Hydrological extremes and feedbacks in the changing water cycle

The prospect of significant climate change over the next decades means that society must urgently face up to the implications for the changing water cycle, in particular increasing risks from extreme floods and droughts. Guidance for policy-makers to support adaptation to these risks and to support mitigation strategies to combat climate change is urgently required. However, the ability of climate scientists and hydrologists to predict the possible magnitudes of floods and droughts, and the frequency with which they are expected to occur, is still limited. This is due largely to gaps in knowledge of how best to use available data and models; of particular concern is the limited ability of climate and hydrological models to produce realistic extremes and changing hydrological behaviour. For example, regional climate models produce data which often requires to be downscaled to finer resolutions, but questions arise about what properties of the downscaled data are critical and how the downscaling should be done. As another example, drought planning requires inter-annual and inter-regional rainfall and evaporation to be represented accurately, however there is little guidance about how this can best be achieved under future climate using available models. In addition, there are weaknesses in the simulation of hydrology (for example, groundwater storage, river flows and evaporation) which act as hurdles to development of next generation climate models; in particular models currently used to simulate feedbacks from the earth surface to the atmosphere neglect important hydrological processes. This proposal will produce the science and models needed to address these questions, integrating climate and hydrological science to take impact modelling beyond the current state of the art. Specifically, the proposal: 1. exploits current generation climate science and statistical methods to improve and enhance projections of potential change in hydrologically-relevant metrics over a time-scale of 10 to 60 years, in particular extremes of heavy precipitation and drought; 2. builds on the analysis of historical data to improve scientific understanding and develop innovative methods for the modelling of extremes and non-stationarity in the hydrological response to climate variability; 3. seeks to improve the representation of hydrological processes in land surface models, in particular, the enhanced modelling of surface and subsurface processes for simulation of land-atmosphere feedbacks.