Main content area

Sensitivity of the PESERA soil erosion model to terrain and soil inputs

Baggaley, Nikki, Potts, Jacqueline
Geoderma regional 2017 v.11 pp. 104-112
landscapes, risk, runoff, sediment yield, soil, soil erosion, soil erosion models, standard deviation, watersheds
Spatially distributed modelling can be a useful tool for identifying areas at risk of soil erosion. The PESERA soil erosion model was developed to be run on a 1km grid at a continental scale. When applying a regional model such as PESERA at a catchment scale it is important to understand how changing the scale of the input data affects the outputs. In PESERA the terrain variable is defined as the Standard Deviation of the elevation values within a block around the grid square of interest. In this paper we show that the scale at which the terrain is represented has an impact on not just the absolute values of predicted sediment yield, but also on the spatial pattern of the modelled erosion risk. The way in which the terrain influence on soil erosion is presented in PESERA means that decreasing the grid and block sizes to better represent changes in slope will decrease the modelled sediment yield and change the spatial pattern of risk. We recommend that, due to the interaction between terrain and soil parameters, if spatially distributed soils data are available, global input parameters such as the runoff fraction should not be used to calibrate the model to compensate for the lesser predicted sediment yields, when the grid size is reduced.