See, then discover


Recently I read Hermann Bondi’s “Relativity and Common Sense,” an improbably slim Dover books with a strange sort of hourglass on its cover. I enjoyed Bondi’s methodical explanation, aimed at the common man–why does it seem like people had more faith in that person back then?–especially with the extremely retro drawings of Adam, Bill, Charles, and occasionally David, flying in their respective UFOs or sitting at home in their giant-receiver-equipped houses, flashing signals at each other and measuring different times on their watches. Bondi points out that, in Einstein’s day, the only way people had of thinking about the speed of light were extremely fast trains passing each other. What a luxury, he says, for those of us in the modern era (the book was written in the early 60s) to be able to think about jet planes and and even space craft.

The phrase “Ah, I see,” encapsulates how important visualization is to the way we understand. But physics often deals in things that we will never “see”–even with microscopes, telescopes, or particle accelerators–and often can only even imagine by analogy(like the fifth dimension, for instance, never mind the other six that might be out there).

The strange hourglass on the cover of Bondi’s book is what’s called a light cone. Shaped by the path of light through space from a single point, it’s the chunk of space-time (mapped from four dimensions onto a three-dimensional space, since that’s all we can visualize in what’s called a Minkowski diagram) representing the chunk of space-time that can ever reach a given point in it. Even as it illustrates the fundamental truth that nothing can travel faster than the speed of light—and thus anything outside this light cone can ever effect the point at its infinitely slender waist—it embodies the hobbles of our three-dimensional sight.

Living in a spatially three-dimensional world may limit what we can understand through seeing, but supercomputers, which can churn through voluminous datasets in at incredibly high speeds, have lately enhanced our imaginations, expanded what we can consider in a concrete form, putting even the most distant, miniscule, or simply mind-bending physical phenomena right before our eyes.


The WIRED Science blog grabbed the 10 best science visualization videos picked by DOE’s SciDAC, which stands for Scientific Discovery through Advanced Computing. This one’s a supernova explosion, rendered in beautiful, vivid technicolor (actually, the colors represent temperatures). WIRED writes:

Type Ia supernovae are thought to be white dwarf stars in binary systems that explode due to a thermonuclear runaway. This movie shows a simulation of Type Ia supernovae exploding from multiple ignition points. When the hot ash breaks through the surface of the star, it spreads rapidly across the stellar surface, converges at the opposite point and produces a jet-like flow that triggers a detonation. The simulation shows that multiple ignition points generate more nuclear burning and produce more expansion of the star than a single ignition point. As a result, less radioactive nickel is produced during the detonation phase, and the explosion is less luminous.

The work comes out of the University of Chicago’s ASC/FLASH center, who first simulated a white dwarf explosion in 2007. The process in nature takes only about three seconds; “just one of the compiling jobs,” states the U Chicago press release, “ran for 75 hours on 768 computer processors, for a total of 58,000 hours.” Understanding how supernovas explode—and what could be a better understanding than visualization?—could help physicists puzzle out why the universe is expanding:

…the scientific payoff for logging these long, stressful hours is potentially huge. Astrophysicists value type Ia supernovas because they all seem to explode with approximately the same intensity. Calibrating these explosions according to their distance reveals how fast the universe has been expanding at various times during its long history.

Cosmic Variance’s Mark Trodden writes:

The current understanding of this has been enough to discover the fact that the universe is accelerating, but our future plans are to exploit it further, to help provide insight into the origin of cosmic acceleration. A detailed understanding of how supernova explosions occur would be a valuable contribution to this quest.

Now, one of my colleague, who will remain anonymous to protect him from hate mail, claims that simulations done entirely on a computer “isn’t real physics.” (In fact, when I was writing about blast waves and brain injury a few days ago, doctors were very hesitant about saying too much about conclusions drawn from a simulation.) But I think the DOE’s terms “Scientific Discovery through Advanced Computing” actually say a lot. If a computer can recreate how a process unfolds based on the laws of physics, it can also reveal features and details that we would never be able to intuit otherwise. At the risk of inciting full-on warfare, I’ll ask you to weigh in. Should we depend as much on computation as we do? Are people, like my colleague, who demand either working entirely theoretically or entirely experimentally, merely unimaginative and stuck in the past?

You may also read these articles

Leave a Reply

Your email address will not be published. Required fields are marked *