The accretion disk size problem in AGN disk reverberation mapping is an obscuration effect: a uniform AGN sample study with Swift

In the past decade, Swift has performed several active galactic nucleus (AGN) high-cadence reverberation mapping campaigns, and it has generally found that the UV/optical interband lags are ∼3 times longer than predicted for a standard thin disk, thus coined “the accretion disk size problem.” Here we present a systematic sample of Swift-monitored AGN. In this analysis, we confirm the accretion disk size problem but find that the lag excess occurs only in the subset of obscured AGN, which show a significantly elevated mean normalization of 5.21 ± 0.47 (p = 0.008), whereas the unobscured AGN exhibit a mean excess consistent with standard disk predictions (1.00 ± 0.31). Correlation and regression analyses similarly reveal X-ray column density as the strongest predictor of lag excess, explaining over 80% of its variance. We interpret these results as line-of-sight obscuration being linked to the too-long lags via additional reprocessed emission from the absorbing material itself. The consistency of lags in the unobscured subgroup with standard disk predictions suggests that the accretion disk size problem is not the result of shortcomings of standard accretion disk theory or contamination by the broad-line region. X-ray to UV lag amplitudes and correlations show more complex and variable behavior in obscured AGN, suggesting that obscuration may disrupt or complicate the connection between high- and low-energy emission potentially through reprocessing, scattering, and/or ionization changes.