X.D. Extended Emission

IRAS Explanatory Supplement
X. The Formats of the IRAS Catalogs and Atlases
D. Extended Emission

Chapter Contents | Authors | References
Table of Contents | Index | Previous Section | Next Section

  1. Introductory Comments
  2. Map Projections and Transformation Equations
    1. Gnomonic Projection - 16.5° Images
    2. Equivalent Cylindrical Projection - Galactic Plane Fields
    3. Aitoff Projection - Low-Resolution All-Sky Maps
  3. 16.5° Images
    1. Prints of 16.5° Images
    2. Tapes of 16.5° Images
  4. Galactic Plane Maps
  5. Low-Resolution All-Sky Maps
  6. Zodiacal Observation History file
  7. Coordinate Overlays

D.1. Introductory Comments

The extended emission data are presented as maps of the infrared sky at two different scales in all four wavelength bands. The entire sky surveyed is mapped with 2' pixels at an effective resolution of 4'-6' in 212 16.5° × 16.5° fields and also with an effective resolution of l° in a single field. A special map of the region within ± 10° of the Galactic plane was also made at 4'-6' resolution. All of these maps are available in both digital and photographic formats. In addition, data from the survey averaged in a ½° × ½° beam is available in time ordered form. The details of the presentation of each of these products are described below. The methods used to produce the products were described in Section V.G. All the map projections used are described below.

The survey covered most parts of the sky several times and the extended emission data have been separated for the reasons outlined in Section V.G.1. The data released in Nov. 1984 includes the 188 fields in the 75% of the sky covered in the third sky coverage, along with the associated Galactic plane and low-resolution all-sky maps. Subsequent releases include a small part of the sky covered by the "minisurvey" done at the beginning of the mission (Section III.C.11) and the first and second coverages, each of about 95% of the sky.

D.2. Map Projections and Transformation Equations

Three separate map projections were used. The descriptions below provide enough information for the user of the IRAS data. More information can be found in Richardus and Adler (1972).

D.2.a Gnomonic Projection - 16.5° Images

The gnonomic projection used for 16.5° images produces a geometric projection of the celestial sphere onto a tangent plane from a center of projection at the center of the sphere. Each individual field has its own tangent projection plane with the tangent point at the center of the field. This projection is neither conformal (angle preserving) nor equivalent (equal area) but does have the property that all straight lines in the projection are great circles on the sphere. All projections were done so that the sky coordinate associated with a pixel refers to the position at the center of the pixel. For the 16.5° fields the maximum distortion of angles is 0.6° and the maximum distortion of area is 6%. The area distortion is approximately proportional to the inverse cube of the cosine of the angular displacement from the center of the field. The distortion is in the sense to make extended areas cover more sq. arcminute pixels than their true solid angles would require. This results in overestimating fluxes when integrating sources within fixed intensity contours.

The transformation equations for conversion between line and sample number in the map and right ascension and declination on the sky are shown below.


scale = 30 pixels/degree
A = cos(X) × cos(X - X)
F = scale × (180/(X))/[sin(X) × sin(X) + A × cos(X)


LINE = -F × [cos(X) × sin(X) -- A × sin(X)]
SAMPLE = -F × cos(X) × sin(X -- X)



X = SAMPLE(scale × 180/(X))
Y = LINE/(scale × 180/(X))
(X) = arctan [(X2 + Y2) ½ ]
(X) = arctan(-X/Y)
XX = sin(X) × sin(X) × cos(X) + cos(X) × cos(X)
YY = sin(X) × sin(X)

(X) = arcsin [sin(X) × cos(X) - cos(X) × sin(X) × cos(X)]
(X) = (X) + arctan(YY/XX)


where (X) and (X) are the declination and right ascension of the field center. The arctangent functions for (X) and (X) must be four quadrant arctangents.

D.2.b Equivalent Cylindrical Projection - Galactic Plane Fields

The Lambert normal equivalent cylindrical project was used to provide an equal area projection of the sky within 10° of the Galactic plane. The projection cylinder is tangent to the celestial sphere at the Galactic equator and the projection proceeds by projecting radially outward from each point on the polar axis of the Galactic coordinate system in a plane parallel to the equatorial plane. The maximum angular distortion (deviation of bearing) is 0.9°. The equal area property of the transformation preserves photometric accuracy when integrating fluxes for an extended source.

The transformation equations are:


scale = 30 pixels/degree
LINE = -scale × 180/(X) × sin(X)
SAMPLE = -scale × (X - X)



(X) = (X) - SAMPLE/scale
(X) = -arcsin [LINE/(scale × 180(X)]


where (X) and (X) are Galactic latitude and longitude and subscript zero denotes the field center.

D.2.c. Aitoff Projection - Low-Resolution All-Sky Maps

The Aitoff equal area projection was used to provide a photometrically correct map of the entire celestial sphere. Galactic coordinates were chosen as a convenient and natural coordinate system. The transformation equations are:


scale = 2 pixels/degree
(X) = arccos [cos(X) × cos(X - X)/2]
0 = arcsin [cos(X) × sin({X - X}/2)/sin(X)]


SAMPLE = -4 × scale × 180/(X) × sin(p/2) × sin(X)
LINE = ± 2 × scale × 180/(X) × sin(X)/2) × cos(X)


where the'+' applies to (X) < 0 and the '-' to (X) = 0.


Y = -LINE/(2 × scale × 180/(X))
X = -SAMPLE/(2 × scale × 180/(X))
A =(4 - X2 - 4 × Y2) ½
(X) = 180/(X) × arcsin(A × Y)
(X) = (X) + 2 × 180/(X) × arcsin[A × X/(2 × cos(X))]


where (X) and (X) are Galactic longitude and latitude and subscript zero denotes the field center.

D.3 16.5° Images

The 16.5° images the high-resolution presentation of the IRAS sky survey in image form. Two hundred twelve (212) 16.5° × 16.5° fields cover the whole sky with field centers spaced by approximately 15°. The three sky coverages of the full mission are presented as three separate sets of 212 maps, with

some maps not included in the third coverage (Table X.D.1). A list of plate numbers vs. plate centers comprises Appendix X.2 and a map of plate locations is given in Fig. X.D.1. An individual map consists of 499 × 499 array 2' × 2' pixels into which the IRAS survey data were mapped using a gnomonic projection (described in Section X.D.2.a). The 16.5° images are available as photographic prints and as digital magnetic tape. The formats of these two forms are described below. Detailed descriptions of the procedures used to produce the maps are given in Section V.G.

Plates Missing From the Third Sky Coverage
Table X.D.1
4795 130
6096 131
7097 132
71107 143
72108 144
83118 154
84119 115
94120 168

D.3.a. Prints of 16.5° Images

Photographic black and white negative transparencies were produced from the digital map data with a film recorder. All four bands of each field were reproduced side by side in a rectangular format approximately 5 inches sq. intended for enlargement to 16 inches by 20 inches. Two hundred fifty-six (256) brightness levels were available with the film recorder and the brightness range in each band of each field was individually compressed, clipped and scaled to fit within these 256 levels.

The compression, clipping and scaling were accomplished by first extracting the fifth root of the surface brightness to compress the dynamic range of the data. A histogram of the the root map was made and the pixel values were shifted and scaled into the 0 - 255 range so as to saturate the lower and upper one percent of the histogram. The approximate surface brightness value of any pixel can be recovered by first comparing the density of the pixel of interest to the grey scales which show every 17th pixel value from zero to 255. The shift and scale are removed using the 0 DN = X Jy sr-1 and 255 DN = Y Jy sr-1 information in the label (see the sample label below).

The complete formula is:

surface brightness (Jy sr-1) = [ (D/255)(Y1/5 - X1/5) + X1/5]1/5


where D is the pixel value determined by comparison with the grey scale and Y and X are obtained from the label.

Final calibration factors were not applied to the third sky coverage data before production of the photo products. Therefore, the intensities obtained using the procedure above on the third coverage may not agree with intensities found on the magnetic tape versions of the maps. The magnetic tape version uses the final IRAS catalog calibration and should be used in case of discrepancy.

Intensities obtained from the third coverage photo products can be approximately corrected to the final calibration by multiplying the intensity derived from a photograph by the following factors:

12 µm 0.84
25 µm 0.80
60 µm 0.80
100 µm 0.69

The first, second and mini-survey coverage photos were corrected to the final calibration.

The coordinate system of each map is arranged so that when viewed with the printing in the label right side up north is up and east is to the left at the field center, which is adopted as pixel (0, 0). In this orientation the horizontal rows of pixels are by convention called lines and the pixels within each line are called samples. Line numbers increase from top to bottom and sample numbers increase from left to right. The line numbers of the top and bottom lines are given in the label as TOP and BOTTOM, respectively. Similarly the left and right extreme sample numbers are given as LEFT and RIGHT. With this information and the tic marks along the sides of the image area the line and sample coordinates of any pixel can be determined for application of the inverse map projection formulae given above in Section X.D.2. The tic marks also allow alignment of the coordinate overlay grids as described in Section X.D.7.

Color composite negative transparencies in 4 × 5 inch format of each sky plate field have been produced by recording the 100 µm map in red (positive), the 60 µm map in green and the 12 µm map in blue. These color versions of the data are not intended for quantitative analysis. The shift and scale information in the label is difficult to read and no attempt has been made to produce a consistent color balance among the plates. In one plate a particular hue will indicate one ratio to 100 µm to 60 µm to 12 µm brightness and in another plate the ratio for that hue will be somewhat different.

Sample Label from 16.5° Image Photograph

     IRAS SKYFLUX HCON:    3    FIELD: 153     DEC:-30 RA:   10:00    25 MICRON

     R.A. & DEC GNOMONIC    PIXEL: 2.00 ARCMIN      JD:2445654.25-2445660.25

     TOP:-249 BOTTOM: 249 LEFT:-249 RIGHT: 249     1 TIC = 5 PIXELS

     0DN= 4.28E+ 2 JY/SR    255DN= 5.38E+ 2 MJY/SR     5TH ROOT DATE:84/08/10

First line:

SKYFLUX: refers to the 212 maps which cover the whole sky with 2' pixels. Can also be ALL SKY for the ½° pixel all-sky maps or GALPLANE for the Galactic plane maps.
HCON: serial number of the sky coverage
FIELD: IRAS field number. See maps in Fig. X.D.1.
DEC, RA: coordinates of field center. DEC in degrees, RA in hours and minutes. Can also be LON and LAT for Galactic coordinate (X) (X) projections.
MICRON: wavelength band of the map.

Second line:

RA & DEC GNOMONIC: Projection type. Can also be LON & LAT CYLINDRICAL or LON & LAT AITOFF
PIXEL: Pixel size.
JD: Julian dates of the earliest and latest data in the map, accurate to ½ day.

Third line:

TOP,BOTTOM,LEFT,RIGHT: Line and sample numbers at the edges of the map.
1 TIC: Spacing between the tic marks around the edge in pixels

Fourth line:

0DN, 255DN: Surface brightness values of the minimum and maximum pixels in the map.
5TH ROOT: Indicates that the fifth root of the actual surface brightness is shown. Can also be LINEAR in which case the true, uncompressed, surface brightness is shown.
DATE: Year/Month/Day on which the map was assembled.

D.3.b Tapes of 16.5° Images

The magnetic tape form of the 16.5° images contains the calibrated surface brightness data in 499x499 arrays of 2' × 2' pixels recorded in the FITS format. The article by Wells et al. (1981) in conjunction with the label records of each tape file gives a detailed description of the format of each map image. A brief description of the format follows and a listing of a sample FITS label can be found in Appendix X.3.

One sky coverage consists of 27 tapes of 6250 bpi (bits per inch) density. The third coverage has only 24 tapes with a total of 188 plates. Each plate consists of four surface brightness maps and four statistical weight maps, one of each for each wavelength band. The plates are ordered on the tapes by plate number and within a plate the image files are ordered: 12 µm brightness, 12 µm weight, 25 µm brightness, 25 µm weight, 60 µm brightness, 60 µm weight, 100 µm brightness and 100 µm weight. The first two records of each file contain the label, then the image appears as a stream of pixel values divided into 2880 byte records without regard for line length. The stream begins with the smallest line and smallest sample number and the sample number increases fastest. The last record is padded to 2880 bytes with zeros. Four-byte integers are used for brightness image data numbers and 2-byte integers for weight images, high order byte first. All this and other information necessary to successfully regenerate a map is contained in the FITS label records described in Appendix X.3.

D.4. Galactic Plane Maps

For convenience in dealing with the Galactic plane the survey data within 10° of the Galactic plane were remapped from the into a set of images in Galactic coordinates to cover the full circle of the Galaxy. This remapping from the 16.5° images resulted in a slight degradation in resolution even though the pixel size was the same in both sets of maps. Twenty-four 16.7° × 20° fields cover the Galactic plane with field centers at integral multiples of 15° Galactic longitude. The three sky coverages of the survey were separated into three sets of maps. The image format is 499 lines of 599 samples each, projected from Galactic coordinates with an equal area cylindrical projection (see
Section X.D.2.b). Galactic plane maps are available in both photographic and FITS tape formats. Two 6250 bpi tapes of 12 maps each hold the 24 Galactic plane maps. The differences in the FITS label between the 16.5° images and the Galactic plane maps are noted in the description and listing of the FITS label in Appendix X.3. No statistical weight images are included for the Galactic plane maps. Statistical weight information may be obtained from the 16.5° maps. Coordinate overlays described in Section X.D.7 are available for the Galactic plane maps.

D.5. Low-Resolution All-Sky Maps

The ½° × ½° beam data contained in the Zodiacal History file described below was split into the three separate sky coverages and assembled into three all-sky maps with an Aitoff equal area projection in Galactic coordinates. Two fields of each sky coverage were produced; one centered on the Galactic center and one centered on the Galactic anti-center. The pixel arrays consist of 325 lines of 649 samples each. Galactic north is in the direction of decreasing line number (up) and Galactic east in the direction of decreasing sample number (left). The all-sky maps are available in both photographic and FITS tape forms. The formats of the photographic and tape forms of all-sky maps are very similar to those of the 16.5° images (see Section X.D.3) with the differences described in the labels of the photographs and tape files. Coordinate overlays described in Section X.D.7 are available for the all-sky maps.

D.6. Zodiacal Observation History file

For convenience in the analysis and treatment of background emission from interplanetary dust (zodiacal emission) and other extremely large scale emission features, the survey data were averaged to ½° × ½° beam size and along with pointing information was preserved as a time ordered data set containing all three sky coverages of the survey. This file is available on magnetic tape written with the format described in Appendix X.4.

D.7. Coordinate Overlays

A set of coordinate overlays for the photographs is available as photographic negative transparencies in the 5 inch sq. format. The scale is identical to the corresponding map product so the overlays will be the correct size if enlarged by the same factor as the map. One overlay is provided for each declination zone from -30° to +30° where the overlays for zones of opposite sign are obtained by rotating the grids through 180°. Five overlays are provided for each declination zone between 45° and 75° to accommodate the fact that integer hour meridians cross the plates in these zones in five different configurations. Again the overlay for the zone of opposite sign is obtained by rotating the grid 180°. All integer hour meridians are labeled 00M. The hour of right ascension should be determined from the position of the plate center given in the label on the photograph. The plate numbers to which a particular overlay pertains are printed in the lower right cornerof the overlay. The overlays are aligned by matching the two triangular fiducials along each edge of the overlay with the two large map border tics which straddle the center of each side of the map. The two overlays for the polar regions are similarly aligned with the fiducials and tics. The correct orientation makes the lettering on the overlay read the same way as the lettering in the plate label.

One overlay is used for all Galactic plane maps. It is aligned with the same method as the sky plates. One orientation of the overlay is used for even numbered fields and has 0° as the center longitude; the other orientation, used for all fields, has 5° as the center longitude. The tens digit of the true longitude should be obtained from label of the picture.

The overlays for the low-resolution all-sky maps come in only two varieties, (X) = 0.0° in the center and (X) = 180.0° in the center. Alignment of the overlays is similar to that for the other maps.

Chapter Contents | Authors | References
Table of Contents | Index | Previous Section | Next Section