Effects of Cloud Seeding in West Texas

Abstract

The effect of randomized seeding with droppable silver iodide (AgI) flares in West Texas during the Southwest Cooperative Program is addressed. Attention is focused on individual convective cells and on the small mesoscale convective clusters that contain the cells. Analysis of three-dimensional, volume-scan, C-band radar data using sophisticated tracking software indicates that AgI seeding increased the areas, durations and rain volumes of the cells. The radar-estimated rainfall volume at bases of the AgI-treated cells was more than double the rain volume from the cells that received simulated treatment. This result is significant at the 3% significance level using rerandomization procedures. The apparent effect of seeding and its significance increases slightly when control cells are incorporated into the analysis. The effect of treatment on maximum cell height, as measured by radar, generally averaged less than 5%. In moving from the cell scale to the larger sca;es, it was found that cell merger occurred twice as often in the AgI-treated cases. Merging was most pronounced for cells treated early in their lifetimes with 9 or more AgI flares. The next step focussed on the areas in which the cells received treatment. This ?focused area? approach involved calculations for radii of 5, 7, 10, 15. 20, 25 and 35 km around each treatment position, providing eight separate analyses. The rainfalls from the seeded cells exceeded the rainfalls from the non-seeded cells in the focused area by over 50% by the end of the analysis period. These results are consistent with a positive effect of AgI treatment on rainfall that begins on the cell scale, where the seeding takes place, and spreads into the overall experimental unit with time. The final step in the study involved examination of the experimental units themselves. The ratios of Seed (S) to No Seed (NS) rainfalls by half-hour interval and cumulatively generally exceed a factor of 1.20 for the two approaches employed in the analyses. The ratios are largest for mean cumulative rainfalls at 2.0 to 2.5 hours after qualification of the experimental units. None of the results have strong P-value support. Because of the small sample and the large natural rainfall variability it is likely that chance has confounded this assignment of the results of treatment in the SWCP. In order to obtain a clearer picture of the effect of cloud seeding in West Texas, it in recommended that the sample be expanded further and that subsequent analyses include the use of predictive equations to reduce the impact of the natural rainfall variability. It is recommended further that cloud microphysical measurements be incorporated into the next stage of the studies in order to better understand how the apparent increases in rainfalls were produced.