The Sun is the source of the energy in our lives as it has been for a few billion years. It’s a reliable and rather benign star. It’s unlike alpha Centauri that caught imagination a decade ago as habitably warm… but prone to bursts that would sterilize any planets.
The Sun has evolved and created a great story with Snowball Earth and the cyanobacteria. It’s calm, but we do get space weather. Solar storms as coronal mass ejections of particles happen often. Some aim at the Earth. We get beautiful auroras. We also get striking impacts on our satellites and our power grid. It’d be nice to predict the solar storms reliably to shut down susceptible bits of our technology.
Astronomers Juie Shetye from NMSU and Mausumi Dikpati from NCAR now present a sophisticated model to predict sunspot numbers, which are associated with solar storms hitting the Earth. It is a statistical or probabilistic model, which is not to say that it make random guesses. It’s well beyond classical statistics. It adds in machine learning, a welcome nontrivial use departing from fake videos and the like.
The model makes very good hindcasts, predicting numbers of sunspots in the past from partial data. Importantly, it also makes forecasts. Expect a peak this December. We’ll see what else comes with sunspots!
This has been an outreach activity of the Las Cruces Academy, viewable at GreatSchools.org
Source: Astrophysical Journal 20 October 2025
Image: Jeff Johnson, Las Cruces
