4.3.1 Case 3: Trunk Response
In this case, because the change in rainfall is spatially variable along the trunk river, the initial change in erosion rate is also variable. The trunk river experiences an approximately 80% increase in erosion rate at the outlet (E > U ), and a decrease of 67% in the headwaters (E < U ), corresponding to the change in upstream average rainfall along its length. Atxsc (located ~27 km upstream from the outlet), Qf = Qi ,Sf = Si , and so immediately following the change in rainfall E = U . Enhanced incision at the outlet produces a concave-up knickpoint; however, as this knickpoint migrates upstream it progressively sharpens and eventually evolves into an oversteepened convex-up knickpoint, contrasting with expectations for the increase in rainfall (e.g. , Case 2). Oversteepening is a consequence of the upstream decrease in erosional efficiency driven by the rainfall gradient that is exacerbated by, but does not depend on, differing modes of adjustment upstream and downstream of xsc related to the complex response. This is analogous to knickpoint behavior described by Forte et al. (2016) and Darling et al. (2020) where modelled lithologic contacts demarcate similar relative variations in erosional efficiency (i.e. hard rocks over soft rocks). Lastly, we note that everywhere upstream ofxsc over-adjusts during the transient response, which is a characteristic of complex responses in general, leading to variable modes of adjustment in time and space. The overadjustment we observe is essentially the whiplash response described by Gasparini et al. (2006, 2007), but notably results here without sediment flux. This continuous evolution of the trunk knickpoint has important consequences for signals passed to tributaries.