RELATING ROCK STRENGTH CONTROLS TO LARGE-SCALE VARIATIONS IN THE COLORADO RIVER’S PROFILE, GLEN AND GRAND CANYONS, UT AND AZ

John Wesley Powell originally hypothesized that channel gradients and widths are directly influenced by river-level bedrock. Subsequent work has cast doubts on his hypothesis because local-scale (0.1-1 km) gradient has been shown to be the result of rapid-pool sequences associated with coarse debris from side canyon tributaries. Profile controls at larger spatial and temporal scales, however, are still under debate and include tectonics, rock mass strength, and bed material size. This study relates bedrock resisting forces to hydraulic driving forces at medium (10 km) and large (100 km) scales in Glen and Grand Canyons, which have contrasting rock types, channel gradients, and widths.

Data collected at 84 outcrop-scale study sites included compressive strengths and joint spacing. Unit stream power was also calculated using gradients interpolated from Birdseye (1924), widths extracted from GIS coverages, and representative downstream discharge distributions.

At the canyon-scale (100 km), Grand Canyon has bedrock with slightly higher, yet significant, compressive strengths, and significantly lower fracture/joint spacing. These observations correspond to the fact that Grand Canyon is four times steeper, one half as wide, and has nine times greater unit stream power than Glen Canyon. Secondly, reach average compressive strengths among 18 reaches of mechanically similar rock-type show a significant, but loose, correlation to gradient and unit stream power. Not surprisingly, reach average channel widths are inversely correlated with compressive strength. Selby RMS scores, devised for hillslopes but more recently used on river studies, show no correlation with any of the three hydraulic properties.

These results support the interpretation that harder and/or less fractured rocks offer greater resistance and that the increased energy required for the long-term incision is provided by a higher gradient. The variability in the data suggests that bedrock is not the lone control; however, the data support the idea that bedrock sets the long-term, large-scale template for the river. There may be an indirect control by bedrock through its influence on hillslope-to-river sediment production and a direct control on the profile by river-level bedrock resistance to incision.