How NASA Makes Scientific Data Beautiful
How do you make education interesting and, more importantly, beautiful? When it comes to the work of NASA, attracting enthusiasts isn’t difficult with the usual visuals of bright stars and colorful planets on hand. Look no further than the recent awe over Mars rover Curiosity’s high-res pictures to see proof of humanity’s fascination with space.
But not all of NASA’s data is packaged into a neat little photos. In fact, some of the organization’s most important findings about space come back in the form of numbers, beamed in by one of the many satellites orbiting our planet. And this information is brought to life by the Scientific Visualization Studio (SVS) — a team of scientists and animators that turns numerical data into a dynamic graphic or video.
The SVS is not only an active and creative tool for NASA outreach — it has even gone viral. Earlier this year, the SVS team received information from a project team called Estimating the Circulation and Climate of the Ocean, or ECCO, which uses mathematical tools to better understand how the ocean’s circulation patterns change over time. The result was Perpetual Ocean, a detailed and moving video interpreting a year’s worth of the ocean’s current patterns in minutes.
“I think scientists have an amazing internal world — they think about these things and how they work,” says Dr. Horace Mitchell, director of SVS. “But, they don’t do the kind of visuals that can be found in a feature film. That’s why we’ve found a niche that works.”
Mashable spoke with Mitchell about Perpetual Ocean and how to bring beauty to educational information. What do you think of the work of SVS and NASA? Let us know in the comments.
Q&A With Dr. Horace Mitchell, Director of NASA Scientific Visualization Studio
How did Perpetual Ocean come about?
We’re tasked to visualize massive results of all kinds for the purposes of public outreach. Sometimes the things we create are very specific — they’re aimed at providing information for a specific thing, like the latest results of sea-surface temperature or some kind of Earth process like the melting sea ice of the North Pole. As part of that, we do a lot of background work, developing processes and methods for doing those visualizations. We’re always looking into the future to say, “What can we do better? What kind of data do we not have tools to visualize for?”
One of those things we thought we couldn’t do was anything that involved a flow field — an ocean current or wind, for example. We weren’t really happy with the way we were doing those visuals, so a number of years ago, I started working with some of my colleagues to come up with new methods of visualization. We would apply those new methods to all kinds of data sets, and one of the more interesting and successful visuals we started to build was based around a 3D ocean current data set — the ECCO 2 Dataset. It’s a 3D model of the ocean, and it represents a kind of reality: A lot of computer models are done this days, and you have models that show how something would work, and how something did work. A lot of work goes into melding those and making them agree — running computer models but always looking at data. Weather forecasts are a similar technique.
We received a data set of the ocean from ECCO 2 that started around 1992 and runs to the present, and we found a very stable field for visualizing. We had done a 20-minute visualization of the ocean as a test for ourselves, where we visualized moving around the whole world. Every year there’s a computer graphics competition and we wanted to enter it, but the pool is limited to five-minute pieces. So, we took the 20-minute piece, we chopped it off at the beginning and end, and then an editor colleague of ours sped up the film to all of the interesting features. Then, we decided that rather than writing a narration to explain what was going on, we would find an evocative piece of music and just play it — allowing people to feel the piece and not need an explanation. And that’s what we did.
We submitted it to the conference, but it didn’t get selected. So, we had shown it to people and we had it hanging around. We actually have a website where every single thing we do and have done for the past 20 years is on display for anyone to use. We had created an animation page describing the video, but we hadn’t tried to advertise. Then a colleague of ours who pushes NASA stuff to social media put it on our Facebook Page, and it just caught like wildfire — it went completely viral at that point.
What’s the process in developing a graphic like Perpetual Ocean? What tools do you use?
The basic technology for all our work is the same kind of tools that the film and animation industries use. We actually use a lot of the same software that Pixar uses to make all of its movies. We use a software called Maya, and a software to render the final product called RenderMan, which is actually the software Pixar sells. So, that’s the basic thing we use, but we take that commercial software, and we add utility to it — specifically the kind of utility that Pixar is not interested in, which is the utility of visualizing data on the globe and data where a visual is a reference to a particular time and place in the universe or on the Earth.
We built various pieces of software because we didn’t have a tool in our toolkit to take data representing moving things, like water or air. We had been working on the software for years, and now it’s a two-stage process: We take a data set that directs where the ocean is moving in 3D, and we then drop virtual particles into the ocean and allow them to be moved along by that current. Then we’ll add particles or take them out as time goes on, to balance it out. We can run that for a virtual year or two of time, then we’ll store the results of where the particles moved in that time. Then, when we go to do our animation, we access those results. So a line in Perpetual Ocean represents the history of a particle. So if you see a long line, that means the particle is moving relatively rapidly. If it’s a very short line, it means the particle didn’t move very far, and it’s sticking in the same place. So we can then go back and choose how long we want the lines to be, how thick we want the arrowheads to be, what color, etc. We iterate over and over and over in an attempt to get something to look the way we want it to look. In this particular case, that went on for a month, off and on, because we were just doing something for ourselves.
People have been comparing the video to works of Vincent Van Gogh. What of you think of that?
I think in this case it’s sort of fair to say it was a fortuitous accident. When you’re looking at the visual, and you’re looking at the blue areas around the lines, the color represents the depth of the ocean. There’s actually a 3D model of the depth of the ocean in there, so that gave it a sort of unique character. I have to give credit to my colleague, Greg Shirah, who really thinks a lot about how these things look. He came up with a wonderful way of having sort of an ocean surface for these arrows to flow along but still be able to see through it to the depth. The depth is exaggerated by a factor of 50, so we really make it look like a basin that water is moving along in. A lot of that was his aesthetic eye, and I don’t think he was aiming for anything Van Gogh-like. But he was looking for something that would be very dramatic.
Did the way the visualization look ultimately inspire the way it was edited?
The visual was done first, so we knew there were certain places in the ocean that we could look at and other places that weren’t quite as interesting. We made the decision interactively, sitting in a room and deciding when to speed it up and which points to focus on — like speeding across the South Atlantic to get to South Africa. By that time, we’d had a big disk crash and we had lost a lot of the information upon which we created this specific visualization. All we had were the frames from the 20-minute version, so we had to take those and individually speed them up and slow them down to get the effect we wanted. We did a number of tests, and we would look at them until we felt it was going to the right places in the right amount of time. It wasn’t any kind of agonizing process.
What are the challenges of creating a data visualization that is both informative and visually striking?
Obviously, if you don’t have interesting data, you don’t do anything. That’s just it. And I have to give all the credit to the fabulous researchers who work in NASA and with NASA to produce these amazing and accurate data sets in all kinds of ways.
Once we have the data set, then we bring two things to bear. I’m a scientist by training — I have a doctorate in physics and I did research for about a decade — but everyone in my group has a level of understanding what the point is of all these animations. We know what we’re doing, and what we’re trying to bring out. Then, the team is amazing — working together, they play off of each other and every person in the team has a separate project they’re working on. They’re always bouncing ideas off of each other. As a result, over the last 10 or 20 years, we’ve gotten very good at figuring out what we can do. We have tools we can bring out at the drop of a hat to do certain kinds of visualizations. But we’re always thinking about the next thing, and what we can’t do yet.
Some of the visuals happen much more successfully than others, and we’re not always aware of which ones will be successful. We do the best job on every visual, and sometimes we strike gold like on this one.
What’s next for SVS?
One of the most recent things we did, which is highlighted on the front page of our website, is called Excerpt from Dynamic Earth. We did a planetarium show to exhibit a solar storm coming out of the sun and hitting the Earth. From there we flew through the atmosphere and into the ocean. Almost everything that’s moving uses the same software, but uses it in a different way.
We also have an app for the iPad called NASA Visualization Explorer. Someone with the app can get two new interesting visuals per week, one on Tuesday and one on Thursday. You can take a look at them, and they’ll point you to more information about NASA data.
And, in the back of our mind, we’re thinking about a sequel to Perpetual Ocean sometime in the future. But right now, it’s just sort of bouncing around.
Image courtesy NASA, Flickr.
Read more: http://mashable.com/2012/08/28/nasa-svs/