 |
 |
||
 |
|
||
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
||||||||
 |
Projects: Stardust@homeStardust@home: Help Find the Elusive Particles from Distant Suns!Learn more about the hunt for Interstellar Dust -- the miniscule building blocks of the universeWe all know SETI@home, the wildly successful distributed computing program that brought together millions in a search for extraterrestrial intelligence. Now, just as SETI@home is transitioning to its new BOINC platform, it has spawned a new and no less remarkable scientific venture: Stardust@home, a project that will draw together users from around the world to search for tiny grains of interstellar dust. The particles are the first samples of distant stars ever brought to Earth from space, and you can help find them! The Long JourneyThe tiny interstellar dust particles were collected by the spacecraft Stardust, which on January 15, 2006, completed its 7-year Odyssey in space by returning its samples to Earth. Two years before, on January 2, 2004, Stardust flew through the hail of rocks and dust that make up comet Wild 2's coma, collecting invaluable samples of cometary particles that will help scientists decipher the history of our solar system. But during its long voyage Stardustalso picked up a different type of sample -- miniscule particles of interstellar dust that arrived at our solar system from distant stars, lightyears away.
Stardust collected these particles between February and May 2000, and again between August and December 2002, while passing through a stream of dust that flows into our solar system from interstellar space. The stream was discovered quite recently -- in 1993 by the spacecraft Galileo, which passed through that region of space on its way to Jupiter. When Stardustflew through the stream,it extended its tennis-racquet shaped aerogel collector, picking up and storing the interstellar particles. No such pristine particles from distant stars have ever been collected before. On January 15, 2006, Stardust swung by the Earth once more and released a sample return capsule, which parachuted safely down onto the Utah desert. Nestled within the capsule's science canister were two sets of samples: cometary particles on one side of the aerogel collector, and interstellar dust on the other. Within days of arrival, mission scientists began extracting the dust grains from Wild 2and preparing them for shipment to scientists around the world. Few and Tiny: Searching for the Interstellar Dust GrainsWhereas the cometary samples are relatively easy to extract from the collector plates, that is not the case for the interstellar dust samples. For one thing, there will be very few of them, probably around 40, compared to the thousands of cometary particles. For another, the interstellar grains are miniscule -- only a few microns in size. These particles, furthermore, are embedded in about 1,000 square centimeters (more than a square foot) of aerogel, which after years in space is likely to be crisscrossed with cracks and flaws. All in all, a very small number of very small particles are scattered in a very messy neighborhood. Before scientists can think about extracting them, they first have to find them. The Planetary Society is an official collaborator of the Stardust@home project. Our role is to reach out and engage the public in this in this remarkable new way of taking part in space exploration. Check on our website for further information and updates on Stardust@home!
Andrew Westphal, an associate director of the Space Sciences Laboratory at the University of California at Berkeley, has spent a great deal of time thinking about how to locate these proverbial needles in the haystack. His first idea was to try an automatic scan: an automated microscope would image and record every tiny portion of the aerogel interstellar dust collector, focusing on varying depths beneath the surface of the collector. The images would then be stored, creating a digital archive of the collector. This in turn would be run through a computer program, designed to detect the tell-tale signs of an impact from an interstellar dust particle. The program would register the locations, and they would then be examined manually by scientists in person. A similar approach had worked well for Westphal in the past, when he and his team developed a method for detecting particle tracks in high-energy astrophysics experiments. In that system the microscope scanned the collector twice, focusing on two different depths. A computer program would then match the two scans and register the locations where both revealed a possible track. This way, local flaws in the collector were excluded, and only "tunnels" deep enough to pass through both levels of the scan would become candidates for actual tracks. In the last stage, Westphal and his team would visually observe the candidates to determine whether they were indeed true tracks. The Stardust interstellar dust collector, however, posed a far more difficult challenge. This is because the miniscule particles are expected to penetrate only the very top layer of the aerogel plates, to a depth of no more than 100 microns. At that depth, it is likely that the aerogel from Stardust, returning from 7 years in space, will be filled with cracks and flaws. As a result, the automated scans will likely be flooded with false identifications, and the 40 or so actual interstellar dust grains may never be found.
To get around this problem, Westphal and his teams considered using sophisticated pattern recognition software that would be able to distinguish between cracks in the aerogel and actual particle tracks. They consulted with Professor Jitendra Malik, a U.C. Berkeley computer scientist, who suggested that such a finely discriminating program was, in principle, possible. In order for it to work, however, they would have to "train" the computer with real images of aerogel containing grains of interstellar dust. But here's the rub: no such particles had ever been collected! Scientists can only approximate what real grains embedded in aerogel would look like. For a computer program, this was simply not good enough, and the plan to automatically scan the aerogel collector seemed to lead to a dead end. How, then, can these precious grains from faraway stars be located? Stardust@homeAlthough sophisticated computer programs could not tell interstellar particles from cracks and dirt, there was yet one instrument available that could do the job: the human eye. From his experience with high-energy physics particles, Westphal knew that unlike computers, humans using microscopes could recognize true tracks quite easily, with only a limited amount of training. If humans could scan the entire surface of the aerogel, then detecting the interstellar dust particles would be easy. This, however, raised a different problem: microscopically covering the entire collector at the required magnification would require millions of separate images. The human eye might be a good tool for identifying particle tracks, but who are the humans who can be expected to scan that many images? And if they do, how long would it take, and can they be expected to stay alert for the chance discovery of a single particle among hundreds of thousands of empty images?
It was at this point that Westphal remembered another project that was being conducted from the very same building where he has his office and laboratory. Only a few doors down the hall at the Space Sciences Laboratory, David Anderson, Dan Wertheimer, and their crew, were running SETI@home -- the largest and most successful distributed computing project in history. With SETI@home, millions of computer users worldwide could join together in the search for extraterrestrial intelligence. Would it be possible to apply a similar approach and have people from around the world join in the search for interstellar dust? It was possible, and with help from the SETI@home veterans, Westphal set out to figure how it could be done. Unlike SETI@home, the new project will be based on actual human observation rather than automated computer processing of data. But like SETI@home, it would rely on the participation of a multitude of users, who would divide a seemingly intractable job into small and manageable parcels. In tribute to the project that inspired it, the new program would be called Stardust@home. Here's how it will work: as in the original plan, the automated microscope will scan the entire surface of the collector, recording digital images of each miniscule portion of the aerogel. Since each image will cover an area of 260 x 340 microns, and since each image will include a 10% overlap with its neighbor, the microscope will need to focus on 1.6 million different locations to cover the entire surface of the collector. For the automated scan of the high-energy particles, the microscope took two images of each location, focusing on two different depths within the detector. This, however, will not be enough to detect the tiny interstellar dust particles, which lie very close to the surface of the collector, among cracks and flaws. The microscope will therefore take 40 separate images of each location, each focusing on a different depth, between 20 microns above the surface to more than 100 microns within the aerogel. Only a track that is visible continuously through a large portion of these images can be considered a serious candidate for the "real thing." For each location, the 40 separate images will be packaged into a "movie," representing a continuous in-depth look at each location through different depths.
Up to this point, Stardust@home seems like a normal scientific project, though its subject matter is definitely unique. But here's where things get really interesting: thousands of users around the world will now log on to the Stardust@home website and use a simple web-based program called a "virtual microscope." The program will contact the Stardust@home server, and download a movie of a single tiny portion of the Stardust collector. Using the virtual microscope, the user will then view the movie and scan it for actual interstellar dust particles. Once the scan is complete, the program will send the results back to the server. The user will then be free to search the next movie, which was downloaded in the meantime. Overall, each user will only view a tiny portion of the collector; but together, thousands of users around the world will be able to survey the entire collector in just a few months. Unlike SETI@home, which only requires users to install the program and let their computers do the work, Stardust@home relies on users' dedication and competence. It is the users themselves, not the computer, which will identify suspected particles. And since it is hard for Westphal and his team to evaluate the competence of each individual user, they will rely on majority opinion to decide whether a particular location deserves a second look: each movie will be sent out to four users, and only if at least two of them report a detection will it be considered a candidate. In that case, it will be sent out again to several more users, who will not know that it has already been flagged by others. If a majority of users in this "second round" also report detections, then professional scientists will observe the location to determine whether is does indeed contain an interstellar dust particle. With Stardust@home and the power of distributed computing, a task beyond the endurance of any scientist and beyond the capacity of even the most sophisticated computer programs will be accomplished in short order by human volunteers. "It's simply the only way we know how to do it" said Westphal.--Amir Alexander |
||||||||||
 |
 |
||||||||||
[ what is this? ]