Unweighted Mean Positions These files are intended to provide a handy source for fairly accurate positions for the NGC/IC objects. They have only one entry per object. An unweighted mean position follows the NGC/IC name as I've adopted it in the position data files. Next comes the number of positions used in calculating that mean, the calculated standard deviations in RA and Dec, and the sources used. The precision to which the position is given reflects the precision of the most precise position in the full list. [Brian Skiff and I are working on a list of "best" positions for each object which will eventually replace these means. Those are ready for NGC objects from 1 to 1999 (with a few beyond 2000 as well), and for IC 1 to 999 (again, with a few beyond); the ones that are ready are flagged with an "s" or a "c" following the number of positions used for the mean. Check back once in a while for these, or check NED for the NGC/IC galaxies.] Note that precision does not necessarily reflect accuracy! especially with unweighted means. However, I've rejected obvious blunders, and many objects have enough accurately-measured positions now that overall accuracy will not suffer much from poor positions out on the wings of the "bell curve". Two sets of files are available, identical except for the equinoxes: B1950.0 and J2000.0. The equinox B1950.0 file is the "root" file, with the equinox J2000.0 file being derived from it. The B1950.0 data in the individual files (of one thousand NGC/IC numbers each, available in the "Miscellaneous Files" section of my page) are identical to the data in the combined B1950.0 files; download whichever set you want. ****************************************************************************** Please note that J2000 positions produced between mid-2001 and 28 October 2005 may have errors of over an arcminute in them due to a bug in GCC 2.96. Most of the precessed positions are all right, but a few are not (see e.g. NGC 1593 where the RA seconds for the incorrect position is listed as 16.333; the correct RA seconds is 06.133). Other positions in later batches were affected by compiler errors which produced many precessed positions containing "NaN" ("Not a Number"), indicating computation errors. The source of those remains a mystery and has disappeared with the latest (November 2010) versions of the files processed with the GCC compiler that I use on workstations at IPAC. The B1950 positions are not affected. My apologies for any problems arising from this. ****************************************************************************** Some Picky Details: I've used standard deviations rather than mean errors as these more accurately reflect the uncertainties in the unweighted mean positions. Errors in published positions (see the Sources file) are almost always underestimated, so quoting the standard deviation increases the qualitative perception of the errors to about what they should be in an ideal world. If I ever get around to calculating weighted means, the mean errors will have more relevance. In the meantime, if you want them, mean errors can be easily calculated by dividing the standard deviations by the square root of the number of positions used. The precession routine that I used came from NED, which I am pleased to acknowledge: This research has made use of the NASA/IPAC Extragalactic Database (NED) which is operated by the Jet Propulsion Laboratory, California Institute of Technology, under contract with the National Aeronautics and Space Administration. See http://nedwww.ipac.caltech.edu/help/calc_help.html for more details about the precession. Finally, these positions are strictly on the FK5 system -- they have not been reduced to the more recently adopted International Celestial Reference Frame (ICRF). However, the FK5 optical system is consistent with ICRF to within the known systematic errors (a few milliarcseconds) of the older system (see Ma et al, Astronomical Journal, Vol. 116, p. 516, 1998 and references therein for more information about ICRF). Given the relatively large accidental errors (on the order of a few tens of arcseconds in the best cases) in the present data, you may also take these positions as being systematically consistent with the ICRF. 19 January 2011