On the impact of helium content on the RR Lyrae distance scale
We constructed new sets of He-enhanced (Y=0.30, Y=0.40) nonlinear, time-dependent convective hydrodynamical models of RR Lyrae (RRL) stars covering a broad range in metal abundances (Z=0.0001–0.02). The increase in He content from the canonical value (Y=0.245) to Y=0.30–0.40 causes a simultaneous increase in stellar luminosity and in pulsation period. To investigate the dependence of the RR Lyrae distance scale on the He abundance we computed new optical () and near-infrared () Period-Luminosity-Metallicity-Helium (PLZY) relations. Interestingly enough, the increase in He content causes a minimal change in the coefficients of both period and metallicity terms, since canonical and He–enhanced models obey similar PLZ relations. On the contrary, the classical - and -band mean magnitude-metallicity relations and the R-band PLZ relation display a significant dependence on the He content. The He-enhanced models are, at fixed metal content, 0.2-0.5 mag brighter than canonical ones. This variation is only marginally affected by evolutionary effects. The quoted distance diagnostics once calibrated with trigonometric parallaxes (Gaia) will provide the opportunity to estimate the He content of field and cluster RRLs. Moreover, the use of either spectroscopic or photometric metal abundances will pave the way to new empirical constraints on the universality of the helium–to–metal enrichment ratio in old (t 10 Gyr) stellar tracers.
During the last century RR Lyrae (RRL) stars have played a crucial role as standard candles and tracers of old stellar populations (Marconi et al., 2015; Madore et al., 2017; Neeley et al., 2017). They are old (t10 Gyr) low–mass radial variables in their central helium burning phase and are observed in the Milky Way (Vivas & Zinn, 2006; Zinn et al., 2014; Drake et al., 2013; Pietrukowicz et al., 2015), in Local Group (Soszyński et al., 2010; Fiorentino et al., 2012; Coppola et al., 2015) and in Local Volume galaxies (Da Costa et al., 2010; Sarajedini et al., 2012).
RRLs are used as standard candles since they obey a relation between absolute visual magnitude and iron abundance (Caputo et al., 2000; Cacciari & Clementini, 2003; Di Criscienzo et al., 2004). This relation, whose linearity has also been questioned in the literature (Caputo et al., 2000; Catelan et al., 2004; Di Criscienzo et al., 2004), suffers from significant intrinsic errors and systematics. RRLs do not obey a Period-Luminosity (PL) relation in the optical bands but, thanks to the characteristic behaviour of near-infrared (NIR) bolometric corrections (Bono et al., 2001, 2003), they obey a PL relation in the NIR regime (Longmore et al., 1990; Braga et al., 2015; Coppola et al., 2015).
The advantages of these relations are the small dependence on reddening and evolutionary effects (Bono et al., 2003) and a milder dependence on metallicity when compared with , magnitudes. Theory and observations indicate that more metal–rich RRLs are fainter than metal-poor ones, but we still lack firm constraints on the coefficient of the metallicity term in the NIR PL relations (Bono et al., 2003; Catelan et al., 2004; Dall’Ora et al., 2004; Sollima et al., 2006; Marconi et al., 2015).
Optical and NIR Period-Wesenheit (PW) relations are solid diagnostics to determine individual RRL distances, but rely on the assumed reddening law (Di Criscienzo et al., 2004; Braga et al., 2015; Coppola et al., 2015; Marconi et al., 2015). These relations are reddening free by construction (Madore, 1982; Ripepi et al., 2012; Riess et al., 2012; Fiorentino et al., 2013; Inno et al., 2013) and include a color term. This means that they mimic a Period-Luminosity-Color (PLC) relation, tracing the position of each variable inside the instability strip (IS). These are the reasons why Period-Wesenheit relations have been widely adopted to trace the 3D structure of highly reddened clusters in the Galactic Bulge (Soszyński et al., 2014; Pietrukowicz et al., 2015).
The main motivations for the current investigations are the following.
a) The helium-to-metals enrichment ratio (/=1.4, with a primordial He abundance of 0.245) adopted in evolutionary (Pietrinferni et al., 2006) and pulsation (Marconi et al., 2015) calculations is still affected by large uncertainties. RRLs are good laboratories for estimating the He content (Caputo, 1998). To provide a new spin on the determination of this parameter we are investigating new pulsation observables together with spectroscopic measurements of the metal content for field and cluster RRLs.
b) Using the S method Walker & Terndrup (1991) found that Bulge RRLs approach solar metallicity. This finding was recently supported by Chadid et al. (2017) using high-resolution spectra, since they found several RRLs at solar chemical compositions. This means a metallicity regime in which RRL pulsation properties are more prone to helium effects (Bono et al., 1995b; Marconi et al., 2011).
c) The RRL distance scale is going to play a crucial role to constrain possible systematics affecting primary distance indicators (Beaton et al., 2016). Sizable samples of RRLs have already been identified in Local Group galaxies (Monelli et al., 2017) and beyond (Da Costa et al., 2010). However, we still lack firm theoretical and empirical constraints on the / ratio in extragalactic systems.
To overcome the limitations of the current theoretical framework we computed new sets of pulsation models with the same metal abundances (Z=0.0001–0.02) adopted in Marconi et al. (2015) but helium enriched (Y=0.30 and Y=0.40111Metals (Z) and helium (Y) abundances by mass fraction., Marconi et al. 2018, in preparation).
2 Impact of helium-enhanced models on RRL distances
Following the same prescriptions as in Marconi et al. (2015) for both evolutionary and pulsation computations, and the same seven metal abundances, new sets of He-enhanced RRL models were computed with Y=0.30 and Y=0.40. The entire set of HB models (Pietrinferni et al., 2006) are available in the BaSTI data base222http://www.oa-teramo.inaf.it/BASTI/. They were computed, for each assumed chemical composition, using a fixed core mass and envelope chemical profile and evolving a progenitor from the pre-main sequence to the tip of the Red Giant Branch with an age of 13 Gyr. For each chemical composition, the mass distribution of HB models ranges from the mass of the progenitors (coolest HB models) down to a total mass of the order of 0.5 (hottest HB models).
The evolutionary phases off the Zero-Age-Horizontal Branch (ZAHB) have been extended either to the onset of thermal pulses, for more massive models, or until the luminosity of the model (along the white dwarf cooling sequence) becomes fainter than -2.5 for less massive structures. The -elements were enhanced with respect to the Grevesse et al. (1993)333Note that new solar abundances by Asplund et al. (2009) provide lower CNO abundances when compared with Grevesse et al. (1993). The difference in metal distributions mainly causes a difference in the zero-point. The impact on the HB mass-luminosity relation is within the different luminosity levels adopted, at fixed mass and chemical composition, to construct pulsation models. solar metal distribution by variable factors (see Table 1 in Pietrinferni et al. (2006)). The overall enhancement—[/Fe]—is equal to 0.4 dex.
Fig. 1 shows the behaviour of HB models in the Hertzsprung-Russell diagram for three assumptions on the helium and on the metal content. In each panel the black solid line shows the location of the ZAHB, the dashed black line corresponds to a central helium exhaustion at level and the long dashed line to the complete exhaustion. Note that the ZAHB becomes dotted for masses higher than the progenitor one, artificially included to populate the IS. The blue and the red vertical lines display the predicted blue and red edge of the IS. The red solid lines show selected evolutionary models of HB structures populating the RRL IS. They range from 0.76 to 0.80 for Z=0.0001, Y=0.245 (top left panel), and from 0.520 to 0.525 for Z=0.0164, Y=0.400 (bottom right panel). Evolutionary prescriptions plotted in Fig. 1 bring forward some relevant properties concerning He-enhanced stellar structures worth being discussed.
i) HB morphology–He-enhanced stellar populations, at fixed metal content and cluster age, are characterized by smaller stellar masses at the main sequence turn off. This means smaller stellar masses at the tip of the red giant branch and a HB morphology dominated by hot and extreme stars. The HB luminosity function is, therefore, dominated by stars that are hotter than the blue edge of the IS. These stellar systems can still produce RRLs, since hot HB stars cross the IS just before or soon after the AGB phase (post early AGB, Greggio & Renzini, 1990; D’Cruz et al., 1996). This means that the red HB and the IS are poorly populated (see middle and right panels of Fig. 1).
ii) Evolutionary time scale inside the instability strip– The evolutionary time spent by a canonical, metal-poor (Z=0.0001, Y=0.245) stellar structure (M=) inside the IS during central He burning phases is 67 Myr. This time decreases by at least a factor of two when moving to He-enhanced models with Y=0.30 (M=, 32 Myr) and by a factor of six for models with Y=0.40 (M=, 11 Myr). The quoted trend marginally changes with the metal content, and indeed, canonical models at solar iron abundance (Z=0.0198, Y=0.273, M=) spend inside the IS an evolutionary time of 29 Myr and this time decreases down to 15 Myr for Y-0.30 (M=) and to 5 Myr for Y=0.40 (M=). The consequence of this difference is that the number of RRLs produced by He-enhanced stellar populations is at least one order of magnitude smaller than canonical ones. The reader is referred to Marconi et al. (2018) for more details.
iii) Evolutionary time-scale to approach the instability strip–The increase in helium content causes, at fixed metallicity, a steady decrease in the evolutionary time scale required to approach the IS. This effect is more severe in the metal-poor regime (see top panels in Fig. 1) where the ZAHB for He-enhanced models located inside the IS is populated by stellar structures that are significantly younger than typical RRLs. Canonical stellar structures with Z=0.0001 and Y=0.245 evolves from the Zero Age Main Sequence (ZAMS) to the ZAHB portion located inside the IS on a time scale of 12.5 Gyr (Valcarce et al., 2012). A He-enhanced stellar structure with Z=0.0001 and Y=0.30 evolves from the ZAMS to the IS in 8.5 Gyr, while for Y=0.40 the same time scale becomes of the order of 4.5 Gyr. A similar trend is also present among stellar structures with Z=0.0006 and Y=0.40, since they approach the portion of the ZAHB located inside the IS with ages younger than 8.5 Gyr. On the other hand, for Z=0.001 and Y=0.40 the ZAHB stellar structures located inside the IS have ages that are only marginally younger (11–12 Gyr) than canonical ones. Note that the current empirical evidence indicates that RRLs stars have only been identified in stellar populations older than 10 Gyrs (Dékány et al., 2018).
On the basis of the quoted evolutionary prescriptions, we computed a set of pulsation models, for each iron and helium abundance, accounting for two values of the stellar mass and three different luminosity levels. The reasons for this choice were already discussed in Marconi et al. (2015). Here we only give the highlights: 1) the ZAHB mass and luminosity level as based on the adopted evolutionary models; 2) the ZAHB mass and a luminosity level 0.1 dex brighter than the ZAHB luminosity; 3) a stellar mass 10 smaller than the ZAHB value and a luminosity level 0.2 dex brighter than the ZAHB luminosity.
The different sets of models were constructed following the same approach discussed in Marconi et al. (2015). The bolometric light curves were transformed into optical () and NIR ()444We adopted the 2MASS——photometric system. bands using static atmosphere models (Bono et al., 1995a) and eventually intensity weighted mean magnitudes and colors were computed. Preliminary results for interpreting Galactic Bulge RRLs were discussed in Marconi & Minniti (2018), while the details of these new helium enriched models are presented in Marconi et al. (2018, in preparation).
2.1 Predicted optical/NIR PL relations
Figure 2 shows the predicted optical/NIR PL distribution for five bands (,,,,) and for three different metal abundances: Z=0.0001 (left), Z=0.001 (middle) and Z=0.02 (right). In each panel are plotted pulsation models constructed assuming a fixed helium-to-metal enrichment ratio Marconi et al. (2015, black circles) together with models constructed assuming two different helium enhancements: Y=0.30 (magenta circles) and Y=0.40 (blue circles). Models plotted in this figure display two well defined trends among canonical and helium enhanced models. i) The period distribution of helium enhanced models is systematically shifted towards longer periods when compared with canonical models. The difference is mainly caused by an evolutionary effect: a decrease in the mean stellar mass populating the RRL IS and an increase in the luminosity level (Marconi et al., 2011). ii) The spread in luminosity between canonical and helium enhanced models steadily decreases when moving from the - to the -band. The quoted spread is mainly caused by a difference in the zero-point, since the slopes are quite similar.
The plotted intensity-weighted mean magnitudes can be used to predict multi-band PL relations. The evolutionary and pulsation parameters of the helium enhanced models will be provided in Marconi et al. (2018, in preparation) together with the bolometric mean magnitude, and the transformation into different optical, NIR and MIR photometric systems. Moreover, we plan to discuss the luminosity amplitudes for both canonical and He-enhanced models and their impact on the Bailey diagram. Finally, we plan to provide for the He-enhanced models the same distance diagnostics provided by (Marconi et al., 2015). Table 1 gives the coefficients of the global555This sample includes both fundamental and first overtone pulsators. The latter group was fundamentalized, i.e. PL relations including both the metallicity and the helium terms (= + + [Fe/H] + Y, where X is the selected photometric band).
A glance at the coefficients listed in this Table and to the models plotted in Fig. 2 disclose three relevant features.
i) The dependence on the period becomes, as expected, systematically steeper when moving from optical to NIR bands. This means that NIR PL relations are intrinsically more accurate than optical ones. The reason is twofold: 1) the standard deviation decreases by a factor of two when moving from the / to the / bands; 2) the coefficient of the metallicity term in the NIR bands attains similar values.
ii) The helium dependence decreases by roughly one dex when moving from the - to the -band. This means that an increase in helium content causes, at fixed period and metal content, an increase in the R-band of the order of a few tenths of a magnitude. The same increase causes a variation of the order of 0.05–0.08 mag in the NIR bands. Such an increase might introduce a mild systematic effect in distance determinations, but it appears negligible because it is similar to the standard deviations.
It has been suggested that the second stellar generation in GCs are made of materials that are enriched in helium, nitrogen and sodium and depleted in carbon and oxygen. The enhancement in helium can be of the order of 0.05–0.10 (Renzini et al., 2015). However, there are reasons to believe that the quoted dependence of NIR PL relations on helium can be considered as a solid upper limit on the RRL distance scale. The reasons are the following.
i)–The second stellar generation appears to be ubiquitous in GCs, but the current spectroscopic evidence indicates that they are very rare in the Galactic field (Gratton, 2016). This means that only a few percent of the galactic stellar content might be helium enhanced.
ii)–He-enhanced stellar populations in the metal-poor and in the metal-intermediate regime cross the IS only during off-ZAHB evolution. This means that the evolutionary time spent inside the IS is at least one order of magnitude smaller compared with the canonical ones (see § 2). Note that the He-enhanced (Y=0.30) ZAHB crosses the IS in the metal-intermediate and in the metal-rich regime. This means that the probability to produce He-enhanced RRLs in the metal-poor regime is quite limited. Moreover, the crossing of the IS at brighter magnitudes (lower surface gravities) causes a systematic shift in the period distribution of He-enhanced RRLs towards longer periods. This also means that He-enhanced stellar populations are more prone to produce type II Cepheids (P 1 day) than RRLs.
2.2 Mean magnitude——metallicity relations
The visual mean-magnitude metallicity () relation was foreseen by Baade (1958) and it was the most popular distance diagnostic for old stellar populations (Sandage, 1990), but it is also prone to a number of potential systematic errors (Caputo et al., 2000; Bono et al., 2003; Di Criscienzo et al., 2004; Marconi et al., 2015).
We have already mentioned that an increase in He content causes, at fixed metallicity, an increase in stellar luminosity, and in turn, in the pulsation period. A glance at the predicted magnitudes plotted in the bottom panel of Figure 3 shows the impact of the He content. Canonical and He-enhanced models, as expected, partially overlap due to off-ZAHB evolution. However, He-enhanced models with Y=0.30 (pink open circles) are on average 0.15 mag brighter than canonical ones, while those with Y=0.40 (blue open circles) are almost half magnitude brighter.
The outcome is the same if we use the -band, but it should be cautiously treated for a possible color dependency (Catelan et al., 2004). Data plotted in the top panel of Figure 3 display that He-enhanced models are 0.2 (Y=0.30) and 0.6 (Y=0.40) mag systematically brighter than canonical ones. To investigate on a more quantitative bases the dependence of the mean magnitude metallicity (MZ) relations, we derived new analytical relations including a He term (MZY), namely:
with an r.m.s=0.21 mag
with an r.m.s=0.22 mag
Data plotted in Figure 3 and the above MZY relations disclose a few relevant predictions concerning the possible occurrence of He-enhanced RRLs.
i) Stellar systems hosting a sizable sample of RRLs covering a broad range in He abundance should show, at fixed metal content, a spread in visual and in -band magnitudes that is on average a factor of two larger than canonical models. A detailed set of synthetic HB models is required to constrain the variation as function of both metal and He content. However, empirical evidence dating back to (Sandage, 1990) indicate that the spread in visual magnitude showed by cluster RRLs ranges from 0.2 mag in the metal-poor regime to 0.6 mag in the metal-intermediate regime. This trend was soundly confirmed by synthetic HB models by Bono et al. (1997)
ii) Stellar systems hosting stellar populations with significantly different He contents should show multi-modal magnitude distributions inside the IS. Indeed, He-enhanced models are characterized by ZAHBs that are systematically brighter. This difference in magnitude cannot be mixed up with a difference in metallicity, since the evolutionary lifetime of He-enhanced models inside the IS is, at fixed metal content, systematically shorter than canonical ones.
3 Final remarks and conclusions
We have presented new sets of He-enhanced (Y=0.30, Y=0.40) nonlinear, time-dependent convective hydrodynamical models of RRLs covering the same range of metal abundances investigated by Marconi et al. (2015). The model mean magnitudes in the bands were used to obtain new Period-Luminosity-Metallicity-Helium (PLZY) relations in these filters (see Table 1). The main effect of an increase in He is an increase in the luminosity level, and in turn, in the predicted pulsation period. Therefore, an increase in primordial He content from the canonical value (Y=0.245) to He-enhanced (Y=0.30,0.40) causes a minimal change in the coefficients of both period and metallicity terms, since the He–enhanced models obey similar PLZ relations. Owing to the sensitivity of the luminosity level to He variations, the classical relations connecting the and mean magnitudes to metallicity and the -band PLZ relation display a significant He dependence. The He-enhanced models models are, at fixed metallicity, 0.20.5 mag brighter than canonical ones.
This is an interesting opportunity, because Gaia is going to provide accurate geometrical distances to calibrate both the zero-point and the slopes of the diagnostics adopted to estimate individual RRL distances. Spectroscopic RRL abundances based on ground-based measurements (Magurno et al., 2018) will pave the way for an empirical calibration of the PLZ relations. This means the opportunity to determine distance, reddening and chemical composition (metal, helium) for field RRLs for which are simultaneously available optical () and NIR () mean magnitudes. Note that this approach applies to RRL in nearby stellar systems, and in turn, the opportunity to investigate the helium-to-metal enrichment ratio currently adopted in evolutionary and pulsation calculations is universal.
- Asplund et al. (2009) Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481
- Baade (1958) Baade, W. 1958, Ricerche Astronomiche, 5, 165
- Beaton et al. (2016) Beaton, R. L., Freedman, W. L., Madore, B. F., et al. 2016, ApJ, 832, 210
- Bono et al. (1997) Bono, G., Caputo, F., Cassisi, S., Castellani, V., & Marconi, M. 1997, ApJ, 479, 279
- Bono et al. (1995a) Bono, G., Caputo, F., Castellani, V., & Marconi, M. 1995a, ApJ, 448, L115
- Bono et al. (2001) Bono, G., Caputo, F., Castellani, V., Marconi, M., & Storm, J. 2001, MNRAS, 326, 1183
- Bono et al. (2003) Bono, G., Caputo, F., Castellani, V., et al. 2003, MNRAS, 344, 1097
- Bono et al. (1995b) Bono, G., Caputo, F., & Marconi, M. 1995b, AJ, 110, 2365
- Braga et al. (2015) Braga, V. F., Dall’Ora, M., Bono, G., et al. 2015, ApJ, 799, 165
- Cacciari & Clementini (2003) Cacciari, C. & Clementini, G. 2003, in Lecture Notes in Physics, Berlin Springer Verlag, Vol. 635, Stellar Candles for the Extragalactic Distance Scale, ed. D. Alloin & W. Gieren, 105
- Caputo (1998) Caputo, F. 1998, A&A Rev., 9, 33
- Caputo et al. (2000) Caputo, F., Castellani, V., Marconi, M., & Ripepi, V. 2000, MNRAS, 316, 819
- Catelan et al. (2004) Catelan, M., Pritzl, B. J., & Smith, H. A. 2004, ApJS, 154, 633
- Chadid et al. (2017) Chadid, M., Sneden, C., & Preston, G. W. 2017, ApJ, 835, 187
- Coppola et al. (2015) Coppola, G., Marconi, M., Stetson, P. B., et al. 2015, ApJ, 814, 71
- Da Costa et al. (2010) Da Costa, G. S., Rejkuba, M., Jerjen, H., & Grebel, E. K. 2010, ApJ, 708, L121
- Dall’Ora et al. (2004) Dall’Ora, M., Storm, J., Bono, G., et al. 2004, ApJ, 610, 269
- D’Cruz et al. (1996) D’Cruz, N. L., Dorman, B., Rood, R. T., & O’Connell, R. W. 1996, ApJ, 466, 359
- Dékány et al. (2018) Dékány, I., Hajdu, G., Grebel, E. K., et al. 2018, ApJ, 857, 54
- Di Criscienzo et al. (2004) Di Criscienzo, M., Marconi, M., & Caputo, F. 2004, ApJ, 612, 1092
- Drake et al. (2013) Drake, A. J., Catelan, M., Djorgovski, S. G., et al. 2013, ApJ, 763, 32
- Fiorentino et al. (2012) Fiorentino, G., Contreras Ramos, R., Tolstoy, E., Clementini, G., & Saha, A. 2012, A&A, 539, A138
- Fiorentino et al. (2013) Fiorentino, G., Musella, I., & Marconi, M. 2013, MNRAS, 434, 2866
- Gratton (2016) Gratton, R. G. 2016, in IAU Symposium, Vol. 317, The General Assembly of Galaxy Halos: Structure, Origin and Evolution, ed. A. Bragaglia, M. Arnaboldi, M. Rejkuba, & D. Romano, 259–265
- Greggio & Renzini (1990) Greggio, L. & Renzini, A. 1990, ApJ, 364, 35
- Grevesse et al. (1993) Grevesse, N., Noels, A., & Sauval, A. J. 1993, A&A, 271, 587
- Inno et al. (2013) Inno, L., Matsunaga, N., Bono, G., et al. 2013, ApJ, 764, 84
- Longmore et al. (1990) Longmore, A. J., Dixon, R., Skillen, I., Jameson, R. F., & Fernley, J. A. 1990, MNRAS, 247, 684
- Madore (1982) Madore, B. F. 1982, ApJ, 253, 575
- Madore et al. (2017) Madore, B. F., Freedman, W. L., & Moak, S. 2017, ApJ, 842, 42
- Magurno et al. (2018) Magurno, D., Sneden, C., Braga, V. F., et al. 2018, ArXiv e-prints [arXiv1807.06681]
- Marconi et al. (2011) Marconi, M., Bono, G., Caputo, F., et al. 2011, ApJ, 738, 111
- Marconi et al. (2015) Marconi, M., Coppola, G., Bono, G., et al. 2015, ApJ, 808, 50
- Marconi & Minniti (2018) Marconi, M. & Minniti, D. 2018, ApJ, 853, L20
- Monelli et al. (2017) Monelli, M., Fiorentino, G., Bernard, E. J., et al. 2017, ApJ, 842, 60
- Neeley et al. (2017) Neeley, J. R., Marengo, M., Bono, G., et al. 2017, ApJ, 841, 84
- Pietrinferni et al. (2006) Pietrinferni, A., Cassisi, S., Salaris, M., & Castelli, F. 2006, ApJ, 642, 797
- Pietrukowicz et al. (2015) Pietrukowicz, P., Kozłowski, S., Skowron, J., et al. 2015, ApJ, 811, 113
- Renzini et al. (2015) Renzini, A., D’Antona, F., Cassisi, S., et al. 2015, MNRAS, 454, 4197
- Riess et al. (2012) Riess, A. G., Macri, L., Casertano, S., et al. 2012, ApJ, 752, 76
- Ripepi et al. (2012) Ripepi, V., Moretti, M. I., Marconi, M., et al. 2012, MNRAS, 424, 1807
- Sandage (1990) Sandage, A. 1990, ApJ, 350, 603
- Sarajedini et al. (2012) Sarajedini, A., Yang, S.-C., Monachesi, A., Lauer, T. R., & Trager, S. C. 2012, MNRAS, 425, 1459
- Sollima et al. (2006) Sollima, A., Borissova, J., Catelan, M., et al. 2006, ApJ, 640, L43
- Soszyński et al. (2010) Soszyński, I., Udalski, A., Szymañski, M. K., et al. 2010, Acta Astronomica, 60, 165
- Soszyński et al. (2014) Soszyński, I., Udalski, A., Szymański, M. K., et al. 2014, Acta Astronomica, 64, 177
- Valcarce et al. (2012) Valcarce, A. A. R., Catelan, M., & Sweigart, A. V. 2012, A&A, 547, A5
- Vivas & Zinn (2006) Vivas, A. K. & Zinn, R. 2006, AJ, 132, 714
- Walker & Terndrup (1991) Walker, A. R. & Terndrup, D. M. 1991, ApJ, 378, 119
- Zinn et al. (2014) Zinn, R., Horowitz, B., Vivas, A. K., et al. 2014, ApJ, 781, 22