A glance at the figure below (of sunspot numbers from Solar Cycles 1-23) reveals a pattern immediately: roughly every eleven years, the sunspot numbers climb to a maximum. In between, they fall to nearly zero. With a longer look, however, a slightly more subtle phenomenon is visible. Not every solar maximum involves a climb to the same height, though the minima are all at similar levels. Comparing the sharp peak around 1957 (up to 200 sunspots per day) and the minor ones in the first half of the 1800s (only about 50), one sees a huge difference. What causes such a difference between solar cycles?

The answer to that question remains unknown, but anyone can see that the difference exists. These differences of 150 sunspots can cause important effects at earth, making the amplitude of the solar maximum a critical number to know. Different solar maxima, depending on their heights, can coincide with a Little Ice Age (as during the Maunder Minimum in the last half of the seventeenth century) or with a period of solar-geomagnetic activity so impressive that an International Geophysical Year is declared (as in 1957-58). People will want to know whether their TV signals will be interrupted because of magnetic storms or, though it's highly unlikely, whether they'll have to bundle up better than usual. With Solar Cycle 23 peaking in 2000, there is a scramble to predict whether the latest max will soar to heights near 200 (like the last few) or rise to only an average 100 or so. But how can such a thing be predicted?

Actually, quite a few methods exist for predicting the amplitude of a solar cycle even before the cycle begins. The first, suggested by Ohl and Ohl in 1979, relates the geomagnetic aa index at the last solar minimum to the amplitude of the next solar maximum (although the data for this are sometimes unavailable until after the cycle begins). A second method, developed by Joan Feynman, uses the same aa data, but relates the "extra data" remaining after the data in phase with the sunspot cycle is removed to the next solar cycle a few years in advance. Also, Richard Thompson found a way to relate the number of days of disturbed conditions in Earth's magnetic field during one solar cycle to the amplitude of the next solar cycle. A very new method found by the Air Force Research Laboratory calculates the time of the solar maximum using emission features seen in Fe XIV that appear at 55 degrees latitude on the Sun, travel toward the poles for three or four years, and disappear at the poles fourteen months after the solar maximum. So what is in store for us this solar maximum?

The latest-prediction graph from the Marshall Space Flight Center is above. The solid, smooth curve shows the predicted values for the entire cycle, with a maximum occurring near June 2000 at around 140. The two smooth, dashed curves on either side of the curve show the error limits of the prediction -- it could be off by that much. Finally, the jagged line climbing up the curve is made up of actual sunspot-number observations. One can see that it follows the prediction fairly closely, at least at the beginning of the cycle. Earlier predictions (such as those of the Solar Cycle 23 Project panel in 1996) suggested a maximum of about 160 in March 2000. In early 1998, the Marshall Space Flight Center predicted a maximum of 170 in June 2000. With the kind of fluctuations visible in the latest-prediction graph, it comes as no surprise that the predictions keep changing. The one thing that seems relatively certain is that the upcoming solar maximum will break no records for height or lack thereof -- it may be slightly above average, but nowhere near devastating.