A multi-coloured survey of NGC 253 with XMM-Newton: testing the methods used for creating luminosity functions from low-count data

TL;DR

This study evaluates various methods for deriving X-ray luminosity functions from low-count data in the galaxy NGC 253, highlighting systematic differences and emphasizing the importance of measuring absorption individually for accurate results.

Contribution

It compares multiple methods for estimating X-ray source luminosities and demonstrates the impact of absorption assumptions on luminosity function accuracy.

Findings

Luminosity estimates vary up to a factor of three between methods.

Assuming Galactic absorption can lead to inaccuracies; measuring absorption is crucial.

Applying a consistent method across galaxy samples could strengthen correlations with star formation rates.

Abstract

NGC 253 is a local, star-bursting spiral galaxy with strong X-ray emission from hot gas, as well as many point sources. We have conducted a spectral survey of the X-ray population of NGC 253 using a deep XMM-Newton observation.NGC 253 only accounts for ~20% of the XMM-Newton EPIC field of view, allowing us to identify ~100 X-ray sources that are unlikely to be associated with NGC\thinspace 253. Hence we were able to make a direct estimate of contamination from e.g. foreground stars and background galaxies. X-ray luminosity functions (XLFs) of galaxy populations are often used to characterise their properties. There are several methods for estimating the luminosities of X-ray sources with few photons. We have obtained spectral fits for the brightest 140 sources in the 2003 XMM-Newton observation of NGC 253, and compare the best fit luminosities of those 69 non-nuclear sources associated with NGC 253 with luminosities derived using other methods. We find the luminosities obtained from these various methods to vary systematically by a factor of up to three for the same data; this is largely due to differences in absorption. We therefore conclude that assuming Galactic absorption is probably unwise; rather, one should measure the absorption for the population. A remarkable correlation has been reported between the XLFs of galaxies and their star formation rates. However, the XLFs used in that study were obtained using several different methods. If the sample galaxies were revisited and a single method were applied, then this correlation may become stronger still.