Laserfiche WebLink
<br />Model reliability was analyzed using satellite observations of clouds <br />observed at 2300 G.m.t. in 1976 and 1977. Results indicated that the model <br />could diagnose convective clouds on nearly all days that they occurred; <br />however, in 10 to 20 percent of the cases the model diagnosed clouds when <br />none were observed within 250 km of the field site. Table 2 summarizes the <br />model reliability in terms of observed convective clouds. In 1976 all <br />convective clouds within 250 km of the rawinsonde were compared. Note <br />that the average separation between National Weather Service rawinsondes <br />is approximately twice this distance in the High Plains. Results in 1977 <br />were divided into two classes - all clouds occurring within 75 km of the <br />site and all clouds within 250 km of the site. For instance, Goodland's <br />(GLD) 1977 results showed that the model missed 14 percent of the cases <br />when all clouds within 250 km were compared; however, it missed 36 percent <br />of the cases within 75 km of the site. These results indicate that the <br />model was able to diagnose the wide area potential for clouds (76 to <br />86 percent correct) better than the site potential. This in part was due <br />to the need for mesoscale triggering to relea~e the convective instability, <br />as discussed by Matthews and Silverman (1980), using similar rawinsonde <br />data from this region. The GPCM model does not explicitly account for this <br />release mechanism. Mesoscale cloud lines were often observed beyond 75 km <br />from GLD, while GLD remained in the clear, hence, poorer verification <br />there. The three-dimensional analysis of mesoscale triggering through <br />low-level convergence, upper-level divergence, local orographic and <br />isentropic lifting should be performed. The lack of this triggering is a <br />critical limitation of one-dimensional, steady-state cloud models. <br /> <br />11 <br /> <br />. !I <br />