For around 67 days, the source is silent. Then, for around 90 days, it wakes up again, spitting out repeated millisecond radio flares before falling silent, and the whole 157-day cycle repeats.
However, fast radio bursts are extremely mysterious, and there was no guarantee that the cycle would continue. So it's pretty exciting that the source has flared up again, right on cue - consistent with predictions of its activity cycle.
This suggests that there's significant value in monitoring known fast radio burst sources - but also in continuing to watch FRB 121102 to try to understand what could be causing the phenomenon.
A quick refresher: fast radio bursts are, as the name suggests, bursts of radio waves that are very fast, just a few milliseconds long, coming from galaxies millions to billions of light-years away. But they're also extremely powerful; within those milliseconds, they can discharge as much power as hundreds of millions of Suns.
Most of the time, they flare once and we have not heard from them since, making them impossible to predict and very difficult to trace. And we don't know what causes them, although recent evidence points pretty strongly to a type of neutron star called magnetars.
But a handful of fast radio burst sources have been detected repeating, and these could be one of the keys that helps at least partially solve the mystery.
Before its cycle was discovered by University of Manchester astronomer Kaustubh Rajwade and his team, FRB 121102 was already famous for being the most active fast radio burst discovered yet, spitting out repeated bursts several times since its discovery in 2012.
Because it repeats, astronomers could watch for activity, and trace it to a source galaxy. It was the first fast radio burst to be localised, to a star-forming region in a dwarf galaxy 3 billion light-years away.
The discovery of periodicity in its activity - based on five years' worth of data - could place some important constraints on what it could be.
For instance, high-mass X-ray binaries in the Milky Way - those that contain neutron stars - can have orbital periods of up to hundreds of days. But there are some types of binary systems with much shorter periods - these could be ruled out for FRB 121102.
And now, periodicity is supported by new sets of observations - although the timing may need revision.
A team led by Marilyn Cruces of the Max Planck Institute for Radio Astronomy detected 36 bursts from FRB 121102 using the Effelsberg 100-m Radio Telescope between September 2017 to June 2020. Combined with the data from Rajwade's research, the team derived a periodicity of 161 days, in a new preprint paper uploaded to arXiv.
This paper gives dates between 9 July and 14 October 2020 for the source's active period.
But Cruces and her team aren't the only ones looking. A team led by Pei Wang of the National Astronomy Observatory of China used the Five-hundred-meter Aperture Spherical radio Telescope to monitor FRB 121102's location on several dates between March and August 2020.
Between mid-March and late July, they didn't detect any bursts. But on 17 August, FAST detected at least 12 bursts from FRB 121102 - suggesting the source is once again in an active phase - although the team calculated a different periodicity from both Rajwade's team and Cruces' team.
"We combine the bursts collected in Rajwade et al. (2020) and Cruces et al. (2020) with these newly detected by FAST in 2019 and 2020, and obtain a new best-fit period of ~156.1 days," they wrote in a notice posted to The Astronomer's Telegram.
According to Wang's team's calculations, the active phase is due to end between 31 August and 9 September 2020. If FRB 121102 continues to show activity beyond these dates, this could suggest that either the periodicity isn't real, or that it has somehow evolved, they noted in their post.
Of course, it's also possible that the periodicity calculations need to be refined. Which means we should continue to keep an eye on FRB 121102.
"We encourage more follow-up monitoring efforts from other radio observatories," the researchers wrote.