For a scan to be included in a 2MASS data release, we currently require that the average seeing shape in the scan be < 1.25, which corresponds to a FWHM of approximately 3.5 arcseconds. For an average seeing shape of 1.25, the seeing can exceed this limit for large fractions of a scan. Both the maximum seeing per scan, and the RMS of the mean shape have been tabuluted for over 40,000 scans that have been processed to date. The summary statistics below illustrate how the maximum and RMS shapes vary with mean seeing.
The plots indicate what has been known for some time: as the mean seeing degrades, the variability in the seeing increases. We may want to consider placing limits on either the maximum seeing shape per scan, or the RMS, along with mean seeing for the releases.
| Band | N/Ntot | Percentage |
| J | 1178/43062 | 2.7 |
| H | 775/43062 | 1.8 |
| Ks | 615/43062 | 1.4 |
worst-case 1196/43062 = 2.8%
| Band | N/Ntot | Percentage |
| J | 2104/41886 | 5.0 |
| H | 1722/42290 | 4.1 |
| Ks | 1640/42449 | 3.9 |
worst-case 2207/41868 = 5.3%
Instances of untracked seeing will roughly correlate with poor seeing, but are not necessarily limited to those periods. The following plots show the J, H and Ks seetracker scores plotted as a function of the maximum seeing shape value in a scan. The first set are for all scans, and the second set is for scans that have mean seeing shapes per scan < 1.25, the current threshold for inclusion in the release catalogs.
The number of scans with seetracker score > 1.0 is smaller in the subset with mean seeing shapes < 1.25, but there are still a significant number of untracked seeing instances.
