Catalogs

In this Section, we discuss the results of our mid-IR photometric survey. Because we required a $>3\sigma$ detection for each source at $3.6\,\micron$ and $4.5\,\micron$, the effective survey area for each galaxy is the overlap of the IRAC $3.6\,\micron$ and $4.5\,\micron$ image mosaics. Table1 lists the effective survey area, gas-phase (H$\alpha$) star formation rates adopted from Khan et al. (2013) and the number of point-sources cataloged in each galaxy followed by the number of matches at $<0.5$ and $0.5-1.0$ pixel, duplicates between the $3.6\,\micron$ and $4.5\,\micron$ source-lists, and the number of $>3\sigma$ counterparts identified at the three longer wavelength images for each galaxy. Tables$2-8$ list the coordinates (J$2000.0$; RA and Dec) of the point-sources followed by their Vega calibrated apparent magnitudes ($m_\lambda$), the associated $1\,\sigma$ uncertainties ( $\sigma_\lambda$), and (for the $3.6-8.0\,\micron$ bands) the differences between the PSF and aperture photometry magnitudes ( $\delta_\lambda$). For the $5.8\,\micron$, $8.0\,\micron$ and $24\,\micron$ bands, $\sigma_\lambda=99.99$ implies that the associated photometric measurement is a $3\,\sigma$ flux upper limit, and $m_\lambda=99.99$ (as well as $\sigma_\lambda=99.99$) indicates that no reliable photometric measurement could be obtained for that location. For the IRAC bands, $\delta_\lambda=99.99$ implies that one or both of the associated photometric measurements did not yield a $>3\sigma$ flux measurement.

Figures 4 and 5 present the $m_{4.5}$vs. $m_{3.6}-m_{4.5}$, $m_{5.8}$vs. $m_{4.5}-m_{5.8}$, and $m_{8.0}$vs. $m_{5.8}-m_{8.0}$ color magnitude diagrams (CMDs) for each galaxy. For comparison, we include the mid-IR CMDs for all sources in a $6$deg$^2$ region (see Khan et al.2013, for details) of the NOAO Bootes Field produced from the Spitzer Deep Wide Field Survey (SDWFS, Ashby et al.2009) data. Our catalogs simply inventory all the sources present on the image mosaics and do not attempt to distinguish between sources actually associated with the galaxies and unrelated contaminants. The contamination is significant for galaxies like NGC247, which is highly inclined and covers a smaller fraction of the Spitzer images, than for the larger and more face-on galaxies (see Figure1). Indeed, the CMDs of NGC247 (Figure5, second column) clearly show two distinct population of sources, with the break appearing near $m_{4.5}\simeq18$. The CMDs of the two closest galaxies (NGC6822 and M33) show distinguishable sequence of bright and red AGB stars (near $m_{4.5}\simeq15$), but for the more distant galaxies this feature is less prominent (see Khan et al.2010 for relevant discussion). The CMDs of NGC7793 has the fewest identified point-ources both due to its lower mass and because of relatively poorer quality (more systematic artifacts) of its image mosaics.

Figure 6 show the apparent magnitude histograms of all sources in the catalog, with the shaded regions showing the sources in M33. It is apparent that the M33 catalog is $\sim 0.5$mag shallower than the other catalogs in all bands. Overall, our source-lists become incomplete at $m_{3.6}\gtrsim 18$, $m_{4.5}\gtrsim 18$, $m_{5.8}\gtrsim 17$ and $m_{8.0}\gtrsim 16$. Figure7 show the mid-IR color histograms of all sources in the catalog, with the shaded regions showing the sources with $1\,\sigma$ uncertainty in color $\lesssim 0.2$. It demonstrates the absence of any surprises in the color distributions and that the mid-color distributions are not due to large uncertainties. All normal stars have the same mid-IR color in the first two IRAC bands, because of the Rayleigh-Jeans tails of their spectra, and we see this from the sharp peak of color distribution at $m_{3.6}-m_{4.5}\simeq 0$. The longer wavelength detections are increasingly dominated by dusty stars, with color distribution peaks at $m_{3.6}-m_{5.8}=m_{4.5}-m_{5.8}\simeq 1$, $m_{3.6}-m_{8.0}=m_{4.5}-m_{8.0}\simeq 2$, and $m_{5.8}-m_{8.0}\simeq 1$. Figures 6 and 7 also show the apparent magnitude and color histograms of the SDWFS point sources (dotted lines). They show that our catalogs are $>1$mag deeper than the SDWFS catalog and the mid-IR color distribution of sources in a random extragalactic field (the NOAO Bootes Field) significantly differs from that of fields containing nearby galaxies. In particular, at longer wavelengths, the galaxies contain more red sources (dusty stars) while the random extragalactic field sources are generally bluer.

Although we report the $24\,\micron$ photometry in the catalogs, these measurements have limited utility due to the lower spatial resolution of this band. The aperture used for this band commonly includes objects other than the intended target and is contaminated by emission from cold interstellar dust. Nevertheless, as we showed in Figure5 of Khan et al. (2013), the spectral energy distributions of the normal stars show the expected negative slope for the Rayleigh-Jeans tail of their SEDs between $8\micron$ and $24\,\micron$. The $24\,\micron$ photometry can be very useful in specific cases, such as for studying evolved massive stars (see Khan et al.2015, for details), despite the resolution limitation.

Where Thompson et al. (2009) identified $\sim53,200$ sources in M33, and Khan et al. (2010) identified $\sim11,200$ sources in NGC300 and $\sim6,000$ sources in M81, now we catalog $\sim78,800$, $\sim21,700$ and $\sim28,600$ sources in these galaxies respectively. This is due to a number of factors. First, both of the earlier studies only cataloged the central regions of these galaxies, while here we analyze the full mosaics. Second, we use a larger matching radius of $1$pixel (rather than $0.5$pixel) to define point-sources and a higher fraction of our cataloged sources can be co-incidental matches between the $3.6\,\micron$ and $4.5\,\micron$. Third, the catalogs presented here are deeper due to improved photometry and search methods based on lessons learned from Khan et al. (2013); Khan et al. (2011); Khan et al. (2015). For example, Thompson et al. (2009) noted that their M$33$ point-source-list becomes incomplete at $m_{3.6}=m_{4.5}>17.1$, while we reach $\sim 0.5$mag deeper for this galaxy and $\sim1$mag deeper for the rest. We have compared our M33, NGC300 and M81 catalogs with the catalogs published by Thompson et al. (2009) and Khan et al. (2010), and found no notable discrepency or significant photometric mismatch.

This catalog is a resource as an archive for studying mid-IR transients and for planning observations with the James Webb Space Telescope. Our survey is being expanded to galaxies with $\sim10$x higher integrated star formation rate than for these seven galaxies. While we have shown that surveys for stellar populations are feasible using archival Spitzer data, JWST will be a far more powerful probe of stars in the mid-IR. The nearly order-of-magnitude higher resolution (Gardner et al.2006) of JWST compared to Spitzer can be used either to greatly reduce the problem of confusion in these galaxies or to greatly expand the survey volume.

This work is based on observations made with the Spitzer Space Telescope, which is operated by the Jet Propulsion Laboratory, California Institute of Technology under a contract with the National Aeronautics and Space Administration (NASA). We extend our gratitude to the SINGS Legacy Survey and the LVL Survey for making their data publicly available. RK is supported through a JWST Fellowship hosted by the Goddard Space Flight Center and awarded as part of the NASA Postdoctoral Program operated by the Oak Ridge Associated Universities on behalf of NASA.