THE IMPORTANCE OF FLOW-CONVERGENCE ROUTING FOR MAINTAINING RIFFLES AND POOLS ON LARGE, REGULATED, GRAVEL-BED RIVERS

Reaches of gravel streams below dams incise, narrow, and lose the bed relief between riffles and pools. A common goal of river rehabilitation in gravel rivers, whether by controlled floods or by gravel placement, is to re-establish bed relief and habitat diversity for aquatic species. However, if relief is restored without addressing geomorphic process that cause degradation, then the benefits of the effort are likely to be short lived. The study reported here determined whether a velocity reversal mechanism in which pools incise during high flows and riffles incise during low flows could yield a self-maintaining bed relief, even as a river systematically incises due to a dam. The study area was the top 10.5 km of the Yuba River below Englebright Dam. Although this dam has blocked bedload transport into the reach for 65 years causing significant channel incision, a residual of 17.7 million m3 of sediment remains in the study reach due to historical hydraulic mining for gold. The approach used was to produce and difference digital elevation models of the reach in 1999 and 2006 to determine the spatial pattern of channel change. The key result was that the relief between pools and riffles has been maintained, even though the reach lost 463,000 m3 of sediment. To explain why the relief has been maintained, a detailed analysis was performed on one pool-riffle-glide channel unit, which was also mapped in 2004 and 2005 between floods with return intervals of 7.7 and 24 years. 2D hydraulic modeling of this channel unit at discharges of 23.4, 158, 328, and 1216 m3 per sec revealed a shifting location of peak velocity, from the riffle crest at low flow to the glide at the second highest flow to the pool at the highest flow. The same shifts were also found by comparing the at-a-station hydraulic geometry of cross-sections at these three locations. The velocity reversal mechanism in the study reach is due to flow-convergence routing, not to shifts in the depth-slope product. The evidence demonstrates that Yuba floods preferentially scour pools. Between floods, knickpoints propagate upstream through riffles and re-establish the long-term sustainable relief. The key lesson for management is that regulated gravel rivers need to be evaluated for velocity reversal dynamics to determine whether proposed rehabilitation actions will be sustainable.