APPROACHES TO QUANTIFYING LONG-TERM CONTINENTAL-SHELF SEDIMENT TRANSPORT WITH AN EXAMPLE FROM THE NORTHERN CALIFORNIA STRESS MID-SHELF SITE

Modeling shelf sediment transport rates and bed reworking depths is pr oblematic when the wave and current forcing conditions are not precise ly known, as is usually the: case when long-term sedimentation pattern s are of interest. Two approaches to modeling sediment transport under such circumstances are considered. The first relies on measured or si mulated time series of flow conditions to drive model calculations. Th e second approach uses as model input probability distribution functio ns of bottom boundary layer flow conditions developed from wave and cu rrent measurements. Sediment transport rates, frequency of bed resuspe nsion by waves and currents, and bed reworking calculated using the tw o methods are compared at the mid-shelf STRESS (Sediment TRansport on Shelves and Slopes) site on the northern California continental shelf. Current, wave and resuspension measurements at the site are used to g enerate model inputs and test model results. An Ii-year record of bott om wave orbital velocity, calculated from surface wave spectra measure d by the National Data Buoy Center (NDBC) Buoy 46013 and verified agai nst bottom tripod measurements, is used to characterize the frequency and duration of wave-driven transport events and to estimate the joint probability distribution of wave orbital velocity and period. A 109-d ay record of hourly current measurements 10 m above bottom is used to estimate the probability distribution of bottom boundary layer current velocity at this site and to develop an auto-regressive model to simu late current velocities for times when direct measurements of currents are not available. Frequency of transport, the maximum volume of susp ended sediment, and average flux calculated using measured wave and si mulated current time series agree well with values calculated using me asured time series. A probabilistic approach is more amenable to calcu lations over time scales longer than existing wave records, but it ten ds to underestimate net transport because it does not capture the epis odic nature of transport events. Both methods enable estimates to be m ade of the uncertainty in transport quantities that arise from an inco mplete knowledge of the specific timing of wave and current conditions . (C) 1997 Elsevier Science Ltd.