The X-ray/UV Connection in NGC 5548: A Rapidly Varying Corona

TL;DR

This work tests whether X-ray reverberation from a dynamically evolving corona can reproduce the UV/optical variability of NGC 5548 observed during the 2014 STORM campaign. It implements a lamp-post reverberation model with time-varying coronal height $h$, energy transfer $L_{\rm transf}/L_{\rm disc}$, and intrinsic photon index $Γ_{\rm int}$, fitting the full UV light curves alongside the X-ray light curve. The results show UV/optical variations can be matched to within 2–5% on average when corona parameters vary on daily timescales, with the inferred parameter evolution indicating rapid changes in coronal geometry and energetics. This supports X-ray reverberation as a major driver of UV variability in NGC 5548, while acknowledging potential contributions from multiple coronal regions and the necessity of extending the analysis to a broader AGN sample to validate the framework.

Abstract

Recent intensive monitoring campaigns of active galactic nuclei (AGN) have provided simultaneous X-ray, UV, and optical data of unprecedented quality. The observations reveal a strong correlation between the UV and optical variability, but a weaker correlation between the X-ray and UV bands, challenging the standard X-ray reprocessing scenario. We revisit the X-ray/UV connection in NGC 5548 by fitting archival 2014 HST and Swift/XRT light curves assuming X-ray reverberation from a dynamically evolving X-ray corona. Our results show that, as long as the corona height, photon index and power vary over time, X-ray reverberation can explain the observed UV and optical variability within 2% and 5%, respectively (on average). The evolution of the best-fit parameters suggests that fast changes in coronal geometry and energetics on a time scale of days are required to explain the observed variability.

Figures (4)

• Figure 1: The Swift/XRT spectra for an observation with low count rate (black points) and an observation with high count rate (red points). The black and red lines show the best-fitting model of Eq.(\ref{['eq:xspec']}). The bottom panel shows the best-fit residuals.
• Figure 2: The grey open squares in panels (a), (b), (c), (d), and (e) show the observed 2--10 keV, H1, H3, H4 and B-band light curves, respectively. The black open squares and the filled coloured circles show the observations we considered in the fitting procedure and the best-fit model predictions, respectively. Panels (f) and (g) show the (data-model)/data and (data-model)/error ratios for the X-ray, the H1, the H3, and H4, and the B-band light curves (black, blue, green, orange, and red circles, respectively).
• Figure 3: The best-fit parameters: height (top panel), $L_{\rm transf}/L_{\rm disc}$ (middle panel), and the photon index (bottom panel). The grey shaded area shows the 1$\sigma$ confidence interval, which we computed by repeating the fitting procedure 100 times, each time resampling the X-ray light curve from Gaussian distributions centred on the measured values with standard deviations equal to their error bar. For computational efficiency, these fits were performed using a coarser parameter grid (i.e., 10 values per parameter instead of 20). The red triangle points show the results of K24.
• Figure 4: Correlations between the best-fit parameters i.e. the height, the photon index, and the energy transferred to the corona) and the X-ray (2--10keV) and UV observed fluxes. The red lines show the best-fit straight line or power law model, and the black squares show the binned data. The numbers in each plot show the significance that each slope is different from zero.