EDP Sciences logo
Subscriber Authentication Point
Search
Menu
Home All issues Volume 687 (July 2024) A&A, 687 (2024) A16 Full HTML
Open Access
Issue
A&A
Volume 687, July 2024
Article Number A16
Number of page(s) 12
Section Catalogs and data
DOI https://doi.org/10.1051/0004-6361/202346504
Published online 25 June 2024

© The Authors 2024

Licence Creative CommonsOpen Access article, published by EDP Sciences, under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

This article is published in open access under the Subscribe to Open model. Subscribe to A&A to support open access publication.

1 Introduction

Quasi-stellar objects (QSOs) are distant galaxies with central supermassive black holes that accrete matter from their surroundings. They are only luminous on a scale of a few hundred parsecs, and they therefore appear as point-like objects.

Quasar candidates are often selected based on optical or infrared (IR) variability (Hook et al. 1994; Carnerero et al. 2023; Treiber et al. 2023) radio emission (Gregg et al. 1996; Tuccillo et al. 2015) X-ray (Shanks et al. 1991; Waddell et al. 2024) or mid–IR colors (Lacy et al. 2004; Sajina et al. 2022). Multi-wavelength selections are also used (DiPompeo et al. 2015; Spiniello & Agnello 2019). Bona fide QSOs are confirmed by their broad emission lines (although the absence of these lines does not imply that the object is not a QSO), which also help to derive their redshifts (Vanden Berk et al. 2001). Millions of these objects are known today (Flesch 2023, 2024; Storey-Fisher et al. 2024; Onken et al. 2023) and the lists continue to be updated with the latest surveys.

Quasars provide insights into the structure and evolution of galaxies and are cosmological probes of the intervening interstellar medium. They can also serve as an absolute astrometric reference frame as distant unmoving point-source objects. This reference frame is required to measure the small proper motions (PMs) within nearby galaxies. However, it is challenging to identify QSOs behind these galaxies because of the rich and diverse stellar content of the galaxies themselves, which blurs the boundaries of the stellar locus. Furthermore, the internal reddening inside the foreground galaxies modifies the QSO colors, especially the optical colors.

It is difficult to find QSOs at the far side of the nearby Magellanic Cloud system because of the foreground contamination. It is particularly difficult to find them behind the Small Magellanic Cloud (SMC) because it exhibits a notable depth and is associated with it stronger source confusion (de Grijs & Bono 2015). Nevertheless, a significant number of QSOs was identified there by studies that specifically targeted this region (Blanco & Heathcote 1986; Dobrzycki et al. 2002, 2003a,b, 2005; Geha et al. 2003; Kozłowski & Kochanek 2009; Kozłowski et al. 2012, 2013, 2011; Véron-Cetty & Véron 2010).

Optical searches can confuse QSOs with stars (van Loon & Sansom 2015; Pennock et al. 2022) or miss more highly obscured QSOs. Searches at radio (Pennock et al. 2021) or near–and mid–IR wavebands can help to alleviate the latter problem. This prompted us to search for QSOs in the VISTA (Visual and Infrared Survey Telescope for Astronomy; Emerson et al. 2006) Survey of the Magellanic Clouds system (VMC; Cioni et al. 2011). The largest single spectroscopically confirmed sample of quasars so far, 758 QSOs, was an OGLE (Optical Gravitational Lensing Experiment Udalski et al. 1992) follow-up by Kozłowski et al. (2013), and more recently, by the Quaia survey based on the low-resolution (R~30) Gaia spectra (Storey-Fisher et al. 2024). Our main goal is to increase the number of known quasars behind the Large and Small Magellanic Clouds, aiming to improve the proper motion reference system. We reported our first results, 37 (34 new) QSOs, also spectroscopically confirmed, in Ivanov et al. (2016). This is the second paper in our series, with further spectroscopy from Very Large Telescope (VLT) and South African Astronomical Observatory (SAAO).

2 Sample selection

2.1 VMC

The VMC is an ESO public survey with a footprint of 184°2 on the sky around the Large Magellanic Cloud (LMC), the SMC, the Magellanic Bridge, and the Magellanic Stream. The photometry reaches a signal-to-noise ratio S/N~10 at Ks=20.3 mag (Vega system) in three epochs in the Y and J bands and in 12 epochs in the Ks band. More epochs are available for some tiles1 and in the areas of overlapping tiles. The time series span nearly eight years. The VMC science is diverse and includes star formation and star formation history, galaxy structure, star clusters, various types of stars (e.g., RR Lyrae and Cepheids), proper motions, background galaxies, and even distant quasars.

The VMC is carried out at VISTA, which is a 4.1m telescope located on Cerro Paranal, equipped with VIRCAM (VISTA InfraRed CAMera; Dalton et al. 2006), which is a near-IR wide-field ~1×1.5°2) camera. The data are reduced with the VISTA Data Flow System (VDFS; Irwin et al. 2004; Emerson et al. 2004) at the Cambridge Astronomical Survey Unit2. Their photometric calibration is described in González-Fernández et al. (2018). The data products are available at the ESO Science Archive Facility3 or at the VISTA Science Archive (VSA; Cross et al. 2012)4.

2.2 Quasar candidate selection

The criteria for identifying QSO candidates from VMC were defined by Cioni et al. (2013) from 117 known QSOs in a (Y–J) versus (J–Ks) color–color diagram (Fig. 1) and from their Ks-band variability, requiring that the absolute value of the light-curve slope exceeds a certain limit of ISlopel>10−4 mag day−1. The light-curve slope was determined from a simple linear fit of the Ks band versus the time of observations in days, and the adopted limit was based on the behavior of the same known quasars. The quasars were selected with the data available at the time of the Phase 2 submission (DR6 plus some additional images, typically one per tile). Since then, some data have been reprocessed and more observations have become available. An inspection of updated light curves and slope measurements from them (Fig. 2) showed that a significant number of the spectroscopically confirmed quasars fall below the adopted limit, which means that a more stringently defined criterion needs to be adopted in the future, for example, based on a damped random walk (Kelly et al. 2009).

Here, as in Ivanov et al. (2016), we applied the same criteria on 18 mostly peripheral tiles: 3 tiles in the LMC, 7 tiles in the SMC, 6 tiles in the Bridge, and 2 tiles in the Stream. We selected 142 candidates: 15 candidates in the LMC, 40 candidates in the SMC, 63 candidates in the Bridge, and 24 candidates in the Stream. The Bridge and Stream tiles yield more candidates per tile because the foreground contamination from the Magellanic Clouds and the extinction are lower. The observations were carried out under vastly diverse weather conditions because the constraints of the program were relaxed. This allowed a seeing of up to 2″ and thin cloud coverage.

During the selection, we excluded any previously known QSOs and gave preference to the brightest candidates in each tile to optimize the spectroscopic follow-up. Therefore, our results cannot be used to draw strict statistical conclusions. Our main goal is to confirm as many QSOs as possible for future astrometric studies.

Our main target selection was aimed at a VLT follow-up, but spectroscopic confirmation with smaller facilities is also possible. To facilitate this, we performed an identical candidate selection in the entire VMC survey footprint with an additional brightness criterion of Y< 18.0 mag to enable a follow-up with a 2 m class telescope. This yielded 36 objects, 15 of which lie in the range Y=16.5–17.4 mag. Seven of these were observed, bringing the total number of objects with follow-up to 151.

Table 1 lists the observed candidates: the VMC identification (Col. 1), coordinates (Cols. 2–3), magnitudes in the YJKS bands and their photometric errors (Cols. 4–9), and the object identification (Col. 10) used in the spectroscopic observations. The last consists of the VMC tile name and a sequential number in the catalog of sources in that tile; the letter g indicates that a source was extended according to the VDFS pipeline. Ivanov et al. (2016) demonstrated that some low- redshift QSOs fall into this category. The location of the newly confirmed QSOs in the (Y–J) versus (J–Ks) color-color diagram is shown in Fig. 1. Their positions on the sky are plotted in Fig. 3, and their Y-band finding charts are shown in Fig. A.1.

Appendix B presents the complete list of all VMC survey quasar candidates in the latest internal release (August 2022) that match our color and variability criteria for the benefit of further studies. The color criterion alone yields 163226 objects, and 3609 of these have statistically significant slopes (Slope/σSlope≥3) that exceed the ISlopel=l0−4 mag day−1 limit.

thumbnail Fig. 1

Color–color diagram showing the color selection of candidates. The dashed lines identify regions (labeled with letters) with known QSOs, and the solid magenta line marks the blue border of the planetary nebulae locus (Cioni et al. 2013). The confirmed QSOs from this work are plotted as red open circles. The open green circles and triangles mark QSOs and non-QSOs from Ivanov et al. (2016), respectively. The blue crosses indicate VMC counterparts to the spectroscopically confirmed QSOs from Kozłowski et al. (2013, VMC photometry selected within a matching radius of 1″). The black dots are randomly drawn LMC objects (with errors in all three bands <0.1 mag) to mark the main stellar locus; some of those in regions B and C are contaminating background galaxies.
thumbnail Fig. 2

Histogram of the absolute slopes from the linear fit to the light curves of the objects in this paper: classified as QSOs (wide bins), remaining unknown because of poor-quality spectra or lack of prominent emission lines (medium-width bins) and confirmed stars (narrow bins). The vertical dotted line is the QSO selection limit defined by Cioni et al. (2013).
thumbnail Fig. 3

Location of the objects with spectroscopic follow-up in this work (red) and confirmed QSOs from Kozłowski et al. (2013, blue) and Ivanov et al. (2016, green). The VMC tiles are shown as rectangles. The grid shows lines of constant right ascension and declination (spaced by 15º and 5°, respectively). The coordinates are offset with respect to (α0, δ0) = (51º, −69º).

3 Spectroscopic follow-up observations

We selected 142 candidates that were among the brightest in each tile for the follow-up. Their locations were scattered throughout the survey footprint for a better chance of observing them in service mode. Two additional objects (BRI 3_4 046_2 and SMC 6_3 141_2) serendipitously fell into the slits, which increased the total number of observed targets to 144. The selection and observations followed the same procedure as in Ivanov et al. (2016). We used FOcal Reducer and low dispersion Spectrograph 2 (FORS2; Appenzeller et al. 1998) at the VLT (Very Large Telescope) between October 2016 and August 2017 in long-slit mode with the 300V+10 grism, GG435+81 order-sorting filter, and a 1.3″ wide slit. The spectra cover λ=445–865 nm with a resolving power R=λ/∆λ~440. Two exposures with integration times between 60 sec and 530 s were taken, depending on the apparent QSO brightness at the time of the VMC observations, which is not necessarily the same as at the moment of the follow-up because the intrinsic QSO variability may render the quasars fainter or brighter. In some cases, we had more exposures because the conditions were too poor for even our relaxed weather constraints or because the observations were interrupted because the telescopes were closed and a full set of observations was obtained on another night. We also exploited the low-quality data as long as emission lines were identifiable. The signal-to-noise ratio typically was ~ 10–20 at the centre of the wavelength range, at continuum level, and higher at the emission lines. The reduction was performed with the FORS2 pipeline (v. 5.3.23). The spectrophotometric calibration was carried out with standards (Oke 1990; Hamuy et al. 1992, 1994; Moehler et al. 2014a,b), observed and processed as the science spectra. The VLT observing log is given in Table 2.

An additional seven objects (bringing the total number of spectra to 151) were observed with the SpUpNIC5 (Spectrograph Upgrade: Newly Improved Cassegrain; Crause et al. 2016, 2019) long-slit spectrograph at the 1.9 m telescope at SAAO in November 2017. Grating 7 with no order-separation filter (to extend the spectral coverage; emission lines from the second-order spectra were ignored) was used with a 1.95″ wide slit, delivering a resolving power of R~700–1700 over λλ=3750–9300 Å. The slit was always oriented East–west, and a single 1800 s exposure was taken for each object. The data reduction was performed in the standard way, with bias and dome flat-field corrections. The wavelength calibration was made with spectra from an internal CuAr lamp, obtained after each science observation to cancel out possible instrument flexures. Feige 110 and LTT 3218 standards (from the same lists as for the VLT) were used to derive the spectral response. The 1.9 m telescope observing log is given in Table 3.

The reduced 148 spectra (for three objects, no 1D spectra could be extracted, usually due to poor weather conditions) are shown in Fig. 4. The emission lines were identified by comparing our spectra with a composite QSO spectrum (Vanden Berk et al. 2001). We typically had several lines; when only one was available, it had to be MgII 2798 Å, because if the line belonged to a different element or if the redshift was different, other emission lines with comparable strength would have been inside the observed wavelength range (because the S/N was sufficient to detect the MgII), and lines like that were missing. The redshifts were measured as in Ivanov et al. (2016), by fitting emission lines with Gaussian profiles using the IRAF6 task splot. The results are listed in Table 4. Some features at the edges of the spectra or features that were strongly contaminated by sky emission lines or affected by intervening absorption were omitted from the analysis. The statistical errors and errors from wavelength calibration were negligible (see Sect. 3 in Ivanov et al. 2016). We measured the redshift error from different lines of the same object. The finding charts of the objects that were followed up are shown in Fig. A.1.

A single line was observed in some spectra. For example, this occurred for quasars at redshifts between ~1 and ~1.4 when the only prominent line within our wavelength range was MgII 2798 Å. The featureless continuum outside of the line excludes the presence of any other lines: When the observed line was CIII] 1909, for instance, then CIV 1549 should have appeared in the blue part of the spectrum. In most cases, the S/N of the spectra was sufficient (we adopted a limit of 10) at the continuum level to exclude the presence of other lines, but this was not always the case. Therefore, some of our redshifts are tentative, and they are marked in Table 4 with colon signs. The table also contains the S/N calculated in the vicinity of each emission line that we used to evaluate the redshift. The S/N refers to resolution elements, and therefore, the real S/N for the broad lines is much higher, and the derived errors for the central wavelength especially for the redshift estimates are more relevant. For narrow-line objects, the estimated S/N is dominated by the continuum level and is a lower limit for the line core (e.g., in LMC 3_5 085g, Bridge 3_4 209g, and Bridge 2_7 053g). Finally, for the single-line objects we tentatively assigned redshift errors of 0.005 for z≤ 1 and 0.015 for z>1. These are typical values at these redshifts.

Table 1

VMC coordinates and photometry of QSO candidates (in order of increasing right ascension).

Table 2

Log of the VLT spectroscopic observations.

Table 3

Log of the 1.9 m SAAO spectroscopic observations.

Table 4

Derived parameters for the objects in this paper.

thumbnail Fig. 4

Spectra of 148 objects (for 3 objects, no spectra could be extracted, usually due to poor weather conditions) sorted by redshift and shifted to the rest-frame wavelength. The spectra were normalized to an average value of one and shifted vertically by offsets of two, four, etc., for display purposes. The SDSS composite QSO spectrum (Vanden Berk et al. 2001) is shown at the top of all panels. Objects without a measured redshift due to a lack of lines or low S/N are plotted assuming z=0 in the fifth panel next to the sky spectrum to facilitate identifying the residuals from the sky emission lines.

4 Results and discussion

4.1 Quasar confirmation

The majority of the candidates, 136 out of 148 observed objects, are bona fide QSOs at z~0.1–2.9 (Fig. 4). They all show some broad emission lines even though some spectra need block averaging, typically by four resolution bins, to make their features clearly evident in the plot (the line measurements were made on the original, not on the smoothed spectra). Lyα is visible in the spectra of the highest-redshift QSOs, and the remaining sample shows other typical emission lines. The newly confirmed QSOs from the VLT-observed sample are distributed as follows: 11 in the LMC, 34 in the SMC, 62 in the Bridge, and 24 in the Stream areas. The 5 QSOs from the bright candidate list are one in the LMC, 3 in the SMC, and another one in the Bridge. The spectra of six objects appear to be star-like without broad emissions, and nine further objects are too noisy for a secure classification.

The VDFS pipeline classifies 55 of our 151 objects as extended sources, and 49 of these are spectroscopically confirmed QSOs. However, their extended nature seems to be more a matter of foreground contamination from Magellanic Cloud sources than a real resolving of the host galaxy because ten extended QSOs have redshift z>1 and their hosts are unlikely to be resolved under atmospheric seeing conditions. Furthermore, one of the extended non-QSOs is a blue LMC star, and its extended nature is likely due to a surrounding star cluster. Therefore, the VDFS classification cannot be a critical constraint when a QSO candidate sample is assembled.

Figure 5 shows that the fraction of confirmed QSOs with z≤ 1 is higher here than in Ivanov et al. (2016). The reason likely is that we selected brighter candidates and based our study on a large number of Bridge tiles, in which the reddening of background objects is lower and the foreground contamination is weaker than behind the LMC and SMC. The colors of lower-redshift QSOs are closer to the stellar locus than those of the higher-redshift QSOs, which makes it harder to identify them in the inner regions of the Magellanic Clouds and easier in the Bridge.

Detections for some of our targets, which are all confirmed quasars, have been reported in the literature: Bridge 3_4 109g and Stream 2_l 174g in UV with GALEX (Galaxy Evolution Explorer; Morrissey et al. 2007); SMC 4_5 060 and SMC 6_3 310 in X-ray with XMM-Newton (X-ray Multi-Mirror Mission; Sturm et al. 2013); SMC 6_3 310 and SMC 4_2 071g are listed as new X-ray identified active galactic nucleus candidates by Maitra et al. (2019), also based on XMM-Newton, and SMC 4_2 071g and again SMC 4_5 060 were detected in X-ray as well, but with Chandra (McGowan et al. 2008). Kim et al. (2012) identified LMC 8_3 039 as a QSO candidate from a combination of light curves and multicolor criteria, and our spectrum confirms it. To summarize, only a handful of our confirmed QSOs have been detected in X-ray so far because of the sparse coverage with sensitive observations. eROSITA, the soft X-ray instrument on board the Spectrum-Roentgen-Gamma (SRG) mission (Predehl et al. 2021), completely covered the region of the Magellanic Clouds during its all-sky surveys. A preliminary investigation of the eROSITA catalog derived from the first all-sky survey data (Merloni et al. 2024) reveals a detection rate up to 30%, depending on matching criteria. More detailed work on the X-ray properties of QSOs behind the Magellanic system is in progress.

thumbnail Fig. 5

Redshift histogram for 136 confirmed QSOs (solid line). For comparison, we plot the 47 objects with measured redshiſt from Ivanov et al. (2016, dashed line).
thumbnail Fig. 6

Parameters of the quasar sample. Top: luminosity functions of quasars from Quaia and the VMC survey in G band. All Quaia quasars (1.3 million objects; blue), Quaia quasars with VMC counterparts with quasar-like colors (2347 objects; orange); spectroscopically confirmed VMC quasars from this paper and from Ivanov et al. (2016) with Gaia counterparts (161 objects; red); the same as the red sample, but without the 5 objects from the bright SAAO sample and those that fall below the Quaia G=20.5 mag limit (green), projected output for a VMC survey full quasar sample (450 objects; black dots), and confirmed Quaia quasars scaled down from the full sky to the VMC survey area (~6300 objects; cyan dots), assuming a loss of 10% in the zone of avoidance. In reality, 7386 Quaia quasars fall in the VMC survey footprint, bottom: redshiſt distribution for all Quaia quasars (1.3 million objects; blue), for Quaia quasars with VMC counterparts with quasar-like colors (2347 objects; orange), and Quaia quasars with spectroscopically confirmed quasars from our VMC selection (161 objects; green). See Sect. 4 for details.
thumbnail Fig. 7

Comparison of the redshifts from Quaia and from our VLT spectra (top) and absolute values of the differences as function of redshiſt in units of sigma (bottom). The labels above each dashed line indicate the numbers of quasars above these lines and the corresponding fractions in percentages. See Sect. 4 for details.

4.2 Completeness

Controlling the completeness is not a main goal of this paper. We aim to increase the number of confirmed quasars with the shortest possible observing time, but to still estimate, we turned to the sample of nearly 1.3 million quasars (with G<20.5 mag; Fig. 6, top panel, blue), with secure redshifts from the low-resolution Gaia spectra from Storey-Fisher et al. (2024; Quaia). This catalog has its own incompleteness and contamination issues, but the spectroscopic nature of the confirmation, the large number of quasars, and the full coverage of the VMC survey footprint makes it most suitable for this purpose.

First, we determined the number of quasars that our search should find that will also be accessible to Quaia. To compare the output of our spectroscopic confirmation campaign from this work and from Ivanov et al. (2016) with the Quaia sample, we scaled our 102 VLT-confirmed quasars up with Quaia counterparts, with the inverse fraction of the observed tiles. Between this work and Ivanov et al. (2016), we have followed-up 25 out of 110 VMC survey tiles (the seven bright additional objects followed-up at the SAAO are scattered across the entire VMC survey footprint and were ignored here). Extrapolating over the entire VMC survey, we scaled these 102 VLT-confirmed quasars up with G in the overlapping range 18–20.5 mag (to be discussed further) by an area ratio factor of 4.4, arriving at ~450 VMC survey quasars that we may be able to confirm with the same selection and following the same observing strategy (black dots in Fig. 6, top). This number must be compared with the expected number of Quaia quasars within the VMC survey footprint over an identical magnitude interval.

On one hand, our estimates are optimistic because the tiles in this study are located in the outermost regions in the system, where crowding and confusion are not as problematic as in the innermost regions. On the other hand, it is known that not all quasars are variable and would meet the variability criterion of Cioni et al. (2013). Furthermore, the observations were taken under highly variable weather conditions because our program was designed as a poor-weather filler, and we only followed-up the brightest candidates to optimize telescope time.

To obtain G-band magnitudes for our candidates and confirmed quasars, we cross-identified them with the Gaia DR3 main catalog (Gaia Collaboration 2022) within a radius of 0.35″. The results for the former are reported in Table B.1, and for the latter, we show them in the red histogram in the top panel of Fig. 6 for 161 objects (quasars confirmed at the 1.9 m SAAO were excluded for statistical reasons). It appears that 42 (~24%) of them cannot ever have Quaia counterparts because they fall below its G=20.5 mag limit, and the bright Quaia quasars fall in the VMC saturation regime. Here and throughout, the matching radius was set to make any spurious contamination unlikely: we repeated the cross-identification with modified coordinates, increasing the Dec by 10′ and found only five matches, three of which are with separations >0.9″ and the other two are at 0.25–0.35″. This keeps the spurious contamination at a level of ~0.1%.

Our quasars are relatively faint because of the saturation of brighter sources in the VMC. Bright quasars are important for studies of intervening absorption lines, for instance, but they only constitute a small fraction: By integrating the Quaia luminosity function above G=16.5mag, we find that quasars without a single VMC survey counterpart constitute ~0.1% of the entire Quaia sample, and those that are brighter than G=18.0mag, which is the brightest limit of our confirmed quasars (dashed black line in the upper panel), constitute only ~2.3%.

Next, we determined the number of Quaia quasars that are accessible to the VMC survey-based quasar search and follow-up We cross-identified Quaia quasars with the VMC source catalog (as of DR6): 7386 Quaia quasars fall within the VMC survey footprint; 6923 have matches within 0.35″ radii, but only for 3149 do we have sufficiently good photometry in all three bands without contamination or error flags and errors <0.15 mag; 2347 matches remain when our color-selection criteria are applied. We considered all objects, regardless of their apparent G-band magnitude. However, for a proper comparison, we should exclude objects that are too bright and would saturate in the VMC. They would reside in the brightest bins with G⪝ 17.5−18 mag, as shown in Fig. 6. Integration of the Quaia luminosity function above this limit indicates that only ~2% of Quaia quasars are lost. We therefore ignored them for the purpose of this comparison.

Finally, our expected yield of 450 confirmed quasars within the Quaia apparent magnitude range from the entire VMC survey constitutes ~6% of all Quaia quasars. This fraction increases to ~ 19% of nearly 2350 Quaia quasars that match our color criteria. We speculate that these low fractions are caused by the weak variability of most quasars, which our survey data cannot recognize. This is consistent with our preliminary investigation of the light curves of Quaia quasars with multi-epoch counterparts in the VMC survey: The preliminary investigation suggested that the vast majority of Quaia quasars are indeed not variable enough to meet our minimum light-curve slope criterion. This warrants further investigation and will be reported in another paper.

This result suggests that the complete list of our candidates presented in Table B.1 probably only contains a small fraction of all quasars within the VMC footprint. However, the success rate of our new spectroscopic follow up is ~90% (~91% for VLT and ~71% for 1.9 m), implying that nearly nine of every ten candidates in the list are likely to be confirmed as quasars when a similar spectroscopic follow-up is carried out.

The distribution of redshifts for the entire Quaia (blue), the Quaia quasars with VMC counterparts (orange), and our spec-troscopically confirmed quasars with Quaia counterparts (green) are shown in the lower panel of Fig. 6. Our sample is dominated by quasars with a lower redshift than Quaia, probably because we tended to select the brightest candidates in each tile for the follow-up to minimize the observing time per object. Storey-Fisher et al. (2024) verified the quality of their redshifts. A comparison with the redshift derived from our 102 VLT spectra suggests that the main disagreements occur in only a few percent of the cases (Fig. 7). This lends additional credibility to the Quaia catalog. The redshifts that differ are not concentrated toward the faintest objects, but occur over a range of apparent magnitudes, and their Quaia redshift errors are typically significantly larger than for other objects with similar magnitudes.

thumbnail Fig. 8

Left: color-color diagram of all spectroscopically confirmed QSOs with VISTA/VIRCAM photometry. For the notation of the regions, see Fig. 1. Right: Y–J and J–KS colors as a function of redshift for all spectroscopically confirmed QSOs with VISTA/VIRCAM photometry. The variations are driven by strong emission lines that enter and exit the band passes of individual filters.

4.3 Color-redshift relations

We explored whether the diagnostic color–color diagram that we used to select candidates can help us to constrain their redshift. To expand the statistics, we added to the VMC QSOs the QSOs from the latest catalog of Véron-Cetty & Véron (2010, 13 edition), which were observed with some other VISTA/VIRCAM surveys: VHS (McMahon et al. 2013, 5719 objects), VIDEO (Jarvis et al. 2013, 339 objects), and VVV (Minniti et al. 2010, 5 objects). The color–color diagram coded by redshift is shown in the left panel of Fig. 8. The lowest-redshift QSOs are clustered in a locus at Y–J~0.45–0.65mag and J–KS ~ 1.1–1.8 mag, but the more distant are scattered over the entire diagram. The reason for this behavior is shown in the panels on the right, which show that the two colors vary with redshift. Sharp color changes are visible as various more prominent emission lines enter or exit the bandpasses of individual filters. Therefore, this diagram has the potential of separating only the nearest QSOs.

5 Summary

We spectroscopically confirmed 136 QSOs within the footprint of the Magellanic system. They were selected from their near-IR colors and variability from the ESO VMC public survey. The uniform VMC observations, spanning nearly 8 yr, proved a reliable resource for QSO selection because nearly 90% of the observed candidates were quasars. However, a comparison with the Quaia catalog indicated that our selection recovered only 6–19% of the quasars identified from the Gaia low-resolution spectra. The fraction depends on the magnitude range, the quality of VMC survey photometry, and the candidate colors. The fraction appears to be relatively low because most quasars are not sufficiently variable to meet our variability criterion. Our quasar candidate list therefore is far from complete, but the candidates in it are quasars with a high degree of certainty. The variability is an important quasar identification tool, especially for radio-quiet quasars. Finally, we reported a list of 3609 candidates that met our criteria from the entire VMC survey footprint. Based on the previous statistics, we expect that nearly 90% of them will be confirmed as quasars if they are subjected to a spectroscopic follow-up similar to the one described here.

Acknowledgements

This paper is based on observations made with ESO telescopes at the La Silla Paranal Observatory under program IDs 092.B-0104(A), 098.B-0229(A) and 099.B-0204(A). We have made extensive use of the SIM-BAD Database at CDS (Centre de Données astronomiques) Strasbourg, the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, CalTech, under contract with NASA, and of the VizieR catalog access tool, CDS, Strasbourg, France. We thank Lisa Crause for efficient support at the SAAO telescope. JEMC acknowledges STFC studentship. We thank the anonymous referee for the comments that helped to improve the paper.

Appendix A Finder charts

VMC survey finder charts of the objects we considered are shown in Fig. A.1.

thumbnail Fig. A.1

Finding charts (Y band, 1′×1′) for all 151 objects (crosses) with follow-up spectroscopy, sorted by right ascension. North is at the top, and east is to the left.

Appendix B Complete VMC quasar candidate sample

Table B.1 lists 3609 objects that match our color and variability selection criteria. The comparison with the Quaia catalog indicates that we identify about 7% of their G≤20,5 mag quasars (see Sect. 4).

A SIMBAD search within 5″ from the VMC positions yielded 117 matches:

  • 97 are known quasars, active galaxies, X-ray, radio, or blue-UV sources that may be consistent with galactic activity, or candidates of any of these classes;

  • 16 are classified as stars. Long-period variables and young stellar objects constitute the largest consistent groups with five and three entries, respectively;

  • 2 are galaxies;

  • 2 are objects of an unknown nature.

Table B.1

A complete list of quasar candidates in the VMC survey.σ

References

  1. Appenzeller, I., Fricke, K., Fürtig, W., et al. 1998, The Messenger, 94, 1 [NASA ADS] [Google Scholar]
  2. Blanco, V. M., & Heathcote, S. 1986, PASP, 98, 635 [NASA ADS] [CrossRef] [Google Scholar]
  3. Carnerero, M. I., Raiteri, C. M., Rimoldini, L., et al. 2023, A&A, 674, A24 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  4. Cioni, M.-R. L., Clementini, G., Girardi, L., et al. 2011, A&A, 527, A116 [CrossRef] [EDP Sciences] [Google Scholar]
  5. Cioni, M.-R. L., Kamath, D., Rubele, S., et al. 2013, A&A, 549, A29 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  6. Crause, L. A., Carter, D., Daniels, A., et al. 2016, SPIE Conf. Ser., 9908, 990827 [NASA ADS] [Google Scholar]
  7. Crause, L. A., Gilbank, D., Gend, C. V., et al. 2019, J. Astron. Telescopes Instrum Syst., 5, 024007 [NASA ADS] [Google Scholar]
  8. Cross, N. J. G., Collins, R. S., Mann, R. G., et al. 2012, A&A, 548, A119 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  9. Dalton, G. B., Caldwell, M., Ward, A. K., et al. 2006, SPIE Conf. Ser, 6269, 30 [Google Scholar]
  10. de Grijs, R., & Bono, G. 2015, AJ, 149, 179 [Google Scholar]
  11. DiPompeo, M. A., Bovy, J., Myers, A. D., & Lang, D. 2015, MNRAS, 452, 3124 [NASA ADS] [CrossRef] [Google Scholar]
  12. Dobrzycki, A., Groot, P. J., Macri, L. M., & Stanek, K. Z. 2002, ApJ, 569, L15 [NASA ADS] [CrossRef] [Google Scholar]
  13. Dobrzycki, A., Macri, L. M., Stanek, K. Z., & Groot, P. J. 2003a, AJ, 125, 1330 [NASA ADS] [CrossRef] [Google Scholar]
  14. Dobrzycki, A., Stanek, K. Z., Macri, L. M., & Groot, P. J. 2003b, AJ, 126, 734 [NASA ADS] [CrossRef] [Google Scholar]
  15. Dobrzycki, A., Eyer, L., Stanek, K. Z., & Macri, L. M. 2005, A&A, 442, 495 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  16. Emerson, J. P., Irwin, M. J., Lewis, J., et al. 2004, SPIE Conf. Ser., 5493, 401 [Google Scholar]
  17. Emerson, J., McPherson, A., & Sutherland, W. 2006, The Messenger, 126, 41 [NASA ADS] [Google Scholar]
  18. Flesch, E. W. 2023, Open J. Astrophys., 6, 49 [NASA ADS] [CrossRef] [Google Scholar]
  19. Flesch, E. W. 2024, Open J. Astrophys., 7, 6 [NASA ADS] [CrossRef] [Google Scholar]
  20. Gaia Collaboration 2022, VizieR Online Data Catalog: I/355 [Google Scholar]
  21. Geha, M., Alcock, C., Allsman, R. A., et al. 2003, AJ, 125, 1 [NASA ADS] [CrossRef] [Google Scholar]
  22. González-Fernández, C., Hodgkin, S. T., Irwin, M. J., et al. 2018, MNRAS, 474, 5459 [Google Scholar]
  23. Gregg, M. D., Becker, R. H., White, R. L., et al. 1996, AJ, 112, 407 [NASA ADS] [CrossRef] [Google Scholar]
  24. Hamuy, M., Walker, A. R., Suntzeff, N. B., et al. 1992, PASP, 104, 533 [NASA ADS] [CrossRef] [Google Scholar]
  25. Hamuy, M., Suntzeff, N. B., Heathcote, S. R., et al. 1994, PASP, 106, 566 [NASA ADS] [CrossRef] [Google Scholar]
  26. Hook, I. M., McMahon, R. G., Boyle, B. J., & Irwin, M. J. 1994, MNRAS, 268, 305 [Google Scholar]
  27. Irwin, M. J., Lewis, J., Hodgkin, S., et al. 2004, SPIE Conf. Ser., 5493, 411 [Google Scholar]
  28. Ivanov, V. D., Cioni, M.-R. L., Bekki, K., et al. 2016, A&A, 588, A93 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  29. Jarvis, M. J., Bonfield, D. G., Bruce, V. A., et al. 2013, MNRAS, 428, 1281 [Google Scholar]
  30. Kelly, B. C., Bechtold, J., & Siemiginowska, A. 2009, ApJ, 698, 895 [Google Scholar]
  31. Kim, D.-W., Protopapas, P., Trichas, M., et al. 2012, ApJ, 747, 107 [NASA ADS] [CrossRef] [Google Scholar]
  32. Kozłowski, S., & Kochanek, C. S. 2009, ApJ, 701, 508 [CrossRef] [Google Scholar]
  33. Kozłowski, S., Kochanek, C. S., & Udalski, A. 2011, ApJS, 194, 22 [CrossRef] [Google Scholar]
  34. Kozłowski, S., Kochanek, C. S., Jacyszyn, A. M., et al. 2012, ApJ, 746, 27 [CrossRef] [Google Scholar]
  35. Kozłowski, S., Onken, C. A., Kochanek, C. S., et al. 2013, ApJ, 775, 92 [CrossRef] [Google Scholar]
  36. Lacy, M., Storrie-Lombardi, L. J., Sajina, A., et al. 2004, ApJS, 154, 166 [Google Scholar]
  37. Maitra, C., Haberl, F., Ivanov, V. D., Cioni, M.-R. L., & van Loon, J. T. 2019, A&A, 622, A29 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  38. McGowan, K. E., Coe, M. J., Schurch, M. P. E., et al. 2008, MNRAS, 383, 330 [NASA ADS] [CrossRef] [Google Scholar]
  39. McMahon, R. G., Banerji, M., Gonzalez, E., et al. 2013, The Messenger, 154, 35 [NASA ADS] [Google Scholar]
  40. Merloni, A., Lamer, G., Liu, T., et al. 2024, A&A, 682, A34 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  41. Minniti, D., Lucas, P. W., Emerson, J. P., et al. 2010, New A, 15, 433 [Google Scholar]
  42. Moehler, S., Modigliani, A., Freudling, W., et al. 2014a, The Messenger, 158, 16 [NASA ADS] [Google Scholar]
  43. Moehler, S., Modigliani, A., Freudling, W., et al. 2014b, A&A, 568, A9 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  44. Morrissey, P., Conrow, T., Barlow, T. A., et al. 2007, ApJS, 173, 682 [Google Scholar]
  45. Oke, J. B. 1990, AJ, 99, 1621 [Google Scholar]
  46. Onken, C. A., Wolf, C., Hon, W. J., et al. 2023, PASA, 40, e010 [NASA ADS] [CrossRef] [Google Scholar]
  47. Pennock, C. M., van Loon, J. T., Filipović, M. D., et al. 2021, MNRAS, 506, 3540 [NASA ADS] [CrossRef] [Google Scholar]
  48. Pennock, C. M., van Loon, J. T., Anih, J. O., et al. 2022, MNRAS, 515, 6046 [CrossRef] [Google Scholar]
  49. Predehl, P., Andritschke, R., Arefiev, V., et al. 2021, A&A, 647, A1 [EDP Sciences] [Google Scholar]
  50. Sajina, A., Lacy, M., & Pope, A. 2022, Universe, 8, 356 [NASA ADS] [CrossRef] [Google Scholar]
  51. Shanks, T., Georgantopoulos, I., Stewart, G. C., et al. 1991, Nature, 353, 315 [NASA ADS] [CrossRef] [Google Scholar]
  52. Spiniello, C., & Agnello, A. 2019, A&A, 630, A146 [EDP Sciences] [Google Scholar]
  53. Stoehr, F., White, R., Smith, M., et al. 2008, in ASP Conf. Ser., 394, Astronomical Data Analysis Software and Systems XVII, eds. R. W. Argyle, P. S. Bunclark, & J. R. Lewis, 505 [Google Scholar]
  54. Storey-Fisher, K., Hogg, D. W., Rix, H.-W., et al. 2024, ApJ, 964, 69 [NASA ADS] [CrossRef] [Google Scholar]
  55. Sturm, R., Haberl, F., Pietsch, W., et al. 2013, A&A, 558, A3 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  56. Treiber, H. P., Hinkle, J. T., Fausnaugh, M. M., et al. 2023, MNRAS, 525, 5795 [CrossRef] [Google Scholar]
  57. Tuccillo, D., González-Serrano, J. I., & Benn, C. R. 2015, MNRAS, 449, 2818 [NASA ADS] [CrossRef] [Google Scholar]
  58. Udalski, A., Szymanski, M., Kaluzny, J., Kubiak, M., & Mateo, M. 1992, Acta Astron., 42, 253 [NASA ADS] [Google Scholar]
  59. van Loon, J. T., & Sansom, A. E. 2015, MNRAS, 453, 2341 [Google Scholar]
  60. Vanden Berk, D. E., Richards, G. T., Bauer, A., et al. 2001, AJ, 122, 549 [Google Scholar]
  61. Véron-Cetty, M.-P., & Véron, P. 2010, A&A, 518, A10 [NASA ADS] [CrossRef] [EDP Sciences] [Google Scholar]
  62. Waddell, S. G. H., Nandra, K., Buchner, J., et al. 2024, A&A, in press, https://doi.org/10.1051/0004-6361/202245572 [Google Scholar]
1

Tiles are contiguous images that combine six pawprints taken in an offset pattern; a pawprint is an image obtained on an individual VIRCAM pointing that generates an image with gaps between the detectors. See Cioni et al. (2011) for a description of the VMC observing strategy.

6

The Image Reduction and Analysis Facility is distributed by the National Optical Astronomy Observatory, which is operated by the Association of Universities for Research in Astronomy (AURA) under a cooperative agreement with the National Science Foundation.

All Tables

Table 1

VMC coordinates and photometry of QSO candidates (in order of increasing right ascension).

In the text
Table 2

Log of the VLT spectroscopic observations.

In the text
Table 3

Log of the 1.9 m SAAO spectroscopic observations.

In the text
Table 4

Derived parameters for the objects in this paper.

In the text
Table B.1

A complete list of quasar candidates in the VMC survey.σ

In the text

All Figures

thumbnail Fig. 1

Color–color diagram showing the color selection of candidates. The dashed lines identify regions (labeled with letters) with known QSOs, and the solid magenta line marks the blue border of the planetary nebulae locus (Cioni et al. 2013). The confirmed QSOs from this work are plotted as red open circles. The open green circles and triangles mark QSOs and non-QSOs from Ivanov et al. (2016), respectively. The blue crosses indicate VMC counterparts to the spectroscopically confirmed QSOs from Kozłowski et al. (2013, VMC photometry selected within a matching radius of 1″). The black dots are randomly drawn LMC objects (with errors in all three bands <0.1 mag) to mark the main stellar locus; some of those in regions B and C are contaminating background galaxies.
In the text
thumbnail Fig. 2

Histogram of the absolute slopes from the linear fit to the light curves of the objects in this paper: classified as QSOs (wide bins), remaining unknown because of poor-quality spectra or lack of prominent emission lines (medium-width bins) and confirmed stars (narrow bins). The vertical dotted line is the QSO selection limit defined by Cioni et al. (2013).
In the text
thumbnail Fig. 3

Location of the objects with spectroscopic follow-up in this work (red) and confirmed QSOs from Kozłowski et al. (2013, blue) and Ivanov et al. (2016, green). The VMC tiles are shown as rectangles. The grid shows lines of constant right ascension and declination (spaced by 15º and 5°, respectively). The coordinates are offset with respect to (α0, δ0) = (51º, −69º).
In the text
thumbnail Fig. 4

Spectra of 148 objects (for 3 objects, no spectra could be extracted, usually due to poor weather conditions) sorted by redshift and shifted to the rest-frame wavelength. The spectra were normalized to an average value of one and shifted vertically by offsets of two, four, etc., for display purposes. The SDSS composite QSO spectrum (Vanden Berk et al. 2001) is shown at the top of all panels. Objects without a measured redshift due to a lack of lines or low S/N are plotted assuming z=0 in the fifth panel next to the sky spectrum to facilitate identifying the residuals from the sky emission lines.
In the text
thumbnail Fig. 5

Redshift histogram for 136 confirmed QSOs (solid line). For comparison, we plot the 47 objects with measured redshiſt from Ivanov et al. (2016, dashed line).
In the text
thumbnail Fig. 6

Parameters of the quasar sample. Top: luminosity functions of quasars from Quaia and the VMC survey in G band. All Quaia quasars (1.3 million objects; blue), Quaia quasars with VMC counterparts with quasar-like colors (2347 objects; orange); spectroscopically confirmed VMC quasars from this paper and from Ivanov et al. (2016) with Gaia counterparts (161 objects; red); the same as the red sample, but without the 5 objects from the bright SAAO sample and those that fall below the Quaia G=20.5 mag limit (green), projected output for a VMC survey full quasar sample (450 objects; black dots), and confirmed Quaia quasars scaled down from the full sky to the VMC survey area (~6300 objects; cyan dots), assuming a loss of 10% in the zone of avoidance. In reality, 7386 Quaia quasars fall in the VMC survey footprint, bottom: redshiſt distribution for all Quaia quasars (1.3 million objects; blue), for Quaia quasars with VMC counterparts with quasar-like colors (2347 objects; orange), and Quaia quasars with spectroscopically confirmed quasars from our VMC selection (161 objects; green). See Sect. 4 for details.
In the text
thumbnail Fig. 7

Comparison of the redshifts from Quaia and from our VLT spectra (top) and absolute values of the differences as function of redshiſt in units of sigma (bottom). The labels above each dashed line indicate the numbers of quasars above these lines and the corresponding fractions in percentages. See Sect. 4 for details.
In the text
thumbnail Fig. 8

Left: color-color diagram of all spectroscopically confirmed QSOs with VISTA/VIRCAM photometry. For the notation of the regions, see Fig. 1. Right: Y–J and J–KS colors as a function of redshift for all spectroscopically confirmed QSOs with VISTA/VIRCAM photometry. The variations are driven by strong emission lines that enter and exit the band passes of individual filters.
In the text
thumbnail Fig. A.1

Finding charts (Y band, 1′×1′) for all 151 objects (crosses) with follow-up spectroscopy, sorted by right ascension. North is at the top, and east is to the left.
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.