Karlstrom et al. (2013) raise two issues in their Comment related to aspects of the discussion in our paper (Pederson et al., 2013) about the Lee’s Ferry record and incision in the Colorado Plateau. We synthesize evidence for the bull’s-eye pattern of incision over the past 105 yr, finding that it does not match spatial patterns of river steepness, stream power, or proposed mantle sources of uplift. Karlstrom et al. (2013) express no concerns with these findings. Instead, their first issue is that longer-time-scale incision may have a different pattern, which may or may not be the case. Their second point expresses doubt that isostatic rebound would significantly increase incision rates in the central plateau, which seems based upon an erroneous geometric argument. Their Comment highlights two of their new papers, and we appreciate the opportunity to contribute to ongoing debates about the patterns of topography and erosion and how they relate to any active deformation, other baselevel controls, and bedrock erodability.
Their first point, taken broadly, is that incision patterns across the Colorado Plateau may be different over longer, 106 yr, time scales. Indeed this may be the case, but further work and validation are required to document the variability over such longer time scales, and this was not within the scope of our study. More specifically, they criticize our narrower approach of comparing the Lee’s Ferry record only to similarly constrained studies in the region. Our “apples to apples” reasoning is explained in the first paragraph of our discussion—but here we can elaborate.
We documented at Lee’s Ferry a record of the Colorado River’s history, constrained by 21 absolute ages from two independent chronometers (optical luminescence and cosmogenic-nuclide profiles) spanning the past 150 k.y. The steady background incision rate over this record is drawn out, and is compared to other regional studies that also specifically have multiple age constraints on a relatively full stratigraphy over 105 yr time scales. This gives us a snapshot of the regional pattern that we can “hang our hat upon” and try to understand. In contrast, Wolkowinsky and Granger (2004), Karlstrom et al. (2008), Darling et al. (2012), and Karlstrom et al. (2012) estimate incision rates at locations based upon single age results (especially burial and isochron cosmogenic techniques), which lack broader chronostratigraphic context. These longer-timespan rate estimates may tend to be slower as they integrate over unsteady processes (Gardner et al., 1987), giving a potentially false impression of rates increasing toward the present. Furthermore, over million-year time scales, potential waves of incision migrating upstream past a location become a significant influence, as explored through modeling by Cook et al. (2009). In a transient landscape such as the Colorado Plateau, the incision rate at a location averaged over a few million years has a higher chance of incorporating both episodes of rapid, transient incision as well as episodes of slower incision before and after. This makes longer-term incision patterns systematically more difficult to deconvolve. The pattern of erosion over longer time scales is indeed an intriguing, difficult, and unresolved problem. Yet, we stand by our “apples to apples” approach, revealing with confidence the bull’s-eye pattern of incision at 105 yr time scales.
The second issue in the Comment is with our hypothesis that isostatic rebound, as a positive feedback on rock uplift, may account for the faster incision occurring in the central Colorado Plateau. First, Karlstrom et al. (2012) and Lazear et al. (2013) confirm earlier work (Pederson et al., 2002; Callahan et al., 2006; Roy et al., 2009) that a domal pattern of isostatic rebound extends across the plateau, peaking through the Canyonlands center and becoming insignificant at the Grand Canyon edge. The longer timespan over which these studies calculate total rebound is not inconsistent with a more recent portion of that rebound following a similar spatial pattern. A contrary question is more pertinent: Why would unloading due to a bull’s-eye of rapid erosion over 105 yr not be matched by an isostatic response at the same time scale? Also, we do not understand their geometric argument that the pattern of such a feedback would be “controlled mainly by the differential rebound between the bull’s-eye center and…the Lee’s Ferry knickpoint.” The broad-wavelength rebound does not stop or start at Lee’s Ferry, partway between the center and the edge of the Colorado Plateau, as if it were a point-source baselevel effect propagating up the river. Instead, this source of rock uplift has the pattern of a smooth arc advecting upward relative to the longitudinal profile of the river, and in proportion to long-wavelength erosional unloading (Callahan et al., 2006; Karlstrom et al., 2012).