Analyses of multiple years of data from WSR-88Ds and rain gauges reveal a distinct diurnal pattern in average frequency of warm-season rainfall between the Rocky Mountains and the Mississippi Valley. During the afternoon, a peak in frequency occurs over the high terrain of the Continental Divide. The peak then moves eastward with remarkable coherency at a speed of 16 m/s, crossing the Great Plains during the evening and night. After sunrise the next morning, the peak frequency in rainfall arrives at the Mississippi Valley.

Unfortunately, the climatological patterns of simulated precipitation in mesoscale forecast models do not look like the observations. Although models do produce an afternoon maximum in frequency over the Continental Divide and a somewhat distorted morning maximum over the Mississippi Valley, they completely fail to produce a realistic pattern of average frequency over the High Plains.

The reasons for the failure are almost certainly complex and will not easily be solved. Even so, some insight into the problem may be gained through case studies and idealized simulations. In his poster, the presenter will show results from tests of the Weather Research and Forecast (WRF) Model in which he systematically investigated specific instances where the model had difficulty capturing observed up-scale organization and propagation of cumulonimbi immediately to the east of the Rocky Mountains. The presenter will focus on issues such as model resolution, microphysical and convective parameterizations, representation of the land surface, and representation of the boundary layer.