For centuries, the scientific process has remained the same: find the issue, propose a hypothesis and perform an experiment to test the claim. But with the rise of supercomputers, this method has become increasingly irrelevant.
On Wednesday, Dimitri Kusnezov, Ph.D and chief scientist for the National Nuclear Security Agency, spoke at the Howard H. Baker Jr. Center for Public Policy about the great capabilities of supercomputers and the immense simulations that were once thought impossible.
Moreover, with faster and more accurate results available, policymakers can now confidently execute decisions that previously relied on gut feeling, he said.
In one example, Kusnezov recalled the Fukushima Daiichi nuclear meltdown in early March 2011. In the midst of the confusion, U.S. leaders wavered over whether or not to order the evacuation of American citizens from Japan. These officials possessed no method for determining which way the contaminated gas would travel.
"The initial estimate was from the simulations that were done, were that Tokyo was not at risk, and we did not have to worry about that," Kusnezov said.
While Kusnezov maintained respect for classic means of scientific investigation, he emphasized the necessity of adaptability in disaster situations.
"It's not easy to do these kinds of scientific problems through the conventional way of peer review," he said.
"You can't pull together a team of your best people in the middle of the night, and say 'why don't you work together and figure this out in an hour.'"
After the Columbia Space Shuttle disintegrated upon its re-entry in February 2003, the national science labs worked to find out what went wrong. In July of that year, after much experimentation, scientists announced that a piece of foam had hit part of a corroded wing, leaving damage that would spell the end of the astronauts.
Through computer simulations however, other researchers arrived at the same conclusion – four months earlier.
While Dr. Kusnezov takes pride in the work he has done, he acknowledges that such work is possible only through public investment, requiring patience and commitment beyond short-term goals.
"Science is long-term, you don't always see what the benefits are immediately, and you have to accept that," Kusnezov said. "You need people with enough time, the luxury of time, to think about the challenges this country has.
"If you constrain them to solve, deliver stuff day in and day out, they never have time to think outside of the box. ... You have to give people the time to think freely, give them time to be scientists and engineers and explore ideas, because those are the kind of people we need in this country."
Taylor Eighmy, the vice-chancellor for research and engagement who helped bring Kusnezov to campus, was surprised by the lecture's relevance.
"I was dumbfounded by the applications of computational simulation in solving really important problems to society," Eighmy said.
"I know about this a fair amount, but the examples he used were very illuminating as to the depth and breadth of simulation as a solution provider; it's pretty profound."