Improving Bedload Transport Predictions by Incorporating Hysteresis

Abstract:

The importance of unsteady flow on sediment transport rates has long been recognized. However, the majority of sediment transport models were developed under steady flow conditions that did not account for changing bed morphologies and sediment transport during flood events. More recent research has used laboratory data and field data to quantify the influence of hysteresis on bedload transport and adjust transport models. In this research, these new methods are combined to improve further the accuracy of bedload transport rate quantification and prediction.

The first approach defined reference shear stresses for hydrograph rising and falling limbs, and used these values to predict total and fractional transport rates during a hydrograph. From this research, a parameter for improving transport predictions during unsteady flows was developed. The second approach applied a maximum likelihood procedure to fit a bedload rating curve to measurements from a number of different coarse bed rivers. Parameters defining the rating curve were optimized for values that maximized the conditional probability of producing the measured bedload transport rate. Bedload sample magnitude was fit to a gamma distribution, and the probability of collecting N particles in a sampler during a given time step was described with a Poisson probability density function. Both approaches improved estimates of total transport during large flow events when compared to existing methods and transport models. Recognizing and accounting for the changes in transport parameters over time frames on the order of a flood or flood sequence influences the choice of method for parameter calculation in sediment transport calculations. Those methods that more tightly link the changing flow rate and bed mobility have the potential to improve bedload transport rates.