Randomized Propane Seeding Experiment: Wasatch Plateau, Utah


  • Arlin B Super St. Cloud, MN
  • James A Heimbach, Jr. Springvale, ME


A randomized experiment to test propane cloud seeding effectiveness was conducted on the Utah's Wasatch Plateau during the 2003/04 winter. Propane dispensers were operated at a previously used high altitude site well up the windward slope. The primary target precipitation gauge was 4.3 km downwind, on top the plateau, where earlier plume tracking demonstrated routine targeting with seeding site winds from the southwest quadrant. Gauges were maintained upwind and downwind of the primary target, along the expected plume trajectory, but no more than 6.5 km downwind of the propane dispensers because of targeting uncertainties. A well-correlated crosswind control gauge provided the covariate used in all statistical testing. The experiment was completely automated with 2 hr experimental periods declared when a specified icing rate was exceeded at the seeding site. Each period contained a pair of experimental units, one seeded and one placebo, with treatment by random decision. Three different statistical tests were applied. Statistically significant seeding-caused precipitation increases were strongly suggested by the exploratory experiment for the entire population of 98 experimental units, for a partition of 69 units with wind directions expected to result in successful targeting, and for the same wind directions but excluding highest natural snowfalls. Intriguing but inconclusive suggestions of possible positive seeding effects resulted when supercooled liquid water (icing) was detected, with warmer temperatures and for lighter wind speeds. Hypothetical estimates suggest about an 8% precipitation increase had all experimental hours been seeded during the 3.5 month project period.




Scientific Papers