The compatibility of erosion data at different temporal scales

Soil erosion processes have been studied intensively throughout the last decades and rates have been measured at the plot scale as well as at larger scales. However, the relevance of this knowledge for the modelling of long-term landscape evolution remains a topic of considerable debate. Some authors state that measurements of current rates are irrelevant to landscape evolution over a longer time span, as they are inconsistent with some fundamental characteristics of landscapes, such as the fact that the long-term sediment delivery ratio needs to be equal to 1 and that extrapolation of current rates would imply that all soils in Europe should have disappeared by now (e.g. Parsons, A.J., Wainwright, J., Brazier, R.E., Powell, D.M., 2006. Is sediment delivery a fallacy? Earth Surface Processes and Landforms 31, 1325–1328). In this study, we investigate if and to what extent estimates of long-term erosion rates are consistent with information obtained over much shorter time spans for the Loam Belt of Belgium.In a first step, observed short-term and long-term patterns in the Belgian loess area are compared statistically by classifying the study area into landscape element classes and comparing average erosion values per class. This analysis shows that the erosion intensities on the two temporal scales are of the same order of magnitude for each landscape element class. Next, the spatially distributed model WaTEM LT (Water and Tillage Erosion Model Long Term) is calibrated based on the available short-term data by optimising average erosion values for the same landscape element classes. Finally, the calibrated model is used to simulate long-term landform evolution, and is validated using long-term data based on soil profile truncation. We found that the model allows simulating landscape evolution on a millennial time scale using information derived from short-term erosion and deposition data. However, it is important that land use is taken into account for the calibration in order to obtain realistic patterns on a longer time scale. Our analysis shows that, at least for the study area considered, data obtained on erosion and deposition rates over various temporal scales have the same orders of magnitude, thereby demonstrating that measurements of current rates of processes can be highly relevant for interpreting long-term landscape evolution.