Solar cycle 25: past the peak, still producing storms

Where cycle 25 stands in 2026

The solar cycle is counted in sunspots. SWPC's Solar Cycle Progression page plots the monthly International Sunspot Number from the Royal Observatory of Belgium, and a smoothed version that averages it over thirteen months to remove the month to month noise. Cycle 24 ended and cycle 25 began in December 2019. The smoothed number climbed through 2022 and 2023 and reached 160.9 in October 2024, the highest value of the cycle in the files read on October 13, 2026.

The smoothing has a cost: it needs six months of data on each side, so the latest smoothed value is always about half a year old. In those files it is 94.5 for March 2026, 59% of the peak. The raw monthly numbers after that are noisier; the last one, for September 2026, is 60.3. SWPC flags the latest six months as preliminary.

The cycle was stronger than expected. In 2019 a panel convened by NOAA, NASA and the International Space Environmental Services predicted a peak of 115 in July 2025, with a range of 105 to 125 between November 2024 and March 2026. The observed peak came a month before that time window and well above the range, and in February 2025 SWPC replaced the panel's curve with an updated prediction, refitted each month to the observed values.

How the cycle is counted

The sunspot number is older than any satellite. Cycles are numbered from cycle 1, which began in 1755, and the count follows a convention set long ago: each sunspot group counts for ten and each individual spot inside it for one, so a Sun with a single small spot already scores 11. SWPC's sunspots page explains the rule and notes that the official daily and monthly numbers come from the World Data Center SILSO at the Royal Observatory of Belgium. The previous cycle gives a sense of the rhythm: cycle 24 began in December 2008, peaked in April 2014 and ended in December 2019, eleven years in all.

SWPC tracks a second index next to the sunspots, the 10.7 cm radio flux of the Sun measured at Penticton in British Columbia. It follows the same cycle and matters to satellite operators, because it tracks the heating of the upper atmosphere that drags on low orbits. For the aurora, the sunspot number is the more intuitive gauge, because spots mark the active regions where eruptions start.

What SWPC predicts for the next years

SWPC's prediction fits a curve with the average shape of past cycles to the observations, taking into account that stronger cycles tend to rise faster. Its uncertainty bands come from running the same method on earlier cycles at the same stage and measuring how far off it was. The table shows the predicted sunspot number for October of each year and the end of the published table, with the band SWPC gives as the likely range.

On this prediction the cycle falls below half its peak around May 2027, and toward single digits by the end of the decade. SWPC expects solar cycle 26 to begin some time between January 2029 and December 2032 and does not yet forecast its strength. The mini-simulator above gives the predicted value and its range for any month to the end of the table.

A sunspot number is not a storm forecast. SWPC writes that its cycle prediction gives a rough idea of how often space weather storms of all types will occur, and it is used that way by satellite operators and power companies. It cannot say whether a given month will bring a G4 storm.

Why the aurora does not stop on the way down

Two different sources drive geomagnetic storms, and they do not follow the cycle in the same way. The first is the coronal mass ejection, a cloud of magnetized plasma blown off the Sun, usually from the complex active regions around sunspot groups. SWPC describes CMEs leaving the Sun at speeds from below 250 km/s to nearly 3,000 km/s; the fastest ones reach Earth in as little as 15 to 18 hours. The geomagnetic storms page credits them with the largest storms. Active regions grow rarer as the cycle declines, but the ones that do appear can still be large and complex, and a single eruption is enough for a G4 or G5 night.

The second source is the high-speed stream of solar wind from coronal holes, which SWPC puts at 500 to 800 km/s against about 400 km/s for the ordinary slow wind. When a fast stream overtakes slow wind, it builds a compressed region of strong magnetic field. SWPC says the storms from these regions are less intense than CME storms but can deposit more energy over a longer interval. Coronal holes reaching toward the Sun's equator are a familiar feature of the declining years, and because the Sun rotates in about 27 days, a stream that produced a G1 or G2 storm once often returns four weeks later.

History backs this up. The storms of October 2003, among the largest of the space age, came in the decline of cycle 23, years after its maximum. In the current cycle, the G5 storm of May 10, 2024 came several months before the smoothed peak. NOAA's storm scale counts about 100 G4 storms and 4 G5 storms per cycle on average, spread unevenly across its eleven years.

What it means for watching in 2026 to 2028

For the northern tier of the United States, where a G1 or G2 storm is usually what it takes, the next two or three years remain a good period. Recurring high-speed streams are well suited to that level, and their 27-day rhythm makes them partly predictable: SWPC's 27-day outlook projects geomagnetic activity day by day over one solar rotation, which helps spot when a known stream may come back. For the rarer G3 to G5 storms that reach the central states, the odds fall slowly as active regions thin out, but they do not reach zero until the Sun is close to minimum.

In practice nothing changes in the method: the forecast for your town reads the Kp forecast and the clouds of the night, whatever the phase of the cycle. The storm scale guide explains what each level brings, the solar wind guide shows how a storm announces itself an hour ahead, and the three-night forecast tells you whether the next storm reaches you.