Every day the sun spews out subatomic particles called neutrinos, and instruments count how many make their way to Earth. But the instruments only detect half as many neutrinos as scientists expected to see. Where did all the neutrinos go? In recent years, scientists worldwide have converged on an answer.

Of the three types of neutrinos named after three low-mass particles -- electron, muon and tau -- the sun produces only electron neutrinos. The current consensus is that matter may turn electron neutrinos into undetectable muon or tau neutrinos as they fly out of the center of the sun. Since the sun's structure is very stable, the flux of solar neutrinos should be constant.

But that's not what physicists at Stanford and NASA Ames Research Center saw when they used novel analytic methods to take a fresh look at the data. Their analysis, funded by NASA and the National Science Foundation, revealed strong evidence that the solar neutrino flux is not constant. It varies as the sun rotates. That finding, published in the March 20 issue of The Astrophysical Journal, puts a new spin on things. Magnetism may play a key role in turning neutrinos into undetectable forms as they traverse the sun's internal magnetic field.

"If neutrinos turn out to have a non-zero (albeit very small) magnetic moment, then magnetic field could have an influence on neutrino propagation," says Peter Sturrock, professor emeritus of applied physics at Stanford who published the finding with NASA statistician and astrophysicist Jeffrey Scargle. "Like human beings, neutrinos may be 'left-handed' or 'right-handed,' with the difference that nature favors left-hand neutrinos: Nuclear reactions in the core produce only left-hand neutrinos, and experiments on Earth detect only left-hand neutrinos! If neutrinos have non-zero magnetic moment, it is possible for magnetic field to convert some left-hand neutrinos into right-hand neutrinos that are not detectable, so you end up with a deficit in the measured flux."

A ball of burning gas, the sun is fueled by nuclear reactions in its core, where hydrogen combines to form helium and gives off neutrinos that stream through the sun -- and everything else -- as if it were transparent.

While scientists have long known that sunspots, regions of strong magnetic field on the surface of the sun, come and go in an 11-year cycle, it's still unclear whether neutrinos fluctuate with this solar cycle. Sturrock's group has been exploring the possibility that the neutrino flux varies on a much shorter time scale -- the roughly 27 days that it takes the sun to rotate.

As the sun turns, so too does its lumpy magnetic field. If no magnetic field comes between the sun's neutrino-producing core and neutrino detectors on Earth, the instruments count all the neutrinos that fly out of the sun. But if there is magnetic field in the way, it may spin neutrinos into undetectable forms, and instruments on Earth will record a temporary neutrino blackout.

The magnetic field may turn neutrinos into "different fish that slip through your net," says postdoctoral researcher Mark Weber.

Earthly experiments for detecting heavenly events
About 100 billion neutrinos pass through your thumbnail each second, but you'd never know it because they hardly ever interact with anything. But once in a great while they do interact. "In the course of your lifetime, you're going to capture one neutrino in your whole body," Sturrock says.

So detecting neutrinos requires herculean feats of engineering. The weirdest of astronomical observatories, neutrino detectors consist of huge tanks of fluid buried thousands of feet underground to reduce background "noise" from cosmic rays, which cannot travel through the rock that neutrinos sail right through.

In chlorine detectors, such as the Homestake detector in South Dakota, an electron neutrino meets up with a chlorine nucleus to produce a radioactive argon nucleus. In gallium detectors -- used in the GALLium EXperiment (GALLEX) and its successor, the Gallium Neutrino Observatory (GNO) experiment in Italy's Gran Sasso mountains, as well as in the Soviet-American Gallium Experiment (SAGE) in the Caucasus -- a neutrino combines with gallium to make radioactive germanium. Scientists then record the radioactive decay of argon back to chlorine, or germanium back to gallium, to find out how many neutrinos they have caught.

"The Homestake detector captures about one neutrino every other day, so in a month you may have only 10 atoms of argon in that huge tank," Sturrock explains. "What is incredible is that the experimenters can find and count those 10 atoms."

Over time, scientists are able to collect enough data to make it possible to discern patterns. In 1997, Sturrock's group analyzed 24 years of data collected from the Homestake experiment and found evidence that the neutrino number varies in a 27- or 28-day cycle. The findings published this March came from a different type of analysis of data from different experiments -- GALLEX/GNO and SAGE -- but are consistent with their earlier results.

Sturrock and Scargle's new analysis employs histograms -- graphs showing how many times an event, such as a neutrino hitting a detector, occurs. These graphs reveal whether or not the flux is varying, but do not tell exactly how it varies. If the neutrino flux were constant, scientists would expect to see only one peak. However, histograms formed from the GALLEX/GNO data revealed a bimodal, or two-peaked, pattern.

Sturrock explains how histograms model events: "Suppose one needs to know whether a crime suspect spends a lot of time away from home. And suppose that the bank supplies us with his monthly telephone payments, for two or three years. If we order the payments by amount, the 'histogram' would be the number of payments between $1 and $10, the number between $11 and $20, etc. If we see that the payment is usually between $51 and $60, less often $41 to $50 or $61 to $70, etc., this points toward his always being at home. On the other hand, if we find that the payments fall into two groups, one around $95 a month and the other around $15 a month, this would point toward his spending some months away from home."

The fact that Sturrock and Scargle find two peaks points to the existence of both a high-flux mode where neutrinos pass unimpeded through the sun and a low-flux mode where they are turned into undetectable forms.

"What you observe is like having the light on a police car going around -- you see flash, flash, flash," Sturrock says. "The most likely explanation is that neutrinos are being lost due to the sun's magnetic field. So it is more like dark, dark, dark."

In a recent report posted in the Los Alamos National Laboratory electronic archive, Sturrock and Weber attempted to locate the region where neutrinos are "lost." They created a colorful map that shows the correlation between the sun's internal rotation and the oscillation in the neutrino flux. Scargle calls the map "a very clever combination of hypothesis and data" that reveals where the neutrinos are being modulated.

Super-K: The final frontier?
A lot is still unknown, making solar neutrino research heady but humbling. When the original experiments were planned, scientists thought that neutrinos, like photons, had neither mass nor magnetic moment. Now, particle theory has evolved, and some theories allow them to have mass and magnetic moment.

"When these neutrino experiments were planned, scientists felt quite sure what they were going to measure," Sturrock says. "At that time, we felt that we knew all about neutrinos and all about the sun, and that we would measure a certain value of the neutrino flux. But we didn't. We learned that we didn't know as much about the sun or neutrinos as we thought we did."

The Stanford-NASA group's analysis of decades of published data from four experiments is a challenge to theorists and experimentalists alike: How should they react to this evidence for a variable neutrino flux in light of the conventional view that the neutrino flux is constant? How much evidence for variability will be required before the conventional model begins to crack?

"We're pooling our resources and literally saying, 'Hey, what we see coming out of the sun isn't what we've been led to expect,'" Sturrock says. "We can't say what the solution is though."

So the researchers continue mining the data for telltale trends. "We're going to keep pouring coal on the fire," Scargle says. "There are many ways of looking at data and seeing what's peculiar in it." The researchers are collaborating with statistics Assistant Professor Guenther Walther, who believes that past claims of variability failed to convince because they were based on flawed statistical analyses. They want to be sure the evidence they present is as convincing as possible.

Hot on the heels of a solar mystery, Scargle and Sturrock hope to access data from the mother of all neutrino observatories, Japan's Super-Kamiokande, or Super-K. In 1998, Super-K's collaborators in Japan and the United States reported evidence that neutrinos have mass, though very little. If Super-K data show that the neutrino flux is varying, that would be the "smoking gun" that would finally settle the argument about whether the solar neutrino flux is variable or constant.

Unlike the chlorine and gallium detectors, Super-K records the exact time that neutrinos arrive. "If you're looking for variability, this is tremendous information to have," says Scargle. If the Super-K experiment does not reveal a variation, however, the issue would not yet be settled, as Super-K responds to neutrinos of much higher energy than do the gallium detectors. And variation in one energy range does not necessarily imply variation in another energy range.

But the Super-K consortium has yet to release their data to outside sources. "We hope that the data will be released soon," Sturrock says. "It is essential that all the relevant solar data be analyzed together. Here is a case where the whole is much greater than the sum of the parts."