WMOD00444 - Laserfiche WebLink

Home Browse Search

, ",I An attempt was made to improve the relationship by excluding sites below 7000 feet elevation. All 14 control sites were above 7000 feet, but only 4 target sites were that high. Seeding might be expected to have the greatest effect on these higher target sites. Little improvement in the R-value resulted from this computation, which produced a value of 0.88l. Although the plot is not shown, the 1993 data point almost reached 2 standard errors of estimate. One nonseeded year had a higher positive departure, and one had almost as high a departure. The low correlation coefficients between the target and control sites, and therefore low predictability, are suspected to partially result from the difference in mountain barrier orientation. The axis of the target area mountains is generally east-west; that of the control site barriers is approximately northwest-southeast. Orographic lift, liquid water condensate production, and subsequent snowfall tend to be greatest when winds are almost perpendicular to mountain barriers. Therefore, the t:arget and control areas can be expected to be most affected by different phases of a storm or even by different storms. A further attempt to improve the degree of association was made by examining R-values between the average of the 9 target sites above 6500 feet and individual control sites. Seeding was expected to have a higher probability of affecting these higher target sites. Each individual control site in tables 1 and 2 was used in these calculations. Individual R-values ranged from 0.67 for the distant site 47 (southwest of Boise) to 0.927 at site 33, west of the target. Surprisingly, four of the five highest values, from 0.876 to 0.900, occurred at control sites northwest of Boise. That region consists of several north-south ridges, dissimilar from the east-west ridge that forms the target area. It is speculated that the high R-values associated with the distant sites northwest of Boise occurred because that region is similar to the target region for storms tracking from the south to southwest. Few significant mountains exist for such storm tracks for a considerable distance upwind from either the northern upper Snake target or the control area northwest of Boise. In contrast, nearby controls west to northwest of the northern upper Snake target are located downwind from the Boise target, and high mountains east of that target, for south to southwest flow. Only three sites initially selected as controls for the northern upper Snake, and listed in table 2, had R-values above 0.865. Several of these nearby potential control sites had R-values in the relatively low 0.67 to 0.80 range. A reasonably high R-value was sought by combining control sites from tables 1 and 2 that had individual R-values exceeding 0.85. The highest value obtained was 0.943 using the four Boise northwest controls (sites 4, 5, 15 and 48), sites 32 and 33 west of the target, and site 36 northwest of the target. The results are shown on figure 10 where the 1993 data point is just above the regression line, equivalent to a 5-percent (1.0-inch SWE) increase. Similar 23