Laserfiche WebLink
<br />CLARK MODEL SIMULATIONS OVER THE GUNNISON BASIN <br />FROM A HIGH RESOLUTION, NESTED MODEL <br /> <br />AN EXAMPLE <br /> <br />The Clark model was adapted to the Gunnison Basin and tested using <br />the terrain shown on Figure 1 and the 1200 UTC December 19, 1983, GJT <br />rawinsonde. This case was selected to test the model simulation of <br />typical airflow during conditions of moderate precipitation over the <br />basin when University of North Dakota's cloud-physics aircraft data were <br />available for comparison. The preliminary sensitivity analyses used a <br />42 by 42 by 24 (x,y,z) grid point domain with a coarse 10-km horizontal <br />grid spacing and an expanding 0.5- to 2.0-km vertical grid spacin~r. The <br />model was run in the non-hydrostatic mode with a time-step of 60 s using <br />Murray-Koenig parameterized microphysics to simulate precipitation. <br />Figure 5 shows the spatial distribution of cloud water SLW <br />(supercooled liquid water) in the surface layer after 120 minutes of <br />integration. Ice crystals (stippled pattern) developed over the major <br />ridges. The vertical structure of ice crystal concentration along line <br />AB is shown on Figure5a. Modeled spatial cloud and precipitation <br />patterns closely matched the terrain and propagated into the wind on the <br />upwind sides of ridges. This cloud evolution is consistent with <br />observations. <br /> <br />Aircraft observations at 3.7 km msl indicated winds veered from <br />290 to 300' and accelerated from 7 to 10 m 5-1 over the Grand Mesa. The <br />modeled winds within the surface layer over the Mesa (marked M in <br />Figure 5b) veered and accelerated. Modeled flow and cloud <br />characteristics ,are consistent with Holroyd and Super's (1984) <br />observations over the Mesa in the winter of 1983. <br /> <br />SUMMARY AND FUTURE STUDIES <br /> <br />This paper presents an overview of Reclamation's approach to <br />nested numerical modeling of precipitation in western Colorado for <br />GCCRP research. This collaborative study with NCAR scientists indicates <br />that the regional MM4 model provides useful climate information for <br />regional-scale climate simulations; however, nested local-scale lTIodels <br />are needed to resolve local effects of terrain on precipitation. Time <br />series of observed high elevation (>2700 m msl) and regional model <br />predicted daily cumulative precipitation matched reasonably well during <br />winter months of 1982-93; however, convective periods diverged <br />significantly. This suggests that MM4's regional climate simulations <br />may provide useful estimates of cumulative precipitation over a month <br />and season. A different pattern occurred in 1988 with a good ~atch of <br />model predictions and observations below 2100 m msl during January and <br />February; however, this pattern diverged from March to December. We are <br />currently examining the sources of this divergent pattern. <br /> <br />The local-scale Clark model simulated strongly forced orographic <br />cloud development that closely resembled observed clouds and airflow. <br />This model will be used to improve the numerical simulation of physical <br />processes within higher resolution nested MM4 and within local-scale, <br />steady-state models. <br />