Water dynamics in the soil-plant-atmosphere ecosystem play an essential role in basic science and practical applications for terrestrial ecosystem and climate impact assessments. An interdisciplinary coupling of models spanning wide spatial-temporal scales is challenging since governing laws, system properties, and paradigms can vary substantially with scale. By using grid nesting, time-splitting, and a flux coupler we are coupling the NCAR mesoscale model (MM5), the USDA Soil Water Assessment Tools (SWAT), and California Environmental Resources Evaluation System (CERES) together to form a two-way coupled soil-plant-atmosphere agroecosystem model. When provided with appropriate initial and boundary conditions, the coupled model system will be run in forecast mode, projecting seasonal crop-available water and thereby allowing evaluation of alternative cropping strategies. The model can also be used for agroecosystem simulations, presenting a powerful tool for understanding physical, biochemical, biophysical, and hydrological processes in agroecosystems.

This fully integrated modeling system brings together the best available tools and understanding for simulating plant physiology, soil science, microclimatology, surface hydrology, regional climate projection. Preliminary results indicate that this coupled modeling system can simulate reasonably well agroecosystem water processes such as evapotranspiration, root water uptake, surface runoff, etc. The model is also able to capture soil moisture temporal variation and spatial distribution, especially soil moisture depletion early in a growing season.