Search for: All records

ABSTRACT Using a month-long X-ray light curve from RXTE/PCA and 1.5 month-long UV continuum light curves from IUE spectra in 1220–1970 Å, we performed a detailed time-lag study of the Seyfert 1 galaxy NGC 7469. Our cross-correlation analysis confirms previous results showing that the X-rays are delayed relative to the UV continuum at 1315 Å by 3.49 ± 0.22 d, which is possibly caused by either propagating fluctuation or variable Comptonization. However, if variations slower than 5 d are removed from the X-ray light curve, the UV variations then lag behind the X-ray variations by 0.37 ± 0.14 d, consistent with reprocessing of the X-rays by a surrounding accretion disc. A very similar reverberation delay is observed between Swift/XRT X-ray and Swift/UVOT UVW2, U light curves. Continuum light curves extracted from the Swift/GRISM spectra show delays with respect to X-rays consistent with reverberation. Separating the UV continuum variations faster and slower than 5 d, the slow variations at 1825 Å lag those at 1315 Å by 0.29 ± 0.06 d, while the fast variations are coincident (0.04 ± 0.12 d). The UV/optical continuum reverberation lag from IUE, Swift, and other optical telescopes at different wavelengths are consistent with the relationship: τ ∝ λ4/3, predicted for the standard accretion disc theory while the best-fitting X-ray delay from RXTE and Swift/XRT shows a negative X-ray offset of ∼0.38 d from the standard disc delay prediction. 

ABSTRACT We broadly explore the effects of systematic errors on reverberation mapping lag uncertainty estimates from javelin and the interpolated cross-correlation function (ICCF) method. We focus on simulated light curves from random realizations of the light curves of five intensively monitored AGNs. Both methods generally work well even in the presence of systematic errors, although javelin generally provides better error estimates. Poorly estimated light-curve uncertainties have less effect on the ICCF method because, unlike javelin , it does not explicitly assume Gaussian statistics. Neither method is sensitive to changes in the stochastic process driving the continuum or the transfer function relating the line light curve to the continuum. The only systematic error we considered that causes significant problems is if the line light curve is not a smoothed and shifted version of the continuum light curve but instead contains some additional sources of variability.