Integrated terrestrial simulations at the continental scale

Abstract

The availability of freshwater under current and future climatic conditions remains one of the main research topics that requires an improved understanding of the terrestrial system and anthropogenic impacts. Yet, hydrologic and atmospheric studies are commonly performed in a disjunct fashion even though hydrologic and atmospheric processes are intrinsically connected through the terrestrial water, energy, momentum and matter cycles. In particular, current continental-scale process and assessment studies do not close these cycles, neglect groundwater dynamics and miss human impacts related to water use. Continental-scale, coupled modeling systems are required to improve process representations of the terrestrial cycles and their feedbacks. In this thesis, a fully-coupled, groundwater-land surface-atmosphere modeling system, the Terrestrial Systems Modeling Platform TerrSysMP, is setup over Europe, in order to simulate the full terrestrial water cycle from groundwater across the land surface into the atmosphere. The model setup is designed to perform sensitivity studies, which assess the impact of (i) groundwater dynamics and (ii) human water use on the continental-scale terrestrial water cycle during the heatwave year in 2003. Beyond that, the simulations are used to define (iii) the associated added value of incorporating human water use.<br /> First, over continental Europe, the impact of groundwater dynamics on associated land-atmosphere feedbacks during the peak of the heatwave is evaluated. Results illustrate the potential of groundwater dynamics to dampen heat extremes, especially in regions of shallow water tables, and uncover potential deficiencies of current assessment studies. Ensemble simulations are performed and address the intrinsic uncertainty of land-atmosphere feedbacks to subsurface characteristics, such as soil texture and the incorporation of an underlying geology. This study shows that the groundwater representation induces variability across the compartments of the terrestrial system, with a decreasing impact from the subsurface over the land surface towards the atmosphere. The impact of groundwater is strongly dependent on the prescribed subsurface characteristics, i.e. the distribution of hydrofacies and soil hydraulic parameters.<br /> Second, the incorporation of human water use related to groundwater abstraction and irrigation, and its impact on water availability and fluxes, is evaluated. Four water use scenarios are constructed to account for the uncertainty of human water use and the associated land-atmosphere feedbacks. The simulations indicate that atmospheric feedbacks to human water use may contribute systematically to continental drying and wetting in Southern Europe. Moreover, the simulated land-atmosphere feedbacks exceed net values of human water use (i.e. irrigation minus groundwater pumping) at the watershed scale, and significantly affect atmospheric moisture transport across watersheds. These results emphasize that effects of local human water use on water availability and sustainability may be of global importance.<br /> Finally, the added value of incorporating human water use in terrestrial simulations and predictions of evapotranspiration and precipitation is examined. Multiple observational data sets are used to assess observation uncertainty and model skills. The results illustrate model deficiencies, which mainly arise from the simulation of precipitation and an overestimation of extreme precipitation events. Observational uncertainties, however, exceed the impact of human water use on precipitation and evapotranspiration at the watershed scale, and impede model validation. At the local scale, human water use significantly affects daily precipitation and evapotranspiration, and potentially improves the model skill through an improved simulation of local precipitation.<br /> Integrated feedbacks from groundwater to the atmosphere, and through the incorporation of human water use, are not yet considered in regional climate simulations, reanalyses and water resource assessment studies, but constitute important processes and additional uncertainties in simulations of the terrestrial water cycle. Therefore, this thesis emphasizes the need and advantages of continental-scale, coupled modeling systems, to advance impact studies across compartments of the Earth systems.