Abstract/Summary

Space weather has long been known to approximately follow the solar cycle, with geomagnetic storms occurring more frequently at solar maximum than solar minimum. There is much debate, however, about whether the most hazardous events follow the same pattern. Extreme events – by definition – occur infrequently, and thus establishing their occurrence behaviour is difficult even with very long space-weather records. Here we use the 150-year aaH record of global geomagnetic activity with a number of probabilistic models of geomagnetic-storm occurrence to test a range of hypotheses. We find that storms of all magnitudes occur more frequently during an active phase, centred on solar maximum, than during the quiet phase around solar minimum. We also show that the available observations are consistent with the most extreme events occurring more frequently during large solar cycles than small cycles. Finally, we report on the difference in extreme-event occurrence during odd- and even-numbered solar cycles, with events clustering earlier in even cycles and later in odd cycles. Despite the relatively few events available for study, we demonstrate that this is inconsistent with random occurrence. We interpret this finding in terms of the overlying coronal magnetic field and enhanced magnetic-field strengths in the heliosphere, which act to increase the geoeffectiveness of sheath regions ahead of extreme coronal mass ejections. Putting the three “rules” together allows the probability of extreme event occurrence for Solar Cycle 25 to be estimated, if the magnitude and length of the coming cycle can be predicted. This highlights both the feasibility and importance of solar-cycle prediction for planning and scheduling of activities and systems that are affected by extreme space weather.