Critical conditions for particle motion in coarse bed materials of nonuniform size distribution

Initiation of particle motion in a bed material of nonuniform size distribution may be quantified by (q(ci)/q(cr)) = (D-i/D-r)(b), where q(ci) is the critical unit discharge at which particle size D-i enters motion, q(cr) is the critical condition for a reference size D-r unaffected by the hiding/exposure effects associated with nonuniform size distributions, i and r refer to percentiles of the distribution and b varies from 0 (equal mobility in entrainment of all particle sizes) to 1.5-2.5 (full size selective transport). Currently there is no generally accepted method for predicting the value of b. Flume and field data are therefore combined to investigate the above relationship. Thirty-seven sets of flume data quantify the relationship between critical unit discharge and particle size for bed materials with uniform size distributions (used here to approximate full size selective transport). Field data quantify the relationship for bed materials of nonuniform size distribution at 24 sites, with b ranging from 0.15 to 1.3. Intersection of the two relationships clearly demonstrates the hiding/exposure effect; in some but not all cases, Dr is close to the median size D-50. The exponent has two clusters of values: b > 1 for sites subject to episodic rain-fed floods and data collected by bedload pit trap and tracers; and b < 0.7 for sites with seasonal snowmelt/glacial melt flow regimes and data collected by bedload sampler and large aperture trap. Field technique appears unlikely to cause variations in b of more than about 0.25. However, the clustering is consistent with possible variations in bed structure distinguishing: for b > 1, sites with relatively infrequent bedload transport where particle embedding and consolidation could reduce the mobility of coarser particles; and, for b < 0.7, a looser bed structure with frequent transport events allowing hiding/exposure and size selection effects to achieve their balance. As yet there is no firm evidence for such a dependency on bed structure but variations in b could potentially be caused by factors outside those determining equal mobility or size selection but appearing to affect b in the same way. (C) 2013 Elsevier B.V. All rights reserved.