Erratum: Finding a faint polarized signal in wide-band radio data

We present a new algorithm for fitting and classifying polarized radiosources, which is based on the QU fitting method introduced by O'Sullivan etal. and on our analysis of pulsars. Then we test this algorithm using MonteCarlo simulations of observations in the 16 cm band of the Australia TelescopeCompact Array (1.3-3.1 GHz), to quantify how often the algorithm identifies thecorrect source model, how certain it is of this identification, and how theparameters of the injected and fitted models compare. In our analysis weconsider the Akaike and Bayesian Information Criteria, and model averaging. Forthe observing setup we simulated, the Bayesian Information Criterion, withoutmodel averaging, is the best way for identifying the correct model and forestimating its parameters. Sources can only be identified correctly if theirparameters lie inside a 'Goldilocks region': strong depolarization makes itimpossible to detect sources that emit over a wide range in RM, whereas sourcesthat emit over a narrow range in RM cannot be told apart from simpler sourcesor sources that emit at only one RM. We identify when emission at similar RMsis 'resolved', and quantify this in a way similar to the Rayleigh criterion inoptics. Also, we identify pitfalls in RM synthesis that are avoided by QUfitting. Finally, we show how channel weights can be tweaked to produceapodized RM spectra, that observing time requirements in RM synthesis and QUfitting are the same, and we analyse when to stop RMClean.Comment: 14 pages, 10 figures, 1 table. Accepted for publication in MNRA