The extragalactic Cepheid distance bias: Numerical simulations

We present numerical simulations for a selection bias in the extragalactic Cepheid method (as detected in Teerikorpi & Paturel 2002), which results in underestimated distances for host galaxies. The selection effect involves the Cepheid detection magnitude limits in V and I, the dispersion in <M(P)>, and importantly, variation amplitude, observable upper period limit, and also extinction. It influences both the bias curve fitting method and the method where short-period Cepheids are removed. When V, I, and P-limits exist, one expects a typical bias pattern in the log H vs. M-lim diagram. A full bias correction is possible only for special favourable situations, but there always appears a nearly unbiased plateau. The simulated bias varies from galaxy-to-galaxy roughly as was revealed by the strong and puzzling observed trend for galaxies in TP02.