So apparently they are all lightcurves of (likely candidates) Active Galactic Nuclei and in some way or the other they all appear to be periodic: enter image description here enter image description here enter image description here I'd like to know the reason for this and the underlying physical mechanism.

  • 3
    $\begingroup$ I'm not sure I see any evidence of periodicity -- can you expand on what you see? Some of the pattern is almost certainly due to gaps between observations. Also, can you add a source for those plots? $\endgroup$
    – HDE 226868
    Commented Aug 7, 2023 at 14:49
  • 5
    $\begingroup$ Who says they are periodic? $\endgroup$
    – ProfRob
    Commented Aug 7, 2023 at 20:47

2 Answers 2


I don't think what you're seeing is periodicity; my bet is that it's just stochastic red noise.

Searches for periodicity or quasi-periodic oscillations (QPOs) in AGN light curves is a difficult and at times contentious task. Plenty of phenomena have been invoked to explain candidates. I kinda like the short summary in the introduction to Krishnan et al. 2021; any list should include interactions with a binary companion, jet precession, and accretion disk hot spots, depending on the wavelength in question.

However, one enormous problem with searching for and trying to verify periodicity claims is that AGN light curves are almost always dominated by stochastic variations. One common model is a damped random walk (see e.g. Kelly et al. 2009), explained as arising from thermal fluctuations within the accretion disk, although there's plenty of debate as to what's really going on. The resulting light curves can, in some cases, look periodic, but there's really nothing there.

Quantitatively, too, this red noise can trip up standard periodicity search techniques. In an appendix, MacLeod et al. 2010 give examples of how naively computing a Lomb-Scargle periodogram can produce spurious candidates because the method compares a periodic signal to white noise, whereas the comparison should really be between a periodic signal and red noise. You need to use far more robust methods to truly find periodicity or QPOs.

That's not to say that all periodicity candidates are spurious! There are increasingly better methods that take this red noise into account, and some AGN, like the famous OJ 287, show strong evidence for periodicity due to a binary companion. But most of the time, variations are just due to red noise. My bet is that that's what you're seeing.


AGN's are divided into roughly 3 types, Quasars, Seyfert galaxies and radio galaxies. All of these produce light, each in their own electromagnetic spectrum and intensity like quasar emits across all, seyfert galaxies have less intensity, and radio galaxies emit in the radio spectrum.

However they oscillate, at an intensity, frequency known as Quasi-periodic oscillation which happens in the case of a compact object like black holes, neutron stars and white dwarfs along the edge of the accretion disk near the congestion zone. This quasi-periodic oscillations are noticed on the lightcurves. Quasiperiodic oscillations occur mainly because of the dynamics of the accretion disk, and the relativistic effects.

So to sum up, the periodic oscillations observed on the light curve are quasi-periodic oscillations which occur mainly because of the dynamics of the accretion disk, and the relativistic effects such as relativistic beaming.

Note: Most of the times, Quasars are not completely periodic and it has signs of slight variability though it is mostly periodic. And it may not be related to QPOs at all, it may be just some background noise and cause stochastic variability, because analysing lightcurves is an extremely difficult and tedious task and may require machine power

Thanks, Hope it helps you!


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .