🔭 astrophysics

Active galactic nuclei do not exhibit strictly sinusoidal brightness variations

By extending the analysis of 181 candidate supermassive black hole binaries identified in Gaia DR3 with longer-baseline data from ZTF and CRTS, this study demonstrates that the reported sinusoidal periodicities are false positives caused by red noise, indicating that strictly periodic AGN variability is exceedingly rare and that short-period binary AGN likely do not exhibit simple sinusoidal light curves.

Original authors: Kareem El-Badry, David W. Hogg, Hans-Walter Rix

Published 2026-01-28

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Kareem El-Badry, David W. Hogg, Hans-Walter Rix

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The Big Idea: Hunting for Cosmic Clocks

Imagine the universe is filled with supermassive black holes at the centers of galaxies. Sometimes, two of these black holes get stuck in a dance, orbiting each other like a pair of ice skaters holding hands. Scientists have been looking for these "binary black holes" because they are the missing link in understanding how galaxies merge and how gravitational waves are created.

One way to spot these dancing pairs is to look at the light they emit. If two black holes are orbiting each other, they might make the galaxy's brightness pulse up and down like a perfect, steady heartbeat. This is called sinusoidal variability—a smooth, predictable wave pattern.

The "False Alarm" in the Data

Recently, a team of researchers (Huijse et al., 2025) looked at data from the Gaia space telescope. Gaia took snapshots of millions of active galaxies over about 1,000 days. They found 181 galaxies that seemed to have this perfect, rhythmic heartbeat. They thought, "Great! We found 181 dancing black hole pairs!"

However, the authors of this new paper (El-Badry, Hogg, and Rix) decided to double-check the work. They asked a simple question: "If the heartbeat is real, will it keep beating the same way in the future?"

The Detective Work: Adding More Data

To answer this, the authors didn't just look at the Gaia snapshots. They went to the "neighborhood watch" and grabbed data from two ground-based telescopes: ZTF (Zwicky Transient Facility) and CRTS.

Think of it this way:

Gaia took a photo of a runner every month for a year. The runner looked like they were jogging in a perfect, straight line.
ZTF took photos of the same runner every few days for several more years.

When the authors combined the Gaia data with the ZTF data, the story changed completely.

The Results: It Was Just "Red Noise"

In almost every single case (116 out of 116 that they could check), the "perfect heartbeat" stopped the moment the Gaia data ended.

The Prediction: If the black holes were truly dancing in a perfect circle, the ZTF data should have shown the light continuing to rise and fall in that exact same rhythm.
The Reality: The ZTF data showed the light behaving erratically—jumping up and down randomly, like static on an old TV or the chaotic movement of a crowd in a mosh pit.

The authors call this "red noise." It's a fancy way of saying that single black holes naturally flicker and vary in brightness in a random, unpredictable way. Sometimes, if you only look at a short window of time (like the 1,000 days Gaia observed), this random flickering looks like a perfect rhythm by pure chance. It's like flipping a coin 10 times and getting "Heads" every time; it looks like a pattern, but it's just luck.

The Conclusion: A Rare Find

The paper concludes that:

The 181 candidates were likely false alarms. The "perfect rhythms" were just random noise that looked like a pattern because the observation window was too short.
Truly perfect rhythms are incredibly rare. The authors estimate that out of one million active galaxies, maybe only a handful show this strict, sinusoidal rhythm.
Binary black holes might still exist, but they are messy. The fact that we didn't find perfect rhythms doesn't mean binary black holes don't exist. It just means that when two black holes dance, their light doesn't beat like a metronome. It's likely mixed with so much chaotic "red noise" that the rhythm is hidden.

The Takeaway

The search for binary black holes needs to change its strategy. Instead of looking for perfect, clockwork rhythms (which are almost non-existent), scientists need to develop new methods to find these pairs even when their light curves are messy, chaotic, and full of noise. The "perfect heartbeat" was a mirage; the real signal is much more complex.

Technical Summary: Active Galactic Nuclei Do Not Exhibit Strictly Sinusoidal Brightness Variations

Problem Statement
Periodic variability in Active Galactic Nuclei (AGN) light curves has long been proposed as a primary observational signature of close supermassive black hole (SMBH) binaries. Recently, Huijse et al. (2025) identified 181 candidate SMBH binaries within the Gaia Data Release 3 (DR3) catalog. These candidates were selected based on light curves spanning approximately 1,000 days that exhibited apparently stable, strictly sinusoidal variability. The selection criteria relied on a Bayesian model comparison between a purely sinusoidal model and a Damped Random Walk (DRW) model, retaining only sources where the logarithmic Bayes factor favored the periodic model significantly ( $\log(B_{PR}) > 3.0423$ ).

However, a fundamental challenge in identifying binary SMBHs is distinguishing genuine periodicity from stochastic "red noise" variability inherent to single quasars. Stochastic variations can mimic periodic signals, particularly when light curves sample only a few cycles. While Huijse et al. (2025) employed rigorous statistical thresholds to minimize false positives, the validity of these candidates remained untested beyond the specific 1,000-day window of the Gaia DR3 data.

Methodology
To validate the periodicity claims, this study supplemented the Gaia DR3 photometry with longer-baseline light curves from ground-based surveys, primarily the Zwicky Transient Facility (ZTF) and the Catalina Real-Time Transient Survey (CRTS).

Data Acquisition: The authors queried Gaia DR3 epoch photometry for the 181 candidates, applying the same pre-processing filters as the original study. They subsequently queried public ZTF $r$ - and $g$ -band data for all 181 candidates, successfully obtaining light curves for 116 objects (64% of the sample). CRTS data were retrieved for three bright sources ( $G < 19$ ).
Data Processing: ZTF data, which offer higher cadence but lower per-epoch precision than Gaia, were binned into 20-day windows to facilitate comparison. For period analysis, $g$ - and $r$ -band ZTF data were normalized and concatenated to create a single light curve per source.
Analysis: The authors tested the predictive power of the sinusoidal models derived from Gaia data. They compared the Gaia-inferred periods and phase against the ZTF/CRTS data. Specifically, they examined whether the sinusoidal fits to the Gaia data could extrapolate to predict the variability observed in the ZTF light curves, which extend the observational baseline to $\sim$ 4,000 days. They also computed Lomb–Scargle periodograms on the ZTF data to determine if the dominant periods remained consistent with the Gaia measurements.

Key Results
The analysis yielded the following critical findings:

Failure of Predictive Models: In all 116 cases with available ZTF data, the sinusoidal models fitted to the Gaia DR3 data failed to predict the subsequent variability observed by ZTF. The ZTF light curves appeared stochastic rather than periodic.
Inconsistency of Periods: There was no correlation between the periods measured in Gaia and the dominant periods measured in ZTF for the same objects five years later. Only three sources showed periods consistent within $1\sigma$ , and six within $2\sigma$ . Statistical shuffling tests indicated that this level of agreement is consistent with random chance, implying the periods are independent and the Gaia detections are likely noise artifacts.
Red Noise Contamination: The results suggest that red noise contamination is the primary source of the false detections. The DRW model used by Huijse et al. (2025) to simulate stochastic variability appears to under-predict the incidence of apparent periodicity on short timescales (100–1,000 days).
Rarity of Strictly Sinusoidal AGN: Based on the failure rate of the candidates, the authors estimate that truly periodic and sinusoidal AGN variability is exceedingly rare. The upper limit for the fraction of AGN exhibiting stable sinusoidal variability on 100–1,000 day timescales is estimated to be at most a few in $10^6$ .

Significance and Claims
The paper concludes that the 181 candidates identified by Huijse et al. (2025) are overwhelmingly false positives. Consequently, the authors assert that strictly periodic, sinusoidal AGN variability is an extremely rare phenomenon, likely occurring in fewer than one in a million AGN on these timescales.

Crucially, the authors do not claim that binary SMBHs do not exist within the Gaia sample. Using merger rate estimates and inspiral timescales, they calculate that the Gaia sample likely contains between 23 and 45 genuine binary SMBHs with orbital periods $P_{orb} \lesssim 600$ days. The absence of strictly sinusoidal light curves in the sample implies that short-period binary SMBHs do not typically produce light curves dominated by simple sinusoidal periodicity. Instead, the stochastic variability inherent to AGN accretion processes likely masks or overwhelms the periodic signal.

The significance of this work lies in its demonstration that searches for binary SMBHs must move beyond the paradigm of identifying simple sinusoidal variability. The authors argue that future detection efforts must embrace the complex, stochastic nature of AGN light curves, potentially utilizing more flexible stochastic models (e.g., those incorporating power-law power spectra or quasi-periodic oscillations) rather than relying on rigid periodic fits that are highly susceptible to red noise contamination.