Implications of processing spatial data from a forested catchment for a hillslope hydrological model

The objective of this study was to investigate with aid of a hillslope hydrological model (1) how landscape forms and land-use types control runoff generation and nitrogen leaching and (2) how the inevitable simplifications made when representing the entire catchment by a single hillslope (characteristic profile) affect modelled water and nitrogen fluxes. Analysis of the effect of hillslope shape parameters (e.g. length and slope) indicated that in forested hillslopes the modelled nitrogen export had a strong linear relationship with the share of subsurface runoff from the total runoff. The fraction of subsurface runoff, however, showed a non-linear relationship to changes in the profile length and slope. Decrease in length, and increase in slope, lead to a greater fraction of subsurface runoff. The non-linearity in the relationship implies that estimation of both the subsurface fraction and the nitrogen load is distorted when sub-areas producing large fractions of subsurface flow are aggregated with areas generating only little subsurface runoff into a single characteristic profile. In constructing the characteristic profile, aggregation of areas having different forest stand/soil properties causes the upslope drainage areas and the flow pathways to the stream to become distorted. The effects of such distortion were found to be particularly high for ammonium as its behaviour – due to efficient retention in soil – is highly non-linear with respect to the distance from the source area into the stream (i.e. size of buffer zone). In case of nitrate, which was considered not to be retained in soil, the role of groundwater table at the source area (clear-cut) was identified to be a more important control than the buffer zone size. As the upslope parts of a catchment have a significant role in determining the groundwater table at the treatment site, a question arises whether in addition to buffer zones one should also turn the attention upslope of the nitrogen source areas. Clearly patterns of landscape, such as distribution of nitrogen source areas in this study, pose challenges in selecting the spatial description for environmental simulation models. In nutrient leaching models, which easily become complex when based on process descriptions, compromises are needed to trade off between model complexity and the way how realistically landscape is described.