The U.S. Army operates test ranges for evaluating materiel that is being considered for procurement. Operational weather analysis and forecast systems are required for test planning, ensuring range safety, satisfying environmental constraints and determining the meteorological conditions at the time and place of specific materiel tests. NCAR scientists and engineers (T. Warner, L. Carson, F.Chen, C. Davis, D. Gill, S. Henry, H.M. Hsu, E. Jones, C. Mueller, S. Low-Nam, W. Myers, T. Sandblom, A. Shantz, S. Swerdlin and R. Weekly) are developing an operational weather analysis and forecast system to satisfy these needs. It consists of the following components.
B. The WDTC Operational System
It is well-known that the time scale for the predictability of atmospheric motions with horizontal scales of a few kilometers is typically less than an hour. Significant exceptions to this rule may be found when external forcing is stationary or regular, and when this forcing exerts a marked influence on flow behavior. Such is frequently the case in regions of complex terrain where channeling and diurnal forcing can dominate the large-scale forcing, leading to the possibility of meso-gamma scale predictability on time scales of many hours. These dynamical assumptions form the foundation of the operational forecast and data assimilation system that is being developed for the WDTC in western Utah which is characterized by steep terrain, arid conditions and complicated diurnal circulations that result from large contrasts in surface characteristics. The high resolution used in our forecast system is essential for capturing the variations of terrain and surface characteristics that drive local circulations.
The forecasting and data assimilation system is based on the Penn State/NCAR mesoscale model, version 5 (MM5), which is a general-purpose, nonhydrostatic system. The computational domain has four two-way-interacting nested grids with grid increments of 30 km, 10 km, 3.3 km and 1.1 km; the four grids are shown in Figure 1. Each day, the run cycle consists of two forecasts, one of 14 hours duration and one of 24 hours duration, and one 24 hour data assimilation cycle that produces an analysis for the previous day. The coarsest grid obtains its lateral boundary conditions from the National Center for Environmental Protection early-Eta model forecast, and the initial conditions are defined based on standard National Weather Service data as well as local surface data, soundings and profiler data. The complexity of the terrain is illustrated in Figure 2 for the 3.3 km grid.
Operational forecasts from the prototype system have been produced since August 1997, and have shown the existence of a remarkably rich array of local mesoscale structures that are forced by the orography and other local variations in the surface characteristics. One example of a distinctive local phenomenon is the "playa breeze," which is caused by differential heating associated with the contrast between salt flats, having a high albedo and high moisture availability, and the surrounding desert surface. Figure 3 shows the models 10 hour forecast, initialized at 1200 UTC 9 September 1997, for the 3.3 km domain; 40 m AGL wind vectors and potential temperature are illustrated. Figure 4 shows the model-predicted 40 m AGL streamlines for this time, and observations from local surface mesonet stations. The flow in the northern half of this domain represents the playa breeze directed from the cooler to the warmer land surface. The correspondence between the model solution and the observations is clearly very good.
Forecasts from the MM5 system will continue to be validated, and the model will be improved based on any shortcomings that are isolated in this validation process. The MM5 system for White Sands Missile Range will undergo development and testing during this same period.
