Catchment area veils land use and topographic controls on runoff generation in Swiss catchments

Distinct effects of land use and topography on runoff metrics are well established at small scales, yet their identification in large sample studies remains challenging, as they are often masked by other dominant catchment features. Here, we systematically investigated the drivers of runoff response across a rich dataset of catchment properties from 152 Swiss catchments using runoff metrics derived from non-linear Ensemble Rainfall-Runoff Analysis (ERRA).

We show that catchment area exhibits the strongest control on both the timing and magnitude of streamflow response to precipitation. We found that across Switzerland runoff response is not related to catchment slope (neither mean slope nor fraction of steep or flat terrain), even when clustering catchments of similar area. Thus, contrary to widely circulating assumptions, steeper catchments do not exhibit faster or stronger runoff response. We also tested the differences between forest and agriculture dominated catchments and found no statistical differences in timing and magnitude of streamflow response to precipitation when using the entire dataset. Thus, our Swiss wide analysis does not show the expected buffering effect of forests and faster responses in agricultural landscapes. Land use effects only emerged when stratifying catchments by area and assessing the runoff response for different precipitation intensities. In agriculture dominated catchments, we observed higher peak flow with intense precipitation compared to forest dominated catchments. Thus, the buffering effects of streamflow response to precipitation in catchments with different land use are non-linear and dependent on rainfall intensity.

Our results caution against generalized assumptions in large sample hydrology, because effects of topography and land use are strongly modulated and often veiled by other catchment properties or climate and dependent on rainfall intensity.