Sources in the All-Sky Release Source Catalog saturate the WISE detectors at characteristic magnitudes of 8.1, 6.7, 3.8 and -0.4 mag in W1, W2, W3 and W4, respectively. WISE profile-fitting photometry reliably extracts measurements of saturated sources using the non-saturated wings of their profiles up to brightnesses of approximately 2.0, 1.5, -3.0 and -4.0 mag in W1, W2, W3 and W4, Brighter than these levels, the reliability and completeness of WISE photometric measurements degrades because there are too few non-saturated pixels available in the measurement area for reliable source extraction.
The profile-fitting photometry module of the WSDS (WPHOT) subsystem characterizes sources that are saturated by fitting the PSF to the non-saturated portions of the source profile. As long as there are sufficient numbers of non-saturated pixels available within the fitting area, the source brightness and position can be extracted. Two features added to WPHOT for second-pass processing that improved the extraction accuracy were a larger PSF model and a dynamically sized fitting radius that increases in response to the presence saturation. However, when the radius of the saturated region exceeds the size of the stored PSF model, equal to 1.84 arcmin in W1, W2 and W3, and 43.5 arcsec in W4, it is not possible to perform a reliable profile-fit and hence there is a maximum brightness limit for good source extraction.
Although WPHOT returns a flux measurement for saturated sources, the accuracy of the measurements may be degraded. Photometric uncertainties are elevated at the onset of saturation, and other biases, such as brightness-dependent PSFs lead to systematic errors in the estimated flux levels as saturation intensifies.
During the measurement of each source, WPHOT determines the fraction of pixels that are saturated within the fitting region, and reports the values in the w?sat fields of the source Catalog record. Saturated pixels are identified by WPHOT using the single-exposure bit-masks that capture the DEB saturated-pixel encoding provided in the raw payload image data. A pixel may saturate because of the presence of a bright astrophysical source, or from the strike of an energetic cosmic ray. We make use of these saturated pixel statistics to determine the characteristic brightness at which point sources reach saturation and at which WPHOT is unable to return a useful magnitude.
Figures 1-4 show differential source counts near the saturations levels for all entries in the All-Sky Release Catalog that are not flagged as being spurious detections of or contaminated by image artifacts (cc_flags[?]='0'). In each of these diagrams, differential source counts are plotted as a function of profile-fit magnitude, w?mpro. The black curves represent all sources, the green curves are sources that have w?sat=0 and the red curves are sources that have w?sat>0. W?sat values are computed in the Multiframe Pipeline version of WPHOT (IV.4.c) using the pixels in all of the frames in the "stack" that contribute to a source measurement in a particular band.
The onset of saturation for point sources is indicated by the transition between the curves for saturated and non-saturated sources. The saturation boundary is well-defined but not perfectly sharp because the sub-pixel variations in the location of a source near the saturation brightness may affect whether or not it saturates a pixel. Approximately 1% of sources fainter than the nominal saturation boundaries have w?sat>0 because of charged particle strikes within the fitting area of a source.
The characteristic saturation level for point sources in each band is defined as the point at which 50% of sources have a non-zero fraction of saturated sources. Figure 5 shows for each band in the Source Catalog the differential fraction of saturated sources as a function of profile-fit magnitude in that band. In Table 1 are listed the magnitudes at which 50% of the sources have non-zero saturated pixel fractions for each band, along with the implied flux density assuming the values for absolute flux calibration given in the notes to the Table.
 |  |
| Figure 1 - W1 Source Catalog differential source counts | Figure 2 - W2 Source Catalog differential source counts |
 |  |
| Figure 3 - W3 Source Catalog differential source counts | Figure 4 - W4 Source Catalog differential source counts |
| Black curves represent total differential source counts for each band from the Catalog. Green curves represent source counts for sources that do not have any saturated pixels in the profile-fitting region. Red curves represent source counts for sources with one or more saturated pixels in the profile-fitting region. | |
| Band | Magnitude | Flux (Jy)a |
|---|---|---|
| W1 | 8.1 | 0.18 |
| W2 | 6.7 | 0.36 |
| W3 | 3.8 | 0.88 |
| W4 | -0.4 | 12.0 |
 |
| Figure 5 - Fraction of Catalog sources with one or more saturated pixels in the profile-fitting region, plotted as a function of profile-fit brightness for multiframe processing. W1, W2, W3 and W4 are represented by the black, blue, green and red curves, respectively. The horizontal magenta line marks the 50% saturated source fraction. |
Figures 6-9 illustrate the differential source counts near the expected saturation brightness levels from the All-Sky Release Single-exposure Source Database. The brightness range for the transition from non-saturated to saturated is broader in the Single-exposure source measurements than the Catalog because the source sub-pixel centers are not "homogenized" when using only one frame in the measurement.
Figure 10 shows for each band the differential fraction of saturated sources in the Single-exposure Source Database as a function of profile-fit magnitude in that band. In Table 2 are listed the approximate magnitudes of 50% saturation fraction for each band, along with the implied flux density assuming the values for absolute flux calibration given in the notes to the Table. As seen in the distributions in Figure 6-9, the observed saturation levels span a range as much as +/- 1 magnitude around these values.
 |  |
| Figure 6 - W1 Single-exposure Source Database differential source counts | Figure 7 - W2 Single-exposure Source Database differential source counts |
 |  |
| Figure 8 - W3 Single-exposure Source Database differential source counts | Figure 9 - W4 Single-exposure Source Database differential source counts |
| Black curves represent total differential source counts for each band from the Level 1b source Working Database. Green curves represent source counts for sources that do not have any saturated pixels in the profile-fitting region. Red curves represent source counts for sources with one or more saturated pixels in the profile-fitting region. | |
| Band | Magnitude | Flux (Jy)a |
|---|---|---|
| W1 | 7.8 | 0.23 |
| W2 | 6.4 | 0.47 |
| W3 | 3.6 | 1.05 |
| W4 | -0.6 | 14.4 |
 |
| Figure 10 - Fraction of sources with one or more saturated pixels in the profile-fitting region, plotted as a function of profile-fit brightness from scan-frame processing. W1, W2, W3 and W4 are represented by the black, blue, green and red curves, respectively. The horizontal magenta line marks the 50% saturated source fraction. |
The source count distributions in Figures 1-4 and 6-9 exhibit cut-offs near 2.0, 1.5, -3.0 and -4.0 mag (49, 43, 460 and 330 Jy) in W1, W2, W3 and W4, respectively. Brighter than these levels, measurements from profile-fitting photometry in second-pass processing becomes increasingly unreliable because the size of the saturated region for a source exceeds the size of the stored PSF model, which is 1.84 arcmin for W1, W2 and W3, and 43.5 arcsec for W4. There are entries in the Source Catalog and Single-Exposure Working Database with brighter extracted magnitudes, but these are increasingly incomplete and unreliable, frequently contaminated by image artifacts.
The IRAS and 2MASS Point Source Catalogs can be used to estimate the number of objects on the sky that may be too bright to have a useful magnitude reported or that may be missing altogether in the WISE Source Catalog and Single-exposure Source Database. The bluest objects in the WISE Catalog and Database should have have [Ks-W1] and [Ks-W2] colors of approximately zero magnitude. There are 2736 and 1450 sources in the full 2MASS PSC that have Ks< 2.0 and 1.5 mag, respectively. There are 92 sources in the IRAS PSC that have F12>460 Jy and 197 that have F25>330 Jy.
Last update: 2012 February 17