larger largest
The reliability of a two HCON source found in regions of the sky observed with two or three HCONs can be deduced from the values in the table and always exceeds 0.99997 in every range of flux density. The reliability of two HCON sources in regions with many more than two coverages, such as the minisurvey, is, however, considerably less than this value. A source seen only twice in a seven HCON area may have a reliability as low as 0.5. Sources with three or more HCONs are, however, almost completely reliable at any depth of HCON coverage, given the values of p^ and q^ deduced above.
|
|
|
Figure VIII.D.1 a) Completeness at 12 µm calculated for a
two HCON survey (top); and b) for a three HCON survey (bottom). larger largest |
The completeness is plotted as a function of flux density in Fig. VIII.D.1a for a two HCON survey and in Fig. VIII.D.1b for a three HCON survey. Note that these figures give the differential completeness, not the cumulative completeness above a given flux density. The completeness is essentially unity above 1.5 Jy for a two HCON survey and above 0.6 Jy for a three HCON survey. It begins to fall sharply at 0.5 Jy for a two HCON survey and at 0.4 Jy for a three HCON survey. Some of the physical reasons for missing HCONs are discussed in Section VII.E.3.
The completeness and reliability values quoted here for 12 µm also apply
to 25 and 60 µm sources with the sharp fall-off in the completeness
occurring at 0.5 and 0.6 Jy at 25 and 60 µm for a three HCON
survey. The reliability is decidedly worse for sources seen only at 25
µm. The 100 µm values cannot be obtained from the minisurvey due
to extensive cirrus contamination.
The completeness of the catalog can be verified by the differential source-counts shown in Fig. VIII.C.1 for the four wavelength bands for |b| > 50°. The counts fall away sharply at 0.4, 0.5, 0.6, 1.0 Jy at 12, 25, 60, 100 µm. The completeness at 100 µm is severely degraded by cirrus below these Galactic latitudes.
A moderate quality flux density is based on at least four detections (Section V.H.5), but the moderate status means that the source also failed to be detected on at least one occasion when it should have been seen. If there were no alibis from dead detectors, the source may have only been detected on four occasions out of a possible eight. If the source had a good quality flux density in another band, the reality of the source is established in that band. In this case there is a reasonable certainty that the source detection in the moderate quality band is real, but the quoted uncertainty in the quoted flux density is larger than that for good quality fluxes, typically by a factor of 1.41. Moderate quality flux densities should be used with caution. Sources with moderate quality flux densities in two adjacent bands are reliable.