A Test for Large-Scale Systematic Errors in Maps of Galactic Reddening

Accurate maps of Galactic reddening are important for a number of applications, such as mapping the peculiar velocity field in the nearby Universe. Of particular concern are systematic errors which vary slowly as a function of position on the sky, as these would induce spurious bulk flow. We have compared the reddenings of Burstein & Heiles (BH) and those of Schlegel, Finkbeiner & Davis (SFD) to independent estimates of the reddening, for Galactic latitudes |b| > 10. Our primary source of Galactic reddening estimates comes from comparing the difference between the observed B-V colors of early-type galaxies, and the predicted B-V color determined from the B-V--Mg_2 relation. We have fitted a dipole to the residuals in order to look for large-scale systematic deviations. There is marginal evidence for a dipolar residual in the comparison between the SFD maps and the observed early-type galaxy reddenings. If this is due to an error in the SFD maps, then it can be corrected with a small (13%) multiplicative dipole term. We argue, however, that this difference is more likely to be due to a small (0.01 mag.) systematic error in the measured B-V colors of the early-type galaxies. This interpretation is supported by a smaller, independent data set (globular cluster and RR Lyrae stars), which yields a result inconsistent with the early-type galaxy residual dipole. BH reddenings are found to have no significant systematic residuals, apart from the known problem in the region 230 < l < 310, -20 < b < 20.