3 The globular cluster Terzan 9

Terzan 9 is also designated GCl B1758-268 and ESO521-SC11, and is located at $\alpha_{1950}$ = 17$^{\rm h}$ 58$^{\rm m}$31.2$^{\rm s}$,$\delta_{1950}=-26^\circ50'23\hbox{$^{\prime\prime}$}$ ($l=3.603^\circ$, $b=-1.989^\circ$). Malkan (1982) derived E(B-V)= 1.7 from integrated infrared photometry, and Zinn (1985) based on the same data estimated a metallicity ${\rm [Fe/H]}=-0.38$. Using near infrared integrated spectroscopy, Armandroff & Zinn (1988) derived ${\rm [Fe/H]}=-0.99$ from CaII triplet lines, and E(B-V)=1.25 from the interstellar band at 8621 Å. By means of the bright giants method Webbink (1985) estimated an horizontal branch level of $V_{\rm HB}=20.3$ which, combined to a reddening of E(B-V)=1.71 based on Malkan's data, led to a distance from the Sun of $d_{\odot}=7.0$ kpc. Webbink (1985) lists a metallicity [M/H]=-0.45 for the cluster. Liu et al. (1994) reported an infrared Colour-Magnitude Diagram (CMD) and derived E(B-V)=1.8, a metallicity ${\rm [Fe/H]}=-1.0$ and a distance from the Sun $d_{\odot}=6.9$ kpc. More recently, by means of near-infrared integrated spectroscopy, Bica et al. (1998) derived E(B-V)=1.60 and ${\rm [Fe/H]}=-1.01$, and pointed out that Terzan 9 was one of the clusters requiring deep photometry for more conclusive results, since bulge contamination might be occurring in the integrated light.

3.1 Colour-magnitude diagrams

Figure 3a shows the V vs. (V-I) CMD for a $400\times 400$ pixel ($2.6'\times2.6'$) extraction centered on the cluster. The main features are a blue disk main sequence (MS), and a red sequence corresponding to the cluster plus bulge giants.

  

Figure 3: V vs. (V-I) CMD for Terzan 9: a) $400\times 400$ pixel extraction ($2.6'\times2.6'$) centered on the cluster, which is very contaminated by the field. b) extraction of r<85 pixel (1.4' radius) showing the cluster sequences. Overimposed is the mean locus of M 30

  

Figure 4: V vs. (V-I) CMD for offset field (4' east of Terzan 9), corresponding to a $500\times500$ pixel area ($3.3'\times3.3'$)

The best extraction for Terzan 9 is shown in Fig. 3b corresponding to a radial extraction of 85 pixels (33$^{\prime\prime}$) centered on the cluster. A vertical giant branch and a blue HB are present. This CMD is very similar to that of Terzan 4 (Ortolani et al. 1997a), an equally central cluster in the Galaxy with blue HB. Overimposed on the Terzan 9 CMD is the mean locus of M 30 (${\rm [Fe/H]}=-2.13$, Zinn 1985). The mean locus of NGC 6752 (${\rm [Fe/H]}=-1.54$, Zinn 1985) could be considered an acceptable fit, but not as good as with M 30. The mean loci of M 30 and NGC 6752 are from Rosino et al. (1997). Like in the case of Terzan 4, Terzan 9 is probably as metal-poor as M 30. Previous integrated methods where metallicities of ${\rm [Fe/H]}\approx -1.0$ or higher were obtained, were certainly contaminated by metal-rich bulge field stars.

We show in Fig. 4 a $500\times500$ pixel extraction ($3.3'\times3.3'$) corresponding to a field CMD located at $\sim 4'$ east of the cluster (see Fig. 1a). Besides the blue disk MS, the red HB of the bulge population is clearly seen at $V \approx 20.5$ and $(V-I)\approx 3.45$. The cluster RGB seen in Fig. 3a corresponds to an underpopulated zone in Fig. 4. On the other hand the bulge late type giants are seen as a curved RGB extending to very red colours, similar to metal-rich globular clusters such as NGC 6553 and NGC 6528, as well as the Baade Window (Ortolani et al. 1995 and references therein). This extension is less clear in the cluster field (Fig. 3a) because of a smaller area, crowding in the cluster area, and possibly a higher extinction.

3.2 Reddening and distance

The brightest part of the blue HB of Terzan 9 is located at $V_{\rm HB}=20.35 \pm 0.15$.

We calculate the cluster reddening taking M 30 as reference. By matching the overall sequences (Fig. 3b) we obtain $\Delta (V-I)=2.49$. Adopting E(V-I)/E(B-V)=1.32 (Dean et al. 1978), this corresponds to $\Delta(B-V)=1.89$, and assuming E(B-V)=0.06 for M 30 (Barbuy et al. 1998a and references therein), we get E(B-V)=1.95, which corresponds to A_V=6.24 (R=A_V/E(B-V)=3.2).

Adopting an absolute magnitude of the HB for Terzan 9 like that of M 30, M_V=0.66 (Barbuy et al. 1998a and references therein), we get the absolute distance modulus $(m-M)_0=13.45\pm0.2$. This corresponds to a distance of $d_{\odot}=4.9\pm0.7$ kpc for Terzan 9.

To derive the Galactocentric coordinates of the cluster, we assume a distance of the Sun to the Galactic center of $R_{\odot}=8.0$ kpc (Reid 1993), to be consistent with Barbuy et al. (1998a) - notice however that Reid (1998), taking into account Hipparcos data, pointed out that $R_{\odot}$ could be as large as 9.3 kpc. Terzan 9 galactocentric coordinates are X=-3.11 (X<0 refers to our side of the Galaxy), Y=0.31 kpc and Z=-0.17 kpc. The distance to the Galactic center is 3.13 kpc, wherefrom we conclude that Terzan 9 is located within the bulge.

3.3 Bulge field parameters

We can calculate parameters for the offset field metal-rich bulge population and compare it with those of the metal-poor cluster. The bulge red HB is located at $V\approx 20.8$ and $(V-I)\approx 3.55$. Using the metal-rich template cluster NGC 6553 as reference (Guarnieri et al. 1998) we get $\Delta (V-I)=1.4$, corresponding to $\Delta (B-V)=1.06$. Assuming E(B-V)=0.73 for NGC 6553 (Barbuy et al. 1998a) we derive E(B-V)=1.79 for the field of Terzan 9, which corresponds to A_V=5.73 (R=3.2). For this low Galactic latitude field ($l=3.6^\circ$, $b=-2.0^\circ$) the extinction is much higher than that of the Baade Window (Terndrup 1988). Assuming M_V=0.95 for a metal-rich HB (Barbuy et al. 1998a), the distance of the bulk of the population for this bulge field results to be 7.0 kpc. However as pointed out by Barbuy et al. (1998a and references therein) the total-to-selective absorption ratio should increase with the metallicity and a more appropriate value could be R=3.47 for this high metallicity field, resulting in a distance of $d_{\odot}=5.1$ kpc. Considering that the distance to the Galactic center is of about 8-9 kpc (Reid 1993, 1998), the possible interpretation is that we are observing a relatively less reddened, foreground part of the bulge, where Terzan 9 would also be embedded (Sect. 3.2).