Summary and Conclusions

We have showed in Section 2 (and appendix B) that we are able to reproduce the GALWORKS fixed circular and elliptical photometry to ~2%. The error can be accounted for by the use of fractional pixels in GALWORKS.

The results of the circular and fixed elliptical petrosian photometry show that in most cases the disagreements between the IRAF and the GALWORKS results are within the quoted errors, but not in all. There are cases where disagreements beyond the quoted errors can arise. These differences are derived from the differences in methods used to determine the petrosian radius. When the variable petrosian apertures are examined they are even more problematic. The differences between the GALWORKS and IRAF photometry do not imply that the GALWORKS values are incorrect. In fact, the GALWORKS determination of the petrosian radius is probably more robust than that used for the IRAF values. The results emphasize the need to examine this parameter carefully for a larger sample of galaxies and to be wary of it. Slightly different methods of determining the petrosian value can produce dramatically differing results.

As with the results of the petrosian photometry, the isophotal photometry points to the necessity for closer study especially once three-channel data is obtained. The differences between the GALWORKS and IRAF results for the circular and fixed elliptical photometry are often larger than the quoted magnitude error, but the difference is smaller than in the petrosian case. Again the method used to determine the isophotal radius must be considered as a factor in these results. The differences in the photometry for the variable ellipse case is similar to that for the fixed case. Most of the differences are less than the quoted values.

One of the most important parameters to understand is that of the axial ratio and the position angles. The axial ratio in particular is important to understand for use in the Tully-Fisher relation. In these cases given it appears that we are able to determine the axial ratios to a few percent in all but the case of the distorted galaxy 1795 for which the axial ratio is not a very well defined parameter. To get a good grasp on the errors involved in the axial ratio a larger sample must be examined. We have even less information about the errors in the position angle in this study. Disregarding galaxy 1869 because it is so close to circular and 1795 because of its strange shape, the position angles agree within a few degrees for the two galaxies remaining.

The main gist of these results seems to indicate that there is nothing grossly wrong with the way GALWORKS is doing photometry (unless for some reason we think that IRAF is making the identical mistake). It will be important to do further investigation of a larger sample of data in order to fully understand the errors inherent in the magnitudes and the utility of parameters like the petrosian radius.