Fluvial landscape models and catchment-scale sediment transport

The need for the ability to make quantitative predictions regarding the transport of sediment within watersheds and from watersheds to the sea has never been more important than it is today. Sediment transport is at the heart of a surprising number of practical problems, ranging from the prevention of soil loss to the tracking of contaminants, and this has resulted in a rich but scattered literature. In many of these contexts, observational data has been used to develop empirical sediment transport laws. For example, the Revised Universal Soil Loss Equation (RUSLE) has been developed for use at the farm-plot scale, and the equation of Syvitski et al. [Water Resour. Res. 36(9) (2000) 2747] was developed for the world's largest river basins. In the context of fluvial geomorphology, several different but closely related sediment transport laws have been used to construct fluvial landscape models. The sediment transport formulas used by these models are a generalized form of several physically based bed-load transport formulas. Based on studies of fluvial landscape models, we now know that many measurable attributes of channelized landscapes, such as junction angles, Horton ratios, hydraulic geometry exponents, valley geometry, and longitudinal profiles, vary in a predictable manner with a few well-constrained, physical parameters. One important consequence of this success in forward modeling is that it is now feasible to solve the inverse problem of calculating regional sediment transport parameters from measurements (such as long profiles) that can easily be made from digital elevation models (DEMs). This may also make it possible to make a rational selection between alternate sediment transport laws. Since fluvial landscape models represent the state of the art in catchment sediment transport modeling, the three-fold purpose of this paper is to: (1) explain the key concepts and simplifying assumptions that are common to these models; (2) explain how several current models differ from one another; and (3) highlight a few of the significant results that have been obtained in recent years. While serving mostly as an expository or overview paper, some new results for a steady state fluvial landscape model are also included.