Hubble Source Catalog (HSC) Use Case # 1 - Point Source Photometry

Reference:- "Deep Optical Photometry of Six Fields in the Andromeda Galaxy"

                   - Brown et al. (2009) ApJS, 184, 152

BACKGROUND: Brown et al. (2009) obtained deep optical images reaching well below the oldest main sequence turnoff in six fields of the Andromeda Galaxy using ACS. The fields are located at four positions on the southeast minor axis, one position in the giant stellar stream, and one position on the northeast major axis (disk). These data were obtained as part of three large observing programs (9453, 10265, 10816) designed to probe the star formation history of the stellar population in various structures of the galaxy.

    Left Fig. 1 (Brown et al. 2009;  Reproduced by permission of the AAS): Shown is stellar density in the Andromeda vicinity, from counts of RGB stars (Ferguson et al. 2002) and fields studied by Brown et al.. Right Fig. 2 (Brown et al. 2009; Reproduced by permission of the AAS) shows the CMD in the six studied  fields.

USE CASE GOAL: Construct the F606W - F814W vs F814W color magnitude diagram (CMD) using the HSC and compare it with the disk field from Brown et al. (2009).

NOTE: Several steps below require the use  of software external to the HSC and HLA (e.g., to make calculations and plot data). This particular document uses TOPCAT to make most of these figures.  Other popular analysis packages are PYRAF, IRAF, IDL, ... . Instructions for these steps will necessarily be less specific than the earlier steps using the HSC and HLA.

Part 1: View the Data

Note: If you use NumImages > 0 in the interactive display you will find several blank sources. These are cosmic rays that have slipped through the processing for one of the 27 images. You will find that the decision of the criteria to use for NumImages is often one of the most important decisions when using the HSC. 

Note: You can view the abstract and papers that have resulted from the10265 data set by clicking on the blue 10265 in the PropID column of the inventory.

Examine the HSC data for the M31 disk via using ‚"Search with Summary Form".

Use the following criteria:
RA = 00:49:08.09
DEC = 42:44:55.0
Radius = 0.5 arcmin
Mag Type = MagAper2
NumImages > 1

 The resulting search returns 537 rows. We can plot their position  by clicking on the "Display numeric columns graphically using VOPlot", which appears near the top of the screen when using the (default) html format to display the results from a HSC Summary form search.

If you click on MatchID entries, you will see all the individual observations for that object. For example MatchID 6390018 includes 28 different images.
Display the properties of MatchID 6390018 by using VOPLOT:

Now make a wider search using the following criteria:

RA = 00:49:08.09
DEC =42:44:55.0
Radius = 2.5 arcmin
Mag Type = MagAper2
NumImages > 1

This will result in N=9053 objects that will be displayed in an html form. A visual inspection of the objects can be made by using the HLA Interactive Display.
Note that in the figure below the HSC (beta) lists 9070 detected objects since the HLA uses the HSC Summary Search form and looks around the image center ra & dec with a Radius = 2.6 arcmin, Mag Type=MagAuto, and NumImages >1.

Note that the first object in the list is MatchID = 6390018. It has 27 detections with mean A_F606W = 25.599 +/-0.022 and A_F814W = 25.084 +/-0.031.  Note that the Concentration Index (CI; difference between the 1 pix and 3 pix radius magnitudes) has a value CI = 1.102, which is typical for a star.

If you look down the list you will find MatchID=6390123 has CI=2.106, which is larger than expected for a star. If you look at the position 00:49:08.10 +42:44:59.04 on the interactive display you will find that, indeed, this is a galaxy (i.e the object near the + in the image above).

If you click on the blue MatchID=6390018 in the first column of the summary search form table you are taken to the detailed search form. This shows that all 27 detections of this source are from program 10265. Hence this is a very uniform data set.

Part 2: Download and Trim the Data

Make a catalog to download by changing:
                                Maximum Records = 10001
                                Output Format = File: < your choice from the 8 options >
You may also want to change the output columns to customize your output (e.g., remove empty columns).
Brown et al. (2009) used DAOPhot (PSF fitting) and the STMAG system for their catalog. They derive the following conversion between the ABMAG system and the STMAG system:       
 ABmag_F606W=STmag_F606W - 0.169
 ABmag_F814W=STmag_F814W - 0.840

Note that Brown et al. also applied a CTE correction for their data.

For our star with match ID = 6390018 using MagAper2, F606W = 25.599 and F814W = 25.084. The HSC uses the ABMAG system.
Hence converting from the HSC photometry system to the Brown system requires:

         HSC F606W = 25.599  (ABMAG and no aperture correction)
               + 0.169   (ABMAG to STMAG)
                - 0.248 aperture  correction (Sirianni et al. 2005, Table 3, i.e. 2.5log[Encircled Energy]=2.5log[0.796])
                  25.520   compared with Brown's 25.490 (so Delta = 0.030 with Brown brighter, probably because he corrects for CTE and the HSC does not)

         HSC F814W = 25.084
            +0.840  (ABMAG to STMAG)
            -0.292 aperture correction (Sirianni et al. 2005, Table 3, i.e. 2.5log[Encircled Energy]=2.5log[0.764])
             25.632 compared with Brown's 25.60  (so Delta = 0.032)

Shown in red is the Brown et al. Disk data set, and in blue the HSC. Note that the HSC provides all Sextractor objects including bright saturated stars which were removed by Brown et al.; in addition, Brown et al. excluded all data along the image edges and the detector gap since the number of available exposures in these area is smaller and hence the noise is larger.

Select stars with errors in MagAper2 for both filters with values less than 0.2 and CI values in the range 1.0 - 1.3, typical of stars. After applying this selection, the total number of stars is 7649.

Part 3: Compare the Data

The figure above shows a comparison between the Brown et al. catalog (green circles) and the HSC catalog (blue circles – before the correction has been made to match positions).  While the correspondence is in general very good, a few bright objects are missing from the HSC in the immediate vicinity of very bright stars.
Note that Brown et al. used the second version of the Guide Star Catalog for absolute astrometric corrections while the HSC use the 2MASS Catalog (see the HSC FAQ), for that a small offset delta ra and delta dec of 0.024  and 0.6 arcsec are needed to adjust the coordinates before matching with the Brown et al.  HLSP catalog.   

Plot the CMDs for both data sets. HSC data are shown in blue and Brown et al. data in red. The right figure shows the data in the AB mag system, and the left figure shows the HSC CMD shifted by 0.64 mag. to enhance the visual comparison. We note that even the fine structure in the "red clump" (AB_F814W~24.8) is very similar.

In general the comparison appears to be good, although there is slightly more scatter in the HSC data set for fainter magnitudes.  If you select a subset of the data with magnitude errors for both filters <=0.01 mag, for F606W the Brown et al is deeper than the HSC by approximately 3 magnitudes, (see figure below) as expected since this come from a mosaic where all 28 visits are co-added compared to the HSC where the measurements are from each of the individual visits. In addition, the HSC is not very aggressive, as discussed in FAQ.


  1. The HSC Beta Version 0.2 catalog has been used to make a detailed comparison with the stellar populations of the deeper Brown et al. (2009) data set.
  2. The HSC and Brown et al. (2009) catalogs agree at about the 0.05 mag level out to 26.5 ABmag in F606W and to 26 ABmag in F814W. The astrometric difference between data sets is relatively small with deltaRA=0.024  and deltaDec=0.6 arcsec.
  3. Relative to the Brown et al. catalog, the HSC is completeness to ~94% for F606W and 97% for F814W.
  4. The HSC Beta 0.2 catalog in this particular field does not go as deep as the data would allow (e.g., the Brown et al. 2009 catalog), due to the use of visit-based measurements in the HLA rather than measurements in a deep mosaic where all the data has been combined (e.g. in the Brown study - see FAQ).
  5. NOTE: In the future, the HLA team is planning to produce deep mosaic images for which Sextractor and DAOPhot catalogs will be available for building the HSC.