1. Introduction
Optical monitoring of many individual quasars, radio-loud and radio-quiet, spanning several decades has revealed that large intensity variations (> 100%) are exclusive to the former category. Prominent examples in this category, showing large outbursts (on month-like time scales), include the blazars AO0235+164: Webb & Smith, Reference Webb and Smith1989; OJ 287: Takalo, Reference Takalo1994; CTA102: Raiteri et al., Reference Raiteri, Villata and Acosta-Pulido2017; 1ES 1927+654: Trakhtenbrot et al., Reference Trakhtenbrot, Arcavi and MacLeod2019. Traditionally, this signature of strongly beamed jets of energetic particles has not been found associated with radio-quiet quasars (RQQs; e.g., Padovani, Reference Padovani2017), for which only a much milder optical variability is observed even on year-like time scales (e.g., MacLeod et al., Reference MacLeod, Ivezić and Sesar2012; Caplar et al., Reference Caplar, Lilly and Trakhtenbrot2017; Sun et al., Reference Sun, Xue, Wang, Cai and Guo2018; Stone et al., Reference Stone, Shen and Burke2022; Arévalo et al., Reference Arévalo, Lira and Sánchez-Sáez2023). The persistent dichotomy in the degree of variability has led to the common perception that among quasars, any transition from the radio-quiet to radio-loud state does not happen on human time scales, although this almost certainly takes place on time scales of
$\sim 10^{5-6}$
years (see, e.g., Reynolds & Begelman, Reference Reynolds and Begelman1997; Czerny et al., Reference Czerny, Siemiginowska, Janiuk and Nikiel-Wroczyński2009; An & Baan, Reference An and Baan2012). Such variability time-scales are also expected by extrapolating from the compact stellar binaries, powered by
$\sim 10 \mathrm{M}_{\odot}$
black holes, for which jet-driven large radio variability is often witnessed on hour/day-like time scales (Mirabel & Rodr guez, Reference Mirabel and Rodrguez1999). However, over the past few years, exceptions to this popular perception about quasars have begun to surface, based on reports of detection of (first-time) switching-on, or restarting, of powerful jet activity in a few quasars, marked by between 6 to 25 fold jump in radio flux at centimetre wavelengths within a time span of at most 15-20 years (Mooley et al., Reference Mooley, Hallinan and Bourke2016; Kunert-Bajraszewska et al., Reference Kunert-Bajraszewska2020; Nyland et al., Reference Nyland, Dong and Patil2020; Wójtowicz et al., Reference Wójtowicz2020; also, Bannister et al., Reference Bannister, Murphy, Gaensler, Hunstead and Chatterjee2011a; Bannister et al., Reference Bannister, Murphy, Gaensler, Hunstead and Chatterjee2011b). The link of this jump in radio emission to newborn jets has been reinforced by the observed Gigahertz-peaked spectrum (GPS) type (evolving) radio spectrum associated with such state transition episodes, as reported in the discovery papers cited above. Indications of changes occurring on such short time scales had earlier come from the targeted Very Large Baseline Interferometry (VLBI) campaigns which revealed nascent radio jets in quasars, with kinematical ages as small as
$\sim$
20 years (Owsianik et al., Reference Owsianik, Conway and Polatidis1998; Owsianik & Conway, Reference Owsianik and Conway1998; Gugliucci et al., Reference Gugliucci, Taylor, Peck and Giroletti2005; Orienti & Dallacasa, Reference Orienti and Dallacasa2021). The detection of such young jets has bolstered the prospects of understanding the nature of the trigger for jet activity in Active Galactic Nuclei (AGN), which sometimes succeeds in pulling them out of the radio-quiet state.
Arguably, the best prospect of AGN hosting the earliest cycles of jet activity is to be found in radio-loud narrow-line Seyfert1 (NLS1) galaxies (Mathur, Reference Mathur2000). The existence of relativistic jets in such sources was indisputably established by their detection at gamma-rays (Abdo et al., Reference Abdo, Ackermann and Ajello2009), supported by independent observations, such as detection of high-amplitude intra-night optical variability (INOV) in several radio-loud NLS1 galaxies (e.g., Liu et al., Reference Liu, Wang, Mao and Wei2010; Paliya et al., Reference Paliya, Stalin and Kumar2013; Kshama et al., Reference Kshama, Paliya and Stalin2017; Gopal-Krishna & Wiita, 2018; Paliya, Reference Paliya2019; Ojha et al., Reference Ojha and Chand2021). A very striking related clue has come from the discovery of strong, recurrent flares at millimetre wavelengths, on day-like time scales, in several NLS1s which are essentially radio-quiet at centimetre wavelengths (Lähteenmäki et al., Reference Lähteenmäki, Järvelä and Ramakrishnan2018)Footnote a. These flares represent jumps in millimetre flux densities by factors of up to 10
$^3$
in a matter of just a few days! The discovery paper poses the question whether this violent behaviour is a manifestation of an early-stage of AGN activity and whether its source, possibly a jittery and intermittent relativistic jet (see, e.g., Czerny et al., Reference Czerny, Siemiginowska, Janiuk and Nikiel-Wroczyński2009; Lalakos et al., Reference Lalakos, Gottlieb and Kaaz2022), is still confined within the broad line region (BLR) region and, consequently, subject to free-free absorption, explaining the jet’s faintness at centimetre/metre wavelengths (e.g., Berton et al., Reference Berton, Järvelä and Crepaldi2020). This raises the possibility that a similarly violent activity on hour/day-like time scale might get triggered during the reported episodes of transition of quasars, from a radio-quiet to radio-loud state (see above). Here we attempt to test such a scenario by searching for strong intranight optical variability (INOV) which has now emerged as a reliable diagnostic of AGN activity driven by Doppler-boosted jet, including even the jets forming in low-mass AGN (e.g., Gopal-Krishna et al., 2023 and references therein). The target chosen for our INOV monitoring campaign is the quasar J0950+5128 which is the optically brightest known candidate for radio-quiet to radio-loud state transition (Nyland et al., Reference Nyland, Dong and Patil2020), out of the discovery of 26 such cases found by these authors to have undergone the radio state transition sometime during the last about two decades. Their radio emergence was unveiled during a comparison of the Very Large Array Sky Survey (VLASS, Lacy et al., Reference Lacy, Baum and Chandler2020) epoch 1 observations (2017-2019) with the existing FIRST (Faint Images of the Radio Sky at Twenty cm; 1993-2011) survey database (Becker et al., Reference Becker, White and Helfand1995). The peak frequencies of the observed GPS spectra of these transition candidates, which have been measured for 14 of them using the Very Large Array (VLA), seem consistent with the very young (radio) ages of these quasars (Nyland et al., Reference Nyland, Dong and Patil2020), as estimated using the radio size-peak frequency relation established for young AGN jets (O’Dea, Reference O’Dea1998; Jeyakumar, Reference Jeyakumar2016).
As noted in Nyland et al. (Reference Nyland, Dong and Patil2020), the
$z = 0.2142$
quasar J0950+5128 (J095036.75+512838.12), which is powered by a black hole with an estimated mass of
$10^{8}$
$\mathrm{M}_{\odot}$
, was not detected in the FIRST survey (hence
$ \lt $
0.44 mJy at 1.4 GHz) in April 1997. In April 2019, however, its flux density was found to have risen to 8.77 mJy at 3 GHz, corresponding to a radio luminosity of 3.7
$\times10^{40}$
erg/s. From VLA observations, its angular size was found to be
$ \lt $
0.16 arcsec at 13.2 GHz (Nyland et al., Reference Nyland, Dong and Patil2020). The SDSS catalogue gives its r-magnitude as 17.35, making it suitable for intranight monitoring with the 1.3-metre optical telescope accessible to us for this pilot experiment. The details of our optical observations, together with the derived differential light curves (DLCs) for 6 sessions are presented in Section 2, followed by a brief discussion of the results in Section 4 and the conclusions summarised in Section 5.
2. Photometric monitoring and data reduction
The (now radio-loudFootnote b) quasar J0950+5128 was observed in the Johnson–Cousins R band over 6 sessions using the 1.3-metre Devasthal Fast Optical Telescope (DFOT; Sagar et al., Reference Sagar, Omar and Kumar2011, see also, Mishra et al., Reference Mishra2021) at Aryabhatta Research Institute of observational sciencES (ARIES), near Nainital (India). In each session, the quasar was monitored continuously for a minimum duration of 4 hr, with each frame having a typical exposure of 4–5 min, recorded on a Peltier-cooled ANDOR CCD, equipped with 2k
$\times$
2k pixels (0.53 arcsec pixel
$^{-1}$
) and providing a field of view of 18.5
$\times$
18.5 arcmin
$^2$
. The CCD detector operates with a gain of 2 e
$^-$
per analogue-to-digital unit and exhibits a readout noise of 7.5 e
$^-$
at a speed of 1 MHz. Pre-processing of the raw images involved bias subtraction, flat-fielding and cosmic-ray removal, all carried out using the standard tasks available in the Image Reduction and Analysis Facility (IRAF)Footnote c. For each frame, the instrumental magnitudes of the quasar and the selected non-varying comparison stars within the same CCD frame were calculated through aperture photometry (Stetson, Reference Stetson1987, Reference Stetson, Worrall, Biemesderfer and Barnes1992), employing the Dominion Astronomical Observatory Photometry II (DAOPHOT II algorithm). The ‘point spread function’ (PSF) for each frame was determined by averaging the full width at half maximum (FWHM) of the profiles of 5 moderately bright stars within the frame. A median of the PSF values found for all the frames in a session was taken as the ‘seeing’ (FWHM of the PSF) for that session. For photometry, an aperture radius of 2
$\times$
FWHM (median) was adopted for that session (see, e.g., Gopal-Krishna et al., 2023, also, Negi et al., Reference Negi2023). The lower panel of Figure 1 displays the PSF variation during the session. Differential light curves (DLCs) for each session were then generated for all the pairs involving the target quasar and the selected three steady comparison stars and these are displayed in Figure 1.
Differential light curves (DLCs) for the six sessions. The name of the target object, together with the date and duration of its monitoring are shown near the top, in the mosaic for each session. The upper three panels in a mosaic display the ‘quasar-star’ DLCs, while the lower three panels display the comparison ‘star-star’ DLCs, as defined in the labels on the right side. The bottom panel for each session shows the seeing (PSF) variation during the session.
3. Statistical Analysis
To assess the presence of INOV in the DLCs, the widely used
$F_{\eta}$
test (de Diego, Reference de Diego2010) was applied, following the basic procedure described in Mishra et al. (Reference Mishra2019), Gopal-Krishna et al. (2023) and Negi et al. (Reference Negi2023). For each session, the three comparison stars were initially selected (Table 1) and the steadier two of them were identified by inspecting the ‘star–star’ DLCs. The
$F_{\eta}$
test was then applied to only the three DLCs involving these two comparison stars and the target quasar (see also, Chand et al., Reference Chand2022). The selected pair of comparison stars used for each session is indicated within parentheses, in column 5 of Table 2. Additionally, the labels of the three DLCs that involve these two comparison stars and the target quasar, are shown within parentheses on the right-hand side in Figure 1.
Basic parameters of the quasar and the selected comparison stars for the 6 monitoring sessions.
The comparison star shown in column 1 has been used for testing the presence of INOV of the quasar on the date marked with an asterisk.
Result of the statistical test for the presence of INOV in the DLCs of the Quasar J0950+5128.
$^a$
V=variable, i.e., confidence level
$\ge 0.99$
; PV = probable variable (
$0.95-0.99)$
; NV = non-variable (
$ \lt 0.95$
).
Variability status identifiers (col. 8), estimated for the ‘quasar- star1’ and ‘quasar - star2’ DLCs are separated by a comma.
It has been found in several independent studies that the photometric errors returned by DAOPHOT are too small, by a factor
$\eta$
ranging between 1.30 and 1.75 and therefore ignoring this factor can sometimes lead to the spurious claim of INOV (Gopal-Krishna et al., 1995; Garcia et al., Reference Garcia, Sodré, Jablonski and Terlevich1999; Sagar et al., Reference Sagar and Stalin2004; Stalin et al., Reference Stalin2004; Bachev et al., Reference Bachev, Strigachev and Semkov2005; Goyal et al., Reference Goyal2012, Reference Goyal2013a). An accurate estimate of
$\eta$
$ = 1.54 \pm 0.05$
was made through an analysis employing a large set of 262 DLCs of pairs of steady comparison stars. These stars had been monitored in 262 intranight sessions of a minimum 3 hr duration, together with the respective target AGN (Goyal et al., Reference Goyal and Mhaskey2013b). This accurate estimate of
$\eta = 1.54$
has been adopted in the present study. The F-values calculated for the selected two quasar DLCs during each session are:
\begin{equation}F_{1}^{\eta} = \frac{Var(q-s1)}{ \eta^2 \sum_{\textbf{i}=\textbf{1}}^{N}\sigma^2_{i,err}(q-s1)/N}, F_{2}^{\eta} = \frac{Var(q-s2)}{ \eta^2 \sum_{\textbf{i}=\textbf{1}}^{N}\sigma^2_{i,err}(q-s2)/N} \end{equation}
where
$Var(q-s1)$
and
$Var(q-s2)$
are the variances of the DLCs of the target quasar, relative to the selected two comparison stars, and
$\sigma_{i,err}(q-s1)$
and
$\sigma_{i,err}(q-s2)$
denote the rms errors returned by DAOPHOT on the
$i^{th}$
data point in the DLCs of the target quasar, relative to the two comparison stars. N denotes the number of data points in the DLCs of the session (Column 3 of Table 2) and the scaling factor
$\eta = 1.54$
, as mentioned above. For each monitoring session, Table 2 provides the values of
$F_{1}^{\eta}$
and
$F_{2}^{\eta}$
computed for the DLCs of the target quasar, relative to the two selected comparison stars. Also shown are the computed critical values of F (
$= F_{c}^{\alpha}$
), taking
$\alpha = $
0.05 and 0.01 which correspond to the INOV detection confidence levels of 95% and 99%, respectively (also listed in columns 6 and 7 of Table 2, for each session). These critical values are meant for comparison with the
$F-$
values computed for the two DLCs of the quasar, using Eq. (1), namely,
$F_{1,2}^{\eta}$
and presented in column 5 of the Table 2. The target quasar is classified as ‘variable’ (V) in a session if its computed
$F-$
value is
$\ge$
$F_{c}$
(0.99). If the
$F-$
value falls between
$F_{c}$
(0.95) and
$F_{c}$
(0.99), the DLC is considered as a ‘probable variable’ (PV). In case the
$F-$
value is found to be less than
$F_{c}$
(0.95), the DLC is designated as ‘non-variable’ (NV). Column 10 of Table 2 lists the ‘Photometric Noise Parameter’ for each session, defined as PNP =
$\sqrt { \eta^2\langle \sigma^2_{i,err} \rangle}$
, where
$\eta=1.54$
, as mentioned above.
4. Discussion
In this section, we shall briefly discuss the results of our search for INOV in the radio-state transition quasar J0950+5128.
The ZTF (r-band) light curves (LCs) of the quasar J0950+5128, for the three sessions. For each session, the top panel displays the LC of the quasar, while the lower 3 panels show the LCs for three selected comparison stars which are close in brightness to the target quasar. Each panel shows the SDSS name of the target (quasar/star) and the date and duration of ZTF monitoring. The INOV parameters determined from the variability analysis are shown below the corresponding LC, for each session (see text).
4.1. The intranight monitoring with DFOT
From Table 2 it is seen that the quasar DLCs for all six sessions are consistent with the non-detection of INOV. For blazars, a duty cycle of INOV detection has been found to be close to 50%, based on a large sample subjected to observations and analysis procedures very similar to that adopted in this work (Goyal et al., Reference Goyal2013a). Therefore, the non-detection of INOV on all six nights is very unlikely to be a chance occurrence. Thus, there is little evidence for blazar-like activity in the wake of the quasar’s change from a radio-quiet to radio-loud state, as found by Nyland et al. (Reference Nyland, Dong and Patil2020). As mentioned in Section 1, our expectation of blazar-like INOV following the radio state-transition stems from the extreme-flaring events detected in the millimetric observations of a few NLS1 galaxies, on day-like time scales, despite their showing a highly muted variability at centimetre wavelengths (Lähteenmäki et al., Reference Lähteenmäki, Järvelä and Ramakrishnan2018). The contrast could either be intrinsic, or due to absorption, or alternatively due to the non-simultaneity of the available data in the two bands (Järvelä et al., Reference Järvelä, Savolainen and Berton2024). In any case, it is expected that, compared to the muted variability at centimetre wavelengths, the (violent) variability seen at millimetre wavelengths would be more closely reflected in the optical band. This, however, is clearly not seen at least in the 6 monitoring sessions of this radio-state transition candidate quasar. Conceivably the INOV non-detection could be because of a misalignment of the jet of this quasar with respect to the line-of-sight, which would then call for an extension of the INOV campaign to other state transition candidate quasars. Moreover, in view of the growing evidence using the VLBI data archives, that the nuclear jets can undergo large directional changes on human time scales (e.g., Britzen et al., Reference Britzen, Krishna and Kun2023a; Britzen et al., Reference Britzen and Zajaček2023b)Footnote d, an INOV search would be worthwhile in future observations of this quasar and, possibly, also in any existing independent datasets available for this and other radio state-transition quasars.
4.2. Intranight light curves from the ZTF survey
For the present quasar J0950+5128, which is currently the brightest known such case, such an opportunity is afforded by the Zwicky Transient Facility (ZTF) surveyFootnote e. A check of the ZTF database, applying the criterion of > 25 data points per session (of minimum 3.5 hr duration), has revealed that such intranight monitoring data are available for 3 nights for the present quasar, all taken during January 2019, i.e., just a few months before its transition to radio-loud state was recorded in the VLASS observations (Section 1). Those 3 ZTF light curves (LCs) are reproduced in Figure 2, together with the results of the statistical test we performed to check for the presence of INOV, again following the method described in Section 3 (note that the rms error on each photometric data point was taken directly from the ZTF database and so the parameter
$\eta$
was set equal to unity). From the quantitative results of the
$F_{\eta}$
test, displayed in Figure 2, it is seen that statistically significant INOV was present during two out of the available three ZTF sessions for this quasar, all during January 2019. Also shown in the same figure are the ZTF LCs of three (simultaneously monitored) comparison stars, which match the quasar to within 0.25-mag (r-band). Neither star showed a statistically significant INOV in any of the 3 sessions, suggesting that the inferred statistical significance of the INOV of the quasar in two of the three ZTF sessions (January 2019) is probably real. This raises the possibility that the present quasar J0950+5128 was going through an INOV phase around the time its transition from a radio-quiet to radio-loud state was discovered (in April 2019, Section 1). However, its optical variability had subsided by the time our INOV campaign began in December 2020 (Table 2).
A sustained follow-up of this pilot experiment is warranted, in order to develop a proper understanding of the relationship between the two extraordinary phenomena, namely the radio-quiet to radio-loud state transition of quasars and the early blazar-like activity hinted by the observations of violent millimetric flaring of (possibly nascent) AGN (Section 1). Such a possibility of activity, although hinted in the present study, is still of a preliminary nature, but a follow-up holds considerable promise of unravelling any underlying pattern.
4.3. A single-point brightness spike
We conclude the present discussion by drawing attention to the only INOV feature detected in the DLCs of J0950+5128 during our campaign reported here. The feature is a
$\sim$
0.15-mag spike lasting
$ \lt $
6 minutes, which was seen at 13.97 UT on 18-March-2021, in all 3 DLCs of the quasar taken relative to the 3 different comparison stars (see Figure 1). For this reason, and also because the seeing disk had remained steady at that time (Figure 1), we conclude that the sharp spike is probably real. It is interesting to recall that 15 highly significant single-point spikes of duration
$ \lt $
5−10 minutes were identified in the DLCs (Gopal-Krishna et al., 2000; Stalin et al., Reference Stalin2004) under the AGN INOV program conducted at ARIES, which typically allowed INOV detection down to the
$1 - 2\%$
level (see, Gopal-Krishna & Wiita, 2018). Interestingly, 6 of the strongest 7 spikes (out of the total of 15), were observed in the DLCs of radio-quiet (or radio lobe dominated) quasars, and not in the concurrently recorded DLCs of the comparison stars. This hints at the possibility of an AGN related (albeit, a non-blazar) origin of the strong optical spikes. With the rapid growth in optical databases on AGN monitoring, it would be interesting to look for similar spikes in their optical light curves, including those of the rare quasars found to have recently transitioned from a radio-quiet to radio-loud state, or vice versa.
5. Conclusions
In this work, we have endeavoured to address the question of whether the transition from a radio-quiet to radio-loud state on a decadal time scale, as noticed a few years ago for a handful of quasars, is accompanied by some change in their activity level in the optical band. Thus, for the quasar J0950+5128, which is the optically brightest such transition case, we have enquired if post-transition it showed signs of a blazar-like activity. This possibility is hinted at, e.g., by the intense flaring of millimetric emission on a day-like time scale, witnessed in a few NLS1 galaxies (which have been proposed to represent the early stage in the formation of radio-loud quasars). Since intranight optical variability (INOV) is a robust marker of blazar activity, we have carried out sensitive and densely sampled intranight optical monitoring of J0950+5128, on 6 nights between December 2020 and April 2021. Down to the 1-2% detection limit, we found no evidence for INOV on any of the 6 nights, and hence our monitoring campaign has not unveiled any INOV activity in this quasar (except for a single
$\sim$
0.15-mag spike lasting < 6 minutes, observed on the night of 18-March-2021). Additionally, we find that for 3 nights, a few months before its radio-state transition was picked in the VLASS survey at 3 GHz, densely sampled ZTF r-band light curves are available for this quasar (during January 2019) and these show statistically significant INOV on two of the nights. This would suggest that around the time its transition to a radio-loud state was first noticed, the quasar exhibited some INOV activity which, however, subsided over the next two years, i.e., by the time of our INOV campaign. This, admittedly marginal evidence for a level change in the INOV activity, could either have an origin intrinsic to the jet, or merely reflect an increase in the jet’s misalignment to the line-of-sight. The observational pointers presented here call for a follow-up of this first-time INOV study of a quasar which exemplifies an exceedingly rare phenomenon of radio-state transition in quasars occurring on a decadal, or possibly even shorter time scale.
Acknowledgements
KC acknowledges the support of SERB-DST, New Delhi, for funding under the National Post-Doctoral Fellowship Scheme through grant no. PDF/2023/004071. G-K acknowledges a Senior Scientist position awarded by the Indian National Science Academy (INSA). The assistance from the scientific and technical staff of ARIES DFOT is thankfully acknowledged. The local hospitality extended to KC by IUCAA, Pune, during the preparation of this manuscript, and IUCAA associateship to HC are also thankfully acknowledged. This work is also based on observations obtained with the Samuel Oschin Telescope 48-inch and the 60-inch Telescope at the Palomar Observatory as part of the Zwicky Transient Facility project. ZTF is supported by the National Science Foundation under Grants No. AST-1440341 and AST-2034437 and a collaboration including current partners Caltech, IPAC, the Oskar Klein Center at Stockholm University, the University of Maryland, University of California, Berkeley, the University of Wisconsin at Milwaukee, University of Warwick, Ruhr University, Cornell University, Northwestern University, and Drexel University. Operations are conducted by COO, IPAC, and UW.
Data availability
The data used in this study will be shared at a reasonable request by the corresponding author.
(Kellermann et al., 
