A channelization model of landscape evolution

Cp. Stark et Gj. Stark, A channelization model of landscape evolution, AM J SCI, 301(4-5), 2001, pp. 486-512
Citations number
Categorie Soggetti
Earth Sciences
Journal title
ISSN journal
0002-9599 → ACNP
Year of publication
486 - 512
SICI code
The geomorphic evolution of many landscapes is fundamentally determined by the evolution of the river channels and their interactions with hillslopes. Consequently, models of landscape evolution ought to track the evolution o f the channel geometry so as to quantify the rate of erosion of channel bot toms and to follow the changes in hillslope-channel coupling over time. Unf ortunately, the spatial resolution required to describe channel morphology adequately is computationally impractical. It is also beyond the resolution of most digital elevation data. What is required is a parameterization sch eme for approximating fine scale channel morphology at a coarse pixel scale . Such a parameterization is already implicitly employed in most models by assuming channel equilibrium, which ties the width and depth of a model cha nnel to the square root of discharge through a pixel. Channel fluxes are th ereby predictable, and a closed form of the governing equations is attained . In reality, mountain river channels do not take a simple equilibrium form and show great spatial variability and evident disequilibrium geometry. Si nce the time scales of changes in channel geometry, bedrock channel erosion , and hillslope response are all closely related, it is reasonable to infer that the spatio-temporal development of the landscape is determined by the ir interaction and that channel disequilibrium is a fundamental factor in t he dynamics of landscape evolution. If this is the case, we need an alterna tive sub-grid scale parameterization that aggregates channel properties suc h as surface morphology, roughness, cross-sectional geometry, so that the t ime dependent behavior of these properties can be estimated at the coarse p ixel scale. We introduce such a parameterization measure, which we term cha nnelization, after extensive investigation of the pixel resolution dependen ce of topographic relief. We focus in particular on the effect of coarse gr aining on digital elevation data for derived measures such as channel slope and upstream area and demonstrate that we can approximately correct for th is effect. We show that a very simple geomorphic model can be constructed a round the channelization parameter and the resolution-invariant topographic measures. This model demonstrates that channel disequilibrium may play a s ignificant role in mountain landscape dynamics. It also shows how geomorphi c models in general could be modified to incorporate such sub-pixel scale c omplexities and to better model these dynamics.