Paper No. 10
Presentation Time: 4:10 PM
A 1-D NUMERICAL MODEL OF POST-GLACIAL VALLEY-FLOOR EVOLUTION OF A LARGE CORDILLERAN RIVER VALLEY
TUNNICLIFFE, Jon, Department of Geography and Environmental Studies, Carleton University, B349 Loeb Building, 1125 Colonel By Drive, Ottawa, ON K1S 5B6, Canada and CHURCH, Michael, Department of Geography, The University of British Columbia, Vancouver, BC V6T 1Z2, Canada, jon_tunnicliffe@carleton.ca
We present a 1-dimensional model of floodplain erosion and sedimentation that integrates a diverse range of data to simulate the postglacial evolution of a large (1 230 km2) river valley in southwestern British Columbia over 13 ka. Chilliwack River is typical of many Cordilleran valleys that have undergone significant degradation of Pleisocene valley fills. Downstream controls on base-level, mainly blockage by glaciers, led to aggradation of glaciofluvial and glaciolacustrine valley fills and fan deposits. A sediment budget indicates that approximately 3.2 km3 of sediment has been eroded from the valley, and roughly 2.4 km3 of sediment has been deposited in an outlet fan (the remainder passing onward to Fraser River). Bed substrate sampling throughout the river network reveals complex trends in downstream grain-size and lithology, related to tributaries, lateral fan sources and coarse-textured moraines. An aerial LiDAR survey of the lower valley highlights the stratigraphy of eroded remnants of glacial fills. We employ an hydraulically-driven, finite-difference morphodynamic gravel bedload transport model to simulate postglacial downcutting of the valley floodplain system and tributary channels. We compare our results to dating evidence recovered from radiocarbon and infrared-stimulated luminescence dating of terrace surfaces. The model incorporates valley stratigraphy and traces the fate of material from glacial deposits and tributaries. Results show an exponential decrease in bedload transport rates over time in response to the deglacial fall in base level. 'Virtual' lithology tracking shows an initially high relative yield of tributary material at the catchment outlet, diminishing in time. Our criteria of model 'success' were to match (1) the contemporary river bed profile concavity, (2) the downstream trend in grain-size and lithology, and (3) the overall downstream trend in sand content in the river substrate. By tuning certain model parameters, it is possible to obtain a satisfactory fit for one or two, but not all three criteria. We review some of the implications for calibrating models of long-term river evolution, and the data requirements for optimum results.
