THE QUANTITY OF INTRACLUSTER LIGHT: COMPARING THEORETICAL AND OBSERVATIONAL MEASUREMENT TECHNIQUES USING SIMULATED CLUSTERS

Using a suite of N-body simulations of galaxy clusters specifically tailored to studying the intracluster light (ICL) component, we measure the quantity of ICL using a number of different methods previously employed in the literature for both observational and simulation data sets. By measuring the ICL of the clusters using multiple techniques, we are able to identify systematic differences in how each detection method identifies the ICL. We find that techniques which define the ICL solely based on the current position of the cluster luminosity, such as a surface brightness or local density threshold, tend to find less ICL than methods utilizing time or velocity information, including stellar particles' density history or binding energy. The range of ICL fractions (the fraction of the clusters' total luminosity found in the ICL component) we measure at z = 0 across all our clusters using any definition spans the range 9%-36%, and even within a single cluster different methods can change the measured ICL fraction by up to a factor of two. Separating the cluster's central galaxy from the surrounding ICL component is a challenge for all ICL techniques, and because the ICL is centrally concentrated within the cluster, the differences in the measured ICL quantity between techniques are largely a consequence of this central galaxy/ICL separation. We thoroughly explore the free parameters involved with each measurement method, and find that adjusting these parameters can change the measured ICL fraction by up to a factor of two. The choice of ICL definition does not strongly affect the ICL's ability to trace the major features of the cluster's dynamical evolution. While for all definitions the quantity of ICL tends to increase with time, the ICL fraction does not grow at a uniform rate, nor even monotonically under some definitions. Thus, the ICL can be used as a rough indicator of dynamical age, where more dynamically advanced clusters will on average have higher ICL fractions.