It turns out that, as the sun rotates on its axis, one gas layer gradually spins faster while the other reduces speed. Scientists are at a loss to explain the phenomenon, which occurs in regular 12-to-16- month cycles.

"It's not what we expected at all," says Stanford research physicist Jesper Schou. "It comes totally out of the blue."

Schou is part of an international team of researchers using satellite and ground-based observatories to monitor the sun.

Writing in the March 31 issue of the journal Science, Schou and postdoctoral fellow Rasmus Larsen point out that these unusual but predictable changes in rotational speed only occur above and below a section of the sun known as the interface layer or tachocline.

Located about 135,000 miles below the solar surface, the tachocline separates the sun's two major regions of gas: the radiative zone, which includes the energy-generating core, and the convection zone near the surface.

Solar experts believe that the tachocline may be the source of powerful magnetic fields that produce strong solar flares and solar winds, and create sunspots that mysteriously appear and disappear during an 11-year cycle.

No one knows how the sun's enormous magnetic fields are generated, or why they reverse polarity from positive to negative every 11 years. But the discovery that the area surrounding the tachocline varies its rotation in a regular pattern could be a clue to solving the mystery.

The research team used independent data from two instruments to detect changes in the solar spin-rate between May 1995 and Nov. 1999.

Stanford team members Schou and Larsen used data from the Michelson Doppler Imager (MDI) on board the Solar and Heliospheric Observatory (SOHO).

Launched in 1995, SOHO is positioned about a million miles from Earth.

The MDI instrument on board SOHO creates a picture of the solar interior by measuring millions of sound waves that constantly ricochet inside the sun.

Independent observations of the sun were made by the Global Oscillation Network Group (GONG), a team of scientists that collects data from six Earth-based solar observatories.

Funded by the National Science Foundation, GONG is managed by researchers from the Tuscon-based National Solar Observatory, including Rachel Howe, lead author of the March 31 Science article.

After analyzing approximately four years of data, the MDI and GONG teams came up with remarkably similar findings.

Their results showed that the convection zone just above the tachocline gradually increased its rotational speed by about 60 feet per second between July 1996 and Feb. 1997, then slowly returned to its original velocity some eight months later.

Meanwhile, the radiative zone just below the tachocline demonstrated the exact opposite behavior, slowing down between July and February, then gradually accelerating eight months later.

This cycle of acceleration and deceleration repeated itself roughly every 16 months, or 1.3 years, at the equator, but only recurred every 12 months at the 60-degree latitude.

The discovery that the inner sun spins at different rates at different latitudes is consistent with earlier studies showing that the surface of the sun also rotates at different speeds.

For example, at the equator, it takes about 25 days for the surface of the sun to rotate on its axis, but at the poles, surface rotation requires roughly 33 days. That's because the sun is made of gas, so different parts of its surface spin independently - unlike the surface of Earth, Mars and other solid planets.

But why are the gas layers above and below the tachocline speeding up and slowing down at opposite rates?

Perhaps this puzzling behavior is somehow related to the mysterious forces that generate the sun's magnetic field and the 11-year sunspot cycle.

"For the interior to change speed every 11 years would make sense," notes Schou. "But a 1.3-year period was unexpected. We don't know what it means, but isn't it interesting!"

And practical, too, because if researchers can determine what drives the sun's magnetic field, they also may be able to forecast solar flares and winds that can knock out satellites, increase the risk of radiation to airline passengers and even cause power outages on Earth.

The ability to predict solar storms could help people avoid such incidents and may even provide researchers insight into the sun's long-term impact on the Earth's climate.

The SOHO spacecraft, which is jointly operated by NASA and the European Space Agency, made headlines in 1998 when it temporarily stopped functioning - an event that caused MDI scientists to lose five months of solar data.

The SOHO mission is scheduled to end in 2003, but Schou and his colleagues would like to continue their observations to see if the 12-to-16-month rotation cycle varies year after year.

NASA has announced plans to replace SOHO with the Solar Dynamics Observatory (SDO).

If launched, SDO would remain in geosynchronous orbit about 22,000 miles above the Stanford campus, allowing researchers to start downloading new acoustical data from the sun beginning in 2006.

Stanford Professor Philip Scherrer, principal investigator of MDI, hopes that NASA will obtain funding for SDO, allowing space- based solar research to continue.

"It's pretty neat to look at that kind of detail inside a star," says Scherrer. "It's really fun!"

The March 31 Science article is co-authored by Rachel Howe, Frank Hill and Rudi Komm of the National Solar Observatory (NSO); Joergen Christensen-Dalsgaard of Aarhus University in Denmark; Michael Thompson of Queen Mary & Westfield College in London; Juri Toomre of the University of Colorado, Boulder; and Rasmus Munk Larsen and Jesper Schou of Stanford.