Paper

“A Population of Short-Period Variable Quasars from PTF as Supermassive Black Hole Binary Candidates” by Charisi et. al.

http://arxiv.org/abs/1604.01020v1

Summary

This paper searched through a sample of 35383 spectroscopically confirmed quasars in the Palomar Transient Factory (PTF) photometric database, for candidate SMBHB binaries. The luminosity of a typical quasar is believed to follow a damped random walk (DRW). They did something similar to Graham et. al. (2015), who looked for periodic variability beyond the DRW in the Catalina real-time transient survey (CRTS) database.

The periodic variability is expected for the following reasons:

• If a SMBHB is embedded in a thin accretion disc, the binary will expel gas from the central region.
• This will leave a cavity in the central region.
• Eventually, the gas from the edge of the cavity is pulled inwards again.
• This process causes the mass accretion rate, and as a proxy the luminosity, to vary with the orbital period, which is backed up by simulations.

Turn to Figure 5 on page 7. Using a periodogram of each quasar, they determined a periodicity. Depending on the significance of the peak, it was either rejected or accepted as a SMBHB candidate. This plot shows how the number of significant sources varies with the significance threshold, for both the data set used, and simulated DRW light curves, using parameters similar to that of the data set. They kept only the objects in the orange region, which have a very high significance, and were observed for at least 1.5 periods.

Turn to Figure 7 on page 10. They show the number of objects they would expect to observe, as a function of the “residence time”, or the time a binary is expected to spend at a particular orbital period. This distribution changes based on the mass ratio of the binary: q. The dotted/solid line shows what is expected without/with observational effects considered. The histograms show the distribution based on their candidates. As you can see, none of the cases fit particularly well, but the low mass ratio (q=0.01) fits the best.

If the majority of SMBHBs have a low mass ratio, then that is problematic, as they would not be in the GW regime, and would not be detectable by eLISA or the PTA. I find this argument problematic, however, as it assumes that all SMBHBs have the same mass ratio.

Turn to Figure 10 on page 12. They repeated the same analysis as before on the SMBHB candidates from Graham et. al. (2015). That paper had assumed equal mass ratios (q=1). As with PTF, this agrees better with q=0.01. They also had to scale down the predicted values by 25%, which they attribute to selection effects, though I suspect that may have influenced their decision that q=0.01 is better.