Stargazers have yet to lay eyes on any of the nearly 240 planets detected outside our solar system. These so-called exoplanets are too faint for current telescopes to distinguish from the stars they orbit*; instead astronomers rely on indirect methods to infer their existence. Yet popular news accounts, supplied by space agency press services, overflow with bold, almost photo-realistic images of distant worlds.
Naturally, people can get confused. When San Francisco artist Lynette Cook painted a particularly striking image of a newly discovered planet passing in front of the star HD 209458 for a 1999 NASA press release, she received e-mail asking what kind of amazing image processing software she had used. "A lot of people didn't understand that it was a rendering," she says.
When astronomers make a sexy new discovery that doesn't lend itself easily to visual inspection, they turn to artists like Cook to bring it to life. A compelling image has the power to transform even the most complex data into an easy-to-grasp mental snapshot, but it runs the risk of fooling people into believing an iffy scientific case is closed.
Well aware of the double-edged sword they wield, artists and astronomers who dream up images of astronomical exotica often spend considerable time deciding how best to illustrate new discoveries. In the case of exoplanets, they are guided by a few key pieces of information and a healthy dose of educated guesswork.
Researchers frequently discover exoplanets by searching for regular variations in the color of a star's light. A planet circling in orbit will tug its star back and forth like a wispy figure skater spinning around a heavier one.
If the system is aligned so that the star wobbles toward and away from us, the light will appear to oscillate between bluer and redder as the wavelength expands or contracts along the line of sight to Earth.
This oscillation, which would not be especially eye-catching plotted as a graph, contains several crucial bits of information. Encoded in the timing of the oscillating light are the shape and distance of the planet's orbit around its star, which in turn determine how warm it is likely to be and, therefore, whether it could have liquid or frozen water, in the case of a rocky, Earth-like planet.
The degree of the color shift indicates the planet's mass and makeup. Astronomers believe that a planet as hefty as the solar gas giants Saturn (95 Earth masses) or Jupiter (318) would almost certainly be gaseous. Smaller planets of up to a few dozen Earth masses might range from those made mostly of rock and ice to larger ones with a solid core enshrouded by a mixture of thick gas.
Orbital distance would also influence a planet's color, says astronomer Geoffrey Marcy of the University of California, Berkeley, whose team has discovered 145 of the known exoplanets.
Jupiter and Saturn have brownish-orangey tints because their atmospheres swirl with carbon-based molecules that reflect light of those colors, Marcy says; a hotter atmosphere would evaporate water clouds and break down those molecules into carbon dioxide and methane, which scatter blue light. The effect might result in "deep blue billiard balls," he says.
Less familiar colors may also be out there. In May astronomers reported that planet HD 149026 b seems to absorb so much light that it should appear nearly black.
Today's instruments have too little resolution to detect details found in some images, including moons, rings and volcanic activity. Cook takes her cue from astronomers when deciding how or whether to depict such enhancements. "Most of them are really pretty open to show what might be there," she says.
Robert Hurt, a self-described visualization scientist for the Spitzer Space Telescope, says he ups the realism of his depictions of distant gaseous planets by matching the size and shape of their roiling clouds with those of known solar gas giants.
Marcy says the artists he works with (including Cook) typically develop two drafts of an image before creating the final masterpiece [click here for examples]. Of course, he doesn't always have final say on the illustration that accompanies press releases. When his team discovered the first Saturn-like exoplanet, for instance, NASA commissioned an image of a reddish gas giant covered in white spiral clouds.
"This makes me cringe," he says. "It looked like the planet was wearing paisley pajamas." But it was out of his hands. "There's not much you can do when NASA says they want a [certain] depiction. You can't stop them."
Some observers say that an emphasis on realism might be a touch deceptive. "My problem with so much of the beautiful stuff that artists do is that we are not privy to the decision-making process," says Felice Frankel, a senior research fellow and science photographer at Harvard University. "The more, quote, 'real' something looks, the more you accept it as fact."
For Frankel, labeling something an artist's rendering is not enough, because an image leaves a lasting impression. She advocates literally showing viewers where the interpretation lies by providing multiple possible views—something Marcy and Cook have tried—or making images sketchier and more interactive.
"My fantasy is, for example, that you see a picture online and, with your cursor, you kind of travel around the artist's concept and you see a rollover of information," she says. "Why not say right at that spot, 'this is supposed to represent this but we're not sure about that and this and that'?"
Nobody disagrees that colorful, detailed images may carry a risk of misconception. But for the image makers—tasked by funding agencies to convey new findings to the public—the benefits outweigh the risk.
"It's alright to have pictures as long as they generate intelligent discussion," Marcy says. "Pictures allow people to imagine, where otherwise their imaginations might be somewhat limited."
Capturing a true image of an exoplanet, Marcy says, may have to wait until NASA scrapes together the money to launch the proposed Terrestrial Planet Finder, which would include a telescope equipped with a tool for blotting out distant stars, thereby creating an artificial eclipse to reveal orbiting planets.
Until then, the problem of realism versus abstract renderings has no easy solution, Hurt says, noting that the goal generally dictates the answer. "If a casual observer notices an artist's concept of a planet in an article in USA Today, [and] thinks 'wow, we're observing planets,' and [then] turns to the sports page, was that a success?" he asks.
Probably so, in his view. "If the image was dull or diagrammatic," Hurt says, "he might not have even noticed the article at all."
* Note: Exo-sticklers may recall that telescopes did image a planet-size red object around the brown dwarf 2M1207, but researchers do not believe it formed like the planets in our solar system did.