Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence

arXiv:2610.30772 · astro-ph.GA, astro-ph.HE, astro-ph.IM · Submitted 2026-10-22 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Astrophysics Radio. Generated commentary on the latest astrophysics papers.

Vera: I'm Vera.

Jocelyn: And I'm Jocelyn, here with Subrahmanyan.

Vera: We're looking at the paper: "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence".

Jocelyn: Here's the gist of the paper. This research investigates how recovery fractions for periodic supermassive black hole binaries in quasar light curves change across different black hole masses and observational cadences.

Vera: How it works The study tests how well the Lomb-Scargle periodogram (LSP) recovers injected sinusoidal and sawtooth signals into damped random-walk (DRW) light curves matched to quasars, using detection thresholds set by 105 noise-only light curves per binary on Page 1.

Jocelyn: The authors inject sinusoidal and sawtooth signals into light curves of mergers in the TNG100-1 cosmological simulation, drawing magnitudes from DR7 quasars of similar mass on Page 1.

Vera: They apply a threshold in each frequency bin, set by 105 noise-only light curves per binary on Page 3.

Title and authors: Vera: So, we’re looking at a paper called "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence," and they’re testing how well we can find periodic signals from supermassive black hole binaries in quasar light curves by changing the mass range and how often we take measurements.

Jocelyn: That’s interesting because they are testing the Lomb-Scargle periodogram against injected sinusoidal and sawtooth signals in damped random-walk light curves across three cadences: PTF-like, idealised, and LSST-like ones.

Subrahmanyan: And by changing the mass range, they are connecting this recovery fraction to how structure forms in the universe because these binaries are key players in that growth.

Vera: The study uses a catalogue from the TNG100-one cosmological simulation, injecting signals into light curves of mergers with magnitudes drawn from DR7 quasars of similar mass <ref:2610.30772#pg1,with magnitudes drawn from DR7 quasars of similar mass>.

Jocelyn: They set their detection thresholds in each frequency bin by using one hundred five noise-only light curves per binary on observed PTF cadences, one Stripe eighty-two cadence, and an LSST-like cadence <ref:2610.30772#pg1,105 noise-only light curves per binary on observed PTF cadences, one>.

Subrahmanyan: They found that the primary masses span log M1 from six point zero seven to ten point zero nine with a median of seven point eight seven, and forty six point four per cent of the systems sit at log M1 greater than or equal to eight.

Vera: They also used quasar magnitudes assigned by a conditional draw from the SDSS Data Release seven catalogue, using its fiducial virial black hole mass and its Mi at redshift z equals two <ref:2610.30772#pg1>.

Jocelyn: The DRW amplitude depends on the absolute i-band magnitude as well as the black hole mass, which means every system needs one of those values to model correctly.

Subrahmanyan: They also looked at three types of cadence: PTF-like, idealised over one epoch per day over the same baselines, and LSST-like modeled on the Legacy Survey of Space and Time.

Vera: The PTF cadences hold six thousand two hundred seven epochs on two thousand eight hundred sixteen nights, which gives a mean of one hundred twenty-seven epochs over a mean baseline of three point nine years.

Jocelyn: The LSST-like cadence is their realization of the LCH26 specification, which is a ten year baseline at five day sampling jittered by zero point zero one day.

Subrahmanyan: They compared their results against the LCH26 values on page nine, and they also compared them against other published recovery fractions on page ten.

Vera: In the high-mass range with equal mass weight, the PTF run recovers thirty-nine point eight plus or minus one point six per cent of sinusoids on significance only detections against LCH26's forty four point seven per cent >

Jocelyn: But then when they switch to the LSST-like cadence, that sinusoid recovery jumps up to seventy-six point nine plus or minus one point one per cent against their published value of twenty two point six per cent >

Subrahmanyan: That jump from twenty two point six to nearly seventy percent for the LSST cadence is a significant result because it suggests our future survey designs might be much more effective at capturing these signals than what was previously estimated >

Vera: It really shows how much cadence matters when you are searching for these periodic signals, and they even noted that in the high-mass range, this sinusoid recovery climbs from seventy two point three per cent in the log M1 bin of eight point zero to eighty two point one per cent in the bin of nine point five to ten >

Jocelyn: That climb is pretty telling because it shows that for brighter and heavier systems, the recovery gets much better on the LSST-like cadence compared to lower mass bins >

Subrahmanyan: And they pointed out that in their forecast for a future archival search, applying the recovery of massive systems at every mass actually overstates the detection efficiency at lower masses >

Vera: It seems like that’s because the PTF sawtooth lands above their predictions, and power spilling onto the harmonics explains two point six points of that four point two excess in their forecast >

Jocelyn: So, to wrap up "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence," the paper shows us how changing mass and cadence really shifts our expectations for finding these periodic black hole signals >

Subrahmanyan: It’s a solid check on the statistical reliability of these detections across different observational setups, which is important for connecting binary evolution to large-scale structure formation >

Vera: We're still looking at how we can use this information to design next-generation surveys that really capture this physics across the whole mass spectrum >

Jocelyn: Next up, we’ll be talking about those dust-origin transitions in very distant stardust galaxies and what that means for galaxy evolution at high redshift.

The paper's summary: Vera: So, to wrap up this paper, "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence," they're showing us how much the way you look at these periodic signals—the mass of the black hole and how often you take measurements—really affects what you can actually detect.

Jocelyn: Exactly, it's about testing the Lomb-Scargle periodogram against those injected sinusoidal and sawtooth signals in damped random-walk light curves across different cadences like PTF-like, idealised, and LSST-like ones.

Subrahmanyan: The big thing here is seeing how that recovery fraction changes depending on the black hole mass and the cadence you use for observation because that ties into how structure forms in the universe because these binaries are key players in that growth.

Vera: Yeah, they found that holding things constant, if you look at high-mass ranges with equal mass weight, their PTF run recovers about thirty-nine point eight plus or minus one point six per cent of sinusoids on significance only detections against LCH26's forty-four point seven per cent.

Jocelyn: But then when they switch to the LSST-like cadence, that sinusoid recovery jumps up to seventy-six point nine percent against their published value of twenty-two point six percent.

Subrahmanyan: That jump from twenty-two point six to nearly seventy percent for the LSST cadence is a significant result because it suggests our future survey designs might be much more effective at capturing these signals than what was previously estimated.

Vera: It really shows how much cadence matters when you are searching for these periodic signals, and they even noted that in the high-mass range, this sinusoid recovery climbs from seventy-two point three percent in the log M1 bin of eight point zero to eighty-two point one percent in the bin of nine point five to ten.

Jocelyn: That climb is pretty telling because it shows that for brighter and heavier systems, the recovery gets much better on the LSST-like cadence compared to lower mass bins.

Jocelyn: Overall, it seems that at the LSST-like cadence, our recovery above one hundred eight solar masses is about three times the published sinusoid fraction and about forty times the published sawtooth fraction.

Subrahmanyan: That disparity remains unexplained because no third source settles it since no amplitude-resolved periodogram recovery curve at LSST-like cadence with a threshold in each frequency bin has been published other than their Figure five.

Vera: So for anyone planning future surveys, the implication is that you should be careful when you apply the recovery of massive systems across all masses because it overstates the actual detection efficiency at lower mass ranges.

Jocelyn: We’ve seen how sensitive these recovery fractions are to both the mass of the system and how quickly we sample it, and this paper gives us a concrete benchmark for that sensitivity.

Subrahmanyan: And this work on "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence" shows us exactly where our observational biases might be leading us when we plan the next generation of searches.

Vera: We're still looking at how we can use this information to design next-generation surveys that really capture this physics across the whole mass spectrum.

Jocelyn: Next up, we’ll be talking about those dust-origin transitions in very distant stardust galaxies and what that means for galaxy evolution at high redshift.

The paper's improvements: Vera: So, we’re looking at what the authors suggest they could do to make this search method even better in their paper, "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence."

Jocelyn: Right, they talk about a few ways to improve how we search for these signals. One key idea is drawing magnitudes from DR7 quasars of similar mass instead of using a fixed absolute magnitude Mi, which allows them to capture how the recovery fraction travels across that distribution.

Subrahmanyan: That’s interesting because it means you're not assuming one specific brightness for every system; you’re letting the brightness itself become part of the statistical test, which should give a more realistic picture of detection efficiency.

Vera: They also discussed how cadence selection could be improved by moving beyond just matching timescales to incorporating a density-based comparison of sampling versus baseline length. They found that the drop from forty-four point seven to twenty-four point four per cent happens at one baseline, but the step up to their LSST-like value is smaller as the baseline goes from three point two years to ten years and photometric error shrinks from zero point zero five magnitude down to zero point zero one magnitude.

Jocelyn: That points toward a more nuanced way of selecting cadences that accounts for both how long you observe and how much noise you are dealing with, which is crucial for real survey planning.

Subrahmanyan: And they wonder if incorporating different light curve generation methods, like comparing the Kelly recursion of Equation (three) against the J Timmer and M Koenig one thousand nine hundred ninety-five method, would give a more robust way to model the underlying noise and better assess detection efficiency.

Vera: That’s a good point because modeling the noise accurately is fundamental, and testing different ways to generate those random walk light curves could reveal systematic biases in how we are currently testing things.

Jocelyn: So, what they’re suggesting is that if we want to get better results on this kind of search, we need to stop assuming fixed brightness and start looking at the noise structure more deeply.

Subrahmanyan: And from a theoretical side, it means we can explore how "the bright, heavy hosts become the easiest to detect" because they are intrinsically brighter and their signal might be stronger in a way that standard assumptions don't capture.

Vera: It really shows that the whole process needs to be flexible; if we want better numbers on detection efficiency, we have to let the data do more of the work by using mass-dependent magnitude assignments.

Jocelyn: And they’re also pushing for a smarter cadence selection strategy based on how long you observe versus how much noise you are dealing with in that observation.

Subrahmanyan: That means when we design future surveys, we need to think about the relationship between the sampling rate and the total time span of our data collection.

Vera: It's about making sure our observational choices—the cadence—are actually optimized for what we’re trying to find in these binary systems.

Jocelyn: And this whole process is basically a check on how sensitive these recovery fractions are to both the mass of the system and how quickly you sample it.

Subrahmanyan: This work shows us exactly where our observational biases might be leading us when we plan the next generation of searches, especially when connecting these binary mergers to large-scale structure formation.

Vera: We're still looking at how we can use this information to design next-generation surveys that really capture this physics across the whole mass spectrum.

Jocelyn: Next up, we’ll be talking about those dust-origin transitions in very distant stardust galaxies and what that means for galaxy evolution at high redshift.

Conclusion: Vera: So, to wrap up "Lomb-Scargle Recovery of Periodic Binary Signals across Black Hole Mass and Cadence," we’ve looked at how recovery fractions for these periodic supermassive black hole binaries change when you shift the black hole mass or the way you observe them.

Jocelyn: Right, it’s about testing the Lomb-Scargle periodogram against those injected sinusoidal and sawtooth signals in damped random-walk light curves across different cadences like PTF-like, idealised, and LSST-like ones.

Subrahmanyan: That jump from twenty-two point six to nearly seventy percent for the LSST cadence is a significant result because it suggests that our future survey designs might be much more effective at capturing these signals than what was previously estimated.

Vera: It really shows how much cadence matters when you are searching for these periodic signals, and they even noted that in the high-mass range, this sinusoid recovery climbs from seventy-two point three percent in the log M1 bin of eight point zero to eighty-two point one percent in the bin of nine point five to ten.

Jocelyn: That climb is pretty telling because it shows that for brighter and heavier systems, the recovery gets much better on the LSST-like cadence compared to lower mass bins.

Vera: We're still looking at how we can use this information to design next-generation surveys that really capture this physics across the whole mass spectrum.

Jocelyn: Next up, we’ll be talking about those dust-origin transitions in very distant stardust galaxies and what that means for galaxy evolution at high redshift.

Karan Akbari, Sayantan Bhattacharya

St. Xavier’s College, Mumbai · Department of Astronomy and Astrophysics, Tata Institute of Fundamental Research

astro-ph.GA, astro-ph.HE, astro-ph.IM

Submitted: 2026-10-22

Updated: 2026-10-22

Comments: 26 pages, 7 figures, 5 tables

License: http://creativecommons.org/licenses/by/4.0/

The gist: The gist: This research investigates how recovery fractions for periodic supermassive black hole binaries in quasar light curves change across different black hole masses and observational cadences.

Key concepts

Lomb-Scargle Periodogram (LSP): This is a mathematical tool used to find periodic signals hidden in time-series data. It analyzes the light curve to determine the most likely frequencies present in the data. The study uses it to check if periodic signals from black hole binaries are successfully recovered from noisy quasar light curves.
Recovery Fraction: This metric measures how often a specific periodic signal (like a sinusoid) is successfully detected by the analysis method. The authors compare these fractions across different black hole masses and observational cadences to see how detection success changes under varying conditions.
Cadence Comparison: Cadence refers to the sampling rate of observations, such as how often data points are taken over time. The study compares three types: PTF-like (one epoch per night), idealized (one epoch per day), and LSST-like (a 10-year baseline with 5-day sampling). This tests if the detection method performs better with different observation schedules.
Black Hole Mass Dependence: The analysis examines how the recovery rate of periodic signals changes depending on the mass of the central black hole. The findings suggest that as black hole masses increase, specific signals become easier to detect under certain conditions, indicating a dependence on system properties.

Summary

The gist: This research investigates how recovery fractions for periodic supermassive black hole binaries in quasar light curves change across different black hole masses and observational cadences.

How it works

The study tests how well the Lomb-Scargle periodogram (LSP) recovers injected sinusoidal and sawtooth signals into damped random-walk (DRW) light curves matched to quasars, using detection thresholds set by 105 noise-only light curves per binary on Page 1. The authors inject sinusoidal and sawtooth signals into light curves of mergers in the TNG100-1 cosmological simulation, drawing magnitudes from DR7 quasars of similar mass on Page 1. They apply a threshold in each frequency bin, set by 105 noise-only light curves per binary on Page 3. The detection rule is defined as a significant signal when the injected light curve clears this threshold in at least one of the 25 bins on Page 3.

Sample and Data

The analysis uses the TNG100-1 black hole merger catalogue of IllustrisTNG, which supplies 18,374 merging black hole pairs with known primary mass and redshift over four decades of mass on Page 2. Primary masses span log M1 ∈ [6.07, 10.09] with median 7.87, and 46.4 per cent sit at log M1 ≥ 8 on Page 3. Quasar magnitudes are assigned by a conditional draw from the Y. Shen et al. (2011) catalogue of SDSS DR7 quasars, using its fiducial virial black hole mass and its Mi(z = 2) on Page 4. The authors assign Mi by a conditional draw from the Y. Shen et al. (2011) catalogue of SDSS DR7 quasars, using its fiducial virial black hole mass and its Mi(z = 2) on Page 4. The DRW amplitude depends on the absolute i-band magnitude as well as the black hole mass on Page 5.

Cadences and Cadence Comparison

Three types of cadence are used: PTF-like (one epoch per observed night of the PTF light curves), idealised (one epoch per day over the same baselines), and LSST-like (modelled on the Legacy Survey of Space and Time, LSST) on Page 5. The PTF cadences hold 6207 epochs on 2816 nights, a mean of 127 epochs over a mean baseline of 3.9 yr on Page 5. The LSST-like cadence is our realisation of the LCH26 specification, a 10-yr baseline at 5-day sampling jittered by 1 d on Page 5.

Results and Comparison with Published Recovery

The study compares recovery fractions across different mass bins and cadences against the LCH26 values on Page 9. In the high-mass range with equal mass weight, the PTF run recovers 39.8±1.6 per cent of sinusoids on significance-only, against 44.7 per cent in LCH26 on Page 9. For the LSST-like cadence, the sinusoid recovers 76.9 ± 1.1 per cent against their 22.6 per cent on Page 10. The authors find that in the high-mass range with equal mass weight, the sinusoid recovery climbs from 72.3 per cent in [8.0, 8.5) to 82.1 per cent in [9.5, 10.0) on Page 9 and Table 3 on Page 14 >

Discussion

The comparison shows that holding the magnitude constant and recovery drops toward the high-mass end, while drawing it from DR7 instead turns over on the archival cadences, so the bright, heavy hosts become the easiest to detect on Page 17. The authors conclude that at LSST-like cadence our recovery above 108 M⊙ is about three times the published sinusoid fraction and about forty times the published sawtooth fraction, and most of that difference remains unexplained on Page 13.

Limitations

The effect of NMC is measured twice: raising NMC from 104 to 105 took the high-mass sinusoid recovery on Stripe 82 from 44.5† to 37.0 per cent and lowered the PTF value by a factor of on Page 16. The authors state that the effect of NMC is measured twice: raising NMC from 104 to 105 took the high-mass sinusoid recovery on Stripe 82 from 44.5† to 37.0 per cent and lowered the PTF value by a factor of on Page 16.

Conclusions

A forecast for an archival search that applies the recovery of massive systems at every mass therefore overstates recovery at lower mass on Page 18. The PTF sawtooth lands above theirs, and power spilling onto the harmonics explains 2.6 points of the 4.2-point excess on Page 18 >

Acknowledgments

This work uses data from the IllustrisTNG project (D. Nelson et al. 2019; A. Pillepich et al. 2018a; V. Springel et al. 2018; D. Nelson et al. 2018; J. P. Naiman et al. 2018; F. Marinacci et al. 2018) and observation times from the Palomar Transient Factory (N. M. Law et al. 2009) on Page 19 >

References

Abazajian, K. N., Adelman-McCarthy, J. K., Agüeros, M. A., et al. 2009, ApJS, 182, 543 on Page 19 >

Agazie, G., Anumarlapudi, A., Archibald, A. M., et al. 2023a, ApJL, 951, L8 on Page 19 >

Agazie, G., Anumarlapudi, A., Archibald, A. M., et al. 2023b, ApJL, 952, L37 on Page 19 >

Akbari, K. 2026, ApJ, 1009, 184 on Page 19 >

Anthropic 2026, Claude (large language model), https://claude.com; Claude Opus 4 and Claude 5 model families; last accessed 2026-10-21 on Page 19 >

Astropy Collaboration, Price-Whelan, A. M., Lim, P. L., et al. 2022, ApJ, 935, 167 on Page 19 >

Astropy Collaboration, Price-Whelan, A. M., Sipocz, B. M., et al. 2018, AJ, 156, 123 on Page 19 >

Astropy Collaboration, Robitaille, T. P., Tollerud, E. J., et al. 2013, A&A, 558, A33 on Page 19 >

Begelman, M. C., Blandford, R. D.

Improvements for AI systems

Improved inference for supermassive black hole binary detection will incorporate mass-dependent magnitude assignment, specifically by drawing magnitudes from DR7 quasars of similar mass rather than using a fixed absolute magnitude Mi. This allows the system to capture how the recovery fraction travels across that distribution and how the bright, heavy hosts become the easiest to detect.
Enhanced cadence selection will move beyond simply matching observed timescales by incorporating a density-based comparison of sampling versus baseline length. The paper shows that the fall from 44.7 to 24.4 per cent happens at one baseline, while the further step to their LSST-like value is under two points as the baseline goes from 3.2 to 10 yr and the photometric error from 0.05 to 0.01 mag.
Advanced noise modeling will allow for a more accurate assessment of detection efficiency by testing different light curve generation methods, such as comparing the Kelly recursion of Equation (3) against the "J. Timmer & M. Koenig (1995) method and examining the impact of padding" on DRW light curves.
Periodogram analysis will utilize a threshold that is conditional on both signal shape and frequency bin, as described by the rule: "max f in Bk p(f) > max j max f in Bk pj (f), (6) in at least one of the 25 bins Bk."
Forecasts for future surveys will be refined by incorporating amplitude-resolved recovery curves. The study found that "the ceiling is the sharpest quantitative tension in the whole comparison, and no third source settles it, since no amplitude-resolved periodogram recovery curve at LSST-like cadence, with a threshold in each frequency bin, has been published other than their Figure 5."

Abstract

Searches for periodic supermassive black hole binaries in quasar light curves rely on recovery fractions, the share of injected signals detected, which forecasts use as detection efficiencies. Lin, Charisi, & Haiman (2026) measured them for the Lomb-Scargle periodogram on damped-random-walk light curves matched to 12,400 bright, high-mass Palomar Transient Factory (PTF) quasars with Sloan Digital Sky Survey Data Release 7 (DR7) masses. To test them over four decades of mass, we inject their sinusoidal and sawtooth signals into light curves of mergers in the TNG100-1 cosmological simulation, with magnitudes drawn from DR7 quasars of similar mass (extrapolated at low mass). We apply their threshold in each frequency bin, set by 10 5 noise-only light curves per binary, on observed PTF cadences, one Stripe 82 cadence and an LSST-like cadence (Legacy Survey of Space and Time). At fixed absolute magnitude M i=-23, sinusoid recovery falls toward high mass, and with DR7 magnitudes it climbs above about 10 8,M, so the magnitude assignment sets the sign of the trend. Above 10 8,M, where their sample lies, PTF sinusoid recovery is 39.8 against their 44.7 per cent. At LSST-like cadence sinusoid recovery is 76.9 per cent, about three times their 22.6 per cent, for reasons mostly unidentified. Denser sampling of the same PTF baselines did not lower recovery in any test. Even at LSST-like cadence we miss most sawtooth signals above 10 8,M. We tabulate recovery per 0.5-dex mass bin and cadence, with low-mass values likely too high for a deeper survey.

Listen

Transcript

Key concepts

Summary

Improvements for AI systems

Abstract

Related papers