A Hubble Space Telescope Catalog of 449 Galaxies Seen Through the Disk of M31
ABSTRACT From inspection of 30 Hubble Space Telescope ACS images of M31, we provide a catalog of 449 galaxies seen through the spiral disk. Measurements of the positions of the galaxies, their integrated magnitudes in two colors and their sizes, determined from isophotometry, are included in the catalog. We discuss the many difficulties of interpreting these data in terms of the effects of intervening extinction by dust in the disk Comment: 19 pages, 3 tables, 7 figures
A HUBBLE SPACE TELESCOPE CATALOG OF 449 GALAXIES
SEEN THROUGH THE DISK OF M31
(Astronomy Department, University of Washington, Seattle, WA, 98195, USA,
Karl Krienke (Seattle Pacific University, Seattle, WA98119, email@example.com)
From inspection of 30 Hubble Space Telescope ACS images of M31, we provide a
catalog of 449 galaxies seen through the spiral disk. Measurements of the positions of
the galaxies, their integrated magnitudes in two colors and their sizes, determined
from isophotometry, are included in the catalog. We discuss the many difficulties of
interpreting these data in terms of the effects of intervening extinction by dust in the
Ever since the era in which it was recognized that galaxies contain dust that causes
extinction of light, it has been tempting to measure this dust by examination of more
distant galaxies seen in the background. The first published account of this technique
that we are aware of was made by Shapley and Nail (1951), who used galaxy counts
made in the direction of the Small Magellanic Cloud to conclude that it is essentially
transparent. As an intimation of this technique’s troubled future, their result was
wrong, apparently because of an unfortunate accident. As related by Hodge (1974),
more recent evidence showed that Shapely and Nail used number counts of non-
stellar objects detected by their assistants on plates taken by the Boyden
Observatory’s 1.5 m telescope. Apparently they didn’t examine these plates
themselves, as a repeat of this experiment showed that inside the boundaries of the
SMC, most of the non-stellar objects marked in ink on the plates were either emission
nebulae or star clusters. The number of such objects approximately equaled the
number of galaxies not seen behind the SMC. The story does not end there, however.
In recent experiments, cited below, it has been shown that the crowding of stellar
images in a galaxy’s disk can obscure background galaxies to a considerable extent.
(Incidentally, Hubble (1934) worried about this problem in his study of the Milky
Way’s dust in the Zone of Avoidance). It is therefore likely that Shapley and Nail’s
result was approximately correct, after all, because of two opposing mistakes.
More recent attempts to use this technique have been more successful. Among
others, the pioneering work of Gonzalez et al. (1998), which has continued in a series
of important papers (see Holwerda et al. 2007), has demonstrated that information on
the opacity of galaxy disks can be obtained from background galaxy counts, when
careful measurements of the various systematic effects are carried out. They say,
however, that the technique should not be used for galaxies in the Local Group, where
the resolved stellar foreground interferes. Somewhat contrary results were published
by us (Krienke and Hodge 2001), who showed that rather weak results could be
obtained for three Local Group galaxies, but, among other things, the distances to the
individual background galaxies were required to be able to make reliable K
corrections to their colors. Another example is the work of Dutra et a. (2001), who
found that the use of background galaxy colors combined with redshifts gave some
useful values for the reddening in parts of the Large and Small Magellanic Clouds.
One can argue that it does not seem to make sense that a procedure that gives
information for distant galaxies does not work for nearby galaxies, where the total
obtainable information is orders of magnitude greater. This is not an uncommon
problem in astronomy, where we are always near the limit of the impossible. For
distant objects the difficulties are smoothed out, permitting the anomalous signals to
be recognized and measured more easily. For the case of the highly-resolved Local
Group galaxies, the same data are available, but their completeness makes the task of
extracting anything from it more complicated and the uncertainties involved are more
obvious and thus more daunting.
One can conceive of using three different sets of data to detect the effects of
extinction by dust in a nearby galaxy’s disk using background galaxies:
a. An excess reddening of the colors,
b. Decreased integrated magnitudes, and
c. A decreased number of galaxies detected
An even larger list can be developed for the difficulties involved in using these data:
d. to measure an excess color, one must know the intrinsic color, which requires
knowing the Hubble type of the galaxy and keeping in mind the intrinsic
spread in the type-color relation.
e. Also one must know the distance and/or radial velocity of the galaxy in order
to apply the K corrections due to the redshift of the galaxies’ SEDs. This must
be known also to correct the effective characteristics of the galaxies because
of relativistic fading at larger values of z.
f. to detect a decrease in the magnitudes of background galaxies, one must know
the distribution of magnitudes of galaxies external to M31 and its dispersion,
which is large, even not including effects of clustering
g. to use the number of galaxies per area behind M31, one must establish a
uniformity of sample, corrected for detection limits and distance (including
h. All of the methods are complicated by the presence of a variable and often
dense field star population, which affects the detection limits as well as the
measurements, in a number of ways.
Because of these many difficulties, the problem is daunting one, but not completely
impossible. We report here a possible method of measuring the extinction of
background galaxy light by the disk of M31, one of the worst environments in the
Local Group in which to use this technique. We were tempted to do this by noting
the wealth of galaxies detected in the M31 fields with which we worked on M31’s
open clusters (Krienke and Hodge 2007, 2008). Perhaps, with the availability of
hundreds of background galaxies, it might be worthwhile to make this attempt. As we
point out below, the problem probably can be solved, but only with additional data,
especially radial velocity data.
Please note that this paper deals only with the galaxy count problem; alternate
measurements of the dust content of M31 are coming from infrared Spitzer data
(Gordon et al., 2004)
2. OBSERVATIONAL MATERIAL
From the Hubble Space Telescope Archives, we have chosen 28 ACS pointings that
cover a variety of positions in the main disk of M31, for which the archives include at
least two colors that can be transformed into standard responses, Johnson-Cousins B,
V and I. In addition we chose two pointings that are near but not in the main disk of
the galaxy. The pointings and their positions are given in Krienke and Hodge (2008).
We chose ACS data and not WFPC2 data because we felt that the increased
resolution and depth of ACS would be important.
We chose to select galaxies by eye. This is defensible in the case of M31, which has
densely-packed field stars, nebulae and star clusters on most of the selected pointings.
We argue that using one of the available automatic selection extractors would
produce so many false candidates that we would have to pick them over by eye in any
case, re-introducing any selection effect that was there before the extractor process.
3. GALAXY PROPERTIES
Tables 1 and 2 (at end of this paper) provide catalogs of 449 galaxies found on the 30
pointings examined. We give measures of the magnitudes and colors (B, V, and I,
transformed from the HST filters), their uncertainties, the J2000 positions of the
centers of the galaxy images and an approximate size parameter, which is the
maximum value of the distance between the center and the largest distance from the
center found on a limiting isophote. The galaxy photometry was carried out with a
program that one of us wrote in IDL that makes special allowance for variable “sky”
values caused by the variable stellar density in the field. It was adapted from that
described in our work on open clusters (Krienke and Hodge 2007).
Galaxy magnitudes ranged from V = 16 to 25, with the majority fainter than V = 19.
Figures 1a and 1b show color-magnitude diagrams for the set with V and I colors and
with B and V colors, respectively. The distribution of points in these diagrams
resembles that found for galaxy surveys in general.
V - I
Fig. 1a. The V, I color-magnitude diagram
00.51 1.52 2.53
B - V
Fig. 1b. The B, V color-magnitude diagram.
The quoted errors are primarily dominated by the variable background. Figure 2
shows the distribution of errors in the V measurements. Average values of the errors
are B(error) = 0.18 ,V(error) = 0.14, I(error) = 0.11.
1516 1718 1920 2122 2324 25
Fig. 2 Measurement errors for V magnitudes
We also scanned three ACS pointings that lie outside of but near the M31 disk. One
of these includes a distant cluster of galaxies and the other two were obtained to study
two outer globular clusters. Figure 3 shows a portion of one the outer fields, including
several galaxies plus the M31 globular cluster Bol 409.
Fig. 3. This image from one of the outer pointings shows four galaxies, plus the
globular cluster Bol409 at J2000 = 12.541258, 41.683533. The globular has a V
magnitude of 16.09 and a V – I of 1.24.
In Figure 4 we have plotted the distribution in color of the galaxies in our M31
sample, for the full sample and for the bright and faint galaxies separately. From this