Malmquist Bias and the Distance to the Virgo Cluster

Abstract

This paper investigates the impact of Malmquist bias on the distance to the Virgo cluster determined by the H_0 Key Project using M100, and consequently on the derived value of H_0. Malmquist bias is a volume-induced statistical effect which causes the most probable distance to be different from the raw distance measured. Consideration of the bias in the distance to the Virgo cluster raises this distance and lowers the calculated value of H_0. Monte Carlo simulations of the cluster have been run for several possible distributions of spirals within the cluster and of clusters in the local universe. Simulations consistent with known information regarding the cluster and the errors of measurement result in a bias of about 6.5%-8.5%. This corresponds to an unbiased distance of 17.2-17.4 Mpc and a value of H_0 in the range 80-82 km/s/Mpc. The problem of determining the bias to Virgo illustrates several key points regarding Malmquist bias. Essentially all conventional astronomical distance measurements are subject to this bias. In addition, the bias accumulates when an attempt is made to construct "distance ladders" from measurements which are individually biased. As will be shown in the case of Virgo, the magnitude and direction of the bias are sensitive to the spatial distribution of the parent poputation from which the observed object is drawn - a distribution which is often poorly known. This leads to uncertainty in the magnitude of the bias, and adds to the importance of minimizing the number of steps in "distance ladders".