1947 Sunspot Shows the Sun Can Produce a Superflare

A team from the Max Planck Institute for Solar System Research (MPS) in Göttingen and the University of Colorado published a paper on September 10, 2026, in the journal Philosophical Transactions of the Royal Society A that, for the first time, calculates the upper limit of flare energy for the largest sunspot groups ever recorded. The conclusion, from a paper led by Natalie Krivova: the sun has the potential to produce a so called superflare, a burst of radiation that would exceed anything ever directly measured on it.

The question itself is not new. In late 2024, MPS researchers showed, using a sample of sun like stars, that such stars on average produce a superflare about once every hundred years per star. Over the seventy years of the space age, meaning the period in which instruments outside Earth’s atmosphere have been able to directly measure flare energy, the sun itself has not produced a single one. An earlier estimate from 2013, for the largest known sunspot group, put its energy at 6 x 10^33 ergs, below the threshold the new study sets for a superflare: 10^34 ergs, or 10^27 joules.

How Was the Limit Calculated?

The new approach does not start from a theoretical model of the sun, but from actual measured data. The researchers analyzed the catalog of the Solar Dynamics Observatory, covering eruptions recorded between 2010 and 2016. According to the German outlet Ingenieur.de, the catalog contains 3,137 solar eruptions, including 306 particularly powerful M and X class flares; the outlet phys.org describes the same sample more broadly, as the 300 strongest flares of that period. The team linked the energy of each eruption to the size of the active region, the part of the sun’s surface carrying a particularly strong and complex magnetic field, from which the eruption originated.

That statistical relationship was then applied to the largest sunspot group ever recorded in the roughly 400 year history of systematic observations. The group designated RGO 14886 reached its largest extent on April 8, 1947, covering about 6,132 millionths of the sun’s visible hemisphere, or roughly 0.6 percent of the visible disk, with a diameter of about 40 Earth diameters. For comparison, the active region from October 2003, known for the so called Halloween storms, covered about 3,500 millionths, roughly half the area of the 1947 group. According to the original records, the 1947 group did not produce any unusual flare at the time.

Applied to that historical case, the new method yields, under typical scenarios, an energy of several times 10^33 ergs, while in the most extreme statistical cases the estimate crosses 10^34 ergs, putting it into superflare territory.

“Our sun has the potential for a superflare. It can create enormous sunspot groups that, in principle, could serve as the starting point for the most extreme releases of radiation,” Krivova said in a statement from the institute.

The paper on the 1947 sunspots is signed, alongside Krivova, by Teodosios Chatzistergos, Marta Kazachenko and Erdal Isik. The same day, in the same journal, a second, companion paper appeared, led by Valery Vasilyev, also an MPS scientist. That paper examines the link between solar magnetism, extreme particle events and superflares on other stars, and is not part of the sunspot study itself.

“According to the current state of research, extreme particle eruptions and especially powerful flares often, but not always, occur together,” Vasilyev said.

Neither paper proves that a superflare has ever actually occurred on the sun, or when, if ever, one might happen again. Krivova stressed this herself:

“There are certain indications that the sun, too, can produce superflares at long intervals. However, there is no direct evidence.”

The estimate for 1947, as the authors themselves note, rests on the assumption that historical sunspots are physically comparable to today’s active regions, and uncertainties compound at every step, from the size of the sunspot to the estimated flare energy. The result is an upper limit on what is possible, not a recorded event, and not a prediction of when such an event might recur.