Modeling of Hydrological Cycle and its Interaction with Vegetation in the Framework of Climate Change

There is a growing interest to extend climate change predictions to smaller, catchment-size scales and identify their implications on hydrological and ecological processes. This thesis presents a blueprint methodology for studying climate change impacts on eco-hydrological dynamics at the plot and catchment scales. A weather generator, AWE-GEN, is developed to produce input meteorological variables to eco-hydrological models. The weather generator is also used for the simulation of future climate scenarios, as inferred from climate models. Using a Bayesian technique, a stochastic downscaling procedure derives the distributions of factors of change for several climate statistics from a multi-model ensemble of outputs of General Circulation Models. The factors of change are subsequently applied to the statistics derived from observations to re-evaluate the parameters of the weather generator. The time series obtained for present and future climates serve as input to a newly developed eco-hydrological model Tethys-Chloris. The methodology is applied to simulate the present (1961-2000) and future (2081-2100) hydrological regimes for the area of Tucson (AZ, U.S.A.). A general reduction of precipitation and a significant increase of air temperature are inferred with the downscaling procedure. The eco-hydrological model is successively used to detect changes in the surface water partition and vegetation dynamics for a desert shrub ecosystem, typical of the semi-arid climate of southern Arizona. An appreciable effect of climate change can be observed in metrics of vegetation performance. The negative impact on vegetation due to amplification of water stress in a warmer and dryer climate is partially offset by the effect of the augment of carbon dioxide concentration. Additionally, an increase of runoff and a depletion of soil moisture with consequence in deep recharge are detected. Such an outcome might affect water availability and risk management in semi-arid systems.