Accounting for Vegetation Dynamics in Landform Evolution Modelling

摘要

Physically-based landform evolution models (LEM) provide a spatial framework to forecast the long-term geomorphic stability of post-niining landscapes. When used to test the relative stability of different landform designs, LEMs may assist with the development of optimal closure designs for rehabilitated mines. However, in order to reduce prediction error, factors controlling landscape evolution must also be accurately represented by models. This requires that the correct sets of processes are represented and that driving parameters used in LEMs are, if necessary, calibrated to the landscape of interest.Vegetation exerts strong controls on the hydrologic and geomorphic processes that drive landscape evolution. Hence, the spatial and temporal scales over which typical patterns in vegetation are represented in different landscapes are important considerations in LEM calibration. However, vegetation is rarely accounted for in LEMs and even when integrated is often represented simply, without representing spatial dynamics of vegetation cover that may impact landform development.This study focuses on characterising vegetation patterns to calibrate a new vegetation module of the The Cellular Automaton Evolutionary Slope and River (CAESAR)-Lisflood LEM to be applied to a conceptual rehabilitated landform of the Ranger uranium mine located in the Alligator Rivers Region (ARR) of northern Australia.. In the ARR the spatial dynamics of vegetation over time are largely driven by monsoonal rains and topography. Extensive dry season fires are also common in the region, temporarily reducing vegetation cover where they occur. Hence, hydrology, topography and fire disturbance interact to shape a distinct, yet dynamic, mosaic of different vegetation types of varying cover density. As a basis for calibrating the new vegetation function, the distribution and dynamics of vegetation are characterised for this savannah landscape by remote sensing, in conjunction with ground data.