Our current land surface energy flux model is SHEELS - Simulator for Hyd rology and Energy Exchange at the Land Surface. The physics of SHEELS are based on those of the Biosphere-Atmosphere Transfer Scheme (BATS) (Dickinson et al., 1993). BATS was developed in order to provide the NCAR Community Climate Model w ith a realistic treatment of surface-atmosphere heat and moisture exchange by ex plicitly formulating the physiological, hydrological and radiative controls. SH EELS has inherited that model's physical treatment of vegetation properties. Th e one-dimensional model uses a single canopy layer but allows for fractional cov erage of the ground by vegetation. The nested soil layer approach of BATS has b een converted to a discrete layer configuration in SHEELS. Other modifications from BATS include a simplification of the radiation scheme to utilize measured r adiative fluxes and a more refined treatment of soil thermal and hydraulic prope rties (Smith et al., 1993, Laymon et al., 1995). Its uniqueness lies in its tre atment of surface variability of vegetation properties obtained from high resolu tion remote sensing and its use of soil properties obtained from the Soil Conser vation Services digital State Soil Geographic (STATSGO) database. Landcover inf ormation is obtained from remotely sensed landcover classification images.

The SHEELS model represents an attempt to describe the entire physical s tate of the surface, planetary boundary layer and subsurface. The system is driv en by eight atmospheric forcing variables, which are supplied by field or throug h interface with an atmospheric model. Separate visible and near-infrared albed os are determined for the canopy based on vegetation type and solar zenith angle . Soil albedo is modeled using a soil moisture-dependent formulation. Other ke y biophysical properties of the canopy are the canopy height and leaf and stem a rea indices. Leaf water quantities consist of the depth of dew on leaves and ca nopy interception of water. Parameters involved in heat and moisture exchange b etween the canopy and the surface layer include leaf and in-canopy temperatures and stomatal and aerodynamic resistances.

The subsurface consists of three discrete layers, each of which is treat ed differently with respect to soil moisture, root distribution and temperature. The two uppermost layers contain the roots and constitute the thermally active layer. The lowest layer has constant temperature, while hydrologic processes a re active. Three interrelated soil moisture quantities are computed for each la yer: depth of water and fractional water contents by volume and by mass.

Universal and model constants which must be specified within SHEELS may by divided into two categories. Those which do not depend on site characteristi cs (such as topography, soil type or vegetation species) are referred to as pre- defined constants. Many of these are well-known universal constants; however, t here are a number of others whose values are less well established. These are p arameters relating to canopy radiative and morphological properties and soil com position and thermal properties, and are typically determined from values stated in the literature. In addition, there are approximately 20 constants which inv olve specific knowledge of vegetation and soil properties representative of the site under study. These parameters may be defined based on published values or through model optimization procedures.

A new, two-dimensional version of SHEELS is currently being developed at MSFC. This new version will account for the effects of topography as well as su rface and vadose zone water and energy fluxes, thereby offering the potential to study and develop parameterization schemes for a range of scales from large cat chment to mesoscale. This model will incorporate pixel-scale variability of sur face and sub-surface properties, as well as more sophisticated and physical trea tments of hydrologic processes, including overland flow and runoff into streams, lateral sub-surface flow, infiltration and percolation. The hydrologic process es will be interactive, i.e. water will be routed from one grid cell to adjacent cells both on and below the surface. The conceptual basis for the surface laye r will remain the same, with atmospheric forcing being used to drive the model. Soil properties such as density, porosity, rooting depth and hydraulic conducti vity will be allowed to vary spatially, allowing for extensive use of soils info rmation. Land cover classification will be used to specify vegetation character istics at the pixel scale.