Asymmetric torus variability in active galactic nuclei driven by global brightening and dimming

TL;DR

This study tests whether mid-infrared torus variability in AGNs shows temporal asymmetry by comparing ensemble structure functions of the MIR (W1/W2) and optical (g) bands using NEOWISE and ZTF data for 0.15 < z < 0.4. The authors carefully correct photometric systematics, subtract host/AGN contributions, and define $SF_+$ and $SF_-$ to quantify asymmetry, revealing a color-dependent MIR asymmetry: blue optical-to-MIR colors exhibit positive asymmetry ($SF_+ > SF_-$) while red colors show negative asymmetry ($SF_- > SF_+$), driven by long-term global brightening/dimming trends. Simulations show this asymmetry can be reproduced by global trends rather than intrinsic stochasticity, supporting a self-regulating torus where hot-dust formation and destruction occur on year timescales, consistent with a wind-driven torus formation scenario. The results furnish new constraints on torus evolution and the coupling between accretion-disk and dusty torus physics, highlighting the role of hot-dust content in shaping long-term MIR variability.

Abstract

Temporal asymmetry in the flux variability of active galactic nuclei (AGNs) offers key insights into the physical mechanisms driving AGN variability. In this study, we investigated the variability of the torus by analyzing temporal asymmetry in the mid-infrared (MIR) continuum. We compared ensemble structure functions between the brightening and dimming phases for AGNs at $0.15<z<0.4$, using monitoring data in the optical from the Zwicky Transient Facility and in the MIR from the Near-Earth Object Wide-field Infrared Survey Explorer. We found that AGNs with bluer optical-to-MIR colors exhibit positive temporal asymmetry in the MIR, indicating that their variability amplitude is larger when brightening. Conversely, those with redder colors show negative asymmetry, exhibiting larger variability amplitude when decaying. However, there is no significant temporal asymmetry in the $g$-band variability driven by the accretion disk, suggesting that the temporal asymmetry in the MIR continuum primarily originates from intrinsic processes in the torus instead of the reflection of the ultraviolet-optical variability from the accretion disk. Analysis of the composite light curves revealed that AGNs with bluer optical-to-MIR colors tend to brighten gradually in the MIR, leading to the observed temporal asymmetry. This finding suggests that hot-dust-rich AGNs evolve with a gradual decline in hot dust emission, while hot-dust-poor AGNs are associated with a steady increase.

Figures (8)

• Figure 1: Conceptual illustration of structure function (SF) asymmetry. Panel (a) shows an example light curve with a fast rise and slower decay, while panel (b) shows the opposite case with a slow rise and fast decay. The corresponding SFs are decomposed into the total (solid line), brightening (${\rm SF_+}$; up triangle), and fading (${\rm SF_-}$; down triangle) components in panels (c) and (d).
• Figure 2: Differences between magnitude pairs for inactive galaxies in the $g$ (left), W1 (middle), and W2 (right) bands. Gray points represent individual magnitude differences. Blue open circles denote the mean values of the magnitude differences with 1$\sigma$ error bars.
• Figure 3: Ensemble SFs for type 1 (left) and type 2 (right) AGNs. Open upward and filled downward triangles denote the ensemble SFs for the brightening and dimming phases, respectively. Ensemble SFs for all phases are shown as solid lines. Blue, red, and black symbols correspond to the W1, W2, and $g$ bands, respectively. The ratio of SF$_{+}$ to SF$_{-}$ is shown in the bottom panels.
• Figure 4: Same as Figure 3, but the subsamples are divided by $g-$W1 color for type 1 (top) and type 2 (bottom) AGNs.
• Figure 5: Same as Figure 4, but the ensemble SFs are computed by applying brighter magnitude cuts to minimize Eddington bias (top), using subsamples divided by host-subtracted $g-$W1 color (middle), and applying magnitude cuts while using the host-subtracted color (bottom).