Geomagnetic activity late in the sunspot cycle has been used successfully to forecast the amplitude of the following cycle. This success is somewhat surprising, however, because the recurrent high-speed wind streams that trigger the activity are not proxies of the Sun's polar fields, whose strength is a critical factor in many solar dynamo models. Instead, recurrent geomagnetic activity signals increases in the Sun's equatorial dipole moment, which decays on the ∼1-2 yr timescale of the surface

more »

... meridional flow and does not survive into the next cycle. In accordance with the original empirical method of Ohl, we therefore argue that solar cycle predictions should be based on the minimum level of geomagnetic activity, which is determined by the Sun's axial dipole strength, not on the peak activity during the declining phase of the cycle. On physical grounds, we suggest that an even better indicator would be the total open flux (or strength of the radial interplanetary field component) at sunspot minimum, which in turn can be derived from the historical aa index by removing the contribution of the solar wind speed. This predictor yields a peak yearly sunspot number R max = 97 ± 25 for solar cycle 24.