Evolution of undercut slopes on abandoned incised meanders in the Eastern Highland Rim of Tennessee, USA

Field surveys, location-for-time reasoning, and computer modeling were used to study the evolution of slopes on valley walls of abandoned bedrock meanders on the Eastern Highland Rim, Tennessee. Hillslopes on the undercut slopes of cutoff incised meanders were ordered as to relative age by the height of their meander floors above the modern stream level. The assumption is that the undercut slope is actively eroded by the stream until abandonment of the meander, at which time the slope begins to evolve to a different form. More-advanced stages of evolution occur on walls of higher meanders that were abandoned earlier. The most rapid change in this initial form is the elimination of a free face, which occurs soon after the Ž . meander is abandoned. In addition, the hillslopes associated with even the lowest youngest cutoff meanders show somewhat gentler overall gradients than the actively undercut slopes. Hillslopes associated with meanders 3 to about 20 m above modern stream level maintain straight segments with angles showing only a slight decrease from the 36–388 associated with the lowest cutoffs; overall angles decrease, however, as the straight segment becomes shorter. The oldest slopes, those on cutoffs 30 m or greater above modern stream level, have developed into convex–concave slopes with maximum slopes of 158. A hillslope evolution model based on previously published algorithms was used to simulate the transition of actively undercut hillslopes into hillslopes on abandoned meanders. Hillslope modeling is particularly useful in this setting. If the valley incision rate is known, an age can be estimated for the cutoff and hence for the hillslope. Alternatively, if hillslope process rates are known, a model age obtained for the hillslope can be used to estimate an incision rate. Even where both incision rates and hillslope process rates are poorly constrained, as in the present setting, modeling allows assumptions about specific rates to be evaluated by determining their implications for other rates. For example, for three cutoff meanders along one stream, best-fit criteria were used to select process rates for the model. Model ages of hillslopes were then obtained and compared with those calculated from a valley-incision rate measured elsewhere in the same physiographic province. For two of the hillslopes, model ages were found to be much younger than those calculated from the incision rate. In order to make the two ages agree, unreasonably low process rates had to be used in the model, thus implying that the incision rate probably underestimates the actual incision rate in this valley. ) Corresponding author. Fax: q1-931-372-3363. Ž . Ž . E-mail addresses: [email protected] H.H. Mills , [email protected] R.T. Mills . 0169-555Xr01r$ see front matter q 2001 Elsevier Science B.V. All rights reserved. Ž . PII: S0169-555X 00 00104-5 ( ) H.H. Mills, R.T. MillsrGeomorphology 38 2001 317–336 318 Experimentation with heights of initial profiles, again using best-fit criteria, suggests that since abandonment of the highest cutoff, the plateau has been downwasting at a rate about one-fourth that of the valley incision rate, a finding in agreement with published rates of chemical denudation in the area. q 2001 Elsevier Science B.V. All rights reserved.