Laserfiche WebLink
<br />Appendix 1 shows more detailed statistical comparisons of modeled proper- <br />ties and satellite observations of cloud structures. <br /> <br />The skill of the GPCM model in predicting the presence of clouds is <br />primarily a function of its ability to select cloud-base height. Initial <br />cloud updraft diameter and the initial impulse of 2 m/s determine the <br />relative intensity or depth of convective growth, but do not affect the <br />model's prediction of cloud occurrence. Cloud-base height is computed <br />from the initial sounding at the eCl usinq a mixing layer of the lowest <br />5 kPa. When the amount of surface temperature rise exceeds the climato- <br />logical extreme ranges of temperature, the model rejects the sounding as <br />requiring excessive surface heating for convective cloud initiation. <br />With this extreme temperature range criteria, the model tended to over- <br />predict cloud occurrence by 10 to 20 percent; however, statistical <br />summaries of model results were limited to those cases where convective <br />clouds were observed. <br /> <br />It should be noted that comparisons between results using the MESOCU <br />model (which incorporated mesoscale lifting as a triggering mechanism <br />when mesoscale lines and clusters were observed in satellite imagery) <br />and the GPCM results in the same sample showed that GPCM predicted clouds <br />in more cases than MESOCU, and MESOCU required lifting to initiate con- <br />vection. Surface heating alone in many cases was not sufficient to <br />trigger convective clouds in MEsoeu (Matthews and Silverman, 1980). <br /> <br />12 <br />