The origin of the UV/optical lags in NGC 5548

The new multiwavelength monitoring campaign on NGC 5548 shows clearly that the variability of the ultraviolet (UV)/optical light curves lags by progressively longer times at longer wavelengths, as expected from reprocessing off an optically thick disc, but that the time-scales are longer than expected for a standard Shakura-Sunyaev accretion disc. We build a full spectral-timing reprocessing model to simulate the UV/optical light curves of NGC 5548. We show that disc reprocessing of the observed hard X-ray light curve produces optical light curves with too much fast variability as well as too short a lag time. Suppressing the fast variability requires an intervening structure preventing the hard X-rays from illuminating the disc. We propose this is the disc itself, perhaps due to atomic processes in the UV lifting the photosphere, increasing the scaleheight, making it less dense and less able to thermalize, so that it radiates low-temperature Comptonized emission as required to produce the soft X-ray excess. The outer edge of the puffed-up Comptonized disc region emits far-UV (FUV) flux, and can illuminate the outer thin blackbody disc, but while this gives reprocessed variable emission that is much closer to the observed UV and optical light curves, the light travel lags are still too short to match the data. We reverse engineer a solution to match the observations and find that the luminosity and temperature of the lagged emission are not consistent with material at the light travel lag distance responding to the irradiating flux (either FUV or X-ray). We conclude that the UV/optical lags of NGC 5548 are neither the light travel time from X-ray reprocessing nor the light travel time from FUV reprocessing, but instead could be the time-scale for the outer blackbody disc vertical structure to respond to the changing FUV illumination.