Abstract
A five-year randomized cloud seeding program conducted in the San Juan Mountains of southwestern Colorado by the Bureau of Reclamation was completed in April 1975. The test design included randomization of the seeding by 24 h experimental days, and other features such as operation only during suitable cloud and wind conditions, and suspension to avoid adverse effects on the public and environment.
Previous experimentation by Grant near Climax, Colorado, during the 1960's had identified conditions under which clouds seeded with silver iodide would produce more snow than similar, untreated clouds. The purpose of the Colorado River Basin Pilot Project was to determine whether the experimental procedure applied at Climax would be effective in an operational mode. The objectives were twofold: 1) to test the physical concepts of weather modification potential, and 2) to test the practical weather modification potential for an operational technology in the continental weather systems that bring snow to the San Juans.
A formal statistical analysis based on precipitation data for 71 experimental treated days and 76 experimental control days found no significant difference between precipitation, gage by gage, on seeded and unseeded days, even after deletion of 22 control days suspected of contamination by previous seeding.
An a posteriori analysis, based on 6 h time blocks (in place of the 24 h day called for in the experimental design), indicates that positive seeding effects may have been achieved during periods of warm cloud-top temperatures, as expected from the Climax experiment. These positive effects may have been overbalanced in the experiment by decreases in snowfall due to seeding unfavorable cloud types.
The results of the a posteriori analysis suggest that an operational seeding program, flawlessly carried out with perfect forecasts and no periods of suspension, could increase 15 October–15 May precipitation by 10–12% in various portions of the drainage basin. The resulting average precipitation increase during the snow accumulation season, put into a hydrologic streamflow model developed as part of the evaluation, yields a potential increase in annual runoff of the San Juan River of 197 000 000 m3. In the Rio Grande Basin, on the downwind side of the crest, the potential increase in annual runoff is 186 000 000 m3.
