Sensitivity of hydrological predictions to ecosystem adaptation in response to climate change: the effect of time-dynamic model parameters

Typically, the future hydrological behavior of a river basin, for example as a result of climate change, is predicted using hydrological models calibrated with historical observations. In reality, hydrological systems, and hence model parameters, experience almost continuous change in time and space. More specifically, there is growing evidence that vegetation adapts to changing conditions by adjusting its root-zone storage capacity, i.e. the amount of water in the unsaturated subsurface which is available to the roots of vegetation for transpiration. Additionally, other species may become dominant under natural and anthropogenic influence. In this study, we test the sensitivity of hydrological model predictions to changes in vegetation parameters that reflect ecosystem adaptation to climate and potential land-use changes. In other words, if the climate changes, how should our models change and what is the effect on the hydrological response? Our methodology directly uses projected climate data to estimate how vegetation adapts its root-zone storage capacity at the catchment scale to changes in hydro-climatic variables and potential land-use change. We test the hypothesis that changes in the hydrological response under global warming are more pronounced when explicitly considering changes reflecting adaptation of the root-zone storage capacity of vegetation. We compare a stationary benchmark model with several non-stationary model scenarios reflecting climate and potential land-use changes in the Meuse basin. We found that the larger root-zone storage capacities (+34%) in response to warmer summers under projected +2K global warming result in up to -15% less streamflow in autumn due to up to +14% higher summer evaporation in the non-stationary scenarios compared to the stationary benchmark scenario. By integrating a time-dynamic representation of changing vegetation properties in hydrological models, we make a potential step towards more reliable hydrological predictions under change.