Projected changes in Rhine River flood seasonality under global warming

Abstract

Abstract. Climatic change alters the frequency and intensity of natural hazards. In order to assess potential future changes in flood seasonality in the Rhine River basin, we analyse changes in streamflow, snowmelt, precipitation and evapotranspiration at 1.5, 2.0 and 3.0 ∘C global warming levels. The mesoscale hydrological model (mHM) forced with an ensemble of climate projection scenarios (five general circulation models under three representative concentration pathways) is used to simulate the present and future climate conditions of both pluvial and nival hydrological regimes. Our results indicate that future changes in flood characteristics in the Rhine River basin are controlled by increases in antecedent precipitation and diminishing snowpacks. In the pluvial-type sub-basin of the Moselle River, an increasing flood potential due to increased antecedent precipitation encounters declining snowpacks during winter. The decrease in snowmelt seems to counterbalance increasing precipitation, resulting in only small and transient changes in streamflow maxima. For the Rhine Basin at Basel, rising temperatures cause changes from solid to liquid precipitation, which enhance the overall increase in precipitation sums, particularly in the cold season. At the gauge at Basel, the strongest increases in streamflow maxima show up during winter, when strong increases in liquid precipitation encounter almost unchanged snowmelt-driven runoff. The analysis of snowmelt events for the gauge at Basel suggests that at no point in time during the snowmelt season does a warming climate result in an increase in the risk of snowmelt-driven flooding. Snowpacks are increasingly depleted with the course of the snowmelt season. We do not find indications of a transient merging of pluvial and nival floods due to climate warming. To refine attained results, next steps need to be the representation of glaciers and lakes in the model set-up, the coupling of simulations to a streamflow component model and an independent validation of the snow routine using satellite-based snow cover maps.