Comprehensive variability analysis of blazars using Fermi light curves across multiple timescales

Abstract

In this study, we conducted a systematic analysis of long-term Fermi-LAT γ-ray data for a sample of blazars, including flat spectrum radio quasars (FSRQs), BL Lacertae objects (BL Lacs), and blazar candidates of unknown type (BCU), to investigate their γ-ray variability. We focused on light curves binned in three-day, seven-day, and 30-day intervals to assess the impact of binning on variability, using data with TS > 4 as a detection threshold. We calculated fractional variability (Fvar) for each category and found that FSRQs exhibit higher mean variability compared to BL Lacs and BCUs, with BCUs displaying intermediate variability closer to BL Lacs. The Kolmogorov-Smirnov test on the variability distributions of FSRQs, BL Lacs, and BCUs indicates that FSRQs differ from both BL Lacs and BCUs, whereas BCUs are more similar to BL Lacs. The observed higher variability in FSRQs is likely linked to more powerful jets and accretion processes. The correlation between γ-ray flux and spectral index suggests a moderate positive correlation for BL Lacs and BCUs, indicating a “softer when brighter” behavior. FSRQs, however, displayed a mild anticorrelation, suggesting a tendency for these sources to become harder as their flux increases. Additionally, the analysis of flux distributions revealed log-normal behavior in many sources, consistent with multiplicative variability processes in blazar jets. Some sources exhibit bimodal distributions, implying transitions between distinct emission states. Moreover, binning effects the observed variability, with longer bins smoothing short-term fluctuations. The power spectral density (PSD) analysis suggested that FSRQs exhibit steeper slopes, reflecting structured variability, while BL Lacs display shallower slopes, dominated by stochastic processes. The absence of PSD breaks suggests no dominant characteristic timescale within the Fermi window. Spectral index distributions further highlight complexity, often requiring multicomponent models.