Indication for Dual Periodic Signatures in PKS 0805-07 from Multi-technique Time Series Analysis

TL;DR

The paper investigates quasi-periodic variability in the gamma-ray emission of the high-redshift blazar PKS 0805-07 using a multi-method time-series approach that accounts for red noise. It identifies two statistically significant QPOs at approximately $P_1\approx254.8$ d and $P_2\approx112.1$ d, confirmed across Lomb-Scargle, WWZ, REDFIT, DCDFT, PDM, and string-length analyses, and validated by phase folding and Monte Carlo significance tests. A double-sine fit captures the dual modulation, and an amplitude-modulated jet precession model (consistent with Doppler boosting in a relativistic jet) provides a physically plausible geometric interpretation, with best-fit parameters yielding a beat pattern that matches the observed light curve. The results suggest that the observed dual QPOs arise from jet dynamics rather than a straightforward binary SMBH orbital period, emphasizing the importance of multi-method, red-noise-aware analyses for understanding complex blazar variability and jet physics.

Abstract

We report the identification of two statistically significant quasi-periodic oscillations in the weekly binned $γ$-ray light curve of the flat-spectrum radio quasar PKS 0805$-$07, observed by Fermi-LAT over the period MJD 59047.5-59740.5. By applying a suite of complementary time-series analysis techniques, we identify periodic signatures at approximately 255 and 112 days. These techniques include the Lomb-Scargle periodogram (LSP), Weighted Wavelet Z-transform (WWZ), REDFIT, Date-Compensated Discrete Fourier Transform (DCDFT), Phase Dispersion Minimization (PDM), and the String-length method. The reliability of these signals is supported by high local significance (> 99) in all methods and reinforced through phase-folding. Model selection using the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) strongly supports a two-component periodic model. The detection of dual QPOs is rare among blazars and suggests complex variability mechanisms. Although a binary supermassive black hole (SMBH) scenario could be considered given the sources high redshift (z = 1.837), the short periodicities are difficult to reconcile with orbital motion unless invoking extreme parameters. Double-sine model fitting reveals that the oscillatory components have comparable amplitudes but are out of phase, suggesting a potential beating phenomenon due to interference. This amplitude-modulated variability is consistent with a geometric origin, most plausibly jet precession driven by Lense-Thirring torques, superimposed with a secondary process such as polar jet oscillation. Doppler factor modulation arising from these effects can account for the observed flux variations without requiring an unrealistically compact binary.

Figures (14)

• Figure 1: The upper panel depicts the weekly binned $\gamma$-ray light curve of PKS 0805-07, integrated over the energy range 0.1-100 GeV [Flux ($E > 100$ MeV)] in units of $10^{-6}\,\text{photons\,cm}^{-2}\,\text{s}^{-1}$ from MJD 54684 to 60264. The lower panel shows the corresponding spectral index values for the same period. Red horizontal lines in both panels represent the flux and index reported in the 4FGL catalog. Shaded regions denote the HOP intervals, with HOP 8 designated as the "active state". Different HOPs are color-coded for clarity. Reproduced from 2024ApJ...977..111A. The time interval MJD 59047.5-59740.5, used for the present QPO study, spans the rise and peak of HOP 8, including the pre-flare phase.
• Figure 2: LSP of the $\gamma$-ray light curve of PKS 0805-07. The two most prominent peaks are observed at frequencies of $0.0039$ and $0.0089$ day$^{-1}$, corresponding to periods of approximately $255$ and $112$ days, respectively. The horizontal lines represent the false alarm probability (FAP) thresholds at 10% (blue dotted), 5% (orange dashed), and 1% (green dash-dotted), indicating the statistical significance of the detected periodicities.
• Figure 3: Top: The $\gamma$-ray light curve of PKS 0805-07 (MJD 59047.5-59740.5), overplotted with the best-fit sinusoidal models corresponding to the two significant periods identified via LSP analysis: $P_1 = 254.8 \pm 26.5$ days and $P_2 = 112.1 \pm 5.8$ days. Bottom: LSP of the same light curve, showing two dominant peaks at frequencies 0.003924 day$^{-1}$ and 0.008920 day$^{-1}$ (periods $\sim$255 and $\sim$112 days), both exceeding the 99.9% confidence level derived from $10^5$ Monte Carlo simulations using the method of emmanoulopoulos2013generating.
• Figure 4: Left: WWZ map of the $\gamma$-ray light curve showing the evolution of power as a function of time (MJD) and frequency. Right: The average WWZ power spectrum with confidence levels derived from $10^5$ Monte Carlo simulations. The dashed green and red lines represent the 99.5% and 99.99% confidence levels, respectively. The dotted and dashed black lines mark the two most significant peaks at frequencies of 0.003743 and 0.008743 day$^{-1}$, corresponding to periods of approximately 267 and 114 days.
• Figure 5: The red noise-corrected REDFIT power spectrum of the $\gamma$-ray light curve of PKS 0805-07. The black curve shows the AR(1) power spectrum, while the blue and cyan curves represent the theoretical and average AR(1) models, respectively. The brown, green, and red lines denote the 90%, 95%, and 99% Monte Carlo significance levels. Two dominant peaks are observed, both exceeding the 99% confidence level.