Laserfiche WebLink
<br />Our study focuses on the Gunnison Basin in western Colorado marked <br />by the bold square in Figure 2 nested within the domain analyzed by MM4. <br />TheMM4 model provides the large-scale (synopti.c) and regional-scale <br />(mesoscale) forcing as shown by the summer monsoon flow over the West in <br />Figure 2a and corresponding 24-h precipitation in Figure 2b. Local- <br />scale Clark model simulations will examine the detailed evolution of <br />clouds and precipitation within the 440-, by 440-km domain (dx=10 km) <br />marked by the bo1d square in Figure 2a. Within this area, the dashed <br />square defines the nested 5-km resolution, 360- by 360-km domain. Small <br />dots show each MM4 grid point and large dots indicate the CCM(Conununity <br />Climate Model) R-15 (rhomboidal' spherical harmonic truncation at wave <br />number 15) 500-km resolution grid points. Clearly, regional-scale <br />models are required to capture valuable mesoscale airflow and <br />thermodynamic information. <br /> <br />Reclamation is using two models to simulate local orographic <br />precipitation: Rhea and Clark models. Medina (1992) has adapted the <br />steady-state Rhea model (Rhea, 1977) to simulate local precipitation <br />. over the Gunnison River Basin in Colorado and has adapted the model to <br />use initial conditions from soundings derived from MM4. This <br />approximates a one-way, local-scale nesting. Comparisons between Rhea <br />model simulations from rawinsondes and MM4-derived soundings are <br />encouraging; however, convective precipitation is not accurately <br />simulated in either the MM4 or Rhea models. <br /> <br />Therefore, we are also using the more sophisticated, three- <br />dimensional, time-dependent, local-scale Clark model to simulate the <br />local evolution of convective clouds and precipitation over complex <br />terrain in the centra'l Colorado Rockies. The Clark model is used <br />extensively throughout the research community to simulate phenomena <br />ranging from airflow over complex terrain to deep convection (Clark, <br />1979). The model is an anelastic, finite difference model with one- and <br />two-way interactive nesting described by Clark (1977), and Clark and <br />Hall (1991). It uses a terrain-following coordinate system to <br />accurately describe orographic effects on airflow. It includes warm <br />rain and ice microphysical parameterizations that explicitly simulate <br />precipitation development. The Clark model will be used to understand <br />the complex three-dimensional airflow and evolution of clouds and <br />precipitation in a limited sample of cases. It will contribute to <br />better parameterizations of airflow structure for the Rhea model, which <br />may be run on a large number of cases. Initial Clark model simulations <br />are examining model sensitivity to initial conditions and increased <br />resolution of complex terrain in western Colorado using rawinsondes and <br />MM4 soundings. Later,' a four-dimensional dynamic data assimilation <br />method, currently under development by scientists at the MMM (Mesoscale <br />and Microscale Meteorology) Division of NCAR, will. be used to assimilate <br />the MM4 dynamics into the Clark model initializati.on process. <br /> <br />INITIALIZATION AND ASSESSMENT OF MODE:L SIMULATIONS <br /> <br />Giorgi and Bates (1989) began a validation of the CCM-MM4 nested <br />model'ing approach using observations for 40 months from January 1982 to <br />December 1983, and January 1988 to April 1989. They used initial <br />conditions from ECMWF (European Center for Medium-range Weather <br />Forecasts) analyses of observed data gridded to a spatial resolution of <br />2.8 by 2.8. and having temporal resolution of 12 h. MM4' s lateral <br />