Distribution of Extended Sources with repeated observations. The Y-axis shows the number of repeated observations. Only sources with S_det > 50 are shown (for S_det < 50, the total number is much greater, roughly an order of magnitude).
2MASS scanned the sky with typically ~10% overlap between adjacent tiles. Consequently, the working database contains roughly ~10% of the total with repeated observations. Additionally, ~40 calibration fields were repeatedly scanned over (in most cases) 1000 times, spanning the duration of the project, resulting in comparable duplicate observations for sources located within these regions of the sky. These repeated observations were "merged" to form deeper images (the so called "deep calibration images"), while the database entries were merged to formed more accurate measurements (e.g., positions, photometry) for the repeated observations.
This section presents some basic statistics and characterization of the merged sources in the extended source working datase and deep calibration fields.
i. Method & Merged Parameters
A select set of repeated measurements for extended sources were
averaged to form more accurate source characterization (assuming
the source is non-variable and that the independent measurements
have roughly the same signal-to-noise); these parameters are
listed below.
Average values were obtained by computing the statistical mean of
a set of repeated measurements whose values were determined to
be nominal and not flagged as corrupted or 'null'. Also reported
are the standard deviation of the distribution mean. No attempt
was made to apply a 'weighted' average (mainly due to the fact that
extended sources have complex flux measurements). For photometric
measurments, an additional 'flag' is reported, which represents
the maximum value of the distribution of photometric flags
for the set of measurements (e.g., for the circular r=7 arcsec apertures,
the flags corresponding to the J, H and Ks measurements).
The user should beware that the mean includes all repeated measurements
in the distribution (excluding those determined to be corrupt or null;
see above), which may include 'outlier' or poorly determined measurements.
Outliers may be due to real variability (e.g., extragalactic supernovae)
or unforseen corruptions in the data (e.g., meteor streaks; artifacts from
bright stars; masked pixels in the nuclear regions).
List of averaged measurements are given
here:
Auto-correlation (WAX) V2.0 SIS Documentation for Extended Sources
ii. Calibration Fields
The extended source database for
the calibration scans includes over
400,000 detections. Of these sources,
approximately 1800 have repeated observations
(S_det > 1). A histogram of the number of
repeated observations for these grouped
measurements is given below (showing only those
with S_det > 50).
|
Distribution of Extended Sources with repeated observations. The Y-axis shows the number of repeated observations. Only sources with S_det > 50 are shown (for S_det < 50, the total number is much greater, roughly an order of magnitude). |
The repeatibility is shown in the next set of plots. Here we show the ratio of the number of detections (S_det) versus the total number of observations. For low confusion regions of the sky, the repeatibility is close to 100% for high S/N galaxies (brighter than 15th mag in the J-band). Source detection drops rapidly at J ~ 15.2 mag.
Repeatibility for low source confusion regions of the sky (|glat| > 20 degrees). Ratio of the number of detections versus the total number of possible detections shown with the average integrated J-band flux. Large points represent sources determined to be galaxies. Yellow crosses denote sources examined visually but whose exact nature remains elusive. |
Repeatibility for high source confusion regions of the sky (0 < |glat| < 20 degrees). Ratio of the number of detections versus the total number of possible detections shown with the average integrated J-band flux. Large points represent sources determined to be galaxies or Galactic fuzz. Yellow crosses denote sources examined visually but whose exact nature remains elusive. |
The standard deviation of the distribution mean is dependent on the distribution variance and the total number of measurements, as shown below
Log-log showing the sqrt of the number of detections versus the standard deviation of the mean. Only S_det > 3 sources are shown. The upper panel shows the J-band results (circular aperture, 7 arcsec radius). Note that the aperture flux does not represent the total flux for a galaxy for large (i.e., bright) galaxies. |
iii. 6X Survey
The extended source database for
the special 6X deep scans includes over
247,000 detections. Of these sources,
approximately 30,252 have repeated observations
(S_det > 1).
Repeatibility for low source confusion regions of the sky (|glat| > 20 degrees). Ratio of the number of detections versus the total number of possible detections shown with the average integrated J-band flux. Large points represent sources determined to be galaxies. Yellow crosses denote sources examined visually but whose exact nature remains elusive. |
Repeatibility for high source confusion regions of the sky (0 < |glat| < 20 degrees). Ratio of the number of detections versus the total number of possible detections shown with the average integrated J-band flux. Large points represent sources determined to be galaxies or Galactic fuzz. Yellow crosses denote sources examined visually but whose exact nature remains elusive. |
iv. Working Database
The extended source database for
the calibration scans includes
~2.6 million detections. Of these sources,
approximately 406,636 have repeated observations
(S_det > 1).
Repeatibility for low source confusion regions of the sky (|glat| > 20 degrees). Ratio of the number of detections versus the total number of possible detections shown with the average integrated J-band flux. Large points represent sources determined to be galaxies. Yellow crosses denote sources examined visually but whose exact nature remains elusive. |
Repeatibility for high source confusion regions of the sky (0 < |glat| < 20 degrees). Ratio of the number of detections versus the total number of possible detections shown with the average integrated J-band flux. Large points represent sources determined to be galaxies or Galactic fuzz. Yellow crosses denote sources examined visually but whose exact nature remains elusive. |
[Last Updated: 2004 Sep 21; by Tom Jarrett]