SEDIMENTARY PROCESSES/Fluvial Geomorphology 653 Figure Relationships between sediment transport and bed velocity/shear stress (A) Hjulstrom’s (1935) critical velocity fields for entrainment, erosion, transport, and deposition Note that for fine grained cohesive materials, much higher velocities may be re quired to erode materials than to transport them in disaggregated form (B) Summary of experimental observations on the forms of sediment transport Reproduced with permission from Bridge JS (2003) Rivers and Floodplains Blackwell: Oxford (C) The more com plex relationships actually observed in field situations Repro duced with permission from Knighton D (1998) Fluvial Forms and Processes Arnold: London bed, but these forces are usually expressed in terms of mean flow parameters (which are more easily estimated) such as mean velocity, bed shear stress, or unit stream power for a given river level (see Figures 2B and 3).These can then be related to movement thresholds, for particles of given size, as bedload or suspension load However, field data show great complexity because of factors such as the presence of mixed grain size and the packing and armouring (in which finer material at depth is protected by coarser material above) of bed materials (Figure 3) The actual entrainment of particles also involves eddy, sweep, and burst flow phenomena in flowing water Nevertheless, in general, turbulent flow (turbulence being expressed in terms of the Reynolds number) is capable of transporting fine material in suspension, with amounts set by supply rates, whereas coarser materials (available to the river from its bed and banks) are more characteristically competence limited, such that a threshold of river flow has to be reached for coarser sediment entrainment and transport For coarser load, it has proved possible to develop a range of bedload formulas (for example, relating transport rate to excess shear stress, discharge, or stream power, the ‘excess’ being the value above the threshold at which particle motion begins) For gravel rivers, such equations have not in general proved very successful when checked against the limited amount of direct bedload transport observations available Equipment may be of the kind installed on the East Fork River, Wyoming, where moving sediment was trapped in a set of open slots on the river bed, moved to the bank on a conveyor belt to be weighed, and then returned to the river further downstream (Figure 4) In general, suspended and solute loads have to be measured following water sampling, and yield measures are much more generally available Data obtained during flow events over periods of years show results that may be complex in detail: for example, sediment concentrations may be higher as waters rise during a flood, compared to when waters recede, and there is no precise and unique relationship between discharge or water stage and concentration However, it appears that events of moderate frequency (around annual) and stage (about bankfull or channel-full) achieve the most work, though for bed materials, large floods may be necessary both to generate slope failure and coarse-material input and to move the larger particles on a river bed Floods that are extreme, compared to the annual average, occur every few years in some environments, whereas in other environments the flood expected, say, every 10 years is not that much larger than the annual one (Figure 5) Globally, sediment yields may vary greatly and are high particularly in high-relief/active tectonic