Dawn Levy, News Service (650) 725-1944; e-mail: firstname.lastname@example.org
Elephant-sized collaboration can't yet budge the Standard Model
Question: How many physicists does it take to calculate one number?
Answer: About 550.
That may sound like a lot, but the number may help answer an important question: Why does matter predominate over antimatter in our universe?
The scientists running the world's largest collaborative particle physics experiment expect to work plenty hard before they are satisfied with the calculation.
Pat Burchat, a Stanford physics professor and last year's analysis coordinator for the enormous experiment, described the research Feb. 16 at the annual meeting of the American Association for the Advancement of Science in San Francisco. Using the Stanford Linear Accelerator Center's BaBar detector, which can track subatomic particles called B mesons, the researchers are measuring a single quantity called sin 2b. It represents an important difference between how matter and antimatter behave.
Scientists believe equal quantities of matter and antimatter were formed in the Big Bang that created the universe. These opposite particles should have completely annihilated each other. But one-billionth of the matter remains, said Stewart Smith, a BaBar project spokesperson. It's as if two armies started with a billion soldiers each, and at the end of the war only one soldier was left standing he was made of matter, said Smith during a tour of the experimental facility: "We're trying to understand why." And doing so requires a lot of people.
From the start, all of the approximately 550 researchers involved in the experiment were required to contribute to some aspect of the project. They came from 73 institutions in nine countries -- the United States, Italy, France, Germany, Canada, Great Britain, Russia, China and Norway. And they had to design, build and maintain the detector and computer system, spend time working shifts in front of a wall of computer screens and do service work from a list of options.
It takes the equivalent of 150 people working full time just to run the experiment and process the data, but the 550 scientists divide up that time among themselves.
And then there's the fun part: analyzing the data. About 15 different groups work on different aspects of the physics. But about 100 people are analyzing the data to boil it down to the single number -- sin 2b -- on which the matter/antimatter question depends.
Things were pretty hectic last month as the scientists worked long hours trying to meet a self-imposed deadline for revealing the new result of their vast experiment.
While the research is going on, the final number is blinded -- altered by a hidden quantity so that the researchers won't adjust the analysis to reach a desired goal. Smith called it "blind-but-not-dumb analysis": Though the real number is unknown, scientists see a phony number so that they can tell whether it is stable.
And every once in a while, they take off the blinders. They did so for the second time on Jan. 29. The first time, in August 2000, the jury was still out on whether the experimental results confirmed physicists' current theory of fundamental particles and how they interact (commonly referred to as the Standard Model). The number itself was ambiguous. More data were needed in order to decrease the degree of error.
So tensions were high when three separate computers simultaneously revealed the numerical result at 5 p.m. Jan. 29 in physicist Riccardo Faccini's office at the Stanford Linear Accelerator Center. More than 25 physicists crammed into the room. Faccini, the current sin 2b analysis coordinator, peeked over the shoulder of another scientist to watch the screen during the minute it took the computer to finish the calculation.
"I was looking at the wrong column and almost had a heart attack because the number was completely impossible," Faccini said. But pretty soon he figured out where to look. And then it was time to spread the news -- first to the core collaborators, and then to the researchers all around the globe.
Burchat described the number as "not the most exciting of possible values." Smith said that the number is halfway between zero and the number predicted by the Standard Model: "Within the errors we have, it's consistent with either. It's the number from hell." So before BaBar breaks down the walls of the Standard Model, scientists will have to collect and analyze a lot more data.
In the years to come, the current trickle of data will become a flood. The accelerator is producing B mesons at a rate faster than the original 30 million per year for which it was designed. By September 2001, it will be producing at a rate of 50 million per year, and the designers expect it to exceed 100 million per year by 2004, Smith said.
In fact, this anticipated flood puts stress on the collaborators. "People built the facility and expected to sit back and use it," Smith said, but now they have to spend more time getting the equipment up to speed so they can handle the flow of data.
The main challenge is computer infrastructure and processing. The entire system needs an upgrade. "We need to be sure that four to five years from now we aren't going to just let data fall on the floor," Smith said. "To all profit from the opportunity, we must all share in the burdens."
Each of the 73 institutions must contribute time based on its fraction of the names on the seven-page author list for publishing the result. Choosing from a spreadsheet of tasks, the principal investigator for each institution must make sure his or her group does its share.
Faccini said he relies on some special skills to coordinate so many individuals' activities: "I've been a Boy Scout in Italy for 20 years, and I am using all the techniques I have learned." Those techniques probably didn't include video and phone conferencing and a Web-based communication system, but these are critical parts of what makes the collaboration successful.
The project's website has more than 10,000 pages and is in constant use. Researchers post graphs and analyses for discussion during telephone meetings across time zones. The site has more than 100 news forums to which members of the collaboration may subscribe.
Burchat said a printed record of what's been done on the sin 2b analysis would be about 2 inches thick. But since it's all posted on the Web, no one has to wade through all that paper. Collaborators can read documents instantaneously.
Smith said, "Without the Web, it would be unthinkable to do this work. We are absolutely dependent on it."
By Katharine S. Miller