Chapter 4 - Imaging Data Products
In this chapter we give the “nuts and bolts” descriptions of the various imaging and catalog data products output by the GALEX pipeline. All of the products described here can be accessed from the the Multimission Archive at the Space Telescope Science Institute (MAST). The web address of the GALEX GR4 archive is: http://galex.stsci.edu/GR4/.
Below is a table listing some of the most frequently used image products that would be sufficient for most users' purposes. Each field has a name followed by the suffix listed in the table. The “f” or “n” denotes whether the data is in the FUV or NUV band whereas the “d” indicates that the data was taken direct image mode. (Grism data is denoted by a “g”). The pixel scales are given in arcsec per pixel.
|Filename||units||pixel scale (arcsec/pixel)||Description|
|fd or nd-cnt.fits.gz||counts/pixel||1.5||The raw number of counts per pixel, not corrected for the exposure time or flat field|
|fd or nd-exp.fits||seconds||6||The exposure time map in sky coordinates, corrected for dead time, but not including the flat field or hot spot masks.|
|fd or nd-rr.fits||seconds||6||The low resolution relative response, defined as the flat field times the exp map and including detector hot spot masks.|
|fd or nd-rrhr.fits.gz||seconds||1.5||The high resolution relative response. This is the rr image linearly interpolated to the same pixel scale as the cnt map.|
|fd or nd-int.fits.gz||counts/sec/pixel||1.5||The “intensity” map in units of counts/sec. This is the cnt map divided by the rrhr map.|
|fd or nd-skybg.fits.gz||counts/sec/pixel||1.5||The sky background map subtracted from the data before identifying sources.|
|fd or nd-flags.fits.gz||flag value||12||Flag map indicating regions of the map likely contaminated by artifacts or regions where various types of artifacts have been removed.|
In this section are sample figures showing the various imaging data products for the MIS field MISDR1_18595_0456 included as a part of the GR4 data release. The first image shows the two color image of the entire field. In all of the remaining figures, the FUV image is shown on the left while the NUV is on the right.
Figure 1 Combined false color FUV+NUV image of MISDR1_18595_0456.
Figure 2 Count (cnt) images of MISDR1_18595_0456 showing the raw number of photons per pixel for the FUV (left) and NUV (right).
Figure 3 High resolution relative response (rrhr) images of MISDR1_18595_0456 for the FUV (left) and NUV (right). These images give the effective exposure time as a function of position including the flat field, the exposure time, and masked hot spots and other artifacts. The small spiral patterns across the field correspond to masked detector hot spots.
Figure 4 Intensity (int) images of MISDR1_18595_0456 for the FUV (left) and NUV (right). These images are equal to the count image divided by the high resolution relative response and have units of counts/sec/pixel.
Figure 5 The sky background (skybg) maps determined by the pipeline for MISDR1_18595_0456 for the FUV (left) and NUV (right). Note that both the FUV and NUV background can vary with position within a field. This field in particular has variable diffuse Galactic cirrus going from the the south-east to the north-west across the field that is visibile in the FUV sky background. The pattern in the NUV background is similar to that seen in the inverse of the relative response maps. This is because the flat fields are defined such that observations of the white dwarf standard star LDS 749B are uniform across the field. The result of this choice is that the NUV sky background is no longer flat. This is probably due to the fact that the primary contributor to the NUV sky background is the zodiacal light which is much redder than the white dwarf.
Figure 6 Flag (flag) image of MISDR1_18595_0456 in the FUV (left) and the NUV (right). These maps flag regions that may be contaminated by image artifacts or where an attempt was made to remove an artifact.
Next, we list some of the other image products that are used in the generation of the source catalogs. The flux measurements are actually made from the “intbgsub” images, which is the background subtracted versions of the “int” images.
|fd or nd-intbgsub.fits.gz||counts/sec/pixel||1.5||Background subtracted image. This is the int image minus the skybg image and is the image used to make the measurements output to the GALEX catalogs.|
|fd or nd-wt.fits.gz||dimensionless||1.5||“Weight” or threshold map input to SExtractor to detect sources.|
|fd or nd-objmask.fits.gz||integer||1.5||SExtractor “segmentation” map indicating those pixels masked out when determining the background map in the first source extraction iteration.|
Here are the most commonly used source catalogs. The “mcat” file contains all of the information in the other single band files and is thus the catalog that would be most useful to most users. This file also forms the basis for most of the columns in the PhotoObjAll table in the MAST GALEX database.
|fd or nd-cat.fits.||SExtractor catalogs for images: Table of sources extracted by GALEX reduction pipeline. The table contains positions, flux, magnitude, and major and minor axes.|
|xd-mcat.fits||Merged source catalog: Band-merged table of extracted sources. Contains all sources contained in "[n/f]d-cat.fits" matched to the best candidate from the other band. Only sources with a S/N greater than 2 are matched. Sources without matches are also listed. For each source, the table contains a global identifier, merged position, calibrated fluxes and magnitudes, etc. All the data from the "[n/f]d-cat.fits" tables are contained in "-xd-mcat.fits" file, as well as fluxes from the "nd-fcat.fits" and "fd-ncat.fits" files.|
|nd-fcat.fits||NUV extractions using FUV positions: A SExtractor catalog of source fluxes and magnitudes from the NUV image, but with source positions taken from the FUV source catalog "fd-cat.fits".|
|fd-ncat.fits||FUV extractions using NUV positions: A SExtractor catalog of source fluxes and magnitudes from the FUV image, but with source positions taken from the NUV source catalog "nd-cat.fits".|
Below are the parameter files, convolution kernel and lists of output parameters used in running SExtractor on the GALEX images. For more detailed descriptions of the meaning of the columns in the SExtractor files, please refer to the SExtractor manual available from http://terapix.iap.fr/rubrique.php?id_rubrique=91/ or to Bertin & Arnouts (1996, A&A, 117, 393). Note that all of the fluxes given here are given in units of counts/sec. Similarly the magnitudes are instrumental and are defined as -2.5*log10(counts/sec).
|fd or nd-gauss_30_7x7.txt||Gaussian convolution kernel used by SExtractor.|
|fd or nd-sexparams.txt||Parameter file used as input to SExtractor.|
|fd or nd-sexcols.txt||List of SExtracot output columns.|
|f or n-sexnnw.txt||Neural network parameter file used by SExtractor to perform star-galaxy classification.|
|fd or nd-sexparams_orig.txt||List of parameter values used in SExtractor during the initial source extraction run.|
|fd or nd-sexcols_orig.txt||List of SExtractor output columns used in the initial source extraction run.|
Once the GALEX pipeline has run on both the FUV and NUV bands for a given field, the SExtractor catalogs are merged together. The sources detected in each band are matched although non-matched sources from either band are retained as well. Uppercase column names originate from the source extractor program (SExtractor) which operates on each band separately. These values are copied verbatim from the source extractor result files (*-n/fd-cat.fits) and have either an NUV_ or FUV_ prepended to the column name. For these columns, -99.0 is used for unfilled (or blank) values, but in some cases blank magnitude values appear as +99.0. Lowercase column names are added by the band merger program. For these columns, -999.0 or -9999.0 are unfilled or blank values. Magnitudes are either AB or GALEX instrumental magnitudes (i.e. -2.5*log10(cts/sec)) as stated. Fluxes are counts/second unless stated to be in micro Janskys.
A detailed description of the columns in the MCAT files is given in Appendix-A.