Laserfiche WebLink
, <br /> April 2009 Silverman 11 <br /> the streamfiow at the control station FRR. The This possible explanation was motivated by the <br /> regression results should be accurate and robust analysis of the Colorado River Basin Pilot Project <br /> since there were no outliers in the data and the by Elliott et al. (1978)who found that under low- <br /> regression residuals exhibited homoscedacity level stable conditions the silver iodide was trans- <br /> (constant variance). ported northwestward parallel to the mountain <br /> barrier instead of northeastward and up into the <br /> clouds over mountain as intended.This and other <br /> Table 2.Linear and multiple regression entity- pommy explanations warrant further Inveetiga- <br /> sis results for GBH against each potential tion and verification through appropriate meteor- <br /> control alone and the indicated combination of logical and hydrological studies. <br /> controls, respectively, for the entire period of <br /> analysis(including both the historical and op- Tabl 3.Results of the Vail evaluation for each <br /> eratianal periods). of the primary targets.Results are given for the <br /> Control Corr.Coed[ Std Dev Res proportional effect of seeding,8(%)=100* <br /> P se(AF) (RR-1),where RRA is the bias-adjusted re- <br /> _ gression ratio,p is the correlation between the <br /> WSF 0.679 1.602 inctcated target and the control(FRR),and <br /> WNF 0.718 1.520 CI90L and CI90U are the lower and upper <br /> bound of the 90%confidence interval,respec- <br /> WDC 0.703 1,581 tively.Statistically significant results In accor- <br /> dance 0.77b 1, dance with a 2-sided level of significance of <br /> 0.10 are shown in bold. <br /> WSF.FRR 0.788 1,420 Target 8 CI90L CI9OU p • <br /> WNF,FRR 0.771 1,407 <br /> WSF,Wt7C 0.732 1,505 PNY +8.3 +0.4 +12.5 0.910 <br /> GBO +9.3 +1.1 +18.1 0.836 , <br /> WNF.WDC 0.716 1• MID +7.9 -0.2 +16.7 0.909 <br /> 4. EVALUATION RESULTS GPT +18.5 +7.3 +30.9 0.812 <br /> The evaluation results for the primary targets are <br /> GBH +28.8 +16.8 +42.2 0.775 <br /> given in Table 3. It can be seen from Fig. 2 that SUP +11.1 +4.7 +18.0 0.837 • <br /> the maximum seeding effect is centered on GBH GBL +4.8 -0.6 +10.1 0.918 <br /> and decreases for targets both northwest and <br /> southeast of t3BH. The seeding effect for GBH TMW -2.0 -11.5 +8.3 0.871 <br /> and all the targets north of it are impressively <br /> large and statistically significant except for MID Si <br /> which is almost,but not quire, statistically signiff- 31 rasa <br /> cant The seeding effect for targets south of GBH <br /> • decreases rapidly with increasing distance from E >a <br /> GBH.The seeding effect Is large and statistically I '0. o OPT <br /> significant for GUP and modest but not quite eta- i 1e our+ gaso <br /> tsticaltyr sign f& ht for GBL, but TMW indicates is �L giro -�-e <br /> no seeding effect at all.Keeping in mind that the e <br /> PNY <br /> choice of the bias adjustment factor gave rise to u Trees <br /> conservative estimates of seeding effects, it is 4' - - ' <br /> possible that the seeding effects for MD and 40 4 0 a n 18 <br /> GBL are statistically significant after all. The fact mum.From GBM(miles) <br /> that the seeding effect changes rapidly over the <br /> very short distances between seeding targets Fig.2. Values of the seeding effect(%)as a <br /> suggests, as one possible explanation, that the function of the distance in miles northwest <br /> silver iodide nuclei from the ground generators (positive)and southeast(negative)from the <br /> • are channeled by the terrain into a focused maximum effect at GBI'� <br /> plume, and not widely dispersed as Intended. <br /> -Scientific Papers- <br />