Larsen et al. (2004) presented data on debris fan reworking in the Green River canyons of Dinosaur National Monument and concluded that discharges that do not exceed the pre-dam 2 yr flood data cause significant fan reworking and mainstem channel adjustment. Kieffer (2004) has two main comments: (1) we did not define the term “constriction ratio” correctly, and (2) we misrepresented the findings of Kieffer (1985) that debis fan reworking only takes places during very large floods.
First, Kieffer (2004) is correct in pointing out our mistake concerning definition of the constriction ratio term. Although we computed constriction ratio in the same way as Kieffer (1985), we described in words the inverse of this ratio, and we thank Kieffer (2004) for pointing this out. The values reported in Larsen et al. (2004) were computed in the same way as those of Kieffer (1985). Channel width at the constriction widened during our study period at both sites.
Second, we agree with Kieffer's (2004) recognition of the role of low and high magnitude floods in reshaping debris fan constrictions. Kieffer (1985) primarily studied the effects of the unprecedented 2700 m3s−1 flood of 1983 on Crystal Creek debris fan that widened the constriction from 0.25 to 0.4. She also estimated the discharge necessary to further widen the constriction to 0.5 and estimated that this discharge is ~11,320 m3s−1. This is a very large discharge; the paleoflood record indicates that one flood exceeding 14,000 m3s−1 occurred 1600–1200 years ago, and 10 floods exceeding 6800 m3s−1 occurred during the past 2000–2300 years (O'Conner et al., 1994). Kieffer (1985) acknowledged conceptually that small floods cause some reworking (Kieffer, 1985, their Fig. 12), but the relative magnitude of reworking by small, moderate, and large floods was not estimated.
Thus, interpretations of Kieffer's (1985) work typically focus on the essential role of large floods in reworking debris fan deposits. For example, the final environmental impact statement for Glen Canyon Dam operations stated, “riverflows within the operational range of [the] Glen Canyon Dam Powerplant will remove some of the new material [from debris flows]. However, floods of 100,000 to 200,000 cfs [2700 to 5400 m3s−1] or more probably would be necessary to remove the largest boulders from some debris fans, to increase the constriction ratio, and to decrease the elevation drop (Kieffer, 1985)” (U.S. Department of the Interior, 1995).
Subsequent studies of debris fan reworking have shown that significant increases in constriction ratio and decreases in elevation drop occur during small and moderate floods, especially if those floods occur soon after debris flow emplacement (Webb et al., 1999). Dam managers may misinterpret Kieffer's (1985) results as a demonstration that significant channel widening can only occur during floods that constitute true hydrologic emergencies. If significant reworking only occurs during emergencies, debris flow reworking need not be considered part of the objectives of regular dam operations. Our work, and that of Webb et al. (1999), demonstrates by field measurement that significant reworking occurs at flows much lower than those at which Kieffer (1985) estimated were necessary to widen constriction ratios to 0.5.