspiration levels, especially at potential demando and in moisture deficits. In contrast, following afforestation, deficits were maintained or enhanced throughout the year, with higher moisture losses to interception than found under heather, due to the higher aerodynamic resistance of the latter. Predictions of actual evapotranspiration, determined from soil moisture models, were generally found to be reliable estimates of those 'observed' from the moorland water balance.Antecedent catchment conditions and storm characteristics were used in analysis of runoff distribution over time, quantified in terms of 'unit hydrographs' and linear regression models. Land-use effects were manifested most significantly in a doubling of hydrograph peak discharge following muirburn, the lower measured soil moisture deficits under a burnt catchment rendering more water available for storm runoff. A secondary, underlying control, that of a slower response from a wet catchment, lent 'support to evidence for the existence of variable source areas.

No description available.

Streams draining coniferous plantations contain higher loadings of hydrogen ion, aluminium, sulphate and nitrate, in comparison with streams in adjacent grasslands. Almost all of this ion-load is transported to streams via subsurface water-pathways. An incontrovertible, physical characterization of these pathways within a natural, layered hillslope, has yet to be presented. This research has sought to provide such a characterization for two hillslopes - one afforested with conifers, the other an improved grassland. Much of the uncertainty associated with the identification of soil-water-pathways stems from an inadequate characterization of the errors imposed by the use of each measurement technique. This research has, therefore, compared the predictions of a number of quasi-independent field and analytical techniques, to attempt to lessen the impact of measurement error upon the observed response of the true hydrological system. The impact of conifers upon the detailed water-pathways and lumped catchment response was monitored to educe any changes in the hydrological response which could account for the increased loading of acidic solutes within forest streams. The results of the analysis, indicated that the pathways of water through hillslopes could be predicted from the response of hydrological properties averaged over control volumes of soil-pores. The accuracy of these solutions was proven by the concordance of the response of all of the properties contained within the Darcy-Richards equation. The marked horizon development within the ferric podzol soil of the instrumented forest hillslope, in particular the presence of an indurated B horizon, deflects most percolation laterally within the 0/A and A/E horizons. This pathway was indicted by the results of techniques which included numerical and approximative calculations, discontinuities between the state-dependent hydraulic conductivity of each soil horizon, and the generation of steep, vertical potential gradients in layered porous media. The instrumented grassland hillslope was ploughed 11 years prior to instrumentation. This greatly increased the conductivity of the controlling B horizon, allowing almost all flow to percolate to depth. During winter-storms, the forest hillslope generated flows smaller than those within the grassland hillslope, concomitant with the 29 percent difference in the rainfall-runoff behaviour of the catchment areas. This increased loss of runoff within the afforested areas, may result from the high losses of wetted-canopy-evaporation (39 percent of gross-precipitation) from the Sitka spruce (Picea sitchensis. Bong. Carr.) trees. Individual conifer trees growing on the steep, ferric podzol hillslope appeared to enhance the lateral deflection of flow within the O/A and A/E horizons, probably as a result of their platy root systems, and the high rates of precipitation input to soil at the stem-base. The enhancement of both lateral near-surface flow and below-canopy ion concentrations could, therefore, generate the chemical signatures characteristic of streams draining coniferous forests.

No description available.

No description available.

No description available.

No description available.

Spatial and temporal variations in channel morphology, near-bed velocity, shear stress, bedload transport rate, pebble tracer movement, and bedload and bed material size distribution were measured in seven different channel patterns in two gravel-bed rivers in the Scottish Highlands (the Dubhaig and Feshie) and a proglacial stream in Norway (the Lyngsdalselva). The results showed that there were discernible links between the channel processes and changes which were consistent for all river types. 169 shear stress estimates from velocity profiles with changing discharge showed that Keller's (1971) velocity-reversal hypothesis holds true in different channel patterns of gravel-bed rivers and can be extended to include subunits of the pool/riffle cycle. At discharges near bankfull there is a decrease in the flow strength and amount of bedload movement from the poolhead down to the pooltail (and then riffle). On a broader scale 72 Helley-Smith bedload samples and the movement of over 3700 pebble tracers showed that the entrainment of different size fractions from heterogeneous bed material is inefficient and is overpredicted by the traditional bedload transport equations. Empirical analyses showed that when the armour is mobile/broken large and small particles have almost equal mobility as first proposed by Parker et al. (1982) and Andrews (1983). However for the majority of flow conditions the armour is static and entrainment is selective to a greater or lesser degree depending on the availability of appropriate-sized sediment at the surface and from bank erosion. The magnitude and direction of flow strength and bedload transport helps to explain the location and mode of channel development as revealed by repeated levelling and mapping. The accelerating convergent/decelerating divergent cells of flow alter the channel morphology in predictable ways. The positions of these cells can change with increasing discharge as the channel becomes generally, rather than locally, competent to move coarse sediment. The rates of bank erosion and volumetric scour and fill decreased from the active multi-braided system through to the stable straight channel type.