CLIMATE CHANGE CONTRIBUTION TO UNCERTAINTY IN VADOSE ZONE THERMAL, HYDROLOGICAL AND GEOCHEMICAL PROPERTIES

The vadose zone is a complex environment in which thermal-hydrological-geochemical (THC) processes couple in non-linear ways to produce an information-rich geological system. Progress in obtaining knowledge from that information reservoir has been steady, but has been hampered by the computationally intensive effort required to interpret the expression of those processes. Simulations are needed to understand the past and future time-dependent responses of these systems to natural and human perturbation, and are important for those instances in which the long term performance of sub-surface engineered facilities may impact future health and safety of the human population and the environment. We evaluated the vadose zone response to climate change for times periods between 100 and 100,000 years. The questions we considered were 1. What is the natural response time of a typical vadose zone system to local perturbations in temperature and infiltration flux at the ground surface, and 2. What is the expression of those perturbations as they propagate through the hydrological and geochemical systems? We conducted this study using a THC reactive transport code (NUFT-C) that simulates multi-component, multi-phase, saturated and unsaturated reactive transport for non-equilibrium, non-isothermal systems. The code runs on a massively parallel IBM SP-2 computer (1200 processors) allowing us to conduct high resolution simulations of complex systems. The characteristic response times for thick vadose zones to achieve steady-state conditions after a change in climate are on the order of thousands of years. This time is much longer than recently reported climate variability in the Southwest [Allen, 2000]. This implies that much of the vadose zone in the Southwest may be in the process of adjusting to relatively recent, post-glacial climate changes, and is not at steady-state. The THC response of such systems is path dependent. Uncertainty in these simulations reflect a paucity of rate constant data for most mineral phases of interest, a lack of knowledge concnering the three dimensional permeability structures, and the initial distribution of secondary mineral phases. These problems make it difficult to obtain accurate and precise initial and boundary conditions for simulations.