W1,W2 Color-Magnitude Diagrams of Globular Clusters in V4 Compared with V3.5

Carl J. Grillmair &

7 June 2011

To gauge the effect of various changes made in V4 (e.g. more relaxed frame inclusion criteria, deeper SNR limts, improved PSFs, photometry, zeropoints), I have examined the color-magnitude diagrams of three globular clusters. These clusters, NGC 4590 (M68), NGC 5139 (Omega Cen), and NGC 5897 were chosen simply because they currently have similar numbers of detections (i.e. same total coverage) in both V4 and V3.5. W1, W2 photometry for all sources within 20 arcmin of these clusters were obtained from the Pass 1 and Pass 2 source catalogs via Gator. To avoid being skewed by extreme crowding in the cores of the clusters, I've selected only stars with 3' < r < 20' (5' < r < 20' for Omega Cen). W1 vs W1 - W2 color magnitude diagrams are shown below.

The first set of CMDs ("_d") show all detections for all stars. Median colors and RMSs were computed for the region extending over 10 < W1 < 13 and and -0.3 < W1 - W2 < 0.3. The widths of the red giant branches are taken as a proxy for relative photometric precision. These values indicate a 1% increase in the RMS in V4, as compared with V3.5, for both NGC 4590 and NGC 5897. Omega Cen shows a much larger increase (31%) but we attribute this to the more conservative inclusion criteria in V3.5. The median color of the giant branch has changed by an average of -0.023 +/- 0.004 mag.

The second set of CMDs show the same data sets, but after averaging the magnitudes of all detections within 2 arcsec of one another to reduce point crowding. The RMSs are once again almost identical in V4 (a 1% increase for NGC 4590 and a 4% decrease for NGC 5897). The median colors of the giant branches have changed by -0.036 +/- 0.009. Given the lack of sophistication in our duplicate resolution procedure, we recommend using the RMSs and offsets measured in the first set of CMDs.

Also shown in the CMDs are theoretical isochrones from the Padova database (Marigo et al. 2008, Girardi et al. 2010), reddened using the known E(B-V)s for these clusters (and A_v = 0.15, 0.10 for W1 and W2, respectively), and offset in magnitude using the appropriate distances. Interestingly, the isochrones match the giant branches in V3.5 better than they do the RGBs of V4. However, we can't put a much stock in this offset until we know what RSRs/zeropoints were used to compute the isochrones.

The results of this analyis demonstrate that, despite the more liberal inclusion criteria in V4, the photometric precision has not been significant compromised.