III. 4. B. Determining color-terms and zero-points



Source code involved:

We use autocalib1.cl to determine the color-terms and zero-points. The program dumps an allstar file into a temporary file, converts the x and y to ra and dec, and then invokes the FORTRAN code "matchem" to match the Mosaic stars with the 42-inch catalog stars. Next, the FORTRAN program "fitem" is invoked to find the zero points and color terms based on these matches, after applying a rather stringent median filter to get rid of mismatched data. The program also queries a data file called kpnocoefs.dat to see if we should actually fit a color term, or adopt some previously determined best value. We are slowly accumulating our list of "good" color terms here . The program "calcoefs" is used to digest the results of the output file "fitresults" in order to determine the average color-terms, an intermediate step.

The results of this process are a zeropoint and color term for transforming Mosaic photometry to the standard system. The transformation for V would look like this: Mosaic(V) = V + zeropt + (B-V)*cterm. The color term for the U image is (U-B); for B is (B-V); for R is (V-R); and for I is (R-I).

All this good stuff gets added to the headers of each image as "zeropt", "zerr" (error in the zeropt), "cterm", and "cerr" (error in the color term). Plots can be generated by uncommenting out a few lines. The zeropt, cterm, and their errors are also put in a file called fitresults and imagename.ans (for use in invertfit). The quantities in fitresults are image name, filter, constant term, its uncertainty, color term, its uncertainy, number of stars used to determine the transformation, and two other numbers.