THE STABILITY OF SUBLITTORAL, FINE-GRAINED SEDIMENTS IN A SUB-ARCTIC ESTUARY

The erodibility of natural estuarine sediments was measured in situ al ong a longitudinal transect of Manitounuk Sound, Hudson Bay, using the benthic flume Sea Carousel. Sedimentation processes along the transec t varied from continuous, rapid, post-glacial sedimentation in the inn er Sound, to glacial outcrops and seabed reworking of the outer Sound. The grain size and physical bulk properties reflect changes in deposi tional environment and correlate with sediment erosion threshold stres s (tau(c)), erosion rate (E), erosion type and still-water mass settli ng rate. There was a steady increase in tau(c) (0 . 8-2 . 0 Pa) with d istance down the Sound in parallel with the decreasing sedimentation r ate (0 . 003-0 . 001 m yr(-1)) and increasing sediment bulk density (1 650-2010 kg m(-3)). The near-surface friction coefficient varied up to 68 degrees in proportion to the clay content of post-glacial material . Glacial sediments were characterized by variable results and general ly higher friction coefficients. Seabed erosion in Sea Carousel began with surface creep of loose aggregates, pellets and organic debris. Th is was followed by Type I bed erosion at rates that varied between 0 . 0002 and 0 . 0032 kg m(-2) s(-1) (mean 0 . 0015). Type I peak erosion rate was inversely related to applied bed shear stress (tau(0)). Type II erosion succeeded Type I, often after a broad transitional period. Simulations of suspended sediment concentration in Sea Carousel were made using four commonly used erosion (E) algorithms. The best results were obtained using Krone's dimensionless ratio relationship: E=M(tau (0)/tau(c)-1). Simulations were highly sensitive to the definition of erosion threshold with sediment depth [tau(c)(z)]. Small errors in def inition of tau(c)(z) caused large errors in the prediction of suspende d sediment concentration which far exceeded differences between the me thods tested.