Since the practice was invented scientists have asked the question: if a cloud had not been seeded, would it still have rained to that extent?
The problem of verifying the effects of cloud seeding is the focus of intense research.
To answer the question, the National Centre for Atmospheric Research in Colorado is beginning to run complicated models through supercomputers, which take into account every known variable in the atmosphere.
"The problem is that in the atmosphere no cloud is the same as another," said Roelof Bruintjes, a meteorologist at the centre. "Once we are able to simulate nature correctly, this will help us to have a baseline with which to evaluate the effects of seeding.
"Up until a few years ago we weren't able to put the physics into models, because we didn't have large enough computers. But with the modern developments now we can do the simulation."