Guidelines for sampling bedload transport with minimum uncertainty

The stability of river channels and their suitability as habitat for aqueous organisms is strongly controlled by the rate of bedload transport. Quantification of bedload transport rates in rivers is difficult, not only because of the temporal variation in transport, but also because of the cross-channel variation in transport. The objectives of this study were: (i) to determine the effect of cross-channel variation in bedload transport on the uncertainty of width-integrated transport rates; and to use this knowledge (ii) to improve guidelines for bedload sampling. This was done through a thorough statistical evaluation of stochastic and systematic uncertainties involved in bed-load transport measurements. Based on this evaluation, new guidelines are presented for determination of the number of samples and sampling positions across the channel that are required for bedload measurements in several types of sand-bed rivers and gravel-bed rivers. The guidelines relate to bedload measurements made with pressure-difference (Helley-Smith type) samplers that require numerous bedload samples of short duration at several positions across the channel. The results show that generally more sampling positions across the channel are required in gravel-bed rivers than in sand-bed rivers. For gravel-bed rivers with unknown cross-channel distribution of transport, at least 10 sampling positions are recommended, whereas for most sand-bed rivers five positions suffice. In addition, at least 12 short-duration samples are required at each position to obtain bedload estimates with uncertainties below 20%. If the same level of uncertainty is desired in the case of high spatial and temporal variation in transport rates, the number of short-durationsamples needed per sampling position increases to 40.