In our paper, we confirmed that the Colorado River has incised the Grand Canyon at significant and spatially variable rates over the middle–late Quaternary and hypothesized that this variability in incision may be due to normal faulting, in that subsidence near the hanging wall of the fault locally reduces incision rates. Hanks and Blair, likewise, have two main comments. The first is about the pattern of offset related to regional normal faulting, which does not seem to conflict with the statements in our paper. The second addresses possible errors in our incision rate calculations related to variations along the Colorado River's long profile. We stand by our method of calculating bedrock incision and believe it provides the best estimates possible at this time (and note that even if Hanks and Blair's criticism is valid, our results and interpretations would not be greatly affected).
Hanks and Blair apparently agree with us that the lower incision rates west of the Hurricane-Toroweap fault system can be explained by the reduction of incision due to west-down slip on these faults. They state that this faulting would not be expected to affect incision rates in the eastern Grand Canyon. We agree, and this was one of the main points of our paper: Contrary to many places discussed in the recent geomorphic literature, the Grand Canyon is a situation where faulting does not drive (increase) incision rates upstream (e.g., eastern Grand Canyon), but instead, through subsidence, decreases incision rates in the hanging wall area downstream of the faults (e.g., Granite Park locality). Hanks and Blair state that incision rates at the locality just a few kilometers upstream of the Toroweap fault should be affected by faulting, though they do not state how or why. It seems they assume that any surface or absolute displacement, regardless of geometry, will steepen river gradient and affect long-term stream incision upstream. This is not always the case—again, one of our main points. Integrated over a time span longer than glacial-interglacial climate changes, the Colorado River in the Grand Canyon has been incising bedrock both upstream and downstream of the faults as a result of effective base-level fall farther downstream.
Their comment about normal fault spacing and seismogenic character, however correct, is not directly relevant to our paper. First, topography only equates to fault displacement if there is zero erosion, which is certainly not the case in the Grand Canyon. Second, our Figure 3 is not a “block motion” tectonic model, but rather a simple geometric illustration of how fault slip may either drive upstream or reduce downstream incision. In the first paragraph of our paper's discussion, we suggested that displacement may decrease away from the fault into the hanging wall, just as Hanks and Blair imply. We had only a single incision rate on the downstream side at the time of the paper, and hence our data could not address this issue. We are continuing to date basalt flows and estimate incision rates across the faults, which should help answer the questions about footwall uplift, possible listric fault geometry, and strain accumulation during faulting that may be underlying Hanks and Blair's comments.
The second part of Hanks and Blair's comment addresses large-scale convexities in the longitudinal profile of the Colorado River through the Grand Canyon. They suggest that the reach-scale convexities in the profile of the river may be the result of spatially variable debris-flow activity creating very broad reaches that are aggrading relative to neighboring reaches. Hanks and Blair point out that if local aggradation is happening, we would be underestimating the depth to bedrock beneath the river and thus our bedrock incision estimate would be too low. We stand by our method as a thoughtful and conservative approach that tries to compare apples to apples (the height of bedrock straths to an estimate of present-day bedrock) and that considers the complexities of fill terraces. There are several possible controls on the Colorado River's long profile depending upon the different time and space scales one is investigating. The Colorado River has notable changes in reach-scale gradient along its entire length, and researchers are investigating this issue at both small and large scales, from how debris fans, pools, and rapids control the channel morphology and bed grain size, and thus the detailed profile of the modern, mixed alluvial-bedrock stream (e.g., Grams and Schmidt, 1999; Webb et al., 1989; Howard and Dolan, 1981), to how bedrock strength, tectonic knickzones, and the long-term history of drainage capture may influence the larger reach-to-canyon-scale profile. The issue becomes more complicated with the evidence that the Colorado River has undergone high amplitude cycles of aggradation-degradation in response to climate changes, as is evident in its sequence of thick Pleistocene fill terraces (Anders, 2003; Lucchitta et al., 2000; Machette and Rosholt, 1991). But the net product of this changing river has been overall bedrock incision, and for significant episodes the river has been a bedrock stream cutting the Grand Canyon, potentially sensitive to other controls on gradient such as bedrock properties.
In summary, it is unclear what is controlling the larger-scale profile of the river, and to what degree the controls on the local gradient of the modern river (debris fans, etc.) are superimposed upon and reflect larger-scale, longer-term controls (cf. Howard et al., 1994). One thing that is clear is that the localized aggradation Hanks and Blair infer, if happening, is at most a Holocene phenomenon superimposed on the longer-term trend of incision. In any case, if we add the upper estimate Hanks and Blair provide of 30 m of alluvium beneath the deeper pools of the river, the eastern Grand Canyon bedrock incision rate would be ~230 m/m.y. instead of our ~150 m/m.y. This somewhat greater rate would not invalidate our interpretations, but only amplify the fact that the Colorado River has actively and variably incised the Grand Canyon over middle–late Quaternary time.