Alpine grassland hydrologic response to climate change from plot to catchment scale

Warming and elevated CO₂ concentrations are expected to alter catchment hydrology through changes in precipitation and evapotranspiration. In particular, warming is expected to enhance evapotranspiration, whereas elevated CO₂ tends to decrease root-water uptake, thus reducing evapotranspiration. Plot-scale Lysimeter Temperature Free Air Carbon Enrichment (Lysi-T-FACE) systems provide in-depth information on the response of soil water fluxes to future climate conditions, particularly evapotranspiration and seepage. Hydrological models of different complexity can be used to extend the findings from the plot scale to the catchment level, allowing the assessment of the discharge response to climate change.

We run a climate change experiment by using lysimeters to study the effect of elevated CO₂ and warming on alpine grassland soil water fluxes. The experiment includes six lysimeters, with a reference lysimeter operating under ambient conditions, two lysimeters are treated with elevated CO₂ concentration of +300 ppm, two lysimeters are operating under constant warming of +3 K, and one operating under a combination of warming and elevated CO₂. Soil water fluxes within each lysimeter were modelled with the process-based hydrological model Hydrus-1D. We observed differences in seepage between the six lysimeters at both the event-based and annual time scale. For some individual events, such as the heavy rainfall event following a dry period in summer 2018, more remarkable differences between the experiments were observed.

To upscale the effects of the lysimeter-based approach to catchment scale, a conceptual lumped-parameter model (GR4J-Cemaneige) was used to model the discharge of a nearby alpine grassland catchment. The GR4J model reproduced discharge well when using lysimeter ET at ambient conditions (NSE>0.75). Evapotranspiration (ET) as input was modified based on the lysimeter ET fluxes representing possible future climate conditions. The effects of different ET inputs on simulated catchment discharge were similar to those on seepage at the plot level on an annual basis. However, no significant effects of different ET input on discharge were observed at individual events, such as the one in 2018. A comparison between a process-based hydrological model and the conceptual lumped-parameter model is planned to further investigate the effect of the hydrological response to climate change at the catchment scale.