Missed matches - spurious pairs

[Michael Sallman <msallman@pro-ns.net>]

Michael,

You wrote regarding the Mark III tenxcat data:

     On spurious close pairs: When observations from different nights and
     sites were entered into the database, we checked to see if each new
     detection matched an existing entry. We used a matching radius of 15
     arcseconds, which is a bit larger than the pixel size. It turns out
     that we occasionally failed to match to an existing star properly, and
     so created another entry. The result was a spurious close pair. I've
     tried to eliminate some of the close pairs by merging isolated pairs
     which satisfy several conditions. There are still some 3000 or so
     spurious pairs, spurious triplets and clumps with more than three
     entries. MWR June 27, 1999

What "conditions" did you use to determine if the pair should be merged?

Just out of curiousity, I re-ran the 291250 objects in the Mark IV
database through my match program.
These are the results:
Start - 291250 stars
After 1st pass - 277621 stars
After 2nd pass - 277573 stars
After 3rd pass - 277572 stars
After 4th pass - 277572 stars
(if nothing else, this would seem to be a very efficient way to deal
with the problem)

The only criterion to match the objects is the 5 arcsec radius.

So, I might have as many as 13678 "spurious pairs".

I can think of lots of ways to deal with this, but I would like to do it
the right way.

Thanks,

Mike

Michael Sallman wrote:

    > I've finished re-processing Tom's observations (using the correct 
match
    > radius this time). The number of stars has dropped from 384288 to
    > 291250. I've already run across a couple of apparently missed matches.
    > Anyone have any suggestions on the best way of dealing with these. I
    > could run the 291250 "stars" through the match program and combine 
them
    > that way. I could also manually fix them as they are encountered. Or I
    > could just leave them as is. Or?
    > It also looks like I fixed the min/max/avg calculations.
    >
    > Tom has requested sigma values instead of/in addition to the min/max
    > values. I also thought I'd add a Welch-Stetson stat to the Objects
list.
    > I'm also thinking about updating the Objects page or adding a new page
    > with more query options.
    >
    > Anyone have any thoughts or wishes regarding the online data?
    >
    > Mike
    >
    >
    >