Planetary Sciences Research Spotlight

Sun's Magnetic Fields Best at Forecasting Solar Cycle Peaks

Models based on the Sun's polar magnetic fields performed best in simulating the solar cycle and predicting solar behavior.

Source: Space Weather


Solar activity level rises and falls every 11 years. The most recent maximum in solar activity level, the 24th since recording began in 1755, was the weakest in almost 100 years and peaked in early 2014.  Solar activity refers to dark regions on the surface of the Sun called sunspots, where the Sun’s magnetic field has become tangled. They can produce sudden explosions of energy in the form of intense radiation and energetic particles. Sometimes, these regions will even kick out a portion of the tangled magnetic field and send it hurdling into space in an event called a coronal mass ejection. When these bursts of radiation, particles, and magnetic field reach Earth, they cause geomagnetic storms, which can interfere with communication satellites and power grids on the ground. These events could also be harmful to astronauts traveling to Mars, exposing them to possibly deadly amounts of radiation.

For each solar cycle scientists publish their forecasts, using their own preferred methods, anticipating how large the peak of solar activity will be and when it will occur. Now that the peak of solar cycle 24 has passed, a new study by Pesnell compares all of the predictions to determine which forecasts were the most accurate.

The two most popular ways to model the solar cycle were (1) to use the historical record of the number of sunspots and (2) to couple the sunspot number to another measurement that varies with the solar cycle. Scientists who used the sunspot number alone argued that past trends can predict statistically what the next solar cycle will look like. This method is particularly appealing because the historical record of sunspots goes back farther than any other space weather measurement. But it turns out that models using this number alone rarely made better predictions than just the average of previous solar maxima.

Scientists who coupled the sunspot number to another measurement predominantly chose either geomagnetic activity level or the strength of the magnetic fields at the Sun’s poles. For past solar cycles, models using geomagnetic activity level made more accurate predictions, but this accuracy may be due to the fact that solar polar magnetic field data were scarce or unreliable. This time, for solar cycle 24, the models using polar magnetic fields made the best predictions by far.

For solar cycle 25, slated to begin as early as 2020, the author predicts that scientists will need more information on the Sun’s magnetic field to increase the accuracy of their models and points out that these predictions will be essential as our society’s  reliance on technology grows and we strive to become an interplanetary species. (Space Weather, doi:10.1002/2015SW001304, 2016)

—Aleida K. Higginson, Freelance Writer

Citation: Higginson, A. K. (2016), Sun’s magnetic fields best at forecasting solar cycle peaks, Eos, 97, doi:10.1029/2016EO046603. Published on 24 February 2016.

© 2016. The authors. CC BY-NC 3.0
  • Philip Judge

    It is tempting to try to predict the Sun’s behavior for at least two reasons. 1. Society wants it. After all, the Sun can mess up our expensive technological infrastructure. 2. It;s behavior looks almost regular. The work you report is interesting, and worth pursuing. Up to a point. There are problems doing this in the name of “science”, though. Personally, I believe that it is premature to ask scientists to predict the outcome of solar eruptions on Earth. The “almost” regularity is fascinating, it is what we might call the emergence of (unreasonable?) order out of a complex system. A few years ago, it was pointed out that the business of predicting the effects of the Sun on the Earth’s space environment is a problem involving two non-linearly coupled, non-linear systems, separated by a large distance in which some odd things can happen to those troublesome solar electrical particles. Thus, the business becomes more of an “art” than a science, as reproducibility becomes rather a subjective thing.

    So I ask, for how long should we pursue such predictions? How do we decide upon success or failure? As scientists, how would we know when to stop spending significant effort on this? It would be interesting to hear some parallels from things such as weather prediction, and the response of immune systems to antagonists, also examples of complex systems.


      The predictive expectation sunspot cycle 24, was very high, which really the Sun emitted. But if we correlate the emission peaks of protons, with the number of sunspots, the approximation is improved. So it is necessary to investigate further the internal radial convective flows of the Sun, and variations in the rotation speed of the surface.

      • Philip Judge

        Dear Eduardo, if I can combine our two comments, I think you are giving a specific example of why such work should still be pursued- (interplanetary?) proton flux vs sunspot number correlation. Indeed, one clear way forward is that so long as statistical correlations continue to be discovered, or perhaps their patterns better understood, one should continue.

        More generally…

        My comment was shared to generate interest but also to try to refine the questions/methodologies in this area. How do we identify the important questions? How do we know when to acknowledge that, like the weather, there are things we will not be able to predict, no matter how much society wants such predictions? Thanks.