The most important difference between the first and second versions of the IRAS Point Source Catalog (hereafter denoted PSC-1 and PSC-2) is the application of a statistical correction to reduce the overestimation of the flux densities of sources near the detection threshold (see XI.J. for a discussion of the effect); changes of as much as a factor of two were made for some weak sources.
A number of other, less significant changes were also made. Flux densities for almost all sources changed by a few percent due to refined calculations of certain calibration factors; a few sources changed by as much as 10%. Corrections to errors in some calibration and confused source processing algorithms resulted in the loss of 6 sources from and the addition of 56 new objects for a revised total of 245,889 individual sources in PSC-2. Table XII.A.5 lists 12 sources which were omitted from both PSC-1 and PSC-2 due to a software error, but which are of sufficient quality to be noted for completeness. A number of new catalogs were added to the list used for associations. Finally, a number of minor errors in calculations of some of the quantities associated with point sources, e.g. the number of neighboring sources, were corrected.
A.1 The Flux Overestimation Correction
As described in Chapter V, sources were extracted from individual detector streams by means of a zero-sum filter followed by a template fit to the data. Only if the detection exceeded thresholds in its signal to noise ratio (SNR) and in its correlation with the template of a "perfect" point source was the detection accepted for subsequent processing. A source was accepted into the catalog if it had enough valid detections to confirm its existence on time scales of seconds, hours and weeks. However, a failing of this strategy (as described in more detail in Section XI.J) is that the flux density of weak sources is overestimated, since the combination of weak source and positive-going noise (Gaussian excursions or spikes due to radiation hits) resulted in a "valid" detection, while the combination of a weak source and negative-going noise (Gaussian excursions) forced the detection below the acceptable thresholds. Thus, for weak sources, the detections used to form the average flux density.
XII-1
Table XII.A.1 IRAS Data Products
Status as March 1987
PRODUCT DESCRIPTION MEDIUM VERSION DATE
Explanatory Detailed description Book 1.0 Nov. 1984
Supplement of hardware, data
processing, and products
Revisions, New Chap XII, Book 2.0 Jun. 1987
Index
Point Source 245,839 point sources Tape, 1.0 Nov. 1984
Catalog (PSC-1) Microfiche,
Tape 1.1 Jan. 1986
245,889 point sources Tape 2.0 Nov. 1986
(PSC-2). Updates described Book 2.0 Jun. 1987
in this Chapter
Ancillary File More detailed information Tape 2.0 Nov. 1984
on point sources 2.1 Jan. 1986
3.0 Nov. 1986
Working Survey More detailed information Tape 2.0 Nov. 1984
Data Base on point sources 3.0 Feb. 1986
4.0 (PSC-2) Nov. 1986
High Source Bins processed by high Tape 1.0 Nov. 1984
Density Bins source density processor
for Catalog
Point Source 372,774 rejected point Tape available at IRAS
Reject File sources data centers only
Reject File More detailed information Tape available at IRAS
Ancillary File on rejected point sources data centers only
Reject File More detailed information Tape available at IRAS
Working Survey on rejected point sources data centers only
Data Base
LRS Spectra Spectra of 5,425 catalog Tape,Hard 1.0 Nov.1984 Catalog point sources (8-22 Copy
µm). See XII.C
Zodiacal History Time-ordered data at Tape 1.0 Nov. 1987
0.5 degree resolution 2.0 May 1986
All Sky Maps All sky images at Tape 1.0 Nov.1984 0.5 degree resolution
XII-2
Table XII.A.1 IRAS Data Products
(Continued)
PRODUCT DESCRIPTION MEDIUM VERSION DATE
Sky Brightness 16 × 16 deg B/W,Color 1.0 Nov.1984 images (HCON 3) 2' Resolution Negatives (Sept-Nov)a
B/W 2.0 May 1986
Negatives (Dec.1985)a
Tapes 2.0 May 1986
(Mar.1986)a
Sky Brightness 16 × 16 deg field B/W 1.0 Aug.1985
Images (HCON 1) 2' Resolution Negatives (Feb-Apr)a
Tapes 1.0 Aug 1985
(May 1985)a
Sky Brightness 16 × 16 deg field B/W 1.0 June 1986
Images (HCON 2) 2' Resolution Negatives (Apr 1986)a
Tapes 1.0 June 1986
Sky Brightness 16 × 16 deg field B/W Negatives 1.0 Nov. 1984
Images (Overlays)
Galactic Plane 2 deg × 15 deg field B/W 1.0 Jan. 1985
Images (HCON 3) 2' resolution Negatives, 2.0 July 1986
Tapes
Galactic Plane 2 deg × 15 deg field B/W
Images (HCON 1) 2' resolution Negatives, 1.0 Oct. 1985
Tapes
Galactic Plane 2 deg × 15 deg field B/W
Images (HCON 2) 2' resolution Negatives, 1.0 Jul.1986
Tapes
Cataloged 11,444 point sources Book 1.0 Feb.1985
Galaxies and associated with cataloged
Quasars galaxies and quasars
Pointed 13,853 images with 0.25'-1.0 Tapes 1.0 Oct.1985
Observations resolution, each covering
~ 1 square degree
Pointed Description of AO's, Book 1.0 Nov.1985
Observations data reduction, and
User's Guide b released grids
Serendipitous Description of Catalog Book 1.0 Dec.1986
Survey Explanatory
Supplement c
Serendipitous 43,866 point sources
Survey Catalog derived from the Tape, micro- 1.0 Dec.1986
Additional Observations fiche
XII-3
Table XII.A.1 IRAS Data Products
(Continued)
PRODUCT DESCRIPTION MEDIUM VERSION DATE
Small Scale Description of Book 1.0 Dec. 1985
Structure Catalog Catalog
Explanatory
Supplement Mar. 1987
Small Scale 16,740 sources Tape, 1.0 Dec. 1985
Structure Catalog with sizes < 8 ' Microfiche,
Book Mar. 1987
CPC Explanatory Detailed description of Book 1.0 Aug. 1985
Supplement CPC hardware, data,
processing, and products
CPC data Chopped photometric 3 tapes 1.0 Jan. 1986
channel image data
IRAS Asteroid IRAS and derived data Tape 1.0 Oct. 1986
and Comet Survey on known asteroids
IRAS Asteroid Description of Catalog Book 1.0 Oct. 1986
and Comet Survey and Summary Information
Explanatory
Supplement
a Internal product dates on tape versions show months in this range
b Young et al. 1985
c Kleinmann et al. 1987
reported in PSC-1 were systematically high. This problem predominantly affected moderate quality fluxes of weak sources; brighter sources detected with high quality fluxes generally had enough detections that flux overestimation was not a large problem.
The most direct way of measuring the overestimation is to compare the brightness of sources in PSC-1 with the values obtained from the more sensitive pointed mode of the satellite (Young et al. 1985). Observations made in the pointed mode were three to five times more sensitive than the scans making up the survey so that the sources detected in the pointed mode would not suffer from threshold effects at the same flux level. The Serendipitous Source Catalog (Kleinmann et al. 1987, hereafter denoted as the SSC) is the catalog of sources extracted from selected pointed mode observations and was used as a "truth table" to determine the magnitude of the magnitude of the flux overestimation effect and to help develop a correction for it. Figures XII.A.1a-4a show the ratio of SSC to PSC-1 flux densities for the four bands as a function of SSC flux density; the presence of an overestimate in the PSC-1 values below about 2 Jy is obvious.
XII-4
Figure XII.A.1 Ratio of SSC to PSC-1 (a) or PSC-2 (b) 12 µm flux densities vs. SSC flux density before (a) and after (b) the correction of the overestimation effect.
XII-5
Figure XII.A.2 Ratio of SSC to PSC-1 (a) or PSC-2 (b) 25 µm flux densities vs. SSC flux density before (a) and after (b) the correction of the overestimation effect.
XII-6
Figure XII.A.3 Ratio of SSC to PSC-1 (a) or PSC-2 (b) 60 µm flux densities vs. SSC flux density before (a) and after (b) the correction of the overestimation effect.
XII-7
Figure XII.A.4 Ratio of SSC to PSC-1 (a) or PSC-2 (b) 100 µm flux densities vs. SSC flux density before (a) and after (b) the correction of the overestimation effect.
XII-8
A simple model for the overestimation described in Section XI.J leads to an algorithm for the correction of the effect. Consider a detector stream with an intrinsic Gaussian noise level (X) and a source with an intrinsic brightness of n(X). The zeroth moment of the Gaussian distribution above the threshold of m (x) for a source with true flux n(x) gives the probability of detection (Eqn. XI.J. 1) and the first moment above m(x) gives the observed flux (Eqn. XI.J.2). These moments were used to relate the flux correction factor to the observed fraction of possible sightings for a given source.
The algorithm was implemented in the following manner. The Working Survey Data Base (WSDB) contains the number, N, of all accepted detections. If the total number of possible detections is denoted by M, then the ratio N/M is an estimate of the probability of detection, denoted by p, which can be used in Equation XI.J.1 to estimate the quantity n--m. This value can be used with Equation XI.J.3 to derive the ratio n'/n which is an estimate of the amount by which the true brightness of the source was overestimated in PSC-1.
For each source fainter than a certain, band-dependent level (2 Jy at 12, 25 and 60 µm and 3 Jy at 100 µm), the number of accepted detections, N, was determined from the WSDB. At the same time the total number of possible detections, M, was determined from a detailed calculation of the satellite's pointing history over the entire mission. If the source passed within the central portion of a working detector then a possible sighting was recorded. The central portion of the detector was defined by a distance in from the edge of the detector and was considered a free parameter, denoted by (X)Z, in each band; the value of this parameter affected the number of possible detections for a given source. The resultant value of N/M was used along with an estimate of the cutoff threshold, m, in that band to derive the correction factor as described above.
The SSC data represented in Figure XII.A.1a-4a were assumed to give the true brightness for faint sources and were used to adjust the parameters (x)Z and m on a band by band basis to make the curves as flat as possible. The algorithm was applied only to those sources which were neither confused nor located in regions of high source density in a given band. Approximately 100,000 sources were adjusted in at least one wavelength. Values of (X)Z and m and the maximum derived correction factors are given in Table XII.A.2.
Table XII.A.2 Flux Overestimation Parameters
Wavelength m (X)Z Max. Correction
(um) (SNR Threshold) (')
12 4 0.2 1.72
25 2 0.1 3.00
60 3 0.7 2.01
100 3 1.0 2.38
XII-9
The photometric uncertainty associated with each flux was also adjusted slightly by examining how uncertainties in the pointing of the telescope could affect the number of possible sightings. Variations in the N/M ratio were propagated through Eqn. XI.J.1 and XI.J.3. and the resulting uncertainty added in quadrature to the uncertainty already given in the catalog.
The above formalism obviously ignores many complications such as differing sensitivities among the detectors within a band and non-Gaussian noise sources like radiation hits and cirrus which could increase a source's flux on some occasions, but not others. Detections could have fallen below thresholds for any number of reasons including sightings by a detector of poor sensitivity or passage over the edge of a detector rather than over its center. A sighting could have been rejected for failing to exceed either the SNR or template thresholds. All of these effects were ignored in the above algorithm. As a result, although the corrections are quite good in a statistical sense, they may be considerably in error in any individual case.
Despite these simplifications, the algorithm worked remarkably well. Figure XII.A.1b-4b shows the comparison of the PSC-2 and SSC fluxes. While some structure including a slight offset in some bands is still evident in the data, the curves are much flatter than those in figure XII.A.1a-4a; the offset may be due to residual uncertainties in the calibration of the photometry in PSC-2 and the SSC.
Figure XII.A.5 shows the effect of the flux overestimation on the photometry for weak sources in PSC-1 and the effectiveness of the correction applied in the generation of PSC-2. The figure plots the 12/25 µm color of stars as a function of their 12 µm brightness. In fig. XII.A.5a the flux density ratio comes from PSC-1 and the 12 µm brightness from the SSC. Above about 2 Jy a distinct populationof stars with colors characteristic of hot (>3000 K) photospheres is apparent, but below 2 Jy the population of stars with photospheric colors disappears and all stars appear to have a 25 µm excess. This astrophysically puzzling result can be understood terms of the 25 µm flux being systematically overestimated for faint stars. The corrected colors from PSC-2 (Fig. XII.A.5.b) reveals a distinct population of stars with normal photospheres at faint levels.
As a result of this correction algorithm, the machine readable version of PSC-2 includes five new fields: MHCON, the number of possible hours confirmed sightings a source could have had and the four band dependent correction factors that were applied to the PSC-1 flux densities to produce the PSC-2 fluxes. These quantities are given only for those sources for which a flux overestimation correction was made. It should be noted that the value of MHCON for any given source could be uncertain by one or two sightings.
The format of the Catalog tape has been revised to incorporate these new quantities. Table XII.A.3 lists the changes to the original format (Table X.B.1, page X-4).
XII-1O
Figure XII.A.5 Ratio of 12 µm to 25 µm flux densities from the PSC-1 (a) or PSC-2 (b) vs.12 µm flux density from the SSC before (a) and after (b) the correction of the overestimation effect.
XII-11
---------------------------------------------------------------------------------------------------------------------------
Table XII.A.3 Revised Portion of Catalog Tape Format
(see Table X.B.1)
Start
Byte Name Description Units Format
139 MHCON Possible number of HCONs --- 12 141 FCOR Flux Correction Factors
Applied (x 1000)
(1 value per band) --- 4I4
149 SPARE 3 Spare bytes --- 3A1
--------------------------
A.2 Additional Flux Density Changes
Flux densities for all sources in the PSC-2 were adjusted by a few percent to make them consistent with the final calibration described in Chapter VI. The differences arose because PSC-1 used a preliminary version of the all-sky intensity maps to correct for the non-linearities of the feedback resistor (Section VI.A.5). Also, as a result of this change, minimum upper limit fluxes are now exactly 0.25, 0.25, 0.40, and 1.0 Jy, instead of varying by a few percent from these values.
In two areas of the sky, from 0 degree to 29 degree ecliptic longitude at -90 degree to 67 degree latitude and at 60 deg flux o 65 degree latitude, flux densities reported in PSC-1 were factors of 1.076, 1.099, 1.013, and 0.967 too high at 12, 25, 60 and 100 um, respectively. Approximately 1% of all sources were affected by this error which has been corrected in PSC-2.
A.3 New and Deleted Sources
A number of effects changed the total number of point sources present in PSC-2. First, six sources present in PSC-l were deleted from PSC-2 because the small flux changes mentioned above pushed the fluxes for these sources below a threshold used by the high source density processor. Second, fiftysix sources were added to PSC-2 because the very near neighbor window was set incorrectly in parts of the production of PSC-l. Thus. the total number of sources in PSC-2 is 245,889. The names of the changed sources are given in Table XII.A.4.
Finally, a software error, discovered after the release of PSC-2, led to the exclusion of a small number of sources from both PSC-1 and PSC-2. Examination of the Working Survey Data Base (WSDB) led to the discovery of sources with similar fluxes located within +/-30" in-scan and +/-90" cross-scan of one another. In each case only a single hours confirmed sighting was reported for each object. Evidently, the weeks confirmation software failed to confirm and merge these objects into a single source which would then have appeared in the catalog.
XII-12
Table XII.A.4 New or Deleted Point Sources
New Sources
01241-7332 05265-6840 08220-4404 18373-0918
04531-6708 05299-6830 08573-4718 19514+4306
04532-6710 05303-6951 09040-7402 21097+4953
04534-6657 05305-6952 09101-5100 21099+4954
05044-7012 05341-6632 09179-7034 21314+5805
05046-7014 05371-6944 09215-7028 21329+5126
05122-6829 05374-6946 09437-6034 21345+5706
05150-6629 05375-6650 10075-5747 21445+5653
05150-6631 05392-6847 10075-5824 21446+5655
05207-6636 05393-6930 10182-5742 21478+5649
05236-6702 05448-6720 12376-6122 22126+6905
05239-6939 05486-7001 13191-6245 22299+6440
05242-6940 08003-5012 18092+4419 22510+7138
05264-6730 08005-5013 18285-0830 22512+7140
Deleted Sources
04526-6951 13046-6222 20250+4316 21503+5105
04538-6952 15574-0052
A procedure was developed for finding these objects systematically in the WSDB. First, the WSDB was searched for all sources with only one hours confirmed sighting, located above an absolute galactic latitude of 10° and above an ecliptic latitude of -80° (to avoid the Large Magellanic Cloud). Second, sources detected only at 25 µm or only at 100 µm were eliminated to avoid contamination by asteroids or cirrus. This sample was examined for objects having companions within +/-30" in-scan and +/-90" cross-scan. Those objects with nearby companions and meeting the final catalog selection criteria (Sections V.H.2 and V.H.7.) were taking as real objects. A final position was derived by simply averaging the position from each sighting; final flux densities were computed by averaging the logarithms of the flux densities without using any weights.
Forty-three objects were selected by this procedure. Examination of the raw data led to the elimination of seven of these 43 as being due to either cirrus or confused objects. Finally, to avoid any problems of source reliability or flux overestimation only sources with a flux density greater than 1 Jy in at least one band were selected for inclusion in Table XII.A.5. The format of the table is similar to that of the printed version of the catalog, as described in Section X.B.2. It should be emphasized that these sources appear only in the table and are not included in the printed or machine readable versions of either PSC-1 or PSC-2.
XII-13
Table XII.A.5
Bright Sources Missing from the Point Source Catalog with |b| > 10 Degrees
-- Name -- Gal ---Flux Density in Janskys-- N C -Association-
X (1950) X X X Coord (Not Color-Corrected) C P S I A -Name & Type MAG
HHMMT DDMM (s) (") 1 b NH 12um 25um 60um 100um 1 W 2 D T M-01-01-060 999
00125-0723 33.5 10 98-68 2 .25L .44L 1.46 4.62 0 1 10 M+04-28-050 999
11434+2042 26.3 50 234+74 2 .25L .46: 3.19: 6.45 1 2 10 U07772 103
12337+2616 47.4 51 230+86 2 .52: .53 4.50 23.31 0 1 7 6 9 D0 20951 1
21492+3716 13.4 1 88-12 2 .36: .25L .48 2.95 15 4 1 2 3746 K0 67
22261+8025 9.1 59 117+20 2 1.01 .28L .40L 1.63L 7 1 13 52130 K0 77
22308+4105 50.3 26 97-14 2 2.05 .54 .40L 6.60L 3 2 13 MKN 914 999
22324+4024 28.4 35 97-15 2 4.43 18.49 29.10 64.39 12 2 27 X2232+408 4
22325+4054 32.0 25 97-15 2 .25L .26L 1.08 7.01 8 1 4 1 32 DG 187 999
22326+4031 37.8 9 97-15 2 .41L .25L 1.10 7.33L 9 1 23
22376+2426 36.8 3 89-29 2 .29L .25L 1.05 2.68 0
23019+3405 57.7 32 99-23 2 .25L .25L 1.45 2.50 0
23132+2449 14.4 12 97-33 2 .25L .25L .48 1.35 4 3 9 U12460 155
A.4 Revised Completeness Estimates for PSC-2
The completeness of PSC-2 was investigated in two different ways. First, PSC-2 was internally checked through analysis of the differential source counts. Second, PSC-2 was externally checked against the SSC which is complete to significantly fainter flux densities than PSC-2. Results from both methods corroborate the general statements made in Chapter VIII that the survey (PSC-1 or PSC-2) is complete, in unconfused regions of the sky, to 0.4, 0.5, 0.6 and 1.0 Jy at 12, 25, 60, and 100 µm.
PSC-1 and PSC-1 differ in the flux density level associated with a given completeness level. Figures XII.A.6a,b show the differential log N/log S curves for 12 and 60 µm for PSC-1 and PSC-2 at high Galactic latitudes covered with two sets of hours confirming scans. Because of the correction for flux overestimation, the peak of the curves is broader in PSC-2 than in PSC-1. The shapes of the curves implies that PSC-2 has more weak sources than PSC-1, but that those sources are very incompletely represented e.g., ~10% completeness at 0.3 Jy at 12um. At the same time, PSC-2 does not achieve a 90% completeness level until slightly higher flux density than PSC-1.
Tables XII.6a,b summarize the flux densities at which a given completeness is reached derived from the source counts in PSC-1 and PSC-2. Values are given for regions with both two or three sets of hours confirming scans (2 or 3 HCONs). To the extent that the flux overestimation algorithm was successful, the table for PSC-2 reflects the true completeness of the IRAS survey as a function of flux density. Note that while it is quite accurate to derive the completeness from the logN/logS curves below completeness levels of about 50%, it is difficult to derive accurate estimates above that level. Thus the flux density values at completeness levels of 90% and 95% are indicated in the tables as uncertain.
XII-14
Figure XII.A.6 a) The differential sources counts versus 12 µm flux density for PSC-1 (dashed line) and PSC-2 (solid line).
b) The differential source counts versus 60 µm flux density for PSC-1 (dashed line) and PSC-2 (solid line).
XII-15
Table XII.A.6a
Flux Densities for a Given Completeness Level
(PSC-1 )
Flux Density (Jy)
Completeness 2 HCON sky 3 HCON sky
(%) 12 µm 60um 12 µm 60 um
95 [.42]* [.59] [.38] [.52]
90 [.40] [.56] [.36] [.50]
50 .33 .47 .32 .44
10 .29 .40 .29 .39
5 .28 .37 .28 .38
* Numbers in brackets are uncertain
Table XII.A.6b
Flux Densities for a Given Completeness Level
(PSC-2)
Flux Density (Jy)
Completeness 2 HCON sky 3 HCON sky
(%) 12 µm 60um 12 µm 60 um
95 [.47]* [.60] [.46] [.60]
90 [.44] [.59] [.44] [.57]
50 .35 .48 .30 .45
10 .24 .38 .23 .33
5 .22 .34 .22 .30
* Numbers in brackets are uncertain
The second way to evaluate the completeness of the survey is to use the SSC, which, as discussed above, reaches three to five times fainter than the survey. A comparison was made of all SSC sources at Galactic latitudes (X) > 30° and (X) </- 50° (to avoid the Magellanic Clouds). The SSC sources were divided into stars [fv(12um) > fv(25um) and fv(12um) > fv(60um)] and galaxies [fv(12um) < fv(60um)], in order to isolate the 12 and 60 µm bands as much as possible. A search radius of 120" was used to find counterparts in PSC-2. Tables XII.A.7a,b summarize the completeness derived from the percentage of SSC sources that were found in
PSC-2 for both the 2 and 3 HCON sky at 12 and 60 µm. Given in the table are the number of actual objects in PSC-2 (denoted N') compared with the possible number of objects (denoted M') found in the more complete SSC.
XII-16
As a check on the consistency of these two approaches and on the agreement of the results in the 2 and 3 HCON sky areas, all the above numbers can be converted to an estimate of the completeness of a single HCON and plotted on the same graph. To do so, recall from Chapter VII that in terms of the single HCON completeness C (= l - p), the completeness of the PSC in the 2 HCON is C 2 and in the 3 HCON sky is C 2(3 -- 2C).
Figures XII.A.7a,b show the single HCON completeness at 12 and 60 µm. There is good agreement between the log N /log S and the SSC vs PSC-2 results. In addition, the numbers from the 2 HCON sky and the 3 HCON sky are also consistent. However, the 12 µm results are significantly different from the results shown in figure VIII.D.1 derived from sources in the minisurvey. Because those sources were in the 7 HCON sky, they suffered tremendously from the effects of flux overestimation. This results in a shift of the minisurvey completeness curves to erroneously higher flux densities. Table XII.A.8 summarizes the Catalog completeness at 12 and 60 µm as derived from the curves shown in Figure XII.A.6a,b.
Table XII.A.7a
12 µm Completeness from SSC
Flux density 2 HCON sky 3 HCON sky
(Jy) N' M' Completeness N' M' Completeness
0.50-0.55 6 6 1.00 16 16 1.00
0.45-0.50 4 5 .80 18 18 1.00
0.40-0.45 13 15 .87 17 17 1.00
0.35-0.40 7 11 .64 15 15 1.00
0.30-0.35 8 17 .47 15 23 .65
0.25-0.30 6 24 .25 24 39 .62
0.20-0.25 3 26 .12 4 50 .08
Table XII.A.7b
60 µm Completeness from SSC
Flux density 2 HCON sky 3 HCON sky
(Jy) N' M' Completeness N' M' Completeness
0.60-0.65 14 17 .82 11 11 1.00
0.55-0.60 10 13 .77 17 20 .85
0.50-0.55 14 25 .56 20 24 .83
0.45-0.50 8 22 .36 18 29 .62
0.40-0.45 5 29 .17 14 35 .40
0.35-0.40 4 36 .11 13 66 .20
0.30-0.35 5 49 .10 10 70 .14
XII-17
Figure XII.A.7 a) The single HCON completeness versus 12 µm flux density. A smooth solid line has been drawn through the data. The other solid line is from the 2 HCON SSC results, the dashed line from the 3 HCON SSC results, the solid points from the 2 HCON logN/logS results, and the open points from the 3 HCON logN/logS results. Uncertain points are in parentheses.
b) The single HCON completeness versus 60 µm flux density. The smooth solid line has been drawn through the data. The other solid line is from the 2 HCON SSC results, the dashed line from the 3 HCON SSC results, the solid points from the 2 HCON logN/logS results, and the open points from the 3 HCON logN/logS results. Uncertain points are in parentheses.
XII-18
Table XII.A.8
Best Estimate of Completeness Level
(PSC-2)
Completeness 2 HCON sky 3 HCON sky
(%) 12 µm 60 µm 12 µm 60 µm
95 .46 .65 .41 .58
90 .45 .64 .38 .56
75 .41 .59 .34 .51
50 .35 .52 .29 .45
25 .29 .45 .25 .38
10 .25 .38 .23 .32
5 .23 .34 .22 .30
A.5 Associations
The PSC-2 contains associations with sources in four additional catalogs not used in the preparation of PSC-1: the IRAS Small Structure Catalog, the IRAS Serendipitous Survey Catalog, the OSU catalog of radio sources and the Michigan Spectral Catalog. Parameters for these new catalogs are given in revised versions of Tables V.H.1 and X.B.4. These catalogs were added at the end of the queue for the printing priority in the printed version of the PSC. However, in the printed version of PSC-2, an asterisk in the column giving the number of associations (NID) denotes an association of a PSC-2 source with an object in the Serendipitous Survey Catalog.
There was an error in associating PSC-1 sources with stars in the Gliese catalog since the proper motion in right ascension was erroneously taken as seconds of arc per year instead of seconds of time per year. Remedying this error resulted in the addition of 37 associations in PSC-2 and the deletion of five associations originally presented in PSC-1. The changed associations are given in Table XII.A.9.
The ESO/Uppsala Catalog, catalog 14, was previously given a "multiple" classification instead of "other". This has been corrected in PSC-2.
XII-19
Table XII.A.9 Changed Gliese Associations
Newly Associated Sources
00027-3737 01365-1812 06562-44l3 16451-4737
00156+4344 01416-1611 08053+6952 16564+4726
00176-6509 01504-2240 11027+4347 17023-0459
00235-773l 02085-5103 13275+l038 20017+2312
00295+6657 02334+0639 13432+1508 20542-4419
00348-2502 03168-6245 13469-2151 21141-3904
00461+5732 03172-6241 14006-4634 21598-5700
00589+7124 03180-4315 14260-6227 22070-0452
01051+5439 05100-4502 14359-6031 23029-3607
23110+5653
Sources No Longer Associated
07483+8023 08355-3958 16267+1831 19145+0505
21362-2732
The following minor changes exist in the associations with all catalogs for sources in the area bounded by ecliptic longitude 260° -280° and ecliptic latitude -60 degree to 60 degrees. The truncation error mentioned in Supplement XI.K.10 has been fixed. One new association with the catalog of Suspected Variables, catalog 16, was made exactly at the maximum 90" radius allowed (source 18237-2417). The latest version (1984) of the SAO catalog was used in the generation of PSC-2; six new associations were made and eight old associations were lost. These sources are given in Table XII.A.10. All 14 of these associations were at the maximum radius allowed.
Table XII.A.1O Changed SAO Associations
Newly Associated Sources
17167-3229 17370-3843 18311-1734 18402-7755
17326-3324 17510-3726
Sources No Longer Associated
16516-6705 17337-0220 17458-0937 17555+3324
17236-2125 17394+2611 17519-3035 17557+3351
A.6 Source Names
Source names were derived from the equatorial position by taking the hours, minutes and tenths of minutes (truncated, not rounded) of right ascension as well as the sign,° and minutes of declination. For example a source at (X) =12h22m15.5s and (X) =-15 deg 20'15" has the name 12222-1520. Separate sources which would have been given the same name based on the above scheme were distinguished from one another by appending letters of the alphabet to their
XII-20
name. It should be noted that since sources were named before the decision to retain them in the catalog was made, it is possible that not all of the sources with names distinguished only by a letter will be present in the Catalog, e.g. 12222-1520B might be in the catalog, but not 1222-1520A.
A.7 Revised Positional Uncertainties for Bright Sources
As described in the Section VII.C.1.b, an additional 3" was supposed to be added in quadrature to the in-scan uncertainty of sources that were both faint and had the minimum possible (1") in-scan uncertainty. That correction was inadvertently applied in PSC-1 to all sources, faint and bright, with the minimum uncertainty. PSC-2 corrects this problem and has 84,287 sources with this increased uncertainty, as opposed to 144,403 sources in PSC-1. The minimum uncertainty ellipse is now 3" × 5", as opposed to 3" × 7".
A.8 Correction of Point Source Neighbor Counts
In PSC-1 the counts of hours and weeks confirmed neighbors, denoted PNEARH and PNEARW, of a given source are in error for objects near which a forced weeks confirmation took place (Section V.H.3). This problem has been fixed for PSC-2.
A.9. Spurious 25 µm Only Sources
The following three sources have been found to be spurious on the basis of examination of co-added detector data, but were left in the catalog since no general rule could be used to delete them. They are: 05570-6722, 08291-6146, and 18021+6556. The first and third sources were produced by noise and radiation hits; the second resulted just from noise.
A.10 Working Survey Data Base, Ancillary File and Reject Files
The version of the WSDB corresponding to PSC-2 is 4.0 (Nov. 1986) and differs from 3.0 only in the revised number of sources in the catalog described above (56 additions, 6 deletions). Version 3.0 (Feb. 1986) corrected a calibration error in the flux densities of individual HCON sightings. Version 2.0 was the first publicly released version of the WSDB.
Two errors were remedied in Version 3.0 of the Ancillary file. First, the "fault" byte in HSDDROC (see Table X.B.7c) was incorrect for a small number of sources due to the incorrect very near neighbor window. This is now corrected. Second, the first publicly released version of the Ancillary File (version 2.0) contained associations with the weeks-confirmed file for the Small Scale Structure Catalog: only about half of these weeks-confirmed sources survived to the final SSS catalog. The hex coding of the bands in which an extended source was detected
XII-21
was calculated incorrectly. These problems have been fixed in version 3.0 of the Ancillary File.
Listings of point sources that failed the catalog selection criteria, e.g., confusion or insufficient number of HCONs, were not publicly released, but are available from the IRAS data center in Pasadena.