Abstract
An examination of hourly rainfall within areas designated small outside (SO) and large outside (LO) reveals that prior to the introduction of a new seeding device (NEI) in 1975, average peninsular rainfall on seed (S) days outside the Florida Area Cumulus Experiment (FACE) target area was slightly less than rainfall on no-seed (NS) days. However, during the NEI period, average station rainfall within the SO area yields S/NS ratios as high as 3 for the 6 h period following the release of the first flare. Moreover, during the 3 h period prior to the time of the first seed, average SO station rainfall on seed days exceeded the rainfall on no-seed days by approximately 60%.
The large S/NS ratios that occurred outside the FACE target area during the NEI period were not associated with anomalously high rainfall on seed days, but rather with anomalously low rainfall on no-seed days. For the SO area, differences between seed and no-seed samples are highly significant, with one-tailed P values as low as 0.001 as determined from a Wilcoxon (Mann-Whitney) test. Natural variability therefore appears to account for statistically significant differences between area-wide rainfall on seed and no-seed days.
