MOUNTAIN VIEW, Calif.--Launched in March 2009, NASA’s Kepler program has been scouring the skies with the aim of exploring the structure and diversity of planetary systems and search for other life-supporting planets in our galaxy.
Using one of NASA's largest space telescopes, a custom-designed photometer and its most powerful supercomputer, Pleiades, scientists are analyzing observational data gathered from the Kepler Mission spacecraft which is searching the constellations of Cygnus and Lyra to monitor over 150,000 stars in the Milky Way.
The three year mission is named after 16th century astronomer Johannes Kepler who discovered that planets orbit the Sun in ellipses, rather than circles, based on the observations of his predecessors and his own unique calculations—without the aid of the yet-to-be invented telescope or the computer.
"The Kepler Mission has already transformed our understanding of how planetary systems work," said Todd Klaus, lead software engineer in the Kepler Science Operations Center (SOC) at NASA's Ames Research Center, Mountain View, Calif.
NASA's Pleiades supercomputer, operated by the NASA Advanced Supercomputing (NAS) Division at Ames, is an "essential component to make these planet searches possible," Klaus said.
Kepler’s photometer boasts a 95-megapixel digital camera—the largest digital camera to ever fly in space. Pixels are downloaded once a month and transferred to the SOC, where they are calibrated, combined to form light curves, corrected for systematic errors introduced in the photometer, and then searched for the signatures of transiting planets.
When a planet passes in front of its host star, it blocks a small fraction of the light from that star that appears as tiny, repeating pulse or beat. By measuring the frequency of these beats and the amount of light blocked, NASA scientists can detect the planets and calculate their size and orbital distance.
The Kepler team has already found over 1,235 candidate planets—yet to be confirmed as true planets—orbiting 997 stars. Of these, the team has already identified 68 candidates approximately the size of Earth, and 54 candidates that are in the “habitable zone”, the distance from a star allowing water to exist on the surface.
Of these 54 potential "sister" planets, five are thought to be near Earth-size.
“What we’re looking for are planets that are both earth size and in the habitable zone,” said Klaus adding “We’re still looking but we have a lot more data to collect.”
In September of this year, the team discovered the rare existence of a planet orbiting two stars, called a circumbinary planet.
Kepler’s planet finding program is extremely compute intensive, however, which is where the computational power of the 112,896-core Pleiades supercomputer comes in.
Once the light curves from the photometer are corrected, automated copies of the data are transmitted between the Kepler science processing/data analysis pipeline and Pleiades to run the transiting planet search and data validation, which are the most computationally intensive portions of the analyses.
“Data has to be folded at various test intervals, which is very computationally intensive,” noted Klaus. "Pleiades has enabled us to do the computationally intensive planetary transit search on Kepler light curves for more than 200,000 observed stars, in less than a day," he added, noting that the same search would take more than a month to complete on NASA’s 500 core Kepler computers.
"Two years into the mission, each of these light curves already contains over 30,000 individual data points, which represent a significant computational challenge for the transit search. Pleiades makes this search possible," he explained.
If Pleiades had 200,000 cores, Klaus posited that the entire search could potentially be done in the space of an hour.
The next challenge for NASA’s Kepler team is in finding terrestrial planets—one-half to twice the size of Earth and composed mainly of rock, positioned in the habitable zones of their stars. As supercomputing potential increases, so too do the chances of finding such planets.
Wow, thanks for this interesting article. It's really exciting to know that there are 54 potential "sister" planets. Let's hope this research will find new facts about earth like planets and may be possible answer to UFO.
Not such a mystery, really. One of may reasons would be, of course, TIME. If some distant civilizations were doing SETI to try to discover us, what would they receive? Nothing. Not unless they were no more than 50 or so lightyears away, which is right in our own backyard.
We need to figure out something along the lines of "subspace communications," as in the Star Trek series. A way of bypassing the speed of light.
This is not only super exciting, but if we end up finding any evidence of life elsewhere, even just microbes, it would be one of those civilization-changing milestones in the history of mankind. It would revolutionize our whole sense of who we are. Similar to when Nicolaus Copernicus figured out that the Earth was not, after all, the center of the universe. And perhaps even more revolutionary than that.
Then again, it may be a whole lot quicker to reach this gigantic milestone if our Mars missions show any evidence of living organisms, even if long expired. Some years ago, when I voiced this hope to a long-time friend, he said, "Why Mars? Why not some other planet in some other solar system?" And my response was, I'm just being pragmatic. If on Mars, we have a hope of reaching this new frontier within our lifetimes. Otherwise, ...?
This sort of search for knowledge is what defines us as humans. I can only hope the funding for these endeavors is not sucked away to more mundane causes.
95 mega pixel camera and the method to detect a planet in the milky way by detecting the blocked light of a star when the planet passes by are quite unique thinking and the finding out of 5 planets of earth size in the same kind of our earth's habitual zone are just exciting to read and understand. Long live NASA.
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.