The discovery of tidal tails around the globular cluster NGC 7492 with Pan-STARRS1

TL;DR

This paper reports the discovery of extensive tidal tails around the globular cluster NGC 7492 using Pan-STARRS1 data, revealing its interaction with the Galactic environment.

Contribution

First detection of tidal tails around NGC 7492 using Pan-STARRS1 data and a matched filter technique on color-magnitude diagrams.

Findings

Tidal tails extend at least 3.5 degrees (~1.5 kpc) from the cluster.

Tails are aligned NE to SE, indicating ongoing tidal disruption.

Detection demonstrates the effectiveness of the matched filter method with Pan-STARRS1 data.

Abstract

We report the discovery of tidal tails around the Galactic globular cluster NGC 7492, based on the Data Release 1 of the Pan-STARRS 1 survey. The tails were detected with a version of the matched filter technique applied to the $(g-r,r)$ and $(g-i,i)$ color-magnitude diagrams. Tidal tails emerging from the cluster extend at least $\sim$3.5 degrees in the North-East to South-East direction, equivalent to $\sim1.5$ kpc in projected length.

Figures

Full text

Galaxy: halo — Galaxy: formation — globular clusters: individual: NGC 7492

1 Introduction

Globular clusters with tidal tails are precious for three reasons. First, there exists a coupling between the strength of the tides experienced by the cluster and its internal dynamics (see e.g. Gnedin et al., 1999). Therefore, through simultaneous modelling of the bound and unbound stellar mass distributions, one can learn the details of the satellite’s evolution in the Galaxy (see e.g. Dehnen et al., 2004). Second, stellar tails grow approximately along the cluster’s orbit (e.g. Eyre & Binney, 2011), thus providing a powerful technique to infer the properties of the host potential itself (e.g. Koposov et al., 2010; Bowden et al., 2015; Küpper et al., 2015). Third, globular cluster tails are fragile enough to be easily perturbed by low-mass objects, for example, dark matter subhalos with masses below $10^{8}$M*⊙* (e.g. Yoon et al., 2011; Erkal & Belokurov, 2015b). Thus globular cluster streams are a unique tool for measuring the lumpiness of the Galactic dark matter distribution (Erkal & Belokurov, 2015a; Erkal et al., 2016).

A large number of the Milky Way globulars are predicted to be undergoing destruction today (see Gnedin & Ostriker, 1997), however so far, tails have been detected only around a handful of objects (e.g. Odenkirchen et al., 2001; Belokurov et al., 2006; Niederste-Ostholt et al., 2010; Sollima et al., 2011; Lauchner et al., 2006). This discrepancy could perhaps be alleviated by taking into account the effects of mass segregation: the stars shed by the cluster in the initial phases of the dissolution are simply too low-mass to light up the tails (see Balbinot & Gieles, 2017). Additionally, this hypothesis may help to explain a substantial number of orphan stellar streams revealed so far (see e.g Grillmair & Dionatos, 2006b; Bonaca et al., 2012; Koposov et al., 2014; Balbinot et al., 2016). Most importantly, it serves to re-invigorate the search for tidal tails around globular clusters, albeit at lower surface brightness levels.

The absolute majority of the discoveries mentioned above have been made using the data from the Sloan Digital Sky Survey (SDSS), which by now has been trawled extensively for the Galactic stellar halo sub-structure. Motivated by the prospects of unearthing new examples of tidal disruption, we have searched for the presence of stellar tails around globular clusters located within the footprint of the Pan-STARRS1 (PS1) 3$\pi$ survey. Recently, after years of anticipation, the PS1 object catalogues were finally released publicly and are currently accessible for download through the MAST archive (see Flewelling et al., 2016). Compared to the SDSS, PS1 i) provides continuous coverage at low Galactic latitudes in the Northern hemisphere and ii) extends $\sim$30*∘* further down under the celestial equator. Note that ours is not the first exploration of the Galactic stellar halo sub-structure with PS1 (see e.g. Bernard et al., 2014, 2016).

In this Letter, we report the discovery of tidal tails around NGC 7492, a sparse outer halo Galactic globular cluster (R${}_{\rm GC}\sim$ 25 kpc; Harris 1996, 2010 edition; see also Figuera Jaimes et al. 2013). Previous studies of the cluster suggested the presence of extra-tidal stellar material. For example, Leon et al. (2000) found a small extension pointing towards the Galactic center, but emphasized the need for high-quality deep CCD data to confirm this low surface-brightness structure unambiguously. Later, studying a field $42\arcmin\times 42\arcmin$ in size, Lee et al. (2004) noticed tail-like structures extending towards the northeast and northwest from the cluster. Given its mass and Galacto-centric distance, NGC 7492 was also tagged as a “tidally-affected” cluster by Carballo-Bello et al. (2012). Other clusters included in this category are already known to have tidal tails, such as Pal 5 (Rockosi et al., 2002; Odenkirchen et al., 2003; Grillmair & Dionatos, 2006a), NGC 5466 (Belokurov et al., 2006), and NGC 5053 (Lauchner et al., 2006).

2 PS1 data

The Panoramic Survey Telescope and Rapid Response System (Pan-STARRS, PS1) survey is observing the sky north of declination $-30\arcdeg$ with a 1.8m optical telescope from the summit of Haleakala, in Hawaii. The telescope has a 7 deg2 field of view, imaged with a mosaic CCD camera with 60 detectors, each with 4800 $\times$ 4800 pixels. The pixel scale is 0.258 arcsec. The images are obtained through five filters: $gP1$, $rP1$, $iP1$, $zP1$ and $yP1$. The first data release (DR1) contains object catalogues produced using both non-stacked and stacked catalogues of the 3$\pi$ Steradian Survey. Details of the data delivered in PS1 DR1 are described in detail by Chambers et al. (2016); Magnier et al. (2016a, b). In the analysis presented here, stars were selected from the PS1 stack detection catalogue considering only sources with (rpsfmag-rkronmag) $<0.05$ (i.e. most likely stellar objects, see Farrow et al., 2014) and $g$ and $r$ STACK_PRIMARY flag-set greater than zero. The faintest magnitude considered was $r=$ 22.5 mag. The apparent magnitudes were corrected for extinction using the dust maps of Schlegel et al. (1998) as calibrated by Schlafly & Finkbeiner (2011).

The left panel of Figure 1 shows the Color-Magnitude diagram (CMD) of the stars located inside the tidal radius of the cluster ($r_{t}\sim 8.3\arcmin$, Harris, 1996). To remove stars suffering from crowding, only objects outside the very center (i.e, $r>1.5\arcmin$) are included. Despite contamination from field stars, the Main Sequence (MS), sub-giant branch, Red Giant Branch (RGB) and Blue Horizontal Branch (BHB) are clearly distinguishable. The CMD of the cluster also reveals the presence of a small number of stars brighter than the MS turn-off (TO), most probably blue straggler stars, as previously claimed by Cote et al. (1991). The RGB appears somewhat broadened: this scatter is difficult to explain given the current constraints on chemical abundance variations in the cluster, but very few giants from the cluster have spectroscopic studies (see Cohen & Melendez, 2005, and references therein).

The three rightmost panels of Figure 1 show the Hess diagrams (stellar density in color-magnitude space) for a field centered on the cluster (middle left panel), for an outer annulus (middle right), and their difference (right panel). The Hess diagram for the cluster field gives a clearer view of the satellite’s stellar populations: an extended BHB at $-0.6<(g-i)<-0.3$ and $i\sim 18$ mag; the MS populated mainly by stars fainter than $i=20$ mag; and the TO point at $(g-i)=0.4$ and $i\sim 20$ mag. While overdensities of stars around these evolution stages are clearly discernible, there is substantial contamination at the faintest and reddest magnitudes, as well as in the RGB region. The decontaminated Hess difference (right panel of the Figure) allows to peel away the foreground layer and thus makes the cluster’s stellar sequence much more evident. Here, reassuringly, the satellite’s MS appears much tidier and even hints of the asymptotic-giant branch are noticeable at magnitudes brighter than $i\sim 17$ mag.

Note that the cluster location on the sky at ($\alpha$, $\delta$) = ($347.1{\arcdeg}$, $-15.6{\arcdeg}$) (J2000), or $l=53\fdg 38$, $b=-63\fdg 48$, NGC 7492 lies within the projected position of the Sagittarius (Sgr) trailing stream as pointed out by Carballo-Bello et al. (2014). Curiously, among all Galactic globular clusters in the vicinity of the stream, it has the lowest probability to be associated with the dwarf according to the model of Law & Majewski (2010). Nonetheless, we conjecture that a good fraction of the field contamination around the NGC 7492 is supplied by the Sgr debris, given that both systems are at similar heliocentric distances and located relatively close on the sky (see Figure 17 from Carballo-Bello et al., 2014).

3 Detection of the tidal tails and discussion

To search for extra-tidal extensions to the cluster’s light distribution, we broadly follow the matched filter methodology as described in Rockosi et al. (2002). Taking advantage of the PS1 multi-band photometry we generate matched filter masks for both $(g-r,r)$, and $(g-i,i)$ CMDs. Only stars with $15.0<i,r<22.5$ were considered. These magnitude cuts avoid possible saturated stars and the faint stars with larger photometric errors. The color-magnitude distribution of the matched filter selected stars was constructed using the background-subtracted Hess diagrams for $(g-i,i)$ (rightmost panel of Figure 1) and the equivalent $(g-r,r)$ Hess subtracted diagram. For the foreground/background color-magnitude distribution, an adjacent area of the sky was used, excluding a window in right ascension around NGC 7492. The background CMD was constructed from four regions, defined at the boundaries: $339\arcdeg<$ R.A. $<342\arcdeg$, $342\arcdeg<$ R.A. $<345\arcdeg$, $349\fdg 25<$ R.A. $<352\fdg 25$, $352\fdg 25<$ R.A. $<355\fdg 25$, and between $-20\arcdeg<$ Dec. $<-10\arcdeg$. The Hess diagrams of the four fields were averaged to obtain a mean background density of stars in the color-magnitude plane.

The ratio of the cluster’s and the Galactic background’s CMD densities (with the pixel size of 0.12$\times$0.2 mag) was used to define the selection region in the $(g-r,r)$ and $(g-i,i)$ planes independently. Any star can then be assigned a weight corresponding to the value of the Hess ratio of the CMD pixel it falls into. For instance, BHB stars tend to have higher weights since the field color-magnitude distribution is less populated in that region. The weight assigned was the sum of the weights in the $(g-r,r)$ and $(g-i,i)$ planes. Using these two matched filters (or three photometric bands) simultaneously improves the signal of the stream significantly. It helps to reduce the impact of small fluctuations in the CMD of the cluster, which is sparsely populated to begin with, and additionally suffers from contamination from field stars even in the cluster innermost area (see Figure 1). Once the weight was assigned, only stars with weights above a certain threshold were considered as most likely cluster members. The weight threshold is determined by maximising the signal-to-noise of the cluster itself. This is the same approach as used by Erkal et al. (2016). The red contour in the left panel of Figure 1 shows the CMD bins where the matched filter selected stars, from Figure 2, fall into. Despite some possible contamination from stars belonging to the Sgr stream, the stars selected seems to follow the MS and RGB of the cluster.

Figure 2 shows the density of the matched filter selected stars in the 4*∘x4∘* (1.8 $\times$ 1.8 kpc) area centered on the cluster. The map was smoothed using a Gaussian filter with $\sigma=0\fdg 12$. The innermost circle marks the tidal radius of the cluster and the contour line confines the pixels with values at 1 $\sigma$ above the background. As evidenced from the Figure, the matched filter reveals two tails on either side of the cluster, extending over $\sim 3.5\arcdeg$ ($\sim 1.5$ kpc length) in the North-South direction. This structure resembles the characteristic “S-shaped” tidal feature found around other disrupting globular clusters, such as Pal 5 and NGC 5466. The contours drawn by Lee et al. (2004) are consistent with the northern branch and the small lobe found at $(\Delta\alpha\cos{\delta},\Delta\delta)\sim(-0\fdg 7,-0\fdg 1)$. Our results extend the northern branch on a factor of $\sim 5$ in length and also recover the southern branch, which was not seen previously. Given that there are no previous measurements of proper motions for this cluster, nor measurements in the Tycho–Gaia Astrometric Solution (TGAS Michalik et al., 2015; Lindegren et al., 2016) catalogue, it is not possible to compare the direction of the stream with the projection of the orbit of the cluster, but in principle, both should be aligned. The white dashed line in Fig. 2 marks the direction of the Sgr stream stars, according to the model of Law & Majewski (2010). Despite the cluster is projected onto the expected location of the Sgr tidal stream, the dwarf’s leading arm’s direction is misaligned with the tidal tails detected in this work, thus making the possibility of an association the two rather tenuous.

As a final check, we investigate whether some of the features seen in Figure 2 could be caused by misclassified galaxies and/or effects of the interstellar extinction. The left panel of Figure 3 shows the logarithmic spatial distribution of galaxies (i.e. sources with (rpsfmag-rkronmag) $\geq 0.05$) around NGC 7492, selected based on the same matching filters as in the stellar density map and using the same bin size and smoothing kernel as in Figure 2. To guide the eye, the tidal radius of the cluster and the 1$\sigma$ contour of the matched filter density are over-plotted. There seems to be no direct correlation between the tails and the distribution of galaxies in the field. The variations in the extinction are unlikely to have affected our results as indicated in the right panel of Figure 3. Clearly, the overall reddening of the Galaxy in the area is rather low (ranging between 0.03 and 0.045 mag) and its variation across the field does not correlate with tidal tails detected. Figure 3 also gives the positions of the BHB candidate stars, shown as light blue circles. These were selected to lie within the blue polygon in the left panel of Figure 1. Despite their overall low number density, BHBs appear to follow the Northern tail, connecting its uppermost tip with NGC 7492 itself. On the other side of the cluster, however, there is a distinct lack of the BHB candidates. This perhaps is not surprising given the low number of BHB stars expected in a given stellar population. While the spatial distribution of the possible BHB candidates is suggestive, deeper wide imaging is required to confirm (and perhaps extend) the discovery presented here.

The Pan-STARRS1 Surveys (PS1) and the PS1 public science archive have been made possible through contributions by the Institute for Astronomy, the University of Hawaii, the Pan-STARRS Project Office, the Max-Planck Society and its participating institutes, the Max Planck Institute for Astronomy, Heidelberg and the Max Planck Institute for Extraterrestrial Physics, Garching, The Johns Hopkins University, Durham University, the University of Edinburgh, the Queen’s University Belfast, the Harvard Smithsonian Center for Astrophysics, the Las Cumbres Observatory Global Telescope Network Incorporated, the National Central University of Taiwan, the Space Telescope Science Institute, the National Aeronautics and Space Administration under Grant No. NNX08AR22G issued through the Planetary Science Division of the NASA Science Mission Directorate, the National Science Foundation Grant No. AST-1238877, the University of Maryland, Eotvos Lorand University (ELTE), the Los Alamos National Laboratory, and the Gordon and Betty Moore Foundation. The research leading to these results has received funding from the European Research Council under the European Union’s Seventh Framework Programme (FP/2007-2013) / ERC Grant Agreement n. 308024. This project is supported by CONICYT’s PCI program through grant DPI20140066. C.N. acknowledges support from CONICYT-PCHA grant Doctorado Nacional 2015-21151643. S.K. thanks Bernie Shiao for the assistance in retrieving the PS1 data. C.N. thanks Francisco Aros for his useful suggestion on the improvements of the figures.