Exploring the crowded central region of 10 Galactic globular clusters using EMCCDs

Exploring the crowded central region of 10 Galactic globular clusters using EMCCDs

Variable star searches and new discoveries1
Key Words.:
crowded fields – globular clusters – NGC 104, NGC 5139, NGC 5286, NGC 6093, NGC 6121, NGC 6541, NGC 6656, NGC 6681, NGC 6723, NGC 6752, variable stars – long-period irregular, semi regular, RR Lyrae, SX Phoenicis, cataclysmic, binary, charge-coupled device – EMCCD, L3CCD, lucky imaging, shift-and-add, tip-tilt

Abstract

Context:

Aims:Obtain time-series photometry of the very crowded central regions of Galactic globular clusters with better angular resolution than previously achieved with conventional CCDs on ground-based telescopes to complete, or improve, the census of the variable star population in those stellar systems.

Methods:Images were taken using the Danish 1.54-m Telescope at the ESO observatory at La Silla in Chile. The telescope was equipped with an electron-multiplying CCD and the short-exposure-time images obtained (10 images per second) were stacked using the shift-and-add technique to produce the normal-exposure-time images (minutes). Photometry was performed via difference image analysis. Automatic detection of variable stars in the field was attempted.

Results:The light curves of 12541 stars in the cores of 10 globular clusters were statistically analysed in order to automatically extract the variable stars. We obtained light curves for 31 previously known variable stars (3 L, 2 SR, 20 RR Lyrae, 1 SX Phe, 3 cataclysmic variables, 1 EW and 1 NC) and we discovered 30 new variables (16 L, 7 SR, 4 RR Lyrae, 1 SX Phe and 2 NC).

Conclusions:

1 Introduction

Globular cluster systems in the Milky Way are excellent laboratories for several fields in astronomy, from cosmology to stellar evolution. As old as our Galaxy, they are stellar fossils that enable astronomers to look back to earlier galaxy formation stages to have a better understanding of the stellar population that is forming the clusters.

Possibly the first official register of globular clusters started with the “Catalog of Nebulae and Star Clusters” made by Messier (1781) and followed by Herschel (1786). Several early studies (i. e. Barnard, 1909; Hertzsprung, 1915; Eddington, 1916; Shapley, 1916; Oort, 1924) of globular clusters can also be found in the literature.

The first variable star found in the field of a globular cluster was T Scorpii (Luther, 1860). Subsequent early studies of variable stars in globular clusters can also be found in the literature (see e. g. Davis, 1917; Bailey, 1918; Sawyer, 1931; Oosterhoff, 1938). Measurements of the variability of these stars were done either by eye or by using photographic plates. It was not until about 1980 (Martínez Roger et al., 1999) that deeper studies in the photometry of globular clusters were done thanks to both the use of CCDs and larger telescopes.

More recently, the implementation of new image analysis tools such as difference image analysis (DIA; Alard & Lupton, 1998), have enabled much more detailed and quantitative studies, but there are still several limiting constraints. The first is saturated stars. It is always important to balance the exposure time to prevent bright stars from saturating the CCD without losing signal-to-noise ratio in the fainter stars. The second difficulty is the crowded central regions. Sometimes it is complicated to measure properly the point spread function (PSF) of the stars in the central regions of globular clusters due to blending caused by the effect of atmospheric turbulence.

To overcome these limitations, we started a pilot study of the crowded central regions of globular clusters using electron-multiplying CCDs (EMCCDs) and the shift-and-add technique. Our globular cluster sample is chosen based mainly on the central concentration of stars while favouring clusters well known to have variable stars. We published the results of globular cluster NGC 6981 in Skottfelt et al. (2013) with the discovery of two new variables and five more globular clusters in Skottfelt et al. (2015a) where a total of 114 previously unknown variables were found. In this paper we report the results on 10 further globular clusters.

Sec. 2 describes the instrumentation, data obtained, pipeline used and the data reduction procedures. Secs. 3, 4 and 5 explain the calibration, the colour magnitude diagrams, and the strategy used to identify variable stars. In Sec. 6 we classify the variable stars. Sec. 7 describes the methodology followed to observe the globular clusters. Sec. 8 presents the results obtained in each globular cluster and Sec. 9 summarises the main conclusions.

2 Data and reductions

2.1 Telescope and instrument

Observations were taken using the 1.54 m Danish telescope at the ESO observatory at La Silla in Chile, which is located at an elevation of 2340 m at 704407.662W 291514.235S. The telescope is equipped with an Andor Technology iXon+897 EMCCD camera, which has a 512 512 array of 16 m pixels, a pixel scale of 0.09 per pixel and a total field of view of arcsec.

For the purpose of this research the camera was configured to work at a frame-rate of 10 Hz (this is 10 images per second) and an EM gain of photon. The camera is placed behind a dichroic mirror which works as a long-pass filter. Considering the mirror and the sensitivity of the camera it is possible to cover a wavelength range between 650 nm to 1050 nm, corresponding roughly to a combination of SDSS filters (Bessell, 2005). More details about the instrument can be found in Skottfelt et al. (2015b).

2.2 Observations

Observations were taken during 2013 and 2014 as part of an ongoing program at the Danish telescope that was implemented from April to September each year. Fig. 1 shows histograms of the number of observations per night per cluster. Data in the left panel corresponds to 2013 and data in the right panel, to 2014. As the time at the telescope was limited, we tried to observe all clusters with the same rate as far as it was possible by taking two observations each night. However, as it is shown in the histograms, it was not possible due to weather conditions or because the telescope was used to monitor microlensing events carried out by the MINDSTEp consortium as part of the program to characterise exoplanets.

It is worth commenting that, depending on the magnitude level of the horizontal branch of each globular cluster, observations with total exposure times between 6 to 10 minutes were produced (see Sec. 7). To do this the camera was continuously taking images (at the rate of 10 images per second) for the total exposure time desired and then all the images obtained during that particular exposure time were stacked to produce a single observation. That is, a 10 minutes exposure time observation is the result of stacking 6000 images that were continuously taken. The technique for image stacking is called “shift-and-add” and it is explained in Sec. 2.3.2.

Figure 1: Histograms with the number of observations per night for each globular cluster studied in this work. In the panels on the left is plotted data taken during 2013 and in the panels on the right, data taken during 2014.

2.3 EMCCD data reduction

EMCCDs also known as low light level charge-coupled devices (L3CCD) (see e. g. Smith et al., 2008; Jerram et al., 2001) are conventional CCDs that are equipped with an extended serial gain register in which each electron produced during the exposure time has a probability to create a second electron (avalanche multiplication) when it is passing through the extended gain register. The process in which the second electrons are created is called impact ionisation.

To do bias, flat-field and tip-tilt corrections, the procedures and algorithms described in Harpsøe et al. (2012) were used.

Tip-tilt correction

As described in Harpsøe et al. (2012), the tip-tilt correction for a sample of bias and flat-field corrected images with respect to a reference image is calculated by finding the global maximum of defined by Eqn. 1,

(1)

which is the cross-correlation image (see Fourier cross correlation theorem). The position of the maximum () is used to shift each of the images in the sample to the position of the stars in the reference image.

Image quality and stacking

The image quality is assessed by defining the quality index

(2)

which represents the ratio of the maximum to the sum of the values of the surrounding pixels. Once this index is calculated, images from a single observation are sorted from the one with the best quality (smallest ) to the one with the worst quality (largest ) and a ten-layer image frame is built by stacking images in the sequence: 1%, 2%, 5%, 10%, 20%, 50%, 90%, 98%, 99% and 100%. This means that the first-layer frame has stacked the top 1% of the best quality images, the second-layer frame has stacked the next 1% of the best quality images, the third-layer frame has stacked the next 3%, the fourth has stacked the next 5% and so on. In this study, we use the layers with the best FWHM over all the frames to build the reference image for each globular cluster and we stack the ten-layers of an observation into a single science image for the photometry (see also Skottfelt et al., 2015a).

2.4 Photometry

To extract the photometry in each of the images we used the DanDIA2 pipeline (Bramich, 2008; Bramich et al., 2013), which is based on difference image analysis (DIA).

As commented in Sec. 2.3.2, for each EMCCD observation, we produced a ten-layer calibrated image cube where each of the images in the cube are sorted by quality. DanDIA builds a high and high-resolution reference frame by selecting and combining the best quality images available in the cubes. This is done by detecting bright stars using DAOFIND (Stetson, 1987) and employing these to align the images using the triangulation technique described in Pál & Bakos (2006). Each of the reference frames used in the analysis of each globular cluster can be found throughout the paper (e. g. Fig. 6) and the mean PSF FWHM for each reference image is listed in Tab. 3. Positions and reference fluxes () in ADU/s for each star are calculated using the PSF photometry package ”STARFINDER” (Diolaiti et al., 2000). The pipeline implements a low-order polynomial degree for the spatial variation of the PSF model (a quadratic polynomial degree was sufficient in our case).

Once the reference image is built, all of the science frames are registered with the reference image by using again the Stetson (1987) and Pál & Bakos (2006) algorithms described before. Image subtraction then determines a spatially variable kernel, modelled as a discrete pixel array, that best matches the reference image to each science image. The photometric scale factor () used to scale the reference frame to each image is calculated as part of the kernel model (Bramich et al., 2015). Difference images are created by subtracting the convolved reference image from each science image. Finally, difference fluxes () in ADU/s for each star detected in the reference image are measured in each of the difference images by optimally scaling the point spread function model for the star to the difference image. The light curve for each star is the total flux () in ADU/s defined as

(3)

We transform this to instrumental magnitudes at each given time :

(4)

A detailed description of the procedures and techniques employed by the pipeline can be found in Bramich et al. (2011).

For a conventional CCD, the noise model used by DanDIA is represented by the pixel variances of the image at the pixel positions ,

(5)

where is the CCD readout noise (ADU), is the master flat-field image, is the CCD gain (/ADU) and is the image model. For an EMCCD, the noise model for a single exposure is different:

(6)

where is the electron-multiplying gain (photons/) and is the Pre Amp gain (/ADU).

If exposures are combined by summation, then the noise model for the combined image is given by Eqn. 7:

(7)

where is the image model of the combined image (see also Skottfelt et al., 2015a).

In Tab. 1, we illustrate the format of the electronic table with all fluxes and photometric measurements as they are available at the CDS3.

Cluster var Filter HJD
id (d) (mag) (mag) (mag) (ADU s) (ADU s) (ADU s) (ADU s)
NGC104 PC1-V12 I 2456472.93210 15.846 6.777 0.014 24971.015 9644.446 -30350.989 1425.567 5.5131
NGC104 PC1-V12 I 2456476.94184 15.729 6.659 0.012 24971.015 9644.446 -16234.626 1138.337 4.9447
NGC5139 V457 I 2456432.58154 15.735 6.649 0.007 20096.379 684.928 +8409.333 673.724 4.6618
NGC5139 V457 I 2456438.70588 15.832 6.746 0.005 20096.379 684.928 -312.498 432.554 4.5021
NGC5286 V37 I 2456445.53773 15.978 6.951 0.005 18451.916 1202.455 -8631.145 385.937 4.6006
NGC5286 V37 I 2456446.53053 15.744 6.717 0.003 18451.916 1202.455 +9806.700 232.129 4.6341
Table 1: Time-series I photometry for all known and new variables in the field of view covered in each globular cluster. The standard and instrumental magnitudes are listed in columns 5 and 6, respectively, corresponding to the cluster, variable star, filter, and epoch of mid-exposure listed in columns 1-4, respectively. The uncertainty on is listed in column 7, which also corresponds to the uncertainty on . For completeness, we also list the quantities , and from Eq. 3 in columns 8, 10 and 12, along with the uncertainties and in columns 9 and 11. This is an extract from the full table, which is available with the electronic version of the article at the CDS.

2.5 Astrometry and finding chart

To fit an astrometric model for the reference images, celestial coordinates available at the “ACS Globular Cluster Survey”4 (see Anderson et al., 2008) were uploaded for the field of the cluster through GAIA (Graphical Astronomy and Image Analysis Tool; Draper (2000)). An (,) shift was applied to all of the uploaded positions until they matched the stars in the fields. Stars lying outside the field of view and those without a clear match were removed, and the (x, y) shift was refined by minimising the squared coordinate residuals. The number of stars used in the matching process ranged from 108 stars to 1034 stars which guaranteed that the astrometric solutions applied to the reference images considered enough stars covering the whole field. The radial RMS scatter obtained in the residuals was 0.1 ( 1 pixel). The astrometrically calibrated reference images were used to produce a finding chart for each globular cluster on which we marked the positions and identifications of all variable stars studied in this work. Finally, a table with the equatorial J2000 celestial coordinates of all variables for each globular cluster is given (see e. g. Tab. 4).

3 Photometric calibration

The photometric transformation of instrumental magnitudes to the standard system was accomplished using information available in the ACS Globular Cluster Survey, which provides calibrated magnitudes for selected stars in the fields of 50 globular clusters extracted from images taken with the Hubble Space Telescope (HST) instruments ACS and WFPC.

By matching the positions of the stars in the field of the HST images with those in our reference images, we obtained photometric transformations for our 10 globular clusters, as shown in Fig. 2. The I magnitude obtained from the ACS (see Sirianni et al., 2005) is plotted versus the instrumental magnitude obtained in this study. The red lines are linear fits with slope unity yielding zero-points labelled in the title of each plot. is the number of stars used in the fit. The correlation coefficient was 0.999 in all cases. A transformation was derived for each globular cluster.

Figure 2: Standard I magnitude taken from the HST observations as a function of the instrumental magnitude. The red lines are the fits that best match the data and they are described by the equations in the titles of each plot. The correlation coefficient is 0.999 in all cases.

4 Colour magnitude diagrams

As our sample has data available for only one filter, we decided to build colour magnitude diagrams (CMD) (see Fig. 5) by using the information available from the HST images at the ACS Globular Cluster Survey. The data used correspond to the V and I photometry obtained in Sirianni et al. (2005). The CMD was useful in classifying the variable stars, especially those with poorly defined light curves such as long-period irregular variables and semi-regular variables, as well as corroborating cluster membership.

5 Variable star searches

During the reductions, the DanDIA pipeline produced a total of 12541 light curves for stars in the cores of the 10 globular clusters studied. We used three automatic (or semi-automatic) techniques to search for variable stars, as described in Secs. 5.1, 5.2 and 5.3.

5.1 Root mean square

Diagrams of root mean square (RMS) magnitude deviation against mean I magnitude (see top Fig. 4) were constructed for each globular cluster. In these diagrams we measure the photometric scatter for each star studied in this work, but also we measure the intrinsic variation of the variable stars over time, which gives them a higher RMS than the none variables. The classification is indicated by the colour as detailed in Tab. 2.

To select possible variable stars, we fit a polynomial model to the RMS values as a function of magnitude and flag all stars with RMS values greater than 3 times the model. All difference images obtained with DanDIA were blinked as well to corroborate the variation of the stars selected.

5.2 statistic

A detailed discussion can be found about the benefits of using the statistic to detect variable stars (Figuera Jaimes et al., 2013) and RR Lyrae with Blazhko effect (Arellano Ferro et al., 2012). The statistic is defined as

(8)

where is the number of data points for a given light curve and is the number of groups formed of time-consecutive residuals of the same sign from a constant-brightness light curve model (e. g. mean or median). The residuals to correspond to the th group of time-consecutive residuals of the same sign with corresponding uncertainties to . The statistic is larger in value for light curves with long runs of consecutive data points above or below the mean, which is the case for variable stars with periods longer than the typical photometric cadence.

Plots of versus mean I magnitude are given for each globular cluster (see bottom Fig. 4) and variable stars are plotted in colour.

To select variable stars, the same technique employed in 5.1 was used, but in this case the threshold was shifted between 3-10 times the model values depending on the distribution of the data in each globular cluster. All stars selected were inspected in the difference images in the same way as explained in Sec. 5.1.

5.3 Stacked Difference images

Based on the results obtained using the DanDIA pipeline, a stacked difference image was built for each globular cluster with the aim of detecting the difference fluxes that correspond to variable stars in the field of the reference image. The stacked image is the result of summing the absolute values of the difference images divided by the respective pixel uncertainty

(9)

where is the stacked image, is the th difference image, is the pixel uncertainty associated with each image and the indexes and correspond to pixel positions.

All of the variable star candidates obtained by using the RMS diagrams or the statistic explained in Secs. 5.1 and 5.2 were inspected visually in the stacked images to confirm or refute their variability. Difference images were also blinked to finally corroborate the variation of the stars selected.

6 Variable star classification

All variable stars found were plotted in the colour-magnitude diagram for the corresponding globular cluster, to have a better understanding of the nature of their variation and their cluster membership. A period search of the light curve variations was undertaken by using the string method (Lafler & Kinman, 1965) and by minimising the in a Fourier analysis fit. To decide the classification of the variable stars, we used the conventions defined in the General Catalogue of Variable Stars (Samus et al., 2009a) and by considering their position in the colour-magnitude diagrams and the periods found.

In Tab. 2, the classification, corresponding symbols and colours used in the plots throughout the paper are shown.

Type Id Point style Color
     Pulsating Variables
RR Lyrae (RRL) RR0 Filled circle Red
RR01 Filled circle Blue
RR1 Filled circle Green
Semi-regular SR Filled square Red
Long-period irregular L Filled square Purple
SX Phoenicis SX Phe Filled triangle Cyan
     Cataclysmic Variables
In general DN, Novae Filled five pointed star Green
     Eclipsing Variables
EW Filled five pointed star Blue
     Unclassified Variables
In general NC Filled square Yellow
Table 2: Convention used in the variable star classification of this work based on the definitions of the general catalogue of variable stars (Samus et al., 2009a).

7 Globular clusters observed

Fig. 3, shows the Galactic plane celestial coordinates (l, b) for the 157 globular clusters listed in (Harris, 1996, 2010 edition). The 10 globular clusters analysed in this work are plotted in red colour where 6 are in the Southern hemisphere and 4 are in the Northern hemisphere. Also plotted (in green) are the 6 globular clusters studied in Skottfelt et al. (2015a, 2013).

Figure 3: Positions of the 157 known globular clusters in the Galaxy are plotted in the Galactic plane system. The 10 globular clusters studied in this work are labelled in red. Green corresponds to the globular clusters studied in Skottfelt et al. (2015a, 2013).

The strategy to select and observe the clusters was the following:

- (1) We take into account the visibility. We selected all globular clusters visible based on the geographical coordinates of the telescope and some pointing restrictions too.

- (2) We focused on the concentration of stars in the central region of the clusters. We used as reference the central luminosity density (pc) listed in (Harris, 1996, 2010 edition). The entire sample available in the catalogue has from 0 to 6. To explore the potential of the EMCCD and the shift-and-add technique for photometry, a sample with different levels of were chosen from a very dense cluster NGC 6752 (=5.04) to a less dense cluster NGC 6723 (=2.79) and some intermediate levels as NGC 5139 (=3.15) and NGC 5286 (=4.10).

- (3) A bibliography review revealed how many variable stars are known from previous work. To do this we used the information available for each globular cluster in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001) and the ADS to do the bibliography review up-to-now.

- (4) Exploration of known colour-magnitude diagrams. As most of the variable stars that are globular cluster members have a particular position in these diagrams, we focused our attention on globular clusters with a high concentration of stars for example at the top of the red giant branch (for semi-regular variables), the instability strip of the horizontal branch (for RR Lyrae) and the blue straggler region (for SX Phoenicis). To do this we used the colour information available in the ACS to build colour-magnitude diagrams for each globular cluster. Additionally we also used the Galactic globular clusters database5.

- (5) The exposure time to be used for a single EMCCD observation for each globular cluster was chosen based on the V magnitude of the horizontal branch. We selected exposure times as follows:

6 min for ¡ 14 mag,

8 min for 14 mag ¡ ¡ 17 mag,

10 min for ¿ 17 mag.

We chose the 10 globular clusters presented in this work. Some of their most relevant physical properties are detailed in Tab. 3.

Section Cluster RA Dec [Fe/H] E(B-V) V c FWHM
NGC J2000 J2000 kpc kpc mag mag mag arcsec s
8.1 104 00:24:05.67 -72:04:52.6 4.5 7.4 -0.72 0.04 14.06 13.37 2.07 4.88 0.61 302.4
8.2 5139 13:26:47.24 -47:28:46.5 5.2 6.4 -1.53 0.12 14.51 13.94 1.31 3.15 0.38 312.0
8.3 5286 13:46:26.81 -51:22:27.3 11.7 8.9 -1.69 0.24 16.63 16.08 1.41 4.10 0.37 316.8
8.4 6093 16:17:02.41 -22:58:33.9 10.0 3.8 -1.75 0.18 16.10 15.56 1.68 4.79 0.45 302.4
8.5 6121 16:23:35.22 -26:31:32.7 2.2 5.9 -1.16 0.35 13.45 12.82 1.65 3.64 0.49 180.0
8.6 6541 18:08:02.36 -43:42:53.6 7.5 2.1 -1.81 0.14 15.35 14.82 1.86 4.65 0.56 340.8
8.7 6656 18:36:23.94 -23:54:17.1 3.2 4.9 -1.70 0.34 14.15 13.60 1.38 3.63 0.50 297.6
8.8 6681 18:43:12.76 -32:17:31.6 9.0 2.2 -1.62 0.07 15.55 14.99 2.50 5.82 0.49 312.0
8.9 6723 18:59:33.15 -36:37:56.1 8.7 2.6 -1.10 0.05 15.48 14.84 1.11 2.79 0.51 360.0
8.10 6752 19:10:52.11 -59:59:04.4 4.0 5.2 -1.54 0.04 13.70 13.13 2.50 5.04 0.59 198.0
Table 3: Some of the physical properties of the globular clusters studied in this work. Column 1 is the section with the individual results for each globular cluster. Column 2, is the name of the cluster as it is defined in the New General Catalogue, Columns 3 and 4 are the celestial coordinates (right ascension and declination), Column 5 is the distance from the Sun, Column 6 is distance from the Galactic centre, Column 7 is metallicity, Column 8 is reddening, Column 9 is V magnitude level of the horizontal branch, Column 10 is V distance modulus, Column 11 is King-model central concentration c = (r/r), Column 12 is central luminosity density , Column 13 is the mean full-width half-maximum (arcsec) measured in the reference image and Column 14 is the exposure time in the reference image.

8 Results

8.1 NGC 104 / C0021-723 / 47 Tucanae

This globular cluster was discovered by Nicholas Louis de Lacaille in 17516. The cluster is in the constellation of Tucana at a distance of 4.5 kpc from the Sun and 7.4 kpc from the Galactic centre. It has a metallicity of [Fe/H]=0.72 dex and a distance modulus of (mM)=13.37 mag. The magnitude of its horizontal branch is V=14.06 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
PC1-V12 00:24:05.921 -72:04:45.20 15.62 1.10 149 NC
WF2-V34 00:24:08.406 -72:04:35.91 10.98 0.03 139 L
LW10 00:24:02.490 -72:05:07.45 9.60 0.24 108 L
LW11 00:24:03.145 -72:04:50.60 2456920.7253 19.37(8) 10.18 0.08 141 SR
LW12 00:24:03.982 -72:05:10.06 9.57 0.09 41 L
EM1 00:24:03.065 -72:04:55.02 2456491.9380 20.28(9) 10.26 0.05 145 SR
EM2 00:24:06.269 -72:04:45.36 2456530.7986 31.50(22) 9.98 0.08 140 SR
EM3 00:24:08.310 -72:04:50.69 2456847.9667 33.41(24) 10.07 0.05 134 SR
EM4 00:24:07.203 -72:04:46.45 2456498.9086 68.07(102) 13.89 0.14 149 SR
EM5 00:24:05.087 -72:04:54.38 10.34 0.06 145 L
EM6 00:24:02.733 -72:05:02.80 10.60 0.07 147 L
EM7 00:24:07.885 -72:05:02.00 146 NC
Table 4: NGC 104: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Column 1 is the id assigned to the variable star, Columns 2 and 3 correspond to the right ascension and declination (J2000), Column 4 is the epoch used, Column 5 is the period, Column 6 is median of the data, Column 7 is the peak-to-peak amplitude in the light curve, Column 8 is the number of epochs and Column 9 is the classification of the variable. The numbers in parentheses indicate the uncertainty on the last decimal place of the period.

In Fig. 4, the root mean square magnitude deviation (top) and S statistic (bottom) are plotted versus the mean magnitude for a total of 575 light curves extracted in this analysis. The marked coloured points correspond to the variable stars studied in this work in comparison with the stars where no variation is found (normal black points).

Figure 4: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 575 stars detected in the field of view of the reference image for NGC 104. Coloured points follow the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.

Known variables

This globular cluster has of the order of 300 known variable sources in the Catalogue of Variable Stars in Galactic Globular Clusters (version of summer 2007; Clement et al., 2001) which include long-period irregular variables, SX Phoenicis, RR Lyrae and binary systems. Most were found by Albrow et al. (2001), Weldrake et al. (2004) and Lebzelter & Wood (2005). There are also 20 millisecond pulsars listed in the literature for this cluster (Freire et al., 2001). However, no visual counterparts have been found at their positions.

In the field of view covered by the reference image (Fig. 6) there are 49 previously known variables. However, for this cluster we were only able to detect stars brighter than =16.1 mag (see RMS in Fig. 4). Due to this, 5 known variable stars brighter than this limit were detected. These stars are labelled in the literature as PC1-V12, WF2-V34, LW10, LW11 and LW12. Our light curves for these variables can be found in Fig. 7.

PC1-V12: This star was discovered by Albrow et al. (2001) and classified as a Blue Straggler star (BSS) with an average magnitude of V=16.076 mag. No period was reported in this case and it was not possible to find a period for this star in the present work.

WF2-V34: The variability of this star was found by Albrow et al. (2001) and was classified as a semi-regular variable with a period of P=5.5 d. However, using this period it was not possible to produce a good phased light curve and it was not possible to find another period for this star. Based on this, we classified this star as long-period irregular variable.

LW10-LW12: These three stars were discovered by Lebzelter & Wood (2005). Their positions in the colour-magnitude diagram (Fig. 5) confirm their cluster membership. Lebzelter & Wood (2005) classified them as long-period irregular variables and suggested periods for these stars to be LW10: 110 d or 221 d; LW11: 36.0 d, and LW12 : 61 d or 116 d. However, LW10 and LW12 are clearly irregular variables (see our Fig. 7, and Fig. 1 from Lebzelter & Wood (2005)). For LW11 we found a period of 19.37 d and we classify it as semi-regular.

New variables

After extracting and analysing all possible variable sources in the field of our images, a total of 7 new variables (EM1-EM7) were found of which 4 are semi-regular variables, 2 are long-period irregular variables and 1 is unclassified.

For this cluster, the nomenclature employed for most of the variable stars discovered in previous works does not correspond to the typical numbering system (e. g. V1, V2, V3, …, Vn). For example, variables discovered by Lebzelter & Wood (2005) and Weldrake et al. (2004) are numbered using their initials LW and W, respectively; Albrow et al. (2001) used a PC or WF nomenclature making reference to the instrument employed during the observations. Due to this, we did not find it practical to assign the typical numbering system to the new variable stars discovered in this work and we decided to use as reference the EMCCD camera used in the observations. That is, new variables are numbered as EM1-EM7.

EM1-EM4: These variable stars are placed at the top of the colour magnitude diagram (Fig. 5), their light curve shape, amplitudes and periods found suggest that these stars are semi-regular variables. Periods found for these variables are listed in Tab. 4. The star EM1 is not over the red giant branch in Fig. 5 but instead it is more toward the asymptotic giant branch, it appears at the top on the left. Finally the star EM4 is located at the bottom part of the red giant branch just below the starting point of the horizontal branch.

EM5-EM6: These stars have amplitudes of 0.06 and 0.07 mag, respectively. They are located at the top of the red giant branch. We were not able to determine a proper period for these stars. Due to this, we classified them as long-period irregular variables.

EM7: This star is placed in the Blue Straggler region. The variation in the difference images is clear for this star. However, the pipeline did not find this star in the reference image and no reference flux is available to convert the difference fluxes to magnitudes. The difference fluxes are plotted in Fig. 7. We were unable to determine a period.

Figure 5: Colour magnitude diagram for the globular cluster NGC 104 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.
Figure 6: Finding chart for the globular cluster NGC 104. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec.
Figure 7: NGC 104: Light curves of the known and new variables discovered in this globular cluster. Red triangles correspond to the data obtained during the year 2013 and blue circles correspond to the data obtained during the year 2014. For EM7, we plot the quantity since a reference flux is not available.

8.2 NGC 5139 / C1323-472/ Omega Centauri

This globular cluster was discovered by Edmond Halley in 16777. The cluster is in the constellation of Centaurus at 5.2 kpc from the Sun and 6.4 kpc from the Galactic centre. It has a metallicity of [Fe/H]=-1.53 dex, a distance modulus of (m-M)=13.94 mag and the level of the horizontal branch is at V=14.51 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V457 13:26:46.246 -47:28:44.81 15.82 0.15 78 NC
V458 13:26:46.103 -47:28:57.05 10.21 0.04 78 L
V459 13:26:46.313 -47:28:40.33 10.98 0.06 78 L
Table 5: NGC 5139: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Columns are the same as in Tab. 4.
Figure 8: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 2616 stars detected in the field of view of the reference image for NGC 5139. Coloured points follow the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.

Known variables

In the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001) there are of the order of 400 variable stars for this globular cluster. Most of these variables were discovered by Bailey (1902), Kaluzny et al. (2004) and Weldrake et al. (2007). All of the known variables are outside the field of view of our reference image. Most recently, Navarrete et al. (2015) made an updated analysis of the variables in this cluster but the variables in their study are also located outside the field of view of our reference image.

New variables

We have found 3 new variables in this globular cluster. Two are long-period irregular variables and one is unclassified.

The new variable stars are plotted in the RMS diagram and statistic (see Fig. 8). Based on the position of the stars in Figs. 8 and 9 there is no evidence of RR Lyrae in the field covered by this work neither in the inspection of the difference images obtained in the reductions.

V457: This variable star has an amplitude of 0.15 mag with a median magnitude 15.82 mag. It lies on the RGB and we were unable to find a period.

V458, V459: Theses two stars are at the top of the red giant branch in Fig. 9 and have amplitudes of 0.04 and 0.06 mag, respectively. No periods were found in this work. V459 is on the red side of the main red giant branch. We classify them as long-period irregular variables.

Figure 9: Colour magnitude diagram of the globular cluster NGC 5139 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.
Figure 10: Finding chart for the globular cluster NGC 5139. The image used corresponds to the reference image constructed during the reduction. The new variables discovered are labelled. Image size is arcsec.
Figure 11: NGC 5139: Light curves of the 3 new variables discovered in this globular cluster. Symbols are the same as in Fig. 7.

8.3 NGC 5286 / C1343-511 / Caldwell 84

This globular cluster was discovered by James Dunlop in 1827 (O’Meara, 2002). It is in the constellation of Centaurus at 11.7 kpc from the Sun and 8.9 kpc from the Galactic centre. It has a metallicity of [Fe/H]=-1.69 dex, a distance modulus of (m-M)=16.08 mag and the magnitude level of the horizontal branch is at V=16.63 mag.

In Fig. 12, the RMS diagram and the statistic are shown for the sample of 1903 stars analysed in this globular cluster. Most of them have 74 epochs. All variable stars studied in this work are plotted using the colour classification given in Tab. 2.

Figure 12: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 1903 stars detected in the field of view of the reference image for NGC 5286. Coloured points follow the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.
Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V37 13:46:27.275 -51:22:51.63 2456847.4835 0.583886(85) 15.92 0.81 52 RR0
V39 13:46:25.458 -51:22:46.81 2456768.6823 0.742099(136) 15.57 0.36 74 RR0
V40 13:46:26.917 -51:22:44.21 2456455.5615 0.365961(33) 15.76 0.30 74 RR1
V41 13:46:26.577 -51:22:42.92 2456822.5630 0.322262(26) 15.88 0.42 74 RR1
V43 13:46:26.730 -51:22:38.08 2456447.5363 0.658478(107) 15.47 0.50 74 RR0
V46 13:46:27.470 -51:22:33.93 2456831.5372 0.682690(115) 15.52 0.56 74 RR0
V50 13:46:25.984 -51:22:30.32 2456457.5131 0.365145(33) 15.71 0.41 74 RR1
V55 13:46:25.403 -51:22:25.79 2456841.5096 0.288925(21) 15.91 0.30 74 RR1
V56 13:46:25.112 -51:22:24.34 2456783.6351 0.283202(20) 15.97 0.22 74 RR1
V57 13:46:27.785 -51:22:20.69 2456833.6352 0.294964(22) 15.75 0.40 74 RR1
V58 13:46:26.627 -51:22:19.29 2456463.5240 0.367064(33) 15.65 0.30 74 RR1
V63 13:46:25.405 -51:22:44.92 2456811.5275 0.0486320(10) 18.52 0.55 74 SXPhe
V64 13:46:26.780 -51:22:26.60 2456833.6900 0.369783(34) 15.72 0.41 74 RR1
V65 13:46:26.763 -51:22:28.93 2456461.4946 0.619491(95) 15.35 0.36 74 RR0
V66 13:46:27.166 -51:22:40.14 2456833.7152 17.55(08) 12.06 0.05 74 SR
V67 13:46:26.121 -51:22:23.76 2456822.5630 21.20(11) 12.29 0.04 74 SR
V68 13:46:26.190 -51:22:26.55 2456844.4857 32.95(27) 11.87 0.10 74 SR
V69 13:46:25.869 -51:22:16.18 12.26 0.05 74 L
V70 13:46:24.334 -51:22:47.14 11.57 0.07 63 L
V71 13:46:27.668 -51:22:35.25 11.66 0.09 74 L
V72 13:46:27.364 -51:22:16.22 11.64 0.14 74 L
V73 13:46:27.008 -51:22:29.88 11.51 0.52 74 L
Table 6: NGC 5286: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Columns are the same as in Tab. 4.

Known variables

This globular cluster has 58 known variable stars listed in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001) of which 52 are RR Lyrae stars. There are only 11 previously known variable stars in the field of view of our reference image (V37, V39, V40, V41, V43, V46, V50, V55, V56, V57, V58). All of them are RR Lyrae discovered by Zorotovic et al. (2010) using DIA on imaging data from a one-week observing run.

The celestial coordinates given in Tab. 1 of Zorotovic et al. (2010) do not match the positions of the known variable stars in the field of our images. We therefore used the finding chart given in their Fig. 1 to do a visual matching of the variables. As pointed out in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001), there is a difference in the position of the variables studied by Zorotovic et al. (2010) with respect to the position of the variables in Samus et al. (2009b) which is 6 arcseconds in declination and 1 arcsecond in right ascension. This difference is corroborated by the position of the variables in our reference image. Celestial coordinates of the positions we used are given in Tab. 6.

Our extended observational baseline has allowed us to greatly improve the periods of the variables discovered by Zorotovic et al. (2010). Our period estimates are listed in Tab. 6 and have typical errors of 0.00002 - 0.00010 d. We confirm the variable star classifications made by Zorotovic et al. (2010).

Note that V41 is a strong blend with a brighter star that is only just resolved in our high-resolution reference image.

New variables

In this globular cluster we found 11 previously unknown variables where 5 are long-period irregular variables, 3 are semi-regular variables, 2 are RR Lyrae, and 1 is a SX Phoenicis.

V63: The star has a median magnitude of 18.52 mag and it is in the blue straggler region. Its amplitude is 0.49 mag and it has a period of P=0.0486320 d. The star is clearly a SX Phoenicis. The variation of this star was not detected using the RMS or statistic, although it is very clear in the difference images.

V64: This star is an RR Lyrae pulsating in the first overtone (RR1) with a period of 0.369783 d and an amplitude of 0.41 mag. In Fig. 14, notice that V64 is very close to a bright star (6.808 pixels or 0.613 arcsec). This could be the reason why this variable was not discovered before and makes a good example of the benefits of using the EMCCD cameras and the shift-and-add technique along with DIA.

V65: This star is another RR Lyrae which is pulsating in the fundamental mode (RR0) with a period of 0.619491 d and an amplitude of 0.36 mag.

V66-V68: These stars are semi-regular variables. As it is seen in Fig. 13, they are at the top of the red giant branch. They have amplitudes that range between 0.04 to 0.10 mag. They have periods between 17 to 33 d. Ephemerides for these stars can be found in Tab. 6.

V69-V73: These five stars are also positioned at the top of the red giant branch with amplitudes of 0.05 to 0.52 mag. It was not possible to find periods for these stars in this work. Due to this, they were classified as long-period irregular variables.

In Fig. 16 the amplitude-period diagram for the RR Lyrae stars studied in this cluster is shown. The filled lines correspond to the Oosterhoff type I (OoI) and the dashed lines correspond to the Oosterhoff type II (OoII) models defined by Kunder et al. (2013). All RR0 variables (with exception of V65) fall on the model for OoII type while the RR1 stars scatter around both models. In this diagram the RR0 stars suggest an OoII type classification for NGC 5286 which is in agreement with the study done by Zorotovic et al. (2010) where they found that their research pointed to an OoII status as well.

Figure 13: Colour magnitude diagram for the globular cluster NGC 5286 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.
Figure 14: Finding chart for the globular cluster NGC 5286. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec. The image stamps are of size arcsec.
Figure 15: NGC 5286: Light curves of the known and new variables discovered in this globular cluster. Symbols are the same as in Fig. 7.
Figure 16: Amplitude-period diagram for the RR Lyrae stars in the globular cluster NGC 5286.

8.4 Ngc 6093 / C1614-228 / M 80

This globular cluster was discovered by Charles Messier in 17818. This cluster is in the constellation of Scorpius at 10.0 kpc from the Sun and 3.8 kpc from the Galactic centre. Its has a metallicity of [Fe/H]=-1.75 dex, a distance modulus of =15.56 mag and the position of its horizontal branch is at V=16.10 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V10 16:17:01.163 -22:58:34.57 2456805.5870 0.614154(92) 14.25 0.29 52 RR0
V17 16:17:02.864 -22:58:32.55 2456831.5535 0.415081(42) 15.31 0.31 62 RR1
V19 16:17:02.107 -22:58:29.10 2456844.5144 0.596064(87) 15.28 0.54 62 RR0
V20 16:17:03.263 -22:58:37.49 2456831.5535 0.745207(136) 15.31 0.46 62 RR0
V34 16:17:02.820 -22:58:33.80 16.44 0.36 62 CV(Nova)
V35 16:17:03.313 -22:58:33.15 11.40 0.09 62 L
V36 16:17:03.145 -22:58:41.92 11.50 0.12 62 L
V37 16:17:02.320 -22:58:30.48 11.51 0.09 62 L
V38 16:17:03.263 -22:58:34.96 11.70 0.04 62 L
V39 16:17:02.201 -22:58:34.49 11.75 0.07 62 L
V40 16:17:03.042 -22:58:25.72 12.36 0.13 62 L
Table 7: NGC 6093: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Columns are the same as in Tab. 4.

A total of 1220 light curves were extracted in the field covered by the reference image for the globular cluster NGC 6093. Most of the stars have 58 epochs in their light curves and variable stars studied in this work are plotted with colour in the RMS and diagrams shown in Fig. 17.

Figure 17: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 1220 stars detected in the field of view of the reference image for NGC 6093. Coloured points follow the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.

Known variables

This globular cluster has 34 known variable sources in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001). Only 7 are in the field of view covered by these observations. There is no visual detection of V11 in our images (discovered by Shara et al., 2005, and classified as a possible U Geminorum-type cataclysmic variable). This is to be expected since it has a U mean magnitude of 19.5 mag. The star V33 was discovered by Dieball et al. (2010) in an ultraviolet survey using the Hubble Space Telescope. They reported a UV magnitude for this target of about 19.8 mag. Again it was not possible to obtain a light curve for this faint target.

For the remaining variable stars in the field of view, three of them are RR0 (V10, V19 and V20), one is a RR1 (V17), and V34 is a Nova. Light curves for these stars are presented in Fig. 20. We have improved on the period estimates published by Kopacki (2013) for the four RR Lyrae variables since we observed them using a time baseline of over 1 year, whereas Kopacki (2013) analysed observations spanning only one week.

V34: This is the Nova discovered by Arthur von Auwers at Koenigsberg Observatory on May 21, 1860 (Luther, 1860). As pointed out in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001), an account of its discovery was given by Sawyer (1938) in which a maximum visual apparent magnitude of 6.8 mag was reported using the data taken by von Auwers and Luther. Another review can also be found in Wehlau et al. (1990). However, no observations of the Nova in outburst have been made until the present study. During our 2013 observing campaign, we caught an outburst of amplitude 0.36 mag starting at around HJD2456500 d which lasted for 50 days. In our 2014 data the Nova continued at its baseline magnitude at 16.44 mag.

Dieball et al. (2010) in their ultraviolet survey assigned the label CX01 to an X-ray source that was found to be associated with the position of the Nova at the coordinates RA(J2000)=16:17:02.817 and DEC(J2000)=-22:58:33.92. These coordinates match with the position of the outburst found in this work and details are given in Tab. 7.

The position in the colour magnitude diagram shown in Fig. 18 suggests that this system is a cluster member. It is located at the bottom part of the red giant branch. Its position is plotted with a green five pointed star.

As the Nova with its outburst has shown variability, we have assigned the label V34 to this object.

Figure 18: Colour magnitude diagram for the globular cluster NGC 6093 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.
Figure 19: Finding chart for the globular cluster NGC 6093. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec.

New variables

A total of 6 new variables were found in this work. All of them are long-period irregular variables.

V35-V40: These stars are long-period irregular variables. They are located at the top of the red giant branch (see Fig. 18). Their amplitudes go from 0.04 to 0.13 mag. The star V40 is the one placed at the blue side of the red giant branch. We did not find any kind of regular periodicity in the variation of these stars.

Figure 20: NGC 6093: Light curves of the known and new variables discovered in this globular cluster. Symbols are the same as in Fig. 7.

8.5 Ngc 6121 / C1620-264 / M4

This globular cluster was discovered by Philippe Loys de Chéseaux in 17469. This cluster is in the constellation of Scorpius at a distance of 2.2 kpc from the Sun and 5.9 kpc from the Galactic centre. It has a metallicity of [Fe/H]=-1.16 dex and a distance modulus of (m-M)=12.82 mag. Its horizontal branch level is at V=13.45 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V21 16:23:35.903 -26:31:33.68 2456477.5041 0.472008(56) 12.55 0.92 64 RR0
Table 8: NGC 6121: Ephemerides and main characteristics of one variable star in the field of this globular cluster. Columns are the same as in Tab. 4.

This globular cluster has of the order of 100 known variables in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001) and only three are in the field of view of the reference image; namely V21, V81 and V101. Furthermore, only V21 is bright enough to be detected. The light curve for this RR0 star is shown in Fig. 22. The known period P=0.4720 d produces a very well phased light curve and it is in agreement with the period P=0.472008 d found in the analysis of this star.

Figure 21: Finding chart for the globular cluster NGC 6121. The image used corresponds to the reference image constructed during the reduction. The only known variable in the field is labelled. Image size is arcsec.

No new variable stars were found in the field covered by the reference image for this globular cluster.

Figure 22: NGC 6121: Light curve of the variable V21 in this globular cluster. Symbols are the same as in Fig. 7.

8.6 Ngc 6541 / C1804-437

This globular cluster was discovered by N. Cacciatore in 182610. It is in the constellation of Corona Australis at 7.5 kpc from the Sun and 2.1 kpc from the Galactic centre. The cluster has a metallicity of [Fe/H]=-1.81 dex, a distance modulus of =14.82 mag and the level of the horizontal branch is at V=15.35 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V14 18:08:01.670 -43:42:52.26 2456908.7354 0.064669(1) 16.38 0.27 43 SXPhe
V21 18:08:02.180 -43:42:52.91 2456782.8812 0.596000(76) 14.39 0.59 43 RR0
V22 18:08:03.743 -43:42:34.46 10.74 0.17 29 L
Table 9: NGC 6541: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Columns are the same as in Tab. 4.

A total of 843 light curves were extracted in this globular cluster. Most of them have 42 epochs. RMS diagram and statistic are shown in Fig. 23.

Figure 23: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 843 stars detected in the field of view of the reference image for NGC 6541. Coloured points follow the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.

Known variables

This globular cluster has 20 known variables. Four of them are in the field of view of our reference image: V12, V14, V15, and V17. Fiorentino et al. (2014) discovered and classified these stars as SX Phoenicis in their study carried out using observations with the HST. We were only able to recover V14 in our data around 2 arcseconds in declination from the reported position for this star in Fiorentino et al. (2014). We did not detect variability at the positions given for V12, V15 and V17 neither in their surrounding areas, more likely due to the fact that the stars are too faint to be detected in our reference image. A reference image with higher signal to noise ratio might be needed in future studies of this globular cluster and longer exposure times for each observation.

V14: For this star we refined the period estimate to 0.064669 d from the previously listed value of 0.0649 d in Fiorentino et al. (2014).

Figure 24: Colour magnitude diagram for the globular cluster NGC 6541 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.
Figure 25: Finding chart for the globular cluster NGC 6541. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec.

New variables

A total of two new variable stars were found in this globular cluster: one RR Lyrae and one long-period irregular variable.

V21: This star is a RR Lyrae pulsating in the fundamental mode. In the colour-magnitude diagram (Fig. 24) for this cluster it falls exactly in the instability strip of the horizontal branch. It is next to a brighter star (1.039 arcsec) and this may be why it was not discovered before.

V22: This star is at the top of the red giant branch in the colour-magnitude diagram of the globular cluster (Fig. 24). From the light curve obtained it was possible to measure an amplitude of =0.17 mag. We classified this star as a long-period irregular variable as it was not possible to find a period.

Figure 26: NGC 6541: Light curves of the known and new variables discovered in this globular cluster. Symbols are the same as in Fig. 7.

8.7 Ngc 6656 / C1833-239 / M 22

This globular cluster was discovered by Abraham Ihle in 166511. It is in the constellation of Sagittarius at 3.2 kpc from the Sun and 4.9 kpc from the Galactic centre. The cluster has a metallicity of [Fe/H]=-1.70 dex, a distance modulus of (m-M)=13.60 mag and the horizontal branch level is at V=14.15 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V35 18:36:24.051 -23:54:29.53 2456472.8818 141(5) 9.43 0.13 19 SR
PK-06 18:36:25.228 -23:54:37.44 2456789.8239 0.140851(4) 17.08 0.65 51 EW
CV1 18:36:24.696 -23:54:35.60 2456877.6355 17.87 3.28 51 CV(DN)
Table 10: NGC 6656: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Columns are the same as in Tab. 4.

Known variables

There are of the order of 100 variable sources known for this globular cluster. Only 4 of these are inside the field of view of the reference image (V35, PK-06, PK-08 and CV1). PK-08 is too faint for us to extract a light curve.

V35: This star is the brightest star in our field of view and it is classified as a semi-regular variable. The star is at the top of the red giant branch in the colour-magnitude diagram shown in Fig. 29. Sahay et al. (2014) found a period of 56 d. However, we found a period of d.

PK-06: This star is classified as an EW eclipsing variable. It was discovered by Pietrukowicz & Kaluzny (2003) and it is their star M22_06. They found a period P=0.239431 d. However, it does not produce a good phased light curve in our data. In the analysis of this variable we found a period of P=0.140851 d which produces a better phased light curve (See Fig. 28). However, our phased light curve is still not as clear as that of Pietrukowicz & Kaluzny (2003).

CV1: The variability of this star was first detected by Sahu et al. (2001) as a suspected microlensing event, but it was not until Anderson et al. (2003) that it was classified as a dwarf nova outburst. The analysis of our light curve for this source shows that it undergoes an outburst of 3 mag around HJD 2456877 which decays over 20d. It is in agreement with previous studies. Anderson et al. (2003) found that an earlier outburst lasted 20-26d with an amplitude of 3 mag peaking at 15 mag. Alonso-Garcia et al. (2015) also observed the 2014 outburst seen in our data and reported a Ks-band brightening of 1 mag. In Fig. 28, we have plotted the light curve for this star.

Figure 27: Finding chart for the globular cluster NGC 6656. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec.

Our search for new variable sources in this cluster did not yield any.

Figure 28: NGC 6656: Light curves of the known variables found in this globular cluster. Symbols are the same as in Fig. 7.
Figure 29: Colour magnitude diagram for the globular cluster NGC 6656 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.

8.8 Ngc 6681 / C1840-323 / M70

This globular cluster was discovered by Charles Messier in 178012. The globular cluster is in the constellation of Sagittarius at 9.0 kpc from the Sun and 2.2 kpc from the Galactic centre. It has a metallicity of [Fe/H]=-1.62 dex, a distance modulus of (m-M)=14.99 mag and the horizontal branch level is at V=15.55 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V6 18:43:12.015 -32:17:29.70 2456814.8334 0.341644(25) 15.09 0.23 50 RR1
Table 11: NGC 6681: Ephemerides and main characteristics of one variable star in the field of this globular cluster. Columns are the same as in Tab. 4.

There are a total of 1315 light curves available for this cluster to be analysed. Most of them have 50 epochs. The RMS and diagrams can be found in Fig. 30. Variable stars studied in this work are plotted in colour.

Figure 30: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 1315 stars detected in the field of view of the reference image for NGC 6681. The coloured point follows the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.

Known variables

This globular cluster has only 5 known variables, none of which are in the field of view of our reference image.

Kadla et al. (1996) reported several RR Lyrae candidates based on the position of the stars in the instability strip of the horizontal branch but none of them matched the position of the only variable star found in the field of view covered by our reference image which is the new variable V6 explained in the next section.

Figure 31: Colour magnitude diagram for the globular cluster NGC 6681 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. One variable star is plotted in colour following the convention adopted in Tab. 2.
Figure 32: Finding chart for the globular cluster NGC 6681. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec.

New variables

One new RR Lyrae was discovered in this globular cluster.

V6: This star has an amplitude of 0.23 mag and a period of P=0.341644d. The star is placed just at the instability strip of the horizontal branch (see Fig. 31). It is clearly a previously unknown RR Lyrae of type RR1. The light curve for this variable is shown in Fig. 33.

Figure 33: NGC 6681: Light curve of the new variable discovered in this globular cluster. Symbols are the same as in Fig. 7.

8.9 Ngc 6723 / C1856-367

This globular cluster was discovered by James Dunlop in 182613. It is in the constellation of Sagittarius at a distance of 8.7 kpc from the Sun and 2.6 kpc from the Galactic centre. It has a metallicity of [Fe/H]=-1.10 dex, a distance modulus of (m-M)=14.84 mag and the horizontal branch level is at V=15.48 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V8 18:59:34.678 -36:37:42.33 2456908.5573 0.480278(49) 15.08 0.71 16 RR0
V34 18:59:33.189 -36:37:58.04 2456435.8962 0.531414(60) 14.85 0.84 56 RR0
V35 18:59:32.963 -36:38:01.46 2456524.8178 0.606451(78) 14.80 0.32 56 RR0
V44 18:59:32.347 -36:37:51.96 2456454.9399 0.440075(41) 15.08 0.93 56 RR0
Table 12: NGC 6723: Ephemerides and main characteristics of the variable stars in the field of this globular cluster. Columns are the same as in Tab. 4.

In Fig. 34, the RMS and statistic diagrams for 1258 stars in the globular cluster NGC 6723 are shown. Most of the light curves have 56 epochs. Variable stars analysed in this work are plotted in colour following the convention adopted in Tab. 2.

Figure 34: Root mean square (RMS) magnitude deviation (top) and statistic (bottom) versus the mean magnitude for the 1258 stars detected in the field of view of the reference image for NGC 6723. Coloured points follow the convention adopted in Tab. 2 to identify the types of variables found in the field of this globular cluster.

Known variables

This globular cluster has 47 known variables stars in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001). 43 are classified as RR Lyrae, two as semi-regular variables, one as a T Tauri star that does not seem to be a cluster member and one as a SX Phoenicis. Four of the known RR Lyrae are in the field of our reference frame (V8, V34, V35 and V44). They are RR0 pulsating stars. The star V8 was discovered by Bailey (1902) and although it is close to the reference image border, it was possible to obtain 16 data points. V34, V35 and V44 were discovered by Lee et al. (2014).

In the colour magnitude diagram for this globular cluster (see Fig. 36) these variables are placed just in the instability strip of the horizontal branch. Light curves for these variables are shown in Fig. 38.

The four RR Lyrae in the field of our images are plotted in the period-amplitude diagram shown in Fig. 35. It is possible to see that even though V34 and V35 have the Blazhko effect, all of them are following the model of fundamental mode pulsating stars for Oosterhoff type I which is consistent with the Oosterhoff classification for this cluster (see Lee et al., 2014; Kovacs et al., 1986, and references therein).

The periods that we derive for the 4 RR Lyrae stars are perfectly consistent with those derived by Lee et al. (2014) using a 10 year baseline. Furthermore, the Blazhko effect in V34 and V35 is also evident in our light curves.

Figure 35: Period-amplitude diagram for the globular cluster NGC 6723. The previously known RR Lyrae are plotted.
Figure 36: Colour magnitude diagram for the globular cluster NGC 6723 built with V and I magnitudes available in the ACS globular cluster survey extracted from HST images. The variable stars are plotted in colour following the convention adopted in Tab. 2.
Figure 37: Finding chart for the globular cluster NGC 6723. The image used corresponds to the reference image constructed during the reduction. All known variables and new discoveries are labelled. Image size is arcsec.

No new variable stars were found in the field covered by the reference image for this globular cluster.

Figure 38: NGC 6723: Light curves of the known variables in this globular cluster. Symbols are the same as in Fig. 7.

8.10 Ngc 6752 / C1906-600 / C93

This globular cluster was discovered by James Dunlop in 182614. It is in the constellation of Pavo with a distance of 4.0 kpc from the Sun and 5.2 kpc from the Galactic centre. It has a metallicity of [Fe/H]=-1.54 dex, a distance modulus of (m-M)=13.13 mag and the level of its horizontal branch is at V=13.70 mag.

Var id RA Dec Epoch Type
J2000 J2000 HJD d mag mag
V26 19:10:51.494 -59:58:56.67 78 CV(DN)
Table 13: NGC 6752: Ephemerides and main characteristics of one variable star in the field of this globular cluster. Columns are the same as in Tab. 4.

Known variables

There are 32 known variable sources in this globular cluster which are listed in the Catalogue of Variable Stars in Galactic Globular Clusters (Clement et al., 2001). Six of them are in the field of view of our reference image. Three are millisecond pulsars (PSRB, PSRD and PSRE) discovered by D’Amico et al. (2001) and D’Amico et al. (2002). We could not find an optical counterpart in our reference image. Three more are Dwarf Novae (V25, V26, V27) discovered by Thomson et al. (2012) of which V26 is the only object in which we could detect variability.

V26: This star is classified as a Dwarf Nova. A finding chart for V26 is already available in Thomson et al. (2012). It was not possible to detect this star in the reference image in its quiescent state. However, it was possible to clearly detect one outburst in the difference images. Due to this, in Fig. 39 the difference fluxes measured for this star are plotted against Heliocentric Julian Day. It is possible to see that in the 2013 campaign the star was in its quiescent state. In 2014 observing season, the outburst started around HJD 2456840 with a maximum at HJD 2456858.8532 and lasted 80 d.

We did not find any evidence in our data for new variable stars in this globular cluster.

Figure 39: NGC 6752: Light curve of the variable V26 in this globular cluster. Symbols are the same as in Fig. 7. We plot the quantity since a reference flux is not available.

9 Conclusions

  1. The central regions of 10 globular clusters were studied.

  2. The benefits of using EMCCDs and the shift-and-add technique were shown.

  3. A total of 12541 stars in the fields covered in the center of each globular cluster were studied for variable star detection.

  4. Light curves for 31 previously known variables are presented (3 L, 2 SR, 20 RR Lyrae, 1 SX Phe, 3 cataclysmic variables, 1 EW and 1 NC). Period improvement and ephemerides are presented for most of them as well.

  5. The discovery of 30 variable stars is presented (16 L, 7 SR, 4 RR Lyrae, 1 SX Phe and 2 NC). Period and ephemerides calculations were also done.

  6. Light curves are available for all variable stars studied in this work.

Acknowledgements.
Our thanks go to Christine Clement for clarifying the known variable star content of each cluster and the numbering systems within each one while we were working on these clusters. This support to the astronomical community is very much appreciated. The Danish 1.54m telescope is operated based on a grant from the Danish Natural Science Foundation (FNU). RFJ thanks Ian Taylor for sorting out all my computational needs on my arrival at St Andrews. RFJ also thanks Katherine White for her comments and suggestions in the improvement of my English language. This publication was made possible by NPRP grant # X-019-1-006 from the Qatar National Research Fund (a member of Qatar Foundation). The statements made herein are solely the responsibility of the authors. KH acknowledges support from STFC grant ST/M001296/1. GD acknowledges Regione Campania for support from POR-FSE Campania 2014-2020. TH is supported by a Sapere Aude Starting Grant from the Danish Council for Independent Research. Research at Centre for Star and Planet Formation is funded by the Danish National Research Foundation. TCH acknowledges support from the Korea Research Council of Fundamental Science & Technology (KRCF) via the KRCF Young Scientist Research Fellowship Programme and for financial support from KASI travel grant number 2013-9-400-00 & 2014-1-400-06. OW and J. Surdej acknowledge support from the Communauté française de Belgique - Actions de recherche concertées - Académie Wallonie-Europe. This work has made extensive use of the ADS and SIMBAD services, for which we are thankful.

Footnotes

  1. thanks: based on data collected by the MiNDSTEp team with the Danish 1.54m telescope at ESO’s La Silla observatory in Chile
  2. DanDIA is built from the DanIDL library of IDL routines available at http://www.danidl.co.uk
  3. http://cds.u-strasbg.fr/
  4. http://www.astro.ufl.edu/~ata/public_hstgc/
  5. http://gclusters.altervista.org/
  6. http://messier.seds.org/xtra/ngc/n0104.html
  7. http://messier.seds.org/xtra/ngc/n5139.html
  8. http://messier.seds.org/m/m080.html
  9. http://messier.seds.org/m/m004.html
  10. http://spider.seds.org/spider/MWGC/n6541.html
  11. http://messier.seds.org/m/m022.html
  12. http://messier.seds.org/m/m070.html
  13. http://spider.seds.org/spider/MWGC/n6723.html
  14. http://spider.seds.org/spider/MWGC/n6752.html

References

  1. Alard, C. & Lupton, R. H. 1998, ApJ, 503, 325
  2. Albrow, M. D., Gilliland, R. L., Brown, T. M., et al. 2001, ApJ, 559, 1060
  3. Alonso-Garcia, J., Minniti, D., Angeloni, R., et al. 2015, The Astronomer’s Telegram, 7238, 1
  4. Anderson, J., Cool, A. M., & King, I. R. 2003, ApJ, 597, L137
  5. Anderson, J., Sarajedini, A., Bedin, L. R., et al. 2008, AJ, 135, 2055
  6. Arellano Ferro, A., Bramich, D. M., Figuera Jaimes, R., Giridhar, S., & Kuppuswamy, K. 2012, MNRAS, 420, 1333
  7. Bailey, S. I. 1902, Annals of Harvard College Observatory, 38, 1
  8. Bailey, S. I. 1918, in Publications of the American Astronomical Society, Vol. 3, Publications of the American Astronomical Society, 291
  9. Barnard, E. E. 1909, ApJ, 29, 72
  10. Bessell, M. S. 2005, ARAA, 43, 293
  11. Bramich, D. M. 2008, MNRAS, 386, L77
  12. Bramich, D. M., Bachelet, E., Alsubai, K. A., Mislis, D., & Parley, N. 2015, A&A, 577, A108
  13. Bramich, D. M., Figuera Jaimes, R., Giridhar, S., & Arellano Ferro, A. 2011, MNRAS, 413, 1275
  14. Bramich, D. M., Horne, K., Albrow, M. D., et al. 2013, MNRAS, 428, 2275
  15. Clement, C. M., Muzzin, A., Dufton, Q., et al. 2001, AJ, 122, 2587
  16. D’Amico, N., Lyne, A. G., Manchester, R. N., Possenti, A., & Camilo, F. 2001, ApJ, 548, L171
  17. D’Amico, N., Possenti, A., Fici, L., et al. 2002, ApJ, 570, L89
  18. Davis, H. 1917, PASP, 29, 260
  19. Dieball, A., Long, K. S., Knigge, C., Thomson, G. S., & Zurek, D. R. 2010, ApJ, 710, 332
  20. Diolaiti, E., Bendinelli, O., Bonaccini, D., et al. 2000, A&AS, 147, 335
  21. Draper, P. W. 2000, in Astronomical Society of the Pacific Conference Series, Vol. 216, Astronomical Data Analysis Software and Systems IX, ed. N. Manset, C. Veillet, & D. Crabtree, 615
  22. Eddington, A. S. 1916, MNRAS, 76, 572
  23. Figuera Jaimes, R., Arellano Ferro, A., Bramich, D. M., Giridhar, S., & Kuppuswamy, K. 2013, A&A, 556, A20
  24. Fiorentino, G., Lanzoni, B., Dalessandro, E., et al. 2014, ApJ, 783, 34
  25. Freire, P. C., Camilo, F., Lorimer, D. R., et al. 2001, MNRAS, 326, 901
  26. Harpsøe, K. B. W., Jørgensen, U. G., Andersen, M. I., & Grundahl, F. 2012, A&A, 542, A23
  27. Harris, W. E. 1996, AJ, 112, 1487
  28. Herschel, W. 1786, Royal Society of London Philosophical Transactions Series I, 76, 457
  29. Hertzsprung, E. 1915, ApJ, 41, 10
  30. Jerram, P., Pool, P. J., Bell, R., et al. 2001, in Society of Photo-Optical Instrumentation Engineers (SPIE) Conference Series, Vol. 4306, Sensors and Camera Systems for Scientific, Industrial, and Digital Photography Applications II, ed. M. M. Blouke, J. Canosa, & N. Sampat, 178–186
  31. Kadla, Z. I., Gerashchenko, A. N., & Malakhova, Y. N. 1996, Information Bulletin on Variable Stars, 4414, 1
  32. Kaluzny, J., Olech, A., Thompson, I. B., et al. 2004, A&A, 424, 1101
  33. Kopacki, G. 2013, Acta Astron., 63, 91
  34. Kovacs, G., Shlosman, I., & Buchler, J. R. 1986, ApJ, 307, 593
  35. Kunder, A., Stetson, P. B., Catelan, M., Walker, A. R., & Amigo, P. 2013, AJ, 145, 33
  36. Lafler, J. & Kinman, T. D. 1965, ApJS, 11, 216
  37. Lebzelter, T. & Wood, P. R. 2005, A&A, 441, 1117
  38. Lee, J.-W., López-Morales, M., Hong, K., et al. 2014, ApJS, 210, 6
  39. Luther. 1860, Astronomical news, 53, 293
  40. Martínez Roger, C., Perez Fournón, I., & Sánchez, F., eds. 1999, Globular clusters. Proceedings.
  41. Messier, C. 1781, Catalogue des Nebuleuses et des amas d’etoiles (Catalog of Nebulae and Star Clusters), Tech. rep.
  42. Navarrete, C., Contreras Ramos, R., Catelan, M., et al. 2015, A&A, 577, A99
  43. O’Meara, S. 2002, The Caldwell Objects, Deep-sky companions (Sky Pub.)
  44. Oort, J. H. 1924, Proceedings of the National Academy of Science, 10, 256
  45. Oosterhoff, P. T. 1938, Bull. Astron. Inst. Netherlands, 8, 273
  46. Pál, A. & Bakos, G. Á. 2006, PASP, 118, 1474
  47. Pietrukowicz, P. & Kaluzny, J. 2003, Acta Astron., 53, 371
  48. Sahay, A., Lebzelter, T., & Wood, P. R. 2014, PASA, 31, 12
  49. Sahu, K. C., Casertano, S., Livio, M., et al. 2001, Nature, 411, 1022
  50. Samus, N. N., Durlevich, O. V., & et al. 2009a, VizieR Online Data Catalog, 1, 2025
  51. Samus, N. N., Kazarovets, E. V., Pastukhova, E. N., Tsvetkova, T. M., & Durlevich, O. V. 2009b, PASP, 121, 1378
  52. Sawyer, H. B. 1931, Harvard College Observatory Circular, 366, 1
  53. Sawyer, H. B. 1938, JRASC, 32, 69
  54. Shapley, H. 1916, The Observatory, 39, 452
  55. Shara, M. M., Hinkley, S., & Zurek, D. R. 2005, ApJ, 634, 1272
  56. Sirianni, M., Jee, M. J., Benítez, N., et al. 2005, PASP, 117, 1049
  57. Skottfelt, J., Bramich, D. M., Figuera Jaimes, R., et al. 2015a, A&A, 573, A103
  58. Skottfelt, J., Bramich, D. M., Figuera Jaimes, R., et al. 2013, A&A, 553, A111
  59. Skottfelt, J., Bramich, D. M., Hundertmark, M., et al. 2015b, A&A, 574, A54
  60. Smith, N., Giltinan, A., O’Connor, A., et al. 2008, in Astrophysics and Space Science Library, Vol. 351, Astrophysics and Space Science Library, ed. D. Phelan, O. Ryan, & A. Shearer, 257
  61. Stetson, P. B. 1987, PASP, 99, 191
  62. Thomson, G. S., Knigge, C., Dieball, A., et al. 2012, MNRAS, 423, 2901
  63. Wehlau, A., Butterworth, S., & Hogg, H. S. 1990, AJ, 99, 1159
  64. Weldrake, D. T. F., Sackett, P. D., & Bridges, T. J. 2007, AJ, 133, 1447
  65. Weldrake, D. T. F., Sackett, P. D., Bridges, T. J., & Freeman, K. C. 2004, AJ, 128, 736
  66. Zorotovic, M., Catelan, M., Smith, H. A., et al. 2010, AJ, 139, 357
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
""
The feedback must be of minumum 40 characters
   
Add comment
Cancel
Loading ...
101196
This is a comment super asjknd jkasnjk adsnkj
Upvote
Downvote
""
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters
Submit
Cancel

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test
Test description