Spatial Variation and Feedbacks in Simulated Ecohydrologic Response to Climate Change

Abstract:

Climate change is altering a range of ecological and hydrologic processes, challenging the assumption of stationarity commonly applied in water resource assessments. Many studies have shown the reliability of widely used catchment-scale hydrologic models for streamflow prediction under climate change. However, less attention has been paid to how nonstationarity in runoff processes might evolve spatially throughout a catchment. We apply a process-based distributed ecohydrologic model (RHESSys) to run a series of virtual catchment experiments assuming different climate forcing and catchment properties, allowing us to examine how terrain, aspect and soil type can influence spatially variable patterns of vegetation and hydrologic responses to climate change. We show that complex interactions between biological and physical processes manifest as distinctive feedbacks that have an important effect on trends in water balance variables across different parts of the landscape. For example, the influence of increasing CO2 and increasing temperature together on vegetation growth and runoff response is shown to be substantially different to the sum of influences from each change separately. We also find that topographic position has a large impact on spatial patterns of runoff generation and subsurface moisture trends. The impacts experienced under a drying scenario show patterns entirely distinct from the impacts under wetting (i.e. the respective trends are not simply the reverse of each other). We find that climate change impacts on vegetation are likely to be different across a catchment, thus requiring detailed ecohydrologic modelling to inform future management. Our work also shows that, regardless of how well we may understand the ecohydrologic impacts of individual climate variables, interactions among forcing variables are important for understanding complex system response. It is unlikely that applying ‘fudge factors’ (e.g. a percentage decrease in transpiration for a given CO2 increase) in conceptual models will robustly improve water management under future climate change.