Nov. 12, 1969 was a rainy day at Cape Canaveral, where astronauts Charles Conrad, Alan Bean, and Richard F. Gordon sat in the Apollo 12 command module, listening to the countdown for their launch to the moon. Finally, they heard “Liftoff!” and felt millions of pounds of thrust pressing them into their seats as the Saturn V rocket rose slowly from the pad.

Just half a minute later, listeners on the ground heard a strange burst of crackling on the audio channel, followed by "What the hell was that?" In the command module, nearly every warning light on the main panel lit up, indicating a serious electrical problem. Commander Conrad asked for help, and after several tension-filled seconds the control room advised "try SCE to auxiliary," an obscure command that only flight engineer Alan Bean recognized. But flipping the right switch worked. The warning lights went off and Commander Conrad soon reported that things were back to normal. Then he said, "I don't know what happened ‹ I'm not sure [but] we even got hit by lightning. . . I think we need to do a little more all-weather testing."

Conrad was right. Apollo 12, the second manned mission to moon, had nearly been aborted by two lightning strikes. They came from a rain cloud which, up to that moment, had seemed to pose no danger. That’s in part because the theory of how lightning worked circa 1969 wasn’t correct. Fortunately, NASA's efforts to avoid a repeat of this near-disaster led them to the one man in the world who understood the truth about lightning: a German meteorologist named Heinz-Wolfram Kasemir.

The scientific understanding of lightning began with Benjamin Franklin's legendary kite experiments that showed lightning was essentially electrical in nature. But many questions about lightning remained unanswered as recently as 1900. How fast does lightning travel, and what are its electrical and physical characteristics? What creates the separate positive and negative charges in thunderclouds that make lightning possible? What triggers lightning bolts? And how do electrical charges move during a lightning strike?

In the 1920s, a South African lightning researcher named Basil F. J. Schonland used a high-speed streak camera to show that a typical cloud-to-ground lightning bolt consisted of two main features: a dart leader, which pioneers a pathway to ground in a series of jumps or darts, and a much stronger return stroke that carries huge currents of several thousand amperes, which does most of the damage.

In a series of papers and books, Schonland presented his experimental discoveries along with a theory of how the lightning charge accumulates and flashes to earth. In developing his theory, Schonland treated the lightning-producing cloud like the body of a person who walks across a carpet on a cold, dry day. The charge accumulated in your body from the carpet is suddenly discharged through a spark when you touch a grounded object such as a doorknob.

Schonland believed thunderclouds worked basically the same way. In his book "Atmospheric Electricity" (1932), he modeled a thundercloud as a charged conducting sphere, and stated, "A single flash…generally completely discharges the whole cloud." Since the inside of a cloud was not accessible to photography or direct measurements at the time, it seemed reasonable to treat the cloud as a single conducting unit that accumulated and discharged electricity.

In the following decades, Schonland's theory worked well for the practical needs of the growing electric-power industry, whose electrical engineers were concerned mainly with the way lightning acted once it hit something on the ground. Exactly how it formed or what was going on in the clouds did not concern them. So the questions of how lightning forms and develops were left to the scientists, who almost all followed Schonland's lead. There was only one problem: Schonland's theory was wrong.

In the late 1940s, Heinz-Wolfram Kasemir (1913-2007) was working for the German weather bureau. He became interested in the electrical nature of storms, but when he compared Schonland's theory with what he knew about the physics of electricity, he found a contradiction. For charges that are widely distributed in a thundercloud to suddenly gather together into a lightning channel and plunge to earth was physically impossible. It was the electrical equivalent of claiming that water could spontaneously flow uphill. If Schonland was wrong, what was the correct theory? After a lot of thought, Kasemir hit on the solution: electrostatic induction.

If you have ever charged a comb by brushing it through your hair and attracted bits of paper with it, you have seen electrostatic induction at work. A piece of paper has no charge, but if the nearby comb has a positive charge, negatively-charged electrons will accumulate on the edge of the paper closest to the comb, leaving positive charges behind at the far end. Though the paper still has no net charge, it has an induced pair of charges at opposite ends, and the induced charge closest to the comb is enough to attract it to the comb.

Kasemir concluded that lightning develops the same way. Once a lightning channel starts, its opposite ends grow toward opposite electrical charges. While one end will have a high positive charge and the other end a high negative charge, the charges are equal and opposite, adding up to zero net charge with respect to the part of the cloud where the channel develops. So in contrast to Schonland's unidirectional cloud-to-ground discharge model, Kasemir claimed that lightning was a bidirectional discharge that grew at both ends.

Kasemir published his ideas in a German-language book in 1950. But when he submitted papers expressing his radical views to English-language journals, the established lightning-research community rejected them. After a few such futile efforts, Kasemir decided to direct his energies elsewhere, although he did promote his bidirectional-growth idea in conference talks from time to time up into the 1970s.

It was at once such talk that Kasemir met a lightning researcher named Vladislav Mazur. Mazur's specialty was making radar images of lightning, and when Mazur presented a paper showing that airplanes can trigger bidirectional lightning strikes, Kasemir was thrilled. Here for the first time was unequivocal experimental confirmation that his idea of bidirectional lightning was correct. Mazur was amazed to find that Kasemir had come up with the idea on a theoretical basis more than thirty years earlier.

Mazur and his colleague Lothar Ruhnke immediately became converts to Kasemir's idea, and in a series of observational and theoretical papers, they gradually convinced the lightning-research community that, at least when it came to lightning development, Schonland was wrong and Kasemir was right. As part of the growing recognition of his expertise, Kasemir was sought out by NASA in the 1970s to help them protect future space flights against lightning. Kasemir conducted numerous research-aircraft flights through thunderstorms to elucidate the ways lightning develops and how to tell if a given storm is capable of producing lightning. 

He also superintended the installation of electric-field sensors at Cape Canaveral since  it was now clear that NASA that they needed to base launch decisions not on just whether a cloud had been generating making lightning previously, but on what the electric field was at the ground beneath the cloud.  

Most lightning researchers today acknowledge the truth of Kasemir's bidirectional theory, but the change has been slow, and as late as the early 2000s some high-profile scholarly works on lightning did not give his idea the prominence it deserves. His simple and elegant concept can be explained to an intelligent high-school student, but many popular treatments of lightning still implicitly endorse the old incorrect Schonland model.

Fortunately, thanks to Kasemir, NASA engineers now know better.