The Great Observatories Origins Deep Survey (GOODS)
Spitzer Legacy Data Products (DR1+)
GOODS-N MIPS 24 micron source list

Table:

1.0 General Information

As part of the interim data release (DR1+) of products from the GOODS Spitzer Legacy Program, we are providing a list of sources for the MIPS 24 micron imaging of the GOODS-N field. This source list is a subset of a more extensive catalog which will be presented by Chary et al. (in preparation). It is limited to flux densities greater than 80 microJy, where the source extraction is highly complete and reliable, and which is well below the flux limit currently explored for IRS spectroscopy. This early source list is intended to facilitate early investigations by members of the scientific community, and to support planning for follow-up observations. Although we believe this source list is reliable and scientifically useful, it is nevertheless preliminary, and will be superseded by the more extensively validated and documented catalogs in Chary et al.

2.0 Source Extraction

Source extraction was performed on version 0.36 of the GOODS-N MIPS 24 micron images, which are described in the documentation for the first interim GOODS data release (DR1+). The measurements were carried out using prior positional information determined from the very deep GOODS IRAC 3.6 and 4.5 micron images of the same field, available from the first GOODS data release (DR1). Chary et al. (2004) have shown that there is excellent correspondence between faint 24 micron sources and objects seen in very deep IRAC imaging like that available for the GOODS fields. We therefore have taken advantage of the higher angular resolution of the IRAC images to define positions and to help deblend sources where necessary in the 24 micron images.

Because the single-epoch GOODS IRAC data in any particular channel cover only roughly half the total GOODS-N field, we created an exposure-weighted average of the data from the 3.6 and 4.5 micron channels. SExtractor (Bertin & Arnouts 1996) was run on this average IRAC image to create a source list. The astrometric alignment of the IRAC and MIPS 24 micron images is excellent, typically about 0.1 arcsec. Therefore, we used the SExtractor positions of the IRAC sources as input to the MIPS source extraction process.

Sources in the IRAC source list were grouped together based on their relative separations, and a region of the MIPS mosaic was cut out to match the group of IRAC sources. The centroid position of each IRAC source in this group was mapped to the appropriate pixel location in the MIPS image. We allowed for the possibility of a small shift in the centroid position of the MIPS source, in order to account for several possibilities: (1) small IRAC to MIPS registration errors during data reduction; (2) physical displacement of the centroid of 24 micron emission from that of the 3.6+4.5 micron light; and (3) IRAC positional errors due to source blending in the IRAC images. A new centroid was estimated from the MIPS image using pixels within 1.8 arcsec radius of the nominal IRAC position. An maximum offset threshold of 0.6 arcsec was applied to prevent the centroid of a faint source from straying too far from the original IRAC position or centroiding on a neighboring, brighter source. If the new centroid was within 0.6 arcsec of the original IRAC position, it was adopted as the new MIPS position. If the MIPS centroid wandered by more than this tolerance, we adopted the original IRAC-defined position.

A MIPS 24 micron PSF was generated from isolated sources in the image, and renormalized based on the aperture corrections published in the MIPS Data Handbook (section 3.7.5, table 3.12). The MIPS PSF was centered on the position of each source in the IRAC list as defined above. The group of sources was then simultaneously fit using an IDL-based singular value decomposition (SVD) routine. The process of creating a group around a source, re-centering, and fitting was repeated for each individual source in the input IRAC source list. A quality of fit parameter was preserved for each group fit based on the residuals of each source and its immediate neighboring sources. The fit which results in the smallest residuals for the source was adopted as the best-fit flux density.

The MIPS PSF was then rescaled by the best fit flux estimate for each source and subtracted from the mosaic. Flux uncertainties are calculated from the resultant residual image using a PSF weighted sum of variance values (see, e.g., section 6.1.2.2 of the Spitzer Observer's Manual v5.0).

A second pass of the source detection and extraction routine was run to fit for 24 micron sources which do not have coverage in the GOODS IRAC images; these are mostly near the edges of the GOODS-N MIPS field. MIPS 24 micron sources whose IRAC counterparts either fall towards the edge of the IRAC mosaic (weight less than 1000 in IRAC exposure map) or do not have any IRAC coverage, have the output parameter IRAC_Tag set to 0. They also correspondingly have -1.0 for their IRAC RA and DEC.

Although nearly all of the faint 24 micron sources in the GOODS images are expected to be unresolved, this is not true for a few of the brighter sources associated with galaxies at relatively low redshift. We have identified 5 such objects by interactive visual inspection. In this DR1+ source list, we have used SExtractor to measure flux densities for these five sources within an isophotally-defined aperture, and have substituted those fluxes for the point-source fits described above. These objects are indicated in the source list by values of 1 in the Extended_Tag column.

3.0 Description of the source list parameters

The GOODS DR1+ MIPS 24 micron source list is released in IPAC Table format, and is available from the Spitzer Archive and the "Popular Products" repository of Legacy Data Products. Table 1 describes the columns provided in this source list.