UNCLOAKING GLOBULAR CLUSTERS IN THE INNER GALAXY^*

The GGCs in our sample were observed over four nights, 2005 May 30 to June 2, with the LCO 6.5 m Magellan Baade Telescope, using the Inamori Magellan Areal Camera and Spectrograph (IMACS) in imaging mode. We used the f/4 camera to image a field of view (FOV) of 1546 × 1546, with a pixel size of 011. All fields were observed using standard Johnson–Cousins B, V, and I filters (Bessell 1979). Two sets of observations with different exposure times (short and long) were taken in the B and V filters, and three (extra-short, short, and long) in I, for every cluster. Although the nights during our observing run were not all completely photometric, the seeing conditions were very good, with average values of ∼06 in V for the whole run. Table 2 lists the details of these ground-based photometry observations.

In order to obtain useful photometry of the inner regions of the more centrally crowded clusters, we supplemented our Magellan images with images taken with the Advanced Camera for Surveys (ACS) on board the HST. These data were obtained in by our group's Snapshot program 10573. The ACS has an FOV of 337 × 337, with a pixel size of 005. Five clusters of our sample were observed using the f435w(B₄₃₅), f555w(V₅₅₅), and f814w(I₈₁₄) filters. Table 3 lists the details of these new HST observations.

To better calibrate our photometry, we also used images available through the HST data archive for all the clusters in our sample. The data taken from the HST data archive are comprised of f439w(B₄₃₉), f555w(V₅₅₅), f606w(V₆₀₆), and f814w(I₈₁₄) images obtained with the Wide Field Planetary Camera 2 (WFPC2), and of f435w(B₄₃₅), f606w(V₆₀₆), and f814w(I₈₁₄) images taken with the ACS. Table 4 lists the different HST programs that we use data from.

For the ground-based data, the initial processing of the raw CCD images was done with the routines in the ccdred package of the Image Reduction and Analysis Facility (IRAF). The images were first corrected for bias, then flat-fielded using a combination of dome and twilight flats. Afterward, since different frames for every cluster were taken with different exposure times in the B, V, and I filters (see Table 2), and they had small spatial offsets between them, the frames of every individual cluster were aligned, using the IRAF task imalign, and average-combined for every exposure time and filter, with the IRAF task imcombine.

Stellar photometry was carried out on the ground-based processed images using an updated version of DoPHOT (Schechter et al. 1993). This version works on any platform and accepts images consisting of real data values. It also provides better aperture corrections than previous versions, allowing for variations in aperture corrections as a function of field position and stellar magnitude (see Appendix A).

We also carried out an astrometric analysis of the cluster fields in our sample. We derived coordinates (right ascension α and declination δ) for all stars detected in our photometric study by comparison with bright stars obtained in each field from the Two Micron All Sky Survey (2MASS) catalog stars available through the Infrared Processing and Analysis Center (IPAC) Web site. In practice, more than 100 stars were available as astrometric references within the fields of every chip of our CCD camera. A third-order polynomial fit, done with the IRAF task mscpeak, produced dispersions of σ ∼ 025, consistent with the catalog precision. Using the astrometric information of the images we could also calculate if there were variations in the pixel area coverage across the images, which could lead to miscalculations in the measured fluxes. But the pixel area changed by less than 1% across any of the chips of the camera, and therefore we did not need to apply any corrections to the measured photometry.

To calibrate our ground-based data, we first transformed the photometry from each of the individual eight CCDs in the camera in IMACS to a common zero point, and then transformed this system to a standard system—in our case, the Johnson–Cousins system (Bessell 1979). We arbitrarily chose one of the CCDs, chip 2, as the one defining the instrumental system and brought the photometry from the other chips to this system (see Appendix B). After this was done, we placed our photometry in the Johnson–Cousins photometric system by observing Landolt (1992) fields over a range of airmasses during the nights of our 2005 observing run (see Appendix C). Since these nights were not completely photometric we used Stetson (2000) photometric standard stars and the photometry from the clusters obtained with the HST to fine-tune the calibration of our ground-based data (see Appendix C).

The photometry of the data from the HST observations was obtained using the programs HSTPHOT for the WFPC2 data (Dolphin 2000) and DolPHOT for the ACS data (Dolphin 2000). These programs have been specifically tailored to analyze data from these cameras. They take the images for every cluster retrieved from the HST in the different filters, and analyze them all at the same time. The programs align the images, combine them and provide the final combined-frame and individual-frame photometry, both in the HST filter system and in the Johnson–Cousins system. We follow the author's prescriptions to give the input parameters for the analysis to both programs. Photometric goodness-of-fit parameters were employed to select only objects with high-quality measurements. Only measurements for which an object was classified as stellar (HSTPHOT and DolPHOT types 1 and 2), and for which −0.1 ⩽ sharpness ⩽ 0.1, crowding ⩽ 0.5, and errors in all measured magnitudes σ ⩽ 0.1 were retained.

The changes in the extinction across the field are significant (ΔE(B − V) = 0.15). The highest extinction is located in the northwestern region of the field, while the northeastern region of the field shows the lowest. There is a small blob of higher extinction (ΔE(B − V) ∼ 0.05) ∼002 south of the center of the cluster, which the SFD map fails to show (see Figure 32). The absolute extinction zero point of our reddening map is E(B − V) = 0.50 from comparing our map with the SFD map. These significant values of extinction, both absolute and differential, are a consequence of the cluster being located behind the Ophiuchus dust complex, about 120 pc away from us.

M 4 is likely the closest GC to the Sun. Because of this, stars in the upper RGB are saturated in our images, even in the shortest exposures, and they do not appear in our CMDs of this cluster. The region studied for this cluster encompasses the whole FOV of our CCD, since the number of field stars present is small (<20% of the stars in our CMD). Due to the proximity of M 4 to the Sun, most of the field stars are located behind the cluster. They are generally dimmer than the cluster stars, and the evolutionary sequences of the field stars are easily differentiable from the cluster sequences. The definition of the different branches in the CMD of the cluster is improved after the dereddening process, especially in the V, V − I CMD. The evolutionary sequences get ∼35% tighter there. A noticeable feature in the CMDs of this cluster is the presence of a bimodal HB, with both red and blue components, though the blue HB (BHB) is not very extended. This bimodality of the HB, noted also in previous studies (e.g., Catelan et al. 1998, and references therein), suggests the presence of several stellar populations in this cluster.

Some recent ground-based optical photometric studies of this cluster include Rosenberg et al. (2000a) in VI, Mochejska et al. (2002) in UBV, and Anderson et al. (2006) in BV. Recent deep optical HST CMDs of this cluster have been presented in U₃₃₆V₅₅₅I₈₁₄ by Bassa et al. (2004), in V₅₅₅I₈₁₄ by Richer et al. (2004), and in V₆₀₆I₈₁₄ by Ferdman et al. (2004) and Hansen et al. (2004). This cluster is also a member of the ACS survey of GGCs obtained in the V₆₀₆I₈₁₄ bands (Anderson et al. 2008b). Recent infrared photometric studies of this cluster are by Cho & Lee (2002) in JK and by Ferraro et al. (2000) in VJK. Pulone et al. (1999) also used the HST for an optical and infrared study of the stellar mass function of this cluster in V₅₅₅I₈₁₄J₁₁₀H₁₆₀. The presence of significant differential reddening in this cluster has been pointed out in many studies, and a map of the differential extinction was provided by Mochejska et al. (2002; see Figure 33). Recently, the presence of distinct populations has been suggested from spectroscopic studies of stars in the RGB (Marino et al. 2008; Carretta et al. 2010) and in the HB (Marino et al. 2011b).

NGC 6144 is also located behind the ρ Ophiuchi dust cloud and very close (∼40' northeast) to NGC 6121 (M 4). Unsurprisingly, NGC 6144 also suffers from high and differential reddening. The variation of the extinction across the field is one of the highest shown in our sample (ΔE(B − V) = 0.5). Extinction is higher in the northern region of the observed field, ∼01 from the center of the cluster, and peaks in the northeastern area. The absolute extinction zero point of our reddening map is E(B − V) = 0.75 from comparing our map with the SFD map. This is one of the few cases in which the differential extinction variation across the field is more important (by a factor of two) in the SFD maps than in ours.

The region studied for this cluster encompasses the whole FOV of our CCD, since there is only a moderate field star contamination (∼50% of the stars in our CMD). Also the field stars are redder on average than the stars in the cluster whose different evolutionary sequences are clearly distinguishable in the CMD. Even if previous works (e.g., Martin & Whittet 1990; Green et al. 1992) give a higher value for the reddening vector toward the ρ Ophiuchi dust cloud, with R_V ∼ 4.3, still with the lower value assumed in our method, R_V ∼ 3.1, the different evolutionary sequences get better defined (∼25% tighter) in the CMD of the cluster after applying the dereddening process. This cluster shows a short BHB, which suggests the presence of a single population, or the dominance of one population.

The most recent ground-based optical photometric study of this cluster we found was done in BVI by Neely et al. (2000) for the RGB and HB stars. From HST data, the first CMD of this cluster was published by Sarajedini et al. (2007) as a part of their V₆₀₆I₈₁₄ ACS survey of GGCs. Lan et al. (2010) also provide CMDs in U₃₃₆V₆₀₆Hα₆₅₆R₆₇₅I₈₁₄ using HST data. Infrared upper RGB photometry has been presented for this cluster by Davidge (2000) in JHK and by Minniti et al. (1995) in JK.

The differential extinction across the field of this cluster is very mild (ΔE(B − V) ∼ 0.04). Extinction is higher in the northern region (∼01 of the cluster center). The absolute extinction zero point of our reddening map is E(B − V) = 0.18 from comparing with the SFD map.

The region studied for this cluster encompasses the whole FOV of our CCD, since the number of field stars present is small (<10% of the stars in our CMD). The mild differential extinction produces only a small improvement in the CMD of the cluster after being dereddened. The evolutionary sequences get ∼10% tighter in V, V − I, a little less in V, B − V. Some of the CMD features, like the RGB bump, are more clearly visible now. A significant feature of this cluster is its BHB with a blue tail, which suggests the presence of several populations of stars.

Ground-based VI photometric data on M 12 down to a few magnitudes under the TO have most recently been presented by von Braun et al. (2002) and Rosenberg et al. (2000b), BV data by Brocato et al. (1996), BVI data by Hargis et al. (2004), and deeper VR data by de Marchi et al. (2006). HST B₄₃₉V₅₅₅ data on M 12 were published by Piotto et al. (2002) using the WFPC2. This cluster is also a member of the ACS survey of GGCs in V₆₀₆I₈₁₄ (Anderson et al. 2008b). We have found no infrared photometrical study of this cluster. von Braun et al. (2002) mapped the extinction in this location, finding little differential reddening in the field (see Figure 33). Recently, Carretta et al. (2007) have suggested the presence of two distinct populations with different helium content from a spectroscopic study of RGB stars. Carretta et al. (2010) also find evidence for different populations from a spectroscopic study of stars in the RGB.

The changes in extinction across the field are moderate (ΔE(B − V) ∼ 0.1), being the southeastern area of the field the region of highest extinction. The absolute extinction zero point of our reddening map is E(B − V) = 0.39 from comparing our map with the SFD map. These moderately high values of extinction, both absolute and differential, are a consequence of the cluster's projected position, close to the Ophiuchus dust complex.

The region studied for this cluster includes only stars located less than 537 away from the cluster center, due to the high number of field stars relative to the GC stars (∼85% of the stars when we look at the whole FOV of IMACS). As with NGC 6144, the field stars are redder on average than the cluster stars. This allows us to easily differentiate the cluster and field evolutionary sequences. These sequences in the cluster get ∼15% tighter after being differentially dereddened. This cluster shows a BHB, not very extended, which suggests the presence of a single population, or the dominance of one population.

The most recent ground-based optical photometric study of this cluster we found was done in BV for RGB and HB stars by Howland et al. (2003), who found a differential extinction relation along the X-axis of their data and used it to correct their photometry. This cluster has also been observed with the WFPC2 on board the HST and its B₄₃₉V₅₅₅ CMD has been published in the GGC survey by Piotto et al. (2002). Infrared upper RGB photometry has been presented for this cluster by Davidge (2000) in JHK and by Minniti et al. (1995) in JK. Deeper JHK RGB near-infrared photometry was presented by Chun et al. (2010).

The extinction variation across the field of this cluster is mild (ΔE(B − V) ∼ 0.1). Extinction is higher toward the west, peaking close (∼002) to the cluster center. The absolute extinction zero point of our reddening map is E(B − V) = 0.28 from comparing our map with the SFD map.

The region studied for this cluster encompasses the whole FOV of our CCD, since the number of field stars present is small (<10% of the stars in our CMD). The CMD of this cluster, with an extended BHB, and its distance to the center of the Galaxy are very similar to those of M 12. The evolutionary sequences get ∼25% tighter after being differentially dereddened, and features like the RGB bump are more clearly visible now. The BHB with a blue tail suggests the presence of several stellar populations.

NGC 6254 has previously been studied using ground-based facilities by Rosenberg et al. (2000b) and von Braun et al. (2002) in VI and by Pollard et al. (2005) in BVI. Piotto & Zoccali (1999) used both ground-based facilities and HST for their study of this cluster in the VI bands. Beccari et al. (2010) also used the HST for their mass segregation study of this cluster in the V₆₀₆I₈₁₄ bands. This cluster is also a member of the ACS survey of GGCs in V₆₀₆I₈₁₄ (Anderson et al. 2008b). Valenti et al. (2004b) studied this cluster in the infrared JK bands. The presence of differential reddening has been previously noted, and von Braun et al. (2002) mapped the extinction in the cluster field, finding a higher differential extinction than in M 12 (see Figure 33). Carretta et al. (2010) find evidence for different populations from a spectroscopic study of stars in the RGB.

A band of material causing a significant increase in extinction (ΔE(B − V) ∼ 0.25) crosses the field in the east–west direction less than 002 south from the cluster center. The absolute extinction zero point of our reddening map is E(B − V) = 0.47 from comparing our map with the SFD map. In the SFD map the presence of this band of extinction is not so clear.

The area studied for this cluster is restricted to the inner 684 of the cluster, due to the number of field stars (∼50% of the stars when we look at the whole FOV of IMACS) and the relatively small tidal radius (r_t = 1128). The definition of the different branches in the CMD of the cluster is highly improved, approximately by a factor of two, after the dereddening process. Also now the presence of an extended blue horizontal branch (EBHB) is clearly visible and the location of the RGB bump is more easily identifiable. A striking feature in the CMD of this cluster is the presence of an extended and well-populated HB, with both red and blue components. The BHB in particular is very extended. The shape of the HB and the fact that this is one of the most massive clusters in the Milky Way make this cluster a clear candidate to have several stellar populations.

Previous work using ground-based data includes that of Rosenberg et al. (2000a) in VI and Gerashchenko & Kadla (2004), Contreras et al. (2005), and Contreras et al. (2010) in BV. Beccari et al. (2006) studied this cluster using a combination of BVI (ground) and U₂₅₅U₃₃₆V₅₅₅ (space) bands. Cocozza et al. (2008) studied a pulsar in M 62 and used their HST data to show the B − R versus R CMD of the cluster. NGC 6266 is also included in the HST GGCs survey in B₄₃₉V₅₅₅ published by Piotto et al. (2002). Valenti et al. (2007) and Chun et al. (2010) studied this cluster in the infrared JHK bands. The presence of differential reddening has been previously noted, and Gerashchenko & Kadla (2004) mapped the extinction toward the central region of the cluster (see Figure 33).

The changes in the extinction across the field are significant (ΔE(B − V) ∼ 0.3). Extinction is higher toward the east, and peaks in a small region located very close, less than 002, to the center of the cluster. The absolute extinction zero point of our reddening map is E(B − V) = 0.32 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 826 of the cluster, due to the number of field stars (∼40% of the stars when we look at the whole FOV of IMACS). The definition of the different branches in the CMD of the cluster is highly improved after the dereddening process. The different sequences get tighter by approximately a factor of two after being differentially dereddened, and now the presence of an EBHB is clearly visible and the location of the RGB bump is more easily identifiable. The EBHB suggests the presence of several stellar populations in this cluster.

The most recent optical photometric study of this cluster we found was done by Piotto et al. (1999) in B₄₃₉V₅₅₅ from HST images. As part of their study, they obtained a map of the differential reddening of the central region of the cluster. Piotto et al. (2002) included the CMD from this work in their HST GGCs survey in B₄₃₉V₅₅₅. In the infrared, this cluster was studied in JHK by Chun et al. (2010), Davidge (2001), and Valenti et al. (2007).

The variation of the extinction across the field is one of the highest shown in our sample (ΔE(B − V) = 0.75). Extinction is higher in the northern region. A narrow band of material causing the highest extinction in the field crosses it in the east–west direction less than 002 from the cluster center. The absolute extinction zero point of our reddening map is E(B − V) = 0.67 from comparing our map with the SFD map. These significant values of extinction, both absolute and differential, are a consequence of the cluster's projected proximity to the Ophiuchus dust complex.

The area studied for this cluster is restricted to the inner 413 of the cluster due to the number of field stars (∼75% of the stars when we look at the whole FOV of IMACS). The definition of the different branches in the CMD of the cluster is highly improved, approximately by a factor of two, after the dereddening process. Now the RGB and SGB are easily identifiable, which was not the case before applying the dereddening technique. A short BHB is observed in the CMD, suggesting the presence of a single population, or the dominance of one population.

Previous photometric studies in the optical were done in BV by Stetson & West (1994) from ground-based observations and in VI by Fullton et al. (1999) from space-based observations. NGC 6287 is included in the HST GGCs survey in B₄₃₉V₅₅₅ published by Piotto et al. (2002). In the infrared, Chun et al. (2010) and Davidge (2001) studied this cluster in JHK and Lee et al. (2001) in J₁₁₀H₁₆₀ using the HST.

The variation of the extinction across the field is moderate (ΔE(B − V) = 0.15). The area located ∼005 east of the center of the cluster is the region of highest extinction. The absolute extinction zero point of our reddening map is E(B − V) = 0.49 from comparing our map with the SFD map.

The area studied for this cluster only includes stars with distances less than 365 away from the cluster center, due to the high number of field stars relative to the GC stars (>90% of the stars when we look at the whole FOV of IMACS). The improvement of the CMD of this cluster is only marginal. The CMD of this cluster shows a very short, red HB (RHB), typical of high-metallicity GGCs.

This cluster has been previously observed and studied using ground-based facilities by Ortolani et al. (2000) in BV and Rosenberg et al. (2000a) in VI. It is one of the clusters in the GGCs survey by Piotto et al. (2002) in B₄₃₉V₅₅₅ and in the ACS GGCs survey in V₆₀₆I₈₁₄ (Anderson et al. 2008b) done using HST data. Valenti et al. (2005) studied this cluster in the infrared JHK bands. Valenti et al. (2007) include results from this study in their infrared compilation of bulge GGCs.

The changes in the extinction across the field are significant (ΔE(B − V) ∼ 0.25). The area located ∼01 southwest of the cluster center is the region of highest extinction. The absolute extinction zero point of our reddening map is E(B − V) = 0.43 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 559 of the cluster due to the number of field stars (∼45% of the stars when we look at the whole FOV of IMACS) and the relatively small tidal radius (r_t = 816). The evolutionary sequences of M 9 get ∼35% tighter after being differentially dereddened. This cluster possesses a BHB with a blue tail, which suggests the presence of several stellar populations in this cluster.

The most recent optical photometric study of this cluster we found is that of Janes & Heasley (1991) in BV for stars brighter than the TO point. Although this cluster has been previously observed with the HST (see Table 4), the first CMD of M 9 using space data is published in this paper in BVI. In the infrared, Davidge (2001) and Chun et al. (2010) studied this cluster in JHK.

The changes in the extinction across the field are significant (ΔE(B − V) ∼ 0.4). The northwestern quadrant of the field, reaching the projected position of the center of the cluster, is the region of highest extinction. The absolute extinction zero point of our reddening map is E(B − V) = 0.59 from comparing our map with the SFD map.

The area studied for this cluster reaches only stars with distances less than 421 away from the cluster center, due to the significant number of field stars relative to GC stars (∼70% of the stars when we look at the whole FOV of IMACS). The definition of the different branches in the CMD of the cluster is highly improved after the dereddening process, getting tighter by a factor of two. Now the RGB and SGB are easily identifiable, along with an RHB, which was not the case before applying the dereddening technique. The RHB is typical of high-metallicity GGCs.

The only recent optical photometric studies found for this cluster are those of Heitsch & Richtler (1999) in VI using ground-based facilities and that of Piotto et al. (2002) in B₄₃₉V₅₅₅ using the HST. In the infrared, Valenti et al. (2004a) studied this cluster in VJHK. Valenti et al. (2007) include results from this study in their infrared compilation of bulge GGCs. A map of the differential reddening was produced by Heitsch & Richtler (1999; see Figure 33).

The differential extinction variation across the field of this cluster is only moderate (ΔE(B − V) ∼ 0.15). Extinction is higher in the northern region of the analyzed field. The absolute extinction zero point of our reddening map is E(B − V) = 0.34 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 561 of the cluster, due to the high number of field stars relative to the GC stars (>90% of the stars when we look at all the available FOV). The evolutionary sequences get ∼30% tighter after being differentially dereddened. The cluster shows an RHB, typical of high-metallicity GGCs.

This cluster was studied by Sarajedini & Norris (1994) in BV and Rosenberg et al. (2000a) and Pancino et al. (2010) in VI from the ground, and from space by Fullton et al. (1995) in VI_C, using pre-COSTAR data, and by Faria & Feltzing (2002) in V₅₅₅I₈₁₄ and by Pulone et al. (2003) in V₆₀₆I₈₁₄. It is also a member of the ACS survey of GGCs in V₆₀₆I₈₁₄ (Anderson et al. 2008b). We have found no infrared photometrical study of this cluster. Pancino et al. (2010) show the presence of two populations of stars with anticorrelated CN and CH band strengths.

The changes in the extinction across the field are significant (ΔE(B − V) ∼ 0.3). The southwestern area of the cluster is the region of highest extinction. The absolute extinction zero point of our reddening map is E(B − V) = 1.20 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 236 of the cluster, due to the high number of field stars relative to the GC stars (>90% of the stars when we look at all of the available FOV). The evolutionary sequences get ∼10% tighter after being differentially dereddened. This cluster possesses a BHB, although the small number of stars in the HB, due to the small studied field, makes it difficult to infer any information about different populations in this cluster.

The only previous recent optical photometric studies found are those by Ortolani et al. (2003) in BVI using ground-based facilities and those by Piotto et al. (2002) in B₄₃₉V₅₅₅ using the HST. In the infrared, Chun et al. (2010), Davidge (2001), and Valenti et al. (2007) studied this cluster in JHK.

The differential extinction in this cluster is only mild (ΔE(B − V) ∼ 0.05). Extinction is not especially concentrated in any area of the field. The absolute extinction zero point of our reddening map is E(B − V) = 0.18 from comparing our map with the SFD map.

This is the cluster with the smallest apparent distance modulus, though it is not the one closest to the Sun. As for M 4, stars in the upper RGB are saturated in our images, even in the shortest exposures, and they do not appear in our CMDs of this cluster. The contamination from field stars is moderate (∼30% of the stars when we look at all the available FOV), mainly affecting MS stars in the cluster. This moderate contamination allows us to study almost all the available FOV, up to 818 from the cluster center. The sequences of NGC 6397 get ∼10% tighter after being differentially dereddened. This cluster shows a short BHB, which suggests the presence of a single population, or the dominance of one population.

Some of the most recent studies of this cluster done with ground-based facilities are those by Rosenberg et al. (2000a) in VI, Anderson et al. (2006) in UBVI, and Kaluzny et al. (2006) in BV. This cluster has also been observed with the HST. A series of papers has been published dedicated to the study of the characteristics of this cluster using very deep photometry obtained with the ACS (Anderson et al. 2008a; Hurley et al. 2008; Richer et al. 2008; Davis et al. 2008). NGC 6397 is also included in the GGC survey from Piotto et al. (2002) in B₄₃₉V₅₅₅ and of the ACS GGC survey in V₆₀₆I₈₁₄ (Anderson et al. 2008b). de Marchi et al. (2000) also used the HST for an optical and infrared study of the stellar mass function of this cluster in V₆₀₆I₈₁₄J₁₁₀H₁₆₀. Recent spectroscopic studies of this cluster have found different chemistry among its stars: Carretta et al. (2010) show that some O–Na anticorrelation is present, although the number statistics is small (only 16 stars studied, the smallest in their sample of GCs); Lind et al. (2011) claim the presence of two different stellar populations that have different chemical compositions of N, O, Na, Mg, and probably Al, and note that the cluster is clearly dominated (75%) by the second generation.

The differential extinction variation in this cluster is moderate (ΔE(B − V) = 0.1), although we should realize that the studied region is also small (see next paragraph). Extinction is higher in the southwestern region of the analyzed field. The absolute extinction zero point of our reddening map is E(B − V) = 0.58 from comparing our map with SFD map.

This cluster is located in Baade's window. The high number of non-cluster member stars (>90% of the stars when we look at all of the available FOV) restricts the study in this region to stars closer than 233 from the cluster center. The improvement of the CMD of this cluster after being dereddened is only marginal. It shows a BHB with a blue tail, which suggests the presence of several stellar populations in this cluster.

The most recent optical photometric studies found, done with ground-based data, are those by Barbuy et al. (1994) in BV and Terndrup & Walker (1994) in BVI. Terndrup et al. (1998) show a differential extinction relation with respect to the center of the cluster and use it to correct their previous photometry. Using the HST, the cluster was studied by Shara et al. (1998) in the U₃₃₆B₄₃₉ bands using pre-COSTAR data. It is also one of the clusters in the Piotto et al. (2002) GCs survey in B₄₃₉V₅₅₅. In the infrared, Davidge (2001) and Valenti et al. (2010) studied this cluster in JHK.

The differential extinction in this cluster is moderate (ΔE(B − V) ∼ 0.1). Extinction is not especially concentrated in any area of the field, although there seems to be an increase toward the southeast area of the observed field. The absolute extinction zero point of our reddening map is E(B − V) = 0.16 from comparing our map with the SFD map.

The region studied for this cluster encompasses most of the FOV of our CCD, up to 863 from the cluster center, since there is only a moderate field star contamination (∼40% of the stars in our CMD). Also the field stars are redder on average than the stars in the cluster, which lets us distinguish NGC 6541 evolutionary sequences unambiguously. The sequences in the cluster get ∼10% tighter after being differentially dereddened. This cluster possesses a BHB with a blue tail, which suggests the presence of several stellar populations.

Ground-based studies of this cluster were presented by Alcaino et al. (1997) in BVRI and by Rosenberg et al. (2000a) in VI. It has also been studied using HST WFPC2 data by Lee & Carney (2006) in VI, and it is one of the clusters in the ACS GGC survey observed in V₆₀₆I₈₁₄ (Anderson et al. 2008b). In the infrared, Kim et al. (2006) studied this cluster in JHK.

There are significant changes in the extinction across the field (ΔE(B − V) ∼ 0.25). In our maps, the highest extinction is located in the northern region of the field, while the southwestern region of the field shows the lowest, which is not the trend shown in the SFD maps, but it does agree well with the trend of the map provided by Heitsch & Richtler (1999; see Figure 33). The absolute extinction zero point of our reddening map is E(B − V) = 1.35 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 391 of the cluster, due to the high number of field stars relative to GC stars (>90% of the stars when we look at all of the available FOV). The upper MS and RGB get ∼30% tighter after being differentially dereddened, and the RGB bump now is more clearly identifiable. The well-populated bump, the very extended upper RGB, and an RHB suggest that this is a metal-rich GGC.

Some of the most recent optical photometric studies of this cluster are that by Sagar et al. (1999) in VI using ground-based data and those by Beaulieu et al. (2001) and Zoccali et al. (2001) in V₅₅₅I₈₁₄ using HST data. Zoccali et al. (2001) calculated the extinction map, not shown in their paper, and applied it to get a clean CMD. Heitsch & Richtler (1999) show an extinction map (see Figure 33) calculated with data from Sagar et al. (1999). Guarnieri et al. (1998) in VIJK and Ferraro et al. (2000) in VJK studied this cluster in the optical and in the infrared. Valenti et al. (2007) include results from this last study in their infrared compilation of bulge GGCs.

The extinction variations in this cluster are significant (ΔE(B − V) ∼ 0.15), especially when we consider the small size of the studied field (see next paragraph). The absolute extinction zero point of our reddening map is E(B − V) = 0.51 from comparing our map with SFD map.

The high number of non-cluster member stars (>90% of the stars when we look at the whole FOV of IMACS) restrict the study of this region to stars closer than 181 from the cluster center. The improvement of the CMD is only marginal. This cluster shows a BHB, although the low definition of the HB due to the small number of stars in this region caused by the small field FOV studied makes it difficult to infer any information about different populations in this cluster.

We found only two previous optical photometric studies using ground-based data, those by Rich et al. (1998) and by Barbuy et al. (2007), both in VI. Although this cluster has been previously observed with the HST, the first CMD of the deep MS stars using space-based data is published in this paper in VI. In the infrared, Chun et al. (2010) studied this cluster in JHK.

The differential extinction across the field is moderate (ΔE(B − V) ∼ 0.15). Extinction is higher in the southeastern region of the analyzed field. The absolute extinction zero point of our reddening map is E(B − V) = 0.25 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 392 of the cluster, due to the high number of field stars relative to the GC stars (>90% of the stars when we look at all of the available FOV). There are also some dim blue field stars in our CMD that belong to the Sagittarius dwarf galaxy. Stars from this galaxy's MS are easily observed in the CMD at magnitudes dimmer than V ∼ 21 and colors bluer than the cluster MS. The improvement of the CMD of the cluster after being dereddened is only marginal. This cluster possesses an RHB, characteristic of a metal-rich GGC.

This cluster has been studied using ground-based data by Sarajedini & Norris (1994), who presented a photometric study of the RGB and HB stars in BV. Richtler et al. (1994) and Rosenberg et al. (2000a) obtained deeper photometry, reaching a few magnitudes below the TO in BV and VI, respectively. Using HST, studies of this cluster were published by Sosin & King (1995) in BV using pre-COSTAR data, and by Guhathakurta et al. (1996) in U₃₃₆B₄₃₉V₅₅₅ and Heasley et al. (2000) in VI. This cluster is also a member of the Piotto et al. (2002) B₄₃₉V₅₅₅ GGCs survey and of the V₆₀₆I₈₁₄ ACS GGCs survey (Anderson et al. 2008b). In the infrared, Valenti et al. (2004a) studied this cluster in VJHK. Valenti et al. (2007) include results from this study in their infrared compilation of bulge GGCs.

The extinction variations across the field of this cluster are significant (ΔE(B − V) ∼ 0.2). Extinction is lower at the projected center and southwestern region of the observed field, which is not the trend observed in the SFD maps. The absolute extinction zero point of our reddening map is E(B − V) = 0.46 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 495 of the cluster, due to the high number of field stars relative to the GC stars (>90% of the stars when we look at all the available FOV). Due to its proximity, the cluster's brightest stars are brighter on average than the field's, and we can identify without much difficulty stars from the RGB. The evolutionary sequences of M 28 get ∼35% tighter after being differentially dereddened. This cluster shows a BHB with a blue tail, which suggests the presence of several stellar populations.

This cluster has been studied in the optical using ground-based data by Rosenberg et al. (2000a) in VI, and using HST data by Testa et al. (2001) and by Golden et al. (2001) in VI. Davidge et al. (1996) used a combination of optical and infrared bands (BVJK) to study this cluster. In the infrared, Chun et al. (2010) studied this cluster in JHK.

The differential extinction across most of the observed field is moderate (ΔE(B − V) ∼ 0.1). Extinction is not especially concentrated in any area of the field. The absolute extinction zero point of our reddening map is E(B − V) = 0.17 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 439 of the cluster, due to the high number of field stars relative to the GC stars (>90% of the stars when we look at all of the available FOV). There are also some dim blue field stars in our CMD that belong to the Sagittarius dwarf galaxy. Stars from this galaxy's MS are easily observed in the CMD at magnitudes dimmer than V ∼ 21 and colors bluer than the cluster MS. The improvement of the CMD of this cluster is only marginal. This cluster possesses an RHB, typical of high-metallicity GGCs.

This cluster has been the subject of several studies, usually along with the metal-rich NGC 6624. Using ground-based data Ferraro et al. (1994), in BVJK, and Sarajedini & Norris (1994), in BV, present a photometric study of the RGB and HB stars of this cluster, and Richtler et al. (1994) and Rosenberg et al. (2000a) got deeper photometry, reaching a few magnitudes below the TO, in BV and VI, respectively. Using HST data, Heasley et al. (2000) studied this cluster in VI. This cluster is also a member of the Piotto et al. (2002) B₄₃₉V₅₅₅ GGCs survey, and of the V₆₀₆I₈₁₄ ACS GGCs survey (Anderson et al. 2008b). In the infrared, this cluster was studied by Ferraro et al. (2000) in VJK and by Valenti et al. (2005) in JHK. Valenti et al. (2007) include results from these studies in their infrared compilation of bulge GGCs.

The differential extinction in this cluster is high (ΔE(B − V) ∼ 0.2). The absolute extinction zero point of our reddening map is E(B − V) = 0.36 from comparing our map with the SFD map.

During the observation of this cluster with the Magellan Telescope, we had the least photometric conditions of our run (see Table 11 in the Appendix). This resulted in the shallowest CMD of our study, reaching only stars down to the very upper parts of the MS. The high number of non-cluster member stars (>90% of the stars when we look to all the available FOV) restricts the study in this area to stars closer than 222 from the cluster center. The improvement of the CMD of this cluster after being dereddened is only marginal. A prominent feature in the CMDs of this cluster is the presence of a bimodal HB, with both red and blue components, though the blue BHB is not very extended. This bimodality of the HB can suggest the presence of several stellar populations.

The most recent optical photometric study of NGC 6642 using ground-based data was done by Barbuy et al. (2006) in BVI. Using HST, Balbinot et al. (2009) studied this cluster in V₆₀₆I₈₁₄. NGC 6642 is also a member of the Piotto et al. (2002) B₄₃₉V₅₅₅ GGCs survey. In the infrared, Davidge (2001), Kim et al. (2006), and Valenti et al. (2007) studied this cluster in JHK.

The extinction variation across the field of this cluster is moderate (ΔE(B − V) ∼ 0.1). Extinction is lower in the southern area of the field. The absolute extinction zero point of our reddening map is E(B − V) = 0.33 from comparing our map with the SFD map.

The relative proximity of M 22 explains why the upper RGB stars are saturated in our observations and are excluded from the CMDs of this cluster. The region studied for this cluster encompasses the whole FOV of our CCD, since there is only a moderate field star contamination (∼40% of the stars in our CMD). Due to the proximity of M 22, most of the field stars are located behind the cluster. They are generally dimmer than the cluster stars, and the evolutionary sequences of the field stars are easily differentiable from the brighter sequences of the cluster. After being dereddened, the sequences of the cluster get ∼25% tighter in V, V − I, a little less in V, B − V. An EBHB suggests the presence of several stellar populations in this cluster.

There are several optical photometric studies for this cluster. Using ground-based data, it was studied by Monaco et al. (2004) in BVI, by Webb et al. (2004) in UV, by Kaluzny & Thompson (2001) in BV, by Rosenberg et al. (2000a) in VI, and by Anthony-Twarog et al. (1995), Richter et al. (1999), and Lee et al. (2009) in Strömgrem uvbyCa, vby, and VbyCa bands, respectively. Piotto & Zoccali (1999) used both ground facilities and the HST for their study of this cluster in VI, and Anderson et al. (2003) show CMDs for the cluster in B₄₃₅V₆₀₆R₆₇₅I₈₁₄ bands. This cluster is also a member of the ACS survey of GGCs in V₆₀₆I₈₁₄ (Anderson et al. 2008b). Piotto (2009) shows a double SGB in his study of M 22 using V₆₀₆I₈₁₄. In the infrared, Minniti et al. (1995), Davidge & Harris (1996), and Cho & Lee (2002) studied this cluster in JK. The presence of a metallicity spread in this cluster has been the subject of a long debate, complicated by the presence of interstellar reddening (see Pancino et al. 2010 for a brief historic review). Some recent studies seem to confirm the presence of at least two stellar populations with different chemical compositions, both photometrically (Piotto 2009) and spectroscopically (Marino et al. 2009, 2011a; Da Costa et al. 2009; Carretta et al. 2010), although the studies by Kayser et al. (2008) and Pancino et al. (2010) do not find clear CH and CN anticorrelations or bimodalities for the stars in their sample.

The differential extinction across most of the observed field is mild (ΔE(B − V) ∼ 0.05). Extinction is not especially concentrated in any area of the field. The absolute extinction zero point of our reddening map is E(B − V) = 0.09 from comparing our map with the SFD map.

The area studied for this cluster is restricted to the inner 521 of the cluster, due to the high number of Galactic field stars relative to the GC stars (∼70% of the stars when we look to all the available FOV). The field stars are redder on average than the cluster stars, which allows us to easily distinguish the different evolutionary sequences. There are also some dim blue field stars in our CMD that belong to the Sagittarius dwarf galaxy. Stars from this galaxy's MS are easily observed in the CMD at magnitudes dimmer than V ∼ 21 and colors bluer than the cluster MS. The improvement of the CMD of this cluster is only marginal. M 70 possesses a BHB with a blue tail, suggesting the presence of several stellar populations in this cluster.

The most recent optical photometric studies we found that used ground-based data are those by Brocato et al. (1996) in BV and by Rosenberg et al. (2000a) in VI. Using HST data, Piotto et al. (2002) included it in their GGCs survey in B₄₃₅V₅₅₅. This cluster is also a member of the V₆₀₆I₈₁₄ ACS survey of GGCs (Anderson et al. 2008b). In the infrared, Kim et al. (2006) studied this cluster in JHK.

The extinction variation across the field of this cluster is only mild (ΔE(B − V) ∼ 0.05). Extinction is higher in the northwestern area of the observed field. The improvement of the CMD of this cluster is only marginal. The absolute extinction zero point of our reddening map is E(B − V) = 0.14 from comparing our map with the SFD map.

The region studied for this cluster encompasses the whole FOV of our CCD, since the number of field stars present is small (<10% of the stars in our CMD), though there is clear contamination from stars in the background Sagittarius dwarf galaxy. Stars from its MS are easily observed in the CMD at magnitudes dimmer than V ∼ 21 and colors bluer than the cluster MS. The tip of the RGB is saturated in our CMDs of this cluster, due to the small apparent distance modulus of M 55. The improvement of the CMD of this cluster is only marginal. The cluster shows a BHB, not very extended, suggesting the presence of a single population, or the dominance of one population.

The latest optical photometric studies of this cluster done using ground-based data are by Kaluzny et al. (2005) in UBVI, by Pych et al. (2001) and by Zloczewski et al. (2011) in BV, and by Rosenberg et al. (2000a) in VI. Using a combination of ground-based and HST data, this cluster was studied by Lanzoni et al. (2007) in BVI and by Bassa et al. (2008) in BVRIH_α. The ACS VI photometry used by Bassa et al. (2008) comes from Anderson et al. (2008b), since this cluster is one of the members of the ACS GGC survey. In the infrared, this cluster was studied by Ferraro et al. (2000) in VJK. The presence of more than one stellar population has been suggested by Carretta et al. (2010) based on the Na–O anticorrelation found in their study, but the studies by Kayser et al. (2008) and by Pancino et al. (2010) do not find clear CH and CN anticorrelations or bimodalities on the stars in their sample.

In this work, we have obtained the photometry using a new version of DoPHOT. DoPHOT is a software created to extract point-spread function (PSF) photometry from astronomical images. A complete and deep explanation of its main characteristics and a description of how it works are given by Schechter et al. (1993).

The main new characteristics of the distribution of DoPHOT we developed for this work are as follows.

• 1.  
  All the program subroutines and functions have been rewritten to make them compatible with the standards of Fortran90.
• 2.  
  We have tried to make the subroutines and functions more easily readable. To achieve this, we have eliminated all the common statements present in previous versions and named all the variables used by the subroutine and function in the call statement. Also with the same spirit of making the program more understandable we have created, combined, or eliminated some of the subroutines, and included, at the beginning of each file, a brief explanation of what every function and subroutine does.
• 3.  
  The program works now with real values for the pixel values, so there is no need to worry about conversions to integers as in previous distributions.
• 4.  
  The stars are masked, obliterated, added, and subtracted across the different subroutines of the program in circular sections, and not in rectangular sections as before. This way fewer usable pixels are obliterated, and fewer pixels outside an interesting radius are dealt with.

In addition to this, this distribution shares some of the characteristics of the program that were present in previous distributions, but that were not mentioned in the original paper.

• 1.  
  DoPHOT allows for a variable PSF model in the image.
• 2.  
  DoPHOT can perform an aperture correction to the magnitudes obtained, according to their positions and magnitudes. The program checks for variations in the aperture correction in magnitude and in position in the image, and corrects the fitted magnitude accordingly.

This new version of DoPHOT is publicly available at http://www.astro.puc.cl:8080/astropuc/software-hecho-en-daa.

The camera in IMACS consists of eight chips, each with its own slightly different mean quantum efficiency and color sensitivity. To calibrate our data, we must first transform the photometry from each of the individual chips to a common zero point, then transform this system to a standard system—in our case, the Johnson–Cousins system (Bessell 1979). In this appendix, we shall focus on the first step of transforming the individual chips to a common instrumental system.

For this purpose, we arbitrarily chose one of the chips, chip 2 (see Figure 34), as the one defining the instrumental system. The goal was to tie all the other chips to this chip's photometric zero point. We attempted to do this in two ways. First, we observed one Landolt (1992) field (PG 1047+003) in every one of the chips and in all three filters (BVI) in rapid succession on every night of our run. There were very few stars in this field, and the scatter we found in the mean photometric offsets—about 0.05 mag between chips—was too large given our need to bring all our photometry to a common system to 1%–2% precision. We surmise that the relatively poor results from this approach reflects the quality of the nights during the 2005 run, and also possible flat-field variations at the 2%–3% level for this run.

Zoom In Zoom Out Reset image size

We tried a second approach during a sequence of observations during a 2006 Magellan run (see Table 2) in which we tried to autocalibrate the IMACS chip array using stars from the cluster NGC 6397. The basic idea was to obtain a minimal number of exposures in each filter so that at least part of all chips overlapped with chip 2, the chip defining our instrumental photometric system. We found that a sequence of three exposures could achieve the desired overlap. Figure 34 illustrates the approach. In the left panel, the darker outline of the IMACS chip shows the location of a first exposure, in this case at the same location as the actual field we observed in 2005 for the cluster NGC 6397. The second exposure, shown as a lighter (red) outline, is offset in both R.A. and decl. so that four of its chips (1, 2, 5, and 6) overlap with the position of chip 2 in the first exposure. The right panel again shows the initial pointing in the darker outline, with a third pointing as a lighter (green) outline. Now, chips 3, 4, 7, and 8 of the last exposure overlap with chip 2 from the first exposure. This sequence, repeated for each filter, generated a large list of stars that have been observed on chip 2 and on every other chip in the array. Because only three exposures were required and because we took these in rapid succession, we can safely ignore airmass variations as we calculate the offsets from a given chip to the system defined by chip 2. In addition, we have exposures from the 2005 and 2006 runs of the same fields (the dark outlines in Figure 34) that allow us to determine any possible variations in chip sensitivities between runs.

The model we used to transform each chip to the chip 2 system has the form

$$\begin{equation} m_2 = m_i + Z_i + \gamma _i c_i, \end{equation} \tag{ B1 }$$

where m_i is the instrumental magnitude of chip i (excluding i = 2), Z_i is the zero-point offset from chip i to chip 2, and γ_i is the color coefficient for instrumental color c_i. In practice, we carried out this analysis separately for each filter, B, V, and I, with the colors defined appropriately for each filter ((B − V) for the B-band transformation and (V − I) for the V- and I-band transformations). We found that the color terms are all negligibly small in the chip-to-chip transformations for each filter, so the final model adopts γ_i = 0 for all chips and all colors. Thus we only had to determine the mean values of the Z_i coefficients as the weighted mean of the differences of the instrumental magnitudes, m_i − m₂, for all stars in common between chip i and chip 2. Weights for each measurement were assigned according to the photometric errors and the dispersion in the offset distribution. A first application of this approach, to the calibration sequence of NGC 6397 obtained in 2006 (Figure 34), allowed us to place all the data from that run on the 2006 chip 2 instrumental system.

Since the relative sensitivities of the chips might have changed between the 2005 and 2006 runs, we then took an additional step to define the chip-to-chip offsets for the 2005 run from which all our ground-based cluster photometry is defined. This process involves the observations of NGC 6397 from our 2006 run (described above) and the observations of the same field for the same cluster obtained in 2005.

For this purpose we adopted the following transformation:

$$\begin{equation} m_{i,06} = m_{i,05} + Y_i + \delta _i c_i + k_{05}X_{05} -k_{06}X_{06}. \end{equation} \tag{ B2 }$$

The zero-point offset from the 2005 observations required to place them on the 2006 system for chip i is given by Y_i (this now includes the case i = 2), for a color term δ_i, and an instrumental color on the 2005 system of c_i. The final two terms account for the different airmasses and possibly different extinction coefficients for each run and each observation. We do not know this product (extinction coefficient k times airmass X) for the 2006 run because we did not observe standard stars on that run. But we know that the sum of these products is constant when comparing the 2005 and 2006 NGC 6397 observations. Thus, we can rewrite the equation above as

$$\begin{equation} m_{i,06} = m_{i,05} + Y^{\prime }_i +\delta _i c_i, \end{equation} \tag{ B3 }$$

in close analogy to Equation (B1). Note that the new zero-point offset now implicitly includes the airmass terms:

$$\begin{equation} Y^{\prime }_i = Y_i + k_{05}X_{05} - k_{06}X_{06}. \end{equation} \tag{ B4 }$$

We can simplify this due to the fact that we found that the color terms in Equation (B3) are all negligibly small. Thus we only had to determine the mean values of the Y'_i coefficients as the weighted mean of the differences of the instrumental magnitudes, m_{i, 06} − m_{i, 05}, for all stars from chip i. As before, weights for each measurement were assigned according to the photometric errors and the dispersion in the offsets distribution.

At this stage, we can transform the 2005 NGC 6397 data to a common instrumental system defined by chip 2 from the 2006 observations of this same cluster. But what we really require are the offsets in the 2005 system that bring all data from the IMACS chips obtained in 2005 to a common system. That is, the final chip-to-chip offsets for the 2005 data are simply

$$\begin{equation} \Delta Y^{\prime }_i = Y^{\prime }_i - Y^{\prime }_2. \end{equation} \tag{ B5 }$$

The values of Z and ΔY' required to tie the various chips systems to a common system defined by chip 2 are listed in Table 9. The errors in the values of Z and ΔY' are σ ∼ 0.001, the error of the mean of the distributions used to obtain them (see Figure 35). The dispersions of the distributions from which these Z and ΔY' are derived, a measure of the internal precision of our photometry, are σ ∼ 0.01.

Zoom In Zoom Out Reset image size

We placed our photometry in the Johnson–Cousins photometric system by observing Landolt (1992) fields over a range of airmasses during the nights of our 2005 observing run. Since the first two nights were the most photometric of our run, we only used the standard fields observed those two nights to obtain the following transformations:

$$\begin{equation} B = 3.417 + b - 0.283 X + 0.130 (b-v), \end{equation} \tag{ C1 }$$

$$\begin{equation} V = 3.706 + v - 0.166 X - 0.017 (b-v), \end{equation} \tag{ C2 }$$

$$\begin{equation} I = 3.065 + i - 0.049 X - 0.020 (v-i), \end{equation} \tag{ C3 }$$

where X is the airmass, b, v, and i are the instrumental magnitudes in the instrumental system and B, V, and I the corresponding magnitudes in the Johnson–Cousins photometric system.

To fine-tune our calibration we decided to compare our photometry with Stetson (2000) photometric standard stars. We found in some cases significant absolute offsets between Stetson's and our photometry. These zero points could be easily calculated (see Table 10 and Figure 36). However, only 13 of our clusters had stars in common with Stetson's catalog, and only 12 had stars calibrated in all three filters.

Zoom In Zoom Out Reset image size

To extend the comparison to more clusters, we looked for members of our sample already observed with at least one of the two HST's optical wide-field imaging instruments, WFPC2 and ACS. We found that all of the clusters have already been observed at least in two of our filters, and most of them in all three (see Table 10). We retrieved the available data from the HST archive and derived photometry from the stars in these data using the programs HSTPHOT for the WFPC2 data (Dolphin 2000) and DolPHOT for the ACS data (Dolphin 2000). These programs have been specifically tailored to analyze data from these cameras.

The comparison between the HST and ground-based data is complicated by the large resolution difference in the data sets and the highly crowded fields. When the offsets in the photometries of individual stars of a cluster are plotted, there is a clear, asymmetric spread in the data (see Figure 37). Due to the higher resolution of the HST, individual stars in the HST data are blends in the ground-based data. To obtain the absolute offset between the two systems eliminating the blended stars effect we calculated a weighted mean of the offsets of the stars in common (see Figure 37), where the weights were assigned according to the photometric errors and the dispersion in the offset distribution. The weighted mean was iteratively calculated, ignoring stars 2σ away from the mean value at each iteration. Whenever we have zero points calculated separately from independent WFPC2 and ACS data, a mean value was adopted.

Zoom In Zoom Out Reset image size

The results were then compared with the ones we found from the Stetson comparison, since our aim is to put all the photometry in the Stetson system (see Figure 38). A small offset for each filter was found and added to the individual cluster zero points (ΔB = −0.045 ± 0.012 mag, ΔV = −0.056 ± 0.007 mag, and ΔI = −0.025  ±  0.008 mag). These offsets may be due to the fact that the HST observations used in our absolute calibration are located closer to the center of the cluster than the Stetson stars used, or they may reflect small system zero-point differences. The final offsets applied to the individual clusters can be seen in Table 11.

Zoom In Zoom Out Reset image size