An empirical test for the metallicity sensitivity of the Cepheid period-luminosity relation

Abstract

We have performed a test to determine whether differences in chemical composition affect the determination of distances using the Cepheid period-luminosity relation. For Cepheids in three different regions in the disk of M31, independent determinations of the mean reddenings and true distance moduli have been made from BVRI CCD observations. These three radially separated regions have metallicities ranging over about a factor of 5, as estimated from abundance analyses of supernova remnants and H II regions. Using multiwavelength period-luminosity relations to determine true distance moduli simultaneously with mean reddenings, consistent distances are found for all three fields. The large observed differences in the apparent moduli can be totally attributed to different mean extinctions among the three fields. That is, we find no statistically significant residual differences in the three true distance moduli. The moduli derived by us are 24.33 +/- 0.12, 24.41 +/- 0.09, and 24.58 +/- 0.12, for Baade's Fields I, III, and IV, respectively (which are situated at 3, 10, and 20 kpc from the center of M31), under the assumption that the same value of 3.1 for the ratio of total-to- selective absorption is valid in all three fields. This conclusion is not materially changed under the assumption of reasonable differences in this ratio among the three fields. From a synthesis of published theoretical models Stothers predicts a metallicity-induced offset of +0.80 mag (Field IV minus Field I) for the extinction-corrected moduli derived from B and V photometry; we find at most a difference of only +0.l9 mag; from BVRI photometry the difference is +0.25 mag, factors of 3-4 smaller. However, the same theoretical models predict that, at longer wavelengths, the effects of metallicity should be very small, in good agreement with our main observational results.