GRAVEL TRANSPORT AND MORPHOLOGICAL CHANGE IN BRAIDED SUNWAPTA RIVER, ALBERTA, CANADA

The relation between morphological change and patterns of variation in bedload transport rate in braided streams was observed by repeated, d aily topographic surveys over a 25 day study period in a 60 m reach of the proglacial Sunwapta River, Alberta, Canada. There are two major p eriods of morphological change, each lasting several days and each inv olving the complete destruction and reconstruction of bar complexes. B ar complex destruction was caused by redirection of the flow and by do wnstream extension of the confluence scour zone upstream. Reconstructi on involved accretion of unit bars on bar head, flank and tail and in one case was initiated by disection of a large, lobate unit bar. High rates of sediment movement, measured from net scour and fill of the cr oss-sections, coincided with these morphological changes. Sediment was supplied from both bed and bank erosion, and patterns and distances o f transfer were highly variable. Rates of transport estimated by match ing upstream erosional volumes with downstream deposition were much gr eater than those estimated from either a step-length approach or a sed iment budget. Measurements of scour and fill and observations of morph ological change indicate that step lengths (virtual transport distance s) were typically 40-100 m during a diurnal discharge cycle. Shorter s tep lengths occurred when transfer was confined to a single anabranch and longer steps involved channel changes at the scale of the entire r each. Sediment budgeting was used to describe the spatial patterns of sediment transport associated with the morphological changes and to es timate minimum daily reach-averaged transport rates. Mean bedload tran sport rates correlate with discharge, but with considerable scatter. T he largest deviations from the mean relation can be tied to phases of channel incision, bank erosion, scour hole migration, bar deposition a nd channel filling apparently controlled by changes and fluctuations i n sediment supply from upstream, independent of discharge. These are i nterpreted as field evidence of 'autopulses' or 'macropulses' in bedlo ad transport, previously observed only in laboratory models of braided streams.