On the Use of Ratio Statistics for the Evaluation of Operational Cloud Seeding Programs


  • Bernard A Silverman Consulting Meteorologist Centennial, CO, USA


The purpose of this study is to describe and demonstrate the capability and merits of using ratio statistics in evaluating the effectiveness of operational (non-randomized) cloud seeding programs. The application of the ratio statistics methodology is illustrated by an independent statistical evaluation of the Kings River operational cloud seeding program over its entire period of operations from water years 1955 to 2004. The effect of seeding in terms of confidence limits was emphasized because they provide information on the strength of the seeding effect whereas null hypothesis significance tests indicate only whether there is any seeding effect at all. The effect of seeding on the Kings River-Pine Flat Dam streamflow station, the primary seeding target in the Kings River Basin, was evaluated using the control that gives the most precise evaluation results possible with the available data. The results of this evaluation study indicate that (i) for the data involved in this study, ratio statistics was found to be a more precise and more reliable evaluation methodology than the traditional historical regression methodology, (ii) evidence for positive, statistically significant and cost effective seeding effects was found at the target site in the Kings River Basin with an estimated increase in streamflow due to seeding of +5.1% with 90% confidence that the true effect of seeding is somewhere between +1.5 and +8.8%, and (iii) it was found that there was a marked improvement in seeding effectiveness that started around 1978, the physical cause(s) of which is worthy of further study.




Scientific Papers