Editing Project Stormfury (section)

==Failure of the working hypothesis==
Multiple eyewalls had been detected in very strong hurricanes before, including [[Typhoon Sarah (1959)|Typhoon Sarah]]<ref name="MWR 1972 p 210">Black, Senn, and Courtright p 210</ref> and [[Hurricane Donna]].<ref name="Jordan and Schatzle">Jordan and Schatzle pp 354-56</ref> Double eyewalls were usually only seen in very intense systems.  They had also been observed post-seeding in some of the seeded storms. At the time, the only observations of rapid changes in eyewall diameter, other than during presumably successful seedings, occurred during rapid changes in storm intensity.<ref name="MWR 1972 p 213">Black, Senn, and Courtright p 213</ref> It remained unclear whether the seedings caused the secondary eyewalls or whether it was just part of a natural cycle<ref name="Journal p. 396">Willoughby, Clos, and Shorebah p 396</ref> (because [[correlation does not imply causation]]). It was initially thought that eyewall changes similar to those observed in seeded but not unseeded systems provided the evidence that Project Stormfury was a success. But if it was later observed that such eyewall changes were common in unseeded systems as well, such observations would throw doubt on the hypothesis and assumptions driving Project Stormfury.<ref name="Bulletin p. 511">Willoughby, Jorgensen, Black, and Rosenthal p 511</ref>

Data and observations did in fact begin to accumulate that debunked Stormfury's working hypothesis. Beginning with Hurricanes [[Hurricane Anita|Anita]] and [[Hurricane David|David]], flights by [[53rd Weather Reconnaissance Squadron|hurricane hunting]] aircraft encountered events similar to what happened in "successfully" seeded storms.<ref name="Bulletin p. 511"/> Anita itself had a weak example of a concentric eyewall cycle, and David a more dramatic one.<ref name="Journal p. 396"/> In August 1980, [[Hurricane Allen]] passed through the [[Atlantic]], [[Caribbean]], and [[Gulf of Mexico]]. It also underwent changes in the diameter of its eye and developed multiple eyewalls. All this was consistent with the behavior that would have been expected of Allen had it been seeded. Thus, what Stormfury thought to have accomplished by seeding was also happening on its own.<ref name=Goldenberg>Goldenberg</ref>

Other observations in Hurricanes Anita, David, [[Hurricane Frederic|Frederic]], and Allen<ref name="Bulletin p. 509">Willoughby, Jorgensen, Black, and Rosenthal p 509</ref> also discovered that tropical cyclones have very little supercooled water and a great deal of ice crystals.<ref name=HRD>Hurricane Research Division</ref> The reason that tropical cyclones have little supercooled water is that the [[updraft]]s within such a system are too weak to prevent water from either falling as rain or freezing.<ref name="Landsea C4">Landsea C4</ref> As cloud seeding needed supercooled water to function, the lack of supercooled water meant that seeding would have no effect.

Those observations called the basis for Project Stormfury into question. In the middle of 1983, Stormfury was finally canceled after the hypothesis guiding its efforts was invalidated.<ref name="Bulletin p. 513">Willoughby, Jorgensen, Black, and Rosenthal p 513</ref>