The Sun is not a perfectly steady lightbulb. Its brightness quivers slightly: it dims a bit, then brightens. For Earth's climate, even slow variations stretched over thousands and millions of years matter. They could have left their mark in the geological record of the last 540 million years.
Astronomer Geller measured the brightness of 170 stars in the open cluster M67. These stars are nearly the same age as the Sun and similar in size and temperature. Their spread on the color–luminosity diagram looks like a smeared band rather than a thin line. Part of that smearing is just measurement noise, like graininess in an old photograph. But when the noise is subtracted, a small intrinsic spread remains: the stars do indeed "breathe" in brightness.
The new analysis shows that the intrinsic brightness variations of such stars do not exceed 5–9 percent. This is an upper limit; the real changes are likely even smaller. For comparison, pure random noise gives about 10 percent. So the Sun probably did not suddenly dim or flare in the past. But even gentle variations could have nudged the climate into warmer or colder epochs.
as written by the authors arXiv:2603.23291
Variations in the luminosity of the Sun on timescales of thousands to millions of years could potentially be responsible for terrestrial climate variations in the Phanerozooic geological period (last 540 million years). In this paper, I consider a method that utilizes observations of an open star cluster with approximately the age of the Sun, specifically M67, with data taken from Geller (2015). The idea is to measure the width of the main sequence in the region of solar-type stars, here generously defined to be about spectral class G0 - K1. This width gives an estimate of the dispersion in absolute magnitude of nominally solar-type stars. The sample used consists of 170 solar-type main sequence stars which are not known to be binaries. With this sample, I form an empirical measurement of the width of the main sequence, which is compared with a theoretical expression from Spangler (2025). The measured spread is consistent with a value of sigma, the normalized Gaussian variability of the primary and (if present) secondary star, in the range of 0.100 - 0.135. However, the expected value of sigma from purely photometric noise is estimated as 0.101. Generous upper limits to the intrinsic variability contribution to the inferred width are sigma in the range 0.058-0.089. These limits are not totally devoid of interest in a paleoclimatic context. However, major improvements in the technique are possible with the use of existing data sets from space astronomy missions such as Gaia and Kepler.
The cluster M67 is older than most open clusters: most of them dissolve within a few hundred million years, but M67 has survived for 4 billion years.