Streamflow recession patterns can help unravel the role of climate and humans in landscape co-evolution

Abstract. Traditionally, long-term predictions of river discharges and their extremes include constant relationships between landscape properties and model parameters. However, due to the co-evolution of many landscape properties more sophisticated methods are necessary to quantify future landscape–hydrological model relationships. As a first step towards such an approach we use the Brutsaert and Nieber (1977) analysis method to characterize streamflow recession behaviour of  ≈  200 Swedish catchments within the context of global change and landscape co-evolution. Results suggest that the Brutsaert–Nieber parameters are strongly linked to the climate, soil, land use, and their interdependencies. Many catchments show a trend towards more non-linear behaviour, meaning not only faster initial recession but also slower recession towards base flow. This trend has been found to be independent from climate change. Instead, we suggest that land cover change, both natural (restoration of natural soil profiles in forested areas) and anthropogenic (reforestation and optimized water management), is probably responsible. Both change types are characterised by system adaptation and change, towards more optimal ecohydrological conditions, suggesting landscape co-evolution is at play. Given the observed magnitudes of recession changes during the past 50 years, predictions of future river discharge critically need to include the effects of landscape co-evolution. The interconnections between the controls of land cover and climate on river recession behaviour, as we have quantified in this paper, provide first-order handles to do so.