Can uncertain landscape evolution models discriminate between landscape responses to stable and changing future climate? A millennial-scale test

In the light of increasing societal interest in the effects of climate change, geomorphologists face the task of discriminating between natural landscape changes and landscape changes that result from human-induced climate change. Landscape Evolution Models (LEMs) are available for this purpose, but their application for prediction of future landscapes is problematic. Calibration of LEMs on a sufficiently long palaeo-record of landscape change solves some of these problems, but large uncertainties in input (e.g. climate) records and process descriptions remain.Using one of the few existing ka-scale LEM studies as starting point, this paper explores how uncertainty in the LEM LAPSUS (LandscApe ProcesS modelling at mUlti dimensions and scaleS, [Schoorl, J.M., Veldkamp, A. and Bouma, J., 2002. Modeling water and soil redistribution in a dynamic landscape context. Soil Science Society of America Journal, 66(5): 1610–1619]) affects its ability to discriminate future one-thousand year landscape change under stable climate from that under human-induced changed climate. Okhombe Valley in South Africa is used as a case study area. LEM uncertainty is characterized by different levels of parameter uncertainty.Results indicate that even under high levels of parameter uncertainty, LEM LAPSUS discriminates between responses to stable and changed climates for some zones in the landscape. Although confidence in model predictions remains limited, some explorative and relative conclusions about the effects of changed climate on future landscape evolution of Okhombe Valley are drawn.Finally, some possibilities and limitations of future studies on landscape evolution under changing climate are discussed.