An important aim in the conversion process is the addition of useful information that was never recorded on the observational record or that was removed before the observations were filed. The information that should be recorded includes the reference catalogue used for the reductions, the telescope used and (where appropriate) the program code. The first and last items have been recorded or retained on the observational record in recent years but the aim is to get all three pieces of information recorded on the bulk of the more than 19 million astrometric observations currently in the MPC archive. This has necessitated, for example, the manual examination of printed copies of the MPCs spanning (as of Aug. 18) 1978 to 1991, and entering details of program codes and telescopes into a datafile for use by conversion process (sometimes having to differentiate between observations of the same object made on the same night at the same observatory by three different observing programs!). This has been tedious in the extreme. As of late on Aug. 18, a newly-written computer program is doing some preliminary scanning of the input files for the 1992 through 1997 batches of MPCs. (With a slight modification to cope with the B1950.0-format observations, it will also pre-process the small number of the remaining unprocessed 1991 batches.) These preliminary scans require manual editing and this has been completed as far as (as of Aug. 31) late-1995 batches of MPCs. The processing of the MPS batches is complicated because we don't archive an ASCII version of this journal, so we will have to dig the ASCII text out of the DVI file. (This is actually quite difficult to do if one wishes to extract some semblance of the original layout, and initial attempts have been only semi-successful.) Once extracted, the resulting file should be able to be run through the above-mentioned program.
But the MPC needs the help of the observers in completing the task.
The datafiles are available from the CF's ftp site:
If you have difficulty accessing the ftp site, please review the notes on the MPCORB page (specifically, the bits about "reverse DNS lookdups" and "anonymous ftp passwords"). Note that these files are not mirrored on the MPCORB mirror sites.These datafile will be updated on an irregular basis as observers report corrections or additions. The telescope datafile has been constructed from examining the observational headers from just two batches of MPCs: it is expected that there will be a large number of additions (and some corrections) required.
These two datafiles are used in preparing a third datafile (the "observer setup information", which is not accessible on-line) which will actually be used during the observation conversion process.
The format of the line(s) in the message should match the format of the on-line datafile. Columns 1-3 contain your observatory code. Column 6 contains the telescope identifier (first telescope = '1', second = '2', etc.): use the next available identifier. The telescope descriptor then begins in column 9. When describing the telescope, be consistent with the form of the existing entries. Note that apertures are generally given to 2 d.p. (e.g., 0.30-m instead of 0.3-m) and that focal ratios are not given for Schmidt-Cassegrain telescopes.
If you are correcting an existing entry, retain the existing identifier and put an asterisk ('*') in column 7. You can specify multiple telescopes for multiple sites in a single e-mail.
There is no specific format required for submission of a new catalogue code. You can request multiple catalogue code additions in the same e-mail. The four-character hex catalogue code (shown in columns 1-4) will be assigned by the MPC.
If you are lucky, you will only have to submit one line of OSI. This will be case if you have only ever observed at one site, used one telescope and reduced your observations with a single reference catalogue. It is more likely that you will have to submit a couple of lines. If you are unlucky, you might have to submit many lines (if you switch between telescopes or reference catalogues frequently, even during one night).
The format of an OSI line is as follows (with a column guide to aid you):
1 2 3 4 5 6 123456789 123456789 123456789 123456789 123456789 123456789 XXX YYYY MM DD.dddddd-YYYY MM DD.dddddd PMRRRR T OBJwhere:
ZZZ 2010 08 07 -2010 08 09 1081 1 K10A00A ZZZ 2010 08 07 -2010 08 09 1022 1 ZZZ 1020 2Note that the order of these lines is important in such a situation. The first line applies to the observations of K10A00A made between 2010/08/07 and 2010/08/09 (inclusive), the second line applies to the non-K10A00A observations made on those same nights and the third line catches everything else. If the order of the lines were changed, at the very least, the observations of K10A00A would not have the correct information added. The last line must be the "catch the remainder" description.
As another example, consider the same object being observed on the same night at the same observatory using different instruments. Let's use the previous example with the complication that a single observation of K10A00A on 2010 Aug. 08.39394 was made with telescope 2 two hours after the last observation of the same object with telescope 1. Now we need four OSI lines:
ZZZ 2010 08 08.39394 1081 2 K10A00A ZZZ 2010 08 07 -2010 08 09 1081 1 K10A00A ZZZ 2010 08 07 -2010 08 09 1022 1 ZZZ 1020 2An alternate way of expressing the same information uses a variant of the OBJ form:
ZZZ 2010 08 07 -2010 08 09 1081 2 K10A00A[2010 08 08.39] ZZZ 2010 08 07 -2010 08 09 1081 1 K10A00A ZZZ 2010 08 07 -2010 08 09 1022 1 ZZZ 1020 2In this form, a date can be placed in square brackets after the designation. This date is interpreted as "only apply these details to the observations whose date string matches '2010 08 08.39'". Multiple dates may be given inside the same brackets, separate each date with a '/'. You need only enter enough of the data string to uniquely identify the observations that are to be modified by this line.
If you frequently switch between different reference catalogues, you can simplify the OSI line by omitting the catalogue code for observations made after August 2001. Then you need distinguish only between different telescopes.
As noted above, most observers will not have to concern themselves with such complexity. But if in doubt, ask!
You should send your OSI line(s) to gwilliams@cfa.harvard.edu with the subject line "OSI".
Our Web policy. Index to the CBAT/MPC/ICQ pages.