next up previous contents
Next: Example 6: DIRBE FITS Up: Examples of FITS Headers Previous: Example 5: ADC FITS

Discussion of Example ADC FITS Table Header for AGK3 Catalog

This ASCII table (XTENSION keyword) header was created by L. Brotzman on September 15, 1988 (DATE keyword) as an illustration of the FITS table headers that would appear on an ADC CD-ROM and were distributed with a request for comments. It is for the AGK3 catalog (HISTORY keywords; also EXTNAME) as compiled by W. Dieckvoss (AUTHOR) and collaborators (HISTORY), published in 1975 by the Hamburg-Bergedorf Observatory (REFERENC). The same catalog had been used for the example header in the FITS ASCII table paper (Harten et al. 1988).

The values of BITPIX, NAXIS, PCOUNT, and GCOUNT are those required by the rules for ASCII tables. Each row consists of 22 fields (TFIELDS) and is 80 characters long (NAXIS1). The sum of the field sizes specified in the TFORMn keywords, where n refers to the field number, is 80, the same as the value of NAXIS1; thus, for this table, the fields fill the entire row. There are 183145 rows (NAXIS2). The values of the TBCOLn keywords show where each column starts, the TTYPEn keywords give a column title, usually explained in the comment field of the same keyword, and the TFORMn keywords the FORTRAN format of the data in the field. Some keywords appear in only some fields: TUNITn keywords provide units, all in lower case in this table although no special significance is to be ascribed to the case, TSCALn and TZEROn keywords show where the value in the FITS table must be transformed to yield the physical value; and TNULLn gives the null string where needed. Blank lines are treated as comments, because the keyword field is blank.

The following list describes the contents of the fields, as described by the header keywords.

1.
AGK3 starts in column 1 and contains a seven-character string giving the declination zone and sequence number within the zone of the star.

2.
AGK3_COMP consists of a single character in column 8 giving a component identification for a member of a multiple star system. The string ``0'' (zero) represents a single star.

3.
PMAG starts in column 9 and contains a three-digit integer giving the photographic magnitude of the star in units of tenths of a magnitude. The value in the field must be multiplied by 0.1, the value of TSCAL3, to obtain the physical quantity in units given by TUNIT3.

4.
SPTYPE starts in column 12 and contains a two-character string giving the spectral type.

5.
RAH starts in column 14 and contains a two-digit integer with the hours of right ascension of the position of the star, equinox 1950, Besselian.

6.
RAM starts in column 16 and contains a two-digit integer, with the minutes of time of right ascension of the position of the star, to be added to the hours of the previous field, equinox 1950, Besselian.

7.
RAS starts in column 18 and contains a five-digit integer, which is 1000 times the number of seconds of time of the right ascension of the position of the star, to be added to the hours and minutes of the previous fields, equinox 1950, Besselian. This design allows storage in integer fields of right ascension accurate to 0.001 second of time.

8.
DECSIGN is a single character in column 23 giving the sign of the declination. Note that, as discussed in connection with the TFORMn keywords in section 3.4.1, the sign is in a separate field, rather than incorporated into the value of the declination, in order to express the distinction between -0 and +0.

9.
DECD starts in column 24 and contains a two-digit integer with the absolute value of the number of degrees of declination of the position the star, equinox 1950, Besselian.

10.
DECM starts in column 26 and contains a two-digit integer, with the minutes of arc of the declination of the star, to be added to the degrees of the previous field, equinox 1950, Besselian.

11.
DECS starts in column 28 and contains a four-digit integer, which is 100 times the number of seconds of arc of declination of the star to be added to the degrees and minutes of the previous fields, equinox 1950, Besselian This design allows storage in integer fields of declination accurate to 0.01 second of arc.

12.
NUMOBS starts in column 32 and contains a two-digit integer giving the number of photographic observations used to determine the position.

13.
EPOCH starts in column 34 and contains a four-digit integer giving the epoch--when the measurement was made--of the AGK position in hundredths of years after 1900; to get the epoch, the number in this field is multiplied by 0.01 and added to 1900. Because the EPOCH keyword has been used in the past to give the equinox of the coordinate system (section 3.1.1.2), it generally better to use a different keyword (DATE-OBS, MJD-OBS) for the time of observation.

14.
PM_RA_COSD starts in column 38 and is a five-digit integer containing the product of the annual proper motion in right ascension and the cosine of the declination, in units of thousandths of arcseconds/year, allowing this five-digit field to contain values accurate to 0.001 arcsec/year. Applying the scaling transformation yields values that are in units of arcseconds/year, the value of TUNIT14.

15.
PM_DEC starts in column 43 and is a five-digit integer containing the annual proper motion in declination, in units of thousandths of arcseconds/year, allowing this five-digit field to contain values accurate to 0.001 arcsec/year.

16.
EPOCH_DIFF starts in column 48 and is a four-digit integer giving the difference between the epoch of the AGK3 position and the epoch of the AGK2 position in hundredths of years.

17.
BD starts in column 52 and contains a seven-character string giving the Bonner Durchmusterung number of the star.

18.
BD_COMP consists of a one digit integer in column 59 giving a numerical code identification of a component of a multiple star system. The string ``0'' is the null value that would be in the field if the star is single. As discussed in section 3.4.2, the value of the TNULL18 keyword is a character string and need not be readable in the integer format specified by the TFORM18 keyword.

19.
DISC_SUM is a two-digit integer starting in column 60 containing a sum that expresses in one number which discrepancy codes are present. A discrepancy code flags a value that is inconsistent with that in another catalog or is otherwise considered unreliable. The sum is \(\sum_{V} 2^V\) where V are the discrepancy codes for the particular position; all values from 0 to 5 are possible. The meaning is explained by the COMMENT keywords immediately following the other column 19 keywords; the DISC_SUM following the COMMENT allows the text to be associated with the appropriate column even if the keywords are rearranged by the reader or a new table is created using some of the data in the current table and the column number changes. You could also think of the DISC_SUM as a hierarchy similar to those in the HIERARCH proposal discussed in section 5.7.

20.
PM_RA is a six-digit integer starting in column 62 containing the annual proper motion in right ascension, in ten-thousandths of seconds of time per year.

21.
RA_RESID is a six-digit integer starting in column 68 containing the difference between the ADK3 right ascension and the ADK2 right ascension in ten-thousandths of seconds of arc, corrected for changes in position due to proper motion.

22.
DEC_RESID is a six-digit integer starting in column 74 containing the difference between the ADK3 declination and the ADK2 declination in ten-thousandths of seconds of arc, corrected for changes in position due to proper motion.


next up previous contents
Next: Example 6: DIRBE FITS Up: Examples of FITS Headers Previous: Example 5: ADC FITS