|dc.description.abstract||Upland rivers control the large-scale topographic form of mountain belts, allow coupling of climate and tectonics at the earth’s surface and are responsible for large scale redistribution of sediment from source areas to sinks. However, the details of how these rivers behave when perturbed by changes to their boundary conditions are not well understood. I have used a combination of fieldwork, remotely sensed data, mathematical analysis and computer modelling to investigate the response of channels to well constrained changes in the forcings upon them, focussing in particular on the effects of glacial remoulding of the catchments draining the south flank of the Ladakh batholith, northwest Indian Himalaya. The last glacial maximum for these catchments is atypically old (~100 ka), and this allows investigation of the response to glaciation on a timescale not usually available. The geomorphology of the catchments is divided into three distinct domains on the basis of the behaviour of the trunk stream – an upper domain where the channel neither aggrades above or incises into the valley form previously carved by glacial abrasion, a middle domain where the channel incises a gorge down into glacial sediments which mantle the valley floor, and a lower domain where the channel aggrades above this postglacial sediment surface. This landscape provides a framework in which to analyze the processes and timescales of fluvial response to glacial modification. The dimensions of the gorge and the known dates of glacial retreat record a time averaged peak river incision rate of approximately 0.5 mm/y; the timescale for the river long profile to recover to a smooth, concave up form must exceed 1 Ma. These values are comparable with those from similarly sized catchments that have been transiently perturbed by changing tectonics, but have never been quoted for a glacially forced basin-scale response.
I have also demonstrated that lowering of the upper reaches of the Ladakh channel long profiles by glacial processes can systematically and nonlinearly perturb the slope-area (concavity) scaling of the channel downstream of the resulting profile convexities, or knickzones. The concavity values are elevated significantly above the expected equilibrium values of 0.3-0.6, with the magnitude controlled by the relative position of the knickzone within the catchment, and thus also by the degree of glacial modification of the fluvial system. This work also documents the existence of very similar trends in measured concavities downstream of long profile convexities in other transiently responding river systems in different tectonoclimatic settings, including those responding to changes in relative channel uplift. This previously unrecognised unity of response across a wide variety of different environments argues that such a trend is an intrinsic property of river response to perturbation. Importantly, it is consistent with the scaling expected from variation in incision efficiency driven by evolving sediment flux downstream of knickzones. The pervasive nature of this altered scaling, and its implications for fluvial erosion laws in perturbed settings, have significant consequences for efforts to interpret past changes in forcings acting on river systems from modern topography.
I follow this by examining in detail the channel hydraulics of the Ladakh streams as they incise in response to the glacial perturbation. I present a new framework under which the style of erosion of a natural channel can be characterized as either detachment- or transport-limited based upon comparison of the downstream distribution of shear stress with the resulting magnitude of incision. This framework also allows assessment of the importance of sediment flux driven effects in studied channels. This approach is then used to demonstrate that fluvial erosion and deposition in the Ladakh catchments is best modelled as a sediment flux dependent, thresholded, detachment-limited system. The exceptional quality of the incision record in this landscape enables an unprecedented calibration of the sediment flux function within this incision law for three different trunk streams. The resulting curves are not compatible with the theoretically-derived parabolic form of this relation, instead showing nonzero erosion rates at zero sediment flux, a rapid rise and peak at relative sediment fluxes of less than 0.5 and a quasi exponential decrease in erosional efficiency beyond this. The position of the erosional efficiency peak in relative sediment flux space and the magnitude of the curve are shown to be both variable between the catchments explored and also correlated with absolute sediment flux in the streams.||en