Laserfiche WebLink
<br />003495 <br /> <br />of these data, which are the best available in terms of areal coverage, <br />spatial resolution, length of record and accessibility, require several <br />assumptions whose validity is difficult to determine. Assumptions include: <br /> <br />1) That all storms can be identified even though soundings are taken <br />only twice each day, <br /> <br />2) That a single sounding can adequately represent all air about to <br />enter the state, and <br /> <br />3) That seedable storms can be identified using statistical criteria <br />based on analysis of experiments conducted in the Climax area of <br />Colorado. <br /> <br />The feasibility of predicting snowfall from storms identified from <br />interpolated soundings is tested by correlating the mnnber of storms <br />identified each winter with observations of water content produced by <br />actual snowfalL The procedure was refined by using a computer model to <br />adjust amounts of snow expected from each storm according to its 700 rob <br />wind direction and 500 rob temperature. Correlations indicate that <br />snowfall can be predicted from interpolated soundings with a fair degree <br />of accuracy. <br /> <br />Dry, normal, and wet winters are identified by comparing water content <br />from snowcourses in four "homogeneous" mountain .sub-regions of the state <br />to long-term averages. Percentages range from 62 percent to 150 percent. <br />Because interpolated soundings are not available for extremely dry winters <br />(like that of 1976-77, when snowpack in the south sub-region was only 35 <br />percent of the long-term average), conclusions cannot be drawn about <br />seeding opportunity during such periods. <br /> <br />The number of seedable events each winter is computed using criteria <br />from Mielke et al., (1981). Even though a few more seedable storms are.identified during wet than dry winters in each sub-region, this difference <br /> <br />36 <br />