Free Access
Issue
A&A
Volume 576, April 2015
Article Number A3
Number of page(s) 7
Section Extragalactic astronomy
DOI https://doi.org/10.1051/0004-6361/201425248
Published online 13 March 2015

© ESO, 2015

1. Introduction

Active galactic nuclei (AGNs) are characterized by their broad band continua, strong emission lines, and fast variability. However, a handful of abnormal AGNs have recently been discovered in the SDSS data, i.e., weak line quasars and radio-quiet BL Lac objects (Diamond-Stanic et al. 2009; Plotkin et al. 2010). The UV/optical emission lines are absent in their UV/optical spectroscopies, though the shape and luminosity of their continua are comparable to the normal AGNs. The fraction of these special AGNs is small (~1/1000) in the SDSS DR 7 sample; however, it could be an important stage in the evolving sequence of AGNs (Hryniewicz et al. 2010; Liu & Zhang 2011). In the early stage of an active cycle of AGNs, the radiative feedback can expel the gas from broad line regions and result in weak or even in the disappearance of broad emission lines (Liu & Zhang 2011). Other models, such as of a cold accretion disk, an extremely high accretion rate, shielding gas, or abnormal BL Lac objects, have also been proposed (Shemmer et al. 2010; Plotkin et al. 2010; Laor & Davis 2011; Wu et al. 2012). The origin of such weak-line AGNs is still not clear. Both the thermal (accretion disk) or non-thermal (jet) component may explain the weak line feature.

The optical continua of classical BL Lac objects are dominated by the synchrotron emission from the relativistic jets, therefore high polarization and fast variability are important characteristics of classical BL Lac objects. Heidt & Nilsson (2011) found the polarization degrees of the radio-quiet BL Lac candidates are low. Plotkin et al. (2010) investigated the long-term variability of radio-quiet BL Lac objects using the data from SDSS stripe 82 and find that the variation amplitude of radio-quiet BL Lac objects is smaller than that of radio-loud BL Lac objects. However, owing to the small size of their sample, this is not very conclusive.

The variation in short time scale (intranight) is another characteristic of classical BL Lac objects. During a very short period, e.g. several hours, the flux of classical BL Lac object can change by several tenths of a magnitude (Wagner & Witzel 1995; Heidt & Wagner 1996; Bai et al. 1998; de Diego et al. 1998). However, it is still unclear whether the intranight variation is frequent in the radio-quiet BL Lac objets, which will be helpful in distinguishing the origin of their continua. Gopal-Krishna et al. (2013, hereafter GJC2013) and Chand et al. (2014, hereafter CKG2014) claim that they have detected a considerable fraction of intranight variation in a sample of weak-line AGNs (duty cycle ~5%), and this fraction can be higher if the signal-to-noise ratio of the light curve is further increased. However, as we discuss in this paper, some galactic sources may contaminate their sample.

In this paper, we report the result of our monitoring campaign of radio-quiet BL Lac objects. Observations and data reduction are described in Sect. 2, and then the significance of the intranight variation is shown in Sect. 3. Section 4 discusses the implication of our results and presents the conclusions.

Table 1

Information on observations.

2. Observations and data reduction

The amplitude of intranight variation is normally several tenths of a magnitude, so the desired error of our observed magnitude is 0.05 mag with the exposure time not longer than 10 min. The corresponding magnitude threshold is R ~ 18.5 for the 2 m class telescopes we used. Seven radio-quiet BL Lac objects were selected from Plotkin et al. (2010), and SDSS J094533.99+100950.1 was selected from Hryniewicz et al. (2010). The above selection criteria are similar to those in GJC2013 and CKG2014. Actually, five sources in our sample are shared with GJC2013 and CKG2014 who based their selection primarily on classification by Plotkin et al. (2010) as a “high-confidence BL Lac candidate”. We additionally included in our sample some low-confidence BL Lac candidates. The sources are classified as low-confidence only because the continuum near the emission line is hard to define, and the equivalent width of emission lines will be larger or smaller than 5 Å depending on the continuum assumptions, which is mainly due to the noisy spectra around some emission lines. We therefore think there should be no systematic difference between high- and low-confidence sources and will investigate the variation property of subsamples in future works.

The observations were carried out by BFOSC (BAO Faint Object Spectrograph and Camera) on the Xinglong (China) 2.16 m telescope and YFOSC (Yunnan Faint Object Spectrograph and Camera) on the Lijiang (China) 2.4 m telescope. All observations were performed in Johnson R band, except for SDSS J094533.99+100950.1 in Johnson V band. The exposure time was 300 s or 600 s depending on the weather conditions. The detailed information about the sample and observations is shown in Table 1. In total, there are 11 sessions of light curves of these eight sources.

Table 2

Information on targets and their companion stars.

The photometric data were reduced with the standard routines in the Image Reduction and Analysis Facility (IRAF) software. The bias frames were extracted from no fewer than ten frames, and the flat frames did not have fewer than five frames in one band for one night of observation. The dark of the CCD is negligible (compared with the readout noise and the flat fluctuation) and therefore not considered. The flat frames for the same band were combined by average, and then the normalized flat frame was generated; the normalized bias frame was generated by median combination. Then the source frames were corrected by the normalized bias frame and flat frame.

thumbnail Fig. 1

Differential light curves of AGN–Star 1, AGN–Star 2, and Star 1–Star 2.

Open with DEXTER

With the corrected source images, we used the package APPHOT to perform aperture photometry. The values of enclosed, moffat, and direct for the comparison stars and the target source were used to estimate the mean full width at half maximum (FWHM). The apertures of the photometry for individual frame were carried with 2.5 ~ 3 times of FWHM. If the value of FWHM significantly changed during one night, we took different values of FWHM even for the same source.

3. Results

To detect the underlying variation of the target, we first calculated the differential light curves (DLCs) between the target and comparison stars. Two nearby comparison stars (noted as Star 1 and Star 2 hereafter) with similar magnitudes to the target were selected and the DLCs of AGNStar 1, AGNStar 2, and Star 1Star 2 are shown in Fig. 1. The position and gr color of targets and comparison stars are shown in Table 2. Since the color difference between target and star pairs is smaller than 1.5, the variation in air mass during the observation has little effect on DLCs (Carini et al. 1992; Stalin et al. 2004). We also tried different companion stars, and the final significance of intranight variation is quite robust. Some exposures with bad weather were excluded from the DLCs, which led to some gaps in the DLCs.

To quantify the significance of the variation of light curves, we performed a scaled F-test, which is more powerful and reliable than the traditional C-test (de Diego 2010). The scaled F value (Howell et al. 1988) is defined as (1)where , and x can stand for AGN–Star 1 or Star 1–Star 2.

The definition of Γ2 is (2)which is the scaled factor to account for the different accuracies between the photometries of the target and comparison stars (Howell et al. 1988). The variables NAGN, NStar 1, and NStar 2 are the total counts (sky-subtracted) of target, Star 1, and Star 2, respectively. The variable P is defined as , where np is the number of pixels in the applied measuring aperture, the variable NS is the sky photons per pixel, and Nr is the readout noise (e/pixel). The value of Γ2 can be calculated frame-by-frame. However, the variation in Γ2 of our observations during one night is no more than 10% owing to the small variation of our targets. Therefore, we have taken the median value of Γ2 for the exposures in one night. Our final result is not sensitive to this choice.

The significance of the variation is determined by the F distribution with NAGN − Star 1 − 1 and NStar 1 − Star 2 − 1 degrees of freedom, where NAGN − Star 1 and NStar 1 − Star 2 are the number of observations in the AGN-Star 1 and Star 1-Star 2 DLCs, respectively.

The results of the significance are listed in Table 3. Since we exchanged the position of Star 1 and Star 2 in Eqs. (1) and (2), there are two values of significance for (AGN − Star 1)/ (Star 1 − Star 2) and (AGN − Star 2)/(Star 2 − Star 1).

As indicated by the results of F-test, we only detect a significant variation (~3σ level) in SDSS J090107.64+384658.8. However, due to the large proper motion of this source (62 ± 11 mas/yr from Monet et al. 2003), its extragalactic nature is doubtable. Therefore, we would like to exclude it from the final sample of radio-quiet BL Lac objects. As a result, there is no significant variation detected in our observations of radio-quiet BL Lac objects.

Table 3

Results of the significance of variations.

4. Discussions and conclusions

Radio-loud AGNs and blazars can exhibit microvariation with a large amplitude up to ~100%. However, some microvariation events are also observed in radio-quiet AGNs with high significance (Stalin et al. 2004; Gupta & Joshi 2005). The mechanism of the microvariation in radio-loud AGNs is believed to be the fluctuation caused by the shocks in jets. However, the instability or flares in the accretion disk can also induce microvariation even for the radio-quiet AGNs (Mangalam & Wiita 1993). A weak blazar component in radio-quiet AGNs is an alternative to microvariation (Czerny et al. 2008). Though the occurrence of microvariations is not a smoking gun of jets, the fraction and amplitude of the microvariations in radio-quiet and radio-loud ones are quite different.

Gupta & Joshi (2005) compiled the microvariations of different classes of AGNs and found the detection fractions of microvariation in radio-quiet and radio-loud (non-blazars) AGNs are ~10% and ~3540%, respectively. For blazars, the fractions are ~6065% and ~8085% for the observations that are less than and more than 6 h, respectively. In addition, they also claim that the amplitude of the microvariation of radio-loud ones is larger than that of radio-quiet ones.

Carini et al. (2007) established a sample of 117 radio-quiet AGNs that have been investigated for microvariations and found a detection rate of microvariations for the entire sample of 21.4%. If the criteria for “radio-quiet” are strengthened to R< 1 (R is the ratio of the radio [5 GHz] flux to optical [4400 Å] flux), the detection rate of microvariations is only 15.9%.

Goyal et al. (2013) analyzed 262 sessions of light curves of 77 AGNs from their uniform AGN monitoring data and found the duty cycles of intranight variation of radio-quiet quasars, radio-intermediate quasars, lobe-dominated quasars, low optical-polarization core-dominated quasars, high optical-polarization core-dominated quasars, and TeV blazars are 10%, 18%, 5%, 17%, 43%, and 45%, respectively.

No significant microvariation is detected in our final sample, in ten sessions of light curves. The 1σ upper limit of the fraction of microvariation is 15% using the method of Cameron (2011)1. In deriving this upper limit, we treated the sources equally. The weights of sources should not be the same owing to different exposure times, signal-to-noise ratios, and observation numbers; however, this potential minor correction will not change our final conclusion. This low fraction in our sample of radio-quiet BL Lac objects is consistent with that of the radio-quiet AGNs but much lower than for the radio-loud ones and blazars. This indicates that the continuum of radio-quiet BL Lac objects is not dominated by the jet component. Actually, the SED of radio-quiet BL Lac objects is similar to the normal radio-quiet AGNs (Lane et al. 2011), which further supports the accretion disk origin of the continuum. Accurate black hole mass measurements can determine the accretion state of radio-quiet BL Lac objects and further distinguish the different models related to the accretion disk origin of their continua, which will be explored in our future works.

GJC2013 and CKG2014 detected significant variations (confidence level >99% for two comparison stars) of SDSS J090843.25+285229.8 and SDSS J121929.45+471522.8 in their 29 light curves. Based on these two events, they derived a duty cycle ~5% of intranight variation from their sample on weak-line AGNs. However, two sources in their sample (SDSS J090107.64+384658.8 and SDSS J121929.45+471522.8) are likely to be galactic sources owing to the large proper motion. The proper motions of SDSS J090107.64+384658.8 and SDSS J121929.45+471522.8 from USNO-B are 62 ± 11 mas/yr and 112 ± 4 mas/yr, respectively (Monet et al. 2003). These two bright sources (V ~ 18) are well above the completeness limit V = 21 of the USNO-B catalog (Monet et al. 2003). The positional error of one epoch is ~200 mas. Thus, the proper motion of ~100 mas/yr can be accurately detected with the epoch difference of ~40 years. The systematic error of the proper motion is comparable to the statistical error. Munn et al. (2004) and Roeser et al. (2010) have further calibrated the USNO-B catalog with SDSS and 2MASS astrometry, and the resulting proper motions of these two sources are consistent with those from the USNO-B catalog. No radio or X-ray counterpart is found near their positions. Actually, Rebassa-Mansergas et al. (2010) lists SDSS J121929.45+471522.8 as a DC white dwarf in their catalog. SDSS J090107.64+384658.8 is classified as an uncertain DC white dwarf by Eisenstein et al. (2006) and as an uncertain DC+M binary system by Kleinman et al. (2013). The variation observed is likely due to the oscillation or accretion of the white dwarfs (Winget & Kepler 2008; Fontaine & Brassard 2008).

The DC white dwarfs are generally difficult to positively identify due to their featureless spectra and the discrepancy indeed exists in different catalogs. However, we think these possible galactic sources should be excluded from the sample of AGNs to be safe. We detected the intranight variation in SDSS J090107.64+384658.8 but excluded it from the sample. After the possible white dwarfs are removed from the sample of GJC2013 and CKG2014, only one significant variation (SDSS J090843.25+285229.8) is detected in the remaining 22 light curves, which is also the only significant event in all 32 light curves (including 10 additional ones of this paper) of weak-line AGNs up to now. This low occurrence rate is consistent with the rate for the normal radio-quiet AGNs.

Our present sample of radio-quiet BL Lac objects is still too small to constrain the duty cycle of the microvariation well. We will enlarge our sample, especially for the monitoring time longer than six hours, and improve the accuracy of photometry to ~0.01 mag to detect smaller variations.

Observations in more bands will help to investigate the color-behavior and further constrain the mechanism of the continuum of the radio-quiet BL Lac objects.


1

Given the sample size n and observed success counts k, the upper limit pu is defined by , where a = k + 1, b = nk + 1, q = 1 − p, and c is the confidence level.

Acknowledgments

The authors thank the referee for useful comments that improved the paper. This work is supported by 973 Program of China under grant 2014CB845802, by the National Natural Science Foundation of China under grant Nos. 11103019, 11133002, and 11103022, and 11373036, and by the Strategic Priority Research Program “The Emergence of Cosmological Structures” of the Chinese Academy of Sciences, Grant No. XDB09000000. We acknowledge the support of the staff of the Xinglong 2.16 m and the Lijiang 2.4 m telescope. Funding for the Lijiang 2.4 m telescope has been provided by CAS and the People’s Government of Yunnan Province. This work was partially supported by the Open Project Program of the Key Laboratory of Optical Astronomy, NAOC, CAS.

References

All Tables

Table 1

Information on observations.

Table 2

Information on targets and their companion stars.

Table 3

Results of the significance of variations.

All Figures

thumbnail Fig. 1

Differential light curves of AGN–Star 1, AGN–Star 2, and Star 1–Star 2.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.