Mausumi Dikpati and her colleagues at the National Center for Atmospheric Research in Boulder, Colo., used new observations of the sun's interior and new computer simulations to model the flow of plasma, or electrically charged gas, that carry the sunspots like a conveyor belt until they become powerful enough to burst free and erupt on the sun's surface. As the spots weaken, the belt then slowly carries them toward the poles and, ultimately, back into the sun's core where they become the foundation of the next sunspot cycle. "The remnants from the past three cycles combine to produce a certain seed for the present cycle," Dikpati explains. "We now know that it takes two cycles to fill half the belt with magnetic field and another two cycles to fill the other half. Because of this, the next solar cycle depends on characteristics from as far back as 40 years previously--the sun has a magnetic memory."
This model proved more than 98 percent effective in predicting the relative strength and duration of the past eight solar storm cycles, which last for roughly 11 years, according to the team's report in the current issue of Geophysical Research Letters. And it calls for the next cycle--so-called cycle 24--to be 30 to 50 percent stronger than the present one. Solar outbursts have effects on everything from satellites to the electrical grid here on Earth and predicting such storms is a continuing effort for NASA, NOAA and other government and scientific groups.
The relatively new science of solar weather prediction remains argumentative, however, not unlike terrestrial meteorology. Another model predicted that cycle 24 would be weaker than recent cycles, but the present model's accuracy in predicting past events and scientists' deeper understanding of the underlying solar physics may give it an edge, according to David Hathaway, a solar astronomer at NASA's Marshall Space Flight Center. But even he disagrees with some of the model's findings. It predicts that the next cycle will be delayed until late 2007 or early 2008. "We have found that large cycles usually start early," he says. "We think the next cycle will start late this year or early next year. We're just anxiously awaiting the appearance of those first spots."