In order to make the best use of limited water resources in a region, it is important to understand the entire hydrologic cycle and to be able to explore the potential effects of increased use and of changes in the regional climate. To calculate water resources within the Rio Grande Basin, we are coupling a series of existing and previously tested models that address the multitude of physical processes and temporal and spatial scales that are important. These models include the Regional Atmospheric Modeling System (RAMS), which provides meteorological variables and precipitation to the Los Alamos Distributed Hydrologic System (LADHS), a hydrology model. The Finite Element Heat and Mass (FEHM) model is being added to the system to include ground water in the simulations.

The RAMS grid configuration includes three nested grids, with 5 km horizontal grid spacing over the upper Rio Grande basin, where the LADHS domain employs 100 m horizontal grid spacing. Thus, one component of the coupled system includes a method to downscale the RAMS predictions to the 100 m grid required by LADHS. Keeping the technique simple yet reflective of the influences of the complex topography in the area, downscaling is accomplished by a linear prediction model (the same model that underlies "kriging").

In earlier work, we have presented evaluations of RAMS predicted precipitation based on a simple bi-linear interpolation of the predictions to observation sites of the Summary of the Day Cooperative observation and SNOTEL networks and a comparison of the interpolated predictions to the observed precipitation at those sites. This paper will evaluate the precipitation fields, after they have been downscaled in LADHS. It will compare the downscaled precipitation to the observed precipitation data and the interpolated RAMS precipitation from the 1992-1993 water year.