Evaluating Bed-material Transport Equations using Field Measurements in a Sandy Gravel-bed Stream, Arbúcies River, NE Spain

Abstract

Bed-material load under a wide range of hydraulic conditions was sampled in a poorly sorted, sandy, gravel-bed river (Arbucies data showed a marked scatter of bed-material discharges, reflecting the high variability of bedload rates as well as of suspended sediment concentrations. Bed-material discharges were used to test five bedload and bed-material formulae. The degree of agreement between observed and predicted values varies greatly. The percentage of observations in which the discrepancy ratio between observed and computed values has a value between 0·5 and 2, range from 25 per cent (van Rijn), to 38 per cent (Brownlie), 52 per cent (Meyer-Peter and Muller), 65 per cent (Engelund and Hansen), and 68 per cent (Ackers and White). The wide range of hydraulic conditions from which the data were obtained and the poor sorting of the bed sediment affected the performance of the van Rijn (1984) and Brownlie (1981) equations. The degree of correlation between observed transport rates and values predicted by the Engelund and Hansen (1967) formula appears to be unaffected by the poorly sorted bed material of the Arbucies River with reasonable accuracy and no bias for transport values under low and intermediate flow conditions. The best agreement with measured values was obtained using the Ackers and White (1973) model, a reflection of its original design for poorly sorted sediment. © 1997 by John Wiley & Sons, Ltd.