We begin by a review of the observational and modeling literature on tornadoes. We note that the recent NAS/NRC Report on Weather Modification Research said little on this issue, on any other aspect of severe storms modification, even though it was in the original NOAA charter for the Panel. We argue that, based on recent devastating tornado outbreaks, including repeated multiple strikes in Oklahoma in 1999, 2002 and 2003, any credible scientific hypothesis for beneficial tornado modification deserves close scrutiny and safe testing. We define “beneficial” as the possible shortening of the tornado life-cycle over what would have occurred naturally (average lifetime of all F3-F5 tornadoes is 20 min).

We propose a conceptual model for the tornado life-cycle, based upon the 1973 Union City, OK case and subsequent 30 years of storm-chasing observations, including Project VORTEX. The key ingredient in the tornado’s decay stage onset is the approach of a nearby gust-front or density-surge line, which gradually undercuts vortex inflow. We would test the hypothesis for accelerating the tornado’s life-cycle, by first doing sensitivity experiments with the numerical tornado simulations of Lee and Wilhelmson and Lewellan and Lewellan. Seeding to produce an early downdraft/gust-front formation relative to the tornado would be simulated in these models. Results of seeded/non-seeded cases would be carefully analyzed and compared. Based on the results of these modeling experiments, field seeding trials on waterspout-producing cumulus congestus cloudlines in the safe working environment of the Florida Keys could be conducted. We base this approach on the recommendations of Davies-Jones(1982).