1997A&A...324..998F

We present spectroscopic determinations of the Fe abundance in a sample of K-type stars selected from the X-ray flux-limited Einstein Extended Medium Sensitivity Survey. These stars have been shown to have a low value of the m₁ Stroemgren photometric index implying low metal abundances, similar to the abundance observed in Pop. II stars. Yet these cool stars are not expected to be metal-poor, being most likely young and thus metal-rich. Based on an equivalent width-based abundance analysis which uses high resolution spectra in the region around the Li I 6707.8Å doublet, we show that none of the stars investigated shows any evidence for low metal abundances. Rather, all have solar or even slightly super-solar abundances. Thus, for very active K stars, photometric metallicity determinations appear to be strongly biased, with the stars appearing to be of much lower metallicity than they actually are. The stars in our sample are estimated to be coeval or younger than those of the Pleiades. If, as we argue, the bias in the photometric indices depends purely on the high activity level of these stars, it is then likely that all stars sufficiently young to have similar activity levels would display similar photometric anomalies. We then predict that a non-negligible fraction of the disk population K-type stars are expected to show comparable anomalies in their photometric indices. These stars will masquerade as Pop. II stars in photometric surveys, and significantly contaminate Pop. II star samples selected on the basis of photometric properties alone, down to quite faint magnitudes.