GALEX observations of quasar variability in the ultraviolet

Abstract

Aims. Using archival observations recorded over a 5+ year timeframe with the NASA GALaxy Evolution eXplorer (GALEX) satellite, we present a study of the ultraviolet (UV) variability of 4360 quasars of redshifts up to z = 2.5 that have optical counterparts in the Sloan Digital Sky Survey DR5 spectroscopic catalog of Schneider et al. (2007, AJ, 134, 102). The observed changes in both the far UV (FUV: 1350−1785 Å) and near UV (NUV: 1770−2830 Å) AB magnitudes as a function of time may help differentiate between models of the emission mechanisms thought to operate in these active galaxies. Methods. A list of NUV and FUV variable quasars was derived from the UV light-curves of sources with 5 or more observational visits by GALEX that spanned a time-frame >3 months. By measuring the error in the derived mean UV magnitude from the series of GALEX observations for each source, quasars whose UV variability was greater than the 3-σ variance from the mean observed value were deemed to be (intrinsically) UV variable. This conservative selection criterion (which was applied to both FUV and NUV observations) resulted in identifying 550 NUV and 371 FUV quasars as being statistically significant UV variable objects. Results. Following the work of Vanden Berk et al. (2004, ApJ, 601, 692), we have performed a structure function (SF) analysis of these data to search for possible correlations between UV variability and parameters such as rest frame time-lag, redshift, luminosity and radio loudness. Firstly, we observe that the amplitudes of variability as a function of time-lag for both the NUV and FUV data are far larger than those observed at visible wavelengths. Secondly, the levels of FUV variability are greater than those observed in the NUV for a given value of time-lag. Also, the amplitudes of both the NUV and FUV variability of quasars increase as a function of rest frame time-lag, irrespective of the value of quasar redshift, for time-lags <200 days. For time-lags >300 days there is a pronounced rollover in the NUV SF for all redshift values, which is also observed (with a lower signiÞcance) in the FUV variability data. Although we find no strong relationship between UV variability and redshift, our data do show that higher redshift quasars appear to be more variable than their low redshift counterparts. Our data also show that, for all values of time-lag, the more luminous quasars tend to be slightly less UV variable, with perhaps the exception of FUV variable quasars for short time-lags. Although our data sample is small, we find that radio-loud quasars are marginally more variable than radio quiet ones by a factor ∼2 in the NUV and by a factor 1−3 in the FUV. Therefore, our present observations support the notion in which the radio properties of quasars have a limited influence on the observed UV variability of these objects. In summation, our present analysis favors a quasar model in which UV variability is mainly due to stochastic changes in the underlying continuum level, rather than models that favor gravitat...

The Extreme Ultraviolet Variability of Quasars

“…Yet, by point 2 and Figure 5, the propensity and magnitude of variability does not increase for time intervals >2.5 10 7 s. This last point does not support a monotonic increase in s rms 2 as the time interval increases unless the rate of increase diminishes drastically for time intervals >2.5 10 7 s. The overall behavior is well described by a broad threshold for measurable variability that occurs typically in the range of time intervals between observations of -1.5 2.0 10 s 7 above which the variability is roughly constant for time intervals less than 1 year. This behavior is similar to what has been seen in UV and optical quasar structure functions that are based on broad band photometry (Welsh et al 2011;Macleod 2012). The photometric structure functions show a steady increase in variability as the time between observations increases to a few hundred days.…”

“…Figure 5 is a histogram highlighting the almost switch-on-like behavior of the variability once the statistical uncertainty of each measurement is accounted for in the definition of variability. We note that s rms 2 averaged over a bin is similar to the structure function used in the UV photometry studies of quasars (Welsh et al 2011;Macleod 2012). However, there is not much evidence in the excess variance analysis of the EUV continuum variability of a gradual increase of variability for small time intervals to larger time intervals that was seen for the UV photometric structure functions.…”

“…This might be a result of the small number statistics in this study, in the range -1.5 2.0 10 7 s, or a fundamental difference associated with the fact that the EUV region is at the inner edge of the accretion disk. We note that a structure function cannot be used on the data presented here since we do not eliminate low-variability sources as in photometric structure functions (Welsh et al 2011). Consequently, the variability can be less than the estimated error (see the negative values of excess variance in Figures 4 and 5) rendering a putative structure function at time intervals, <2.0 10 7 s, to values that are imaginary numbers (i.e., basically the square root of the negative excess variance).…”

“…SDSS photometry and Palomar Sky Survey photometry were compared in order to study the time dependent variability in quasars in the UV and optical bands (Macleod 2012). Broad band photometry was also implemented in Welsh et al (2011) using GALEX data in order to explore far-UV and near-UV variability in quasars. Broad band photometry suffers from broad line contamination so one does not study the continuum variability explicitly.…”

A multi-epoch spectroscopic study of the BAL quasar APM 08279+5255

“…It is interesting to compare these slope values with typical values found for optical and UV variability, where the slopes are in the range 0.3 to 0.45 (Vanden Berk et al 2004;Wilhite et al 2008;Bauer et al 2009;MacLeod et al 2012;Welsh et al 2011). If the absorption variations that we are measuring are in fact due to a change in C 3+ column density, caused by variations in the A91, page 8 of 10 Table 2 are shown as dotted lines.…”