Nearby galaxies: modelling star formation histories and contamination by unresolved background galaxies
Abstract
Galaxies are complex systems of stars, gas, dust, and dark matter which evolve over billions of years, and one of the main goals of astrophysics is to understand how these complex systems form and change. Measuring the star formation history of nearby galaxies, in which thousands of stars can be resolved individually, has provided us with a clear picture of their evolutionary history and the evolution of galaxies in general.
In this work, we have developed the first public Python package, SFHPy, to measure star formation histories of nearby galaxies using their colour-magnitude diagrams. In this algorithm, an observed colour-magnitude diagram is modelled as a linear combination of many simple stellar populations with different ages and metallicities. This package treats metallicity as a free parameter, and the uncertainties are estimated by bootstrapping. This algorithm was tested on two different simulated populations and successfully recovered the input parameters. We have also measured the star formation history of the galaxy IC 1613 and found that the measured star formation history agrees with previous measurements for this galaxy.
Observing fainter phases of stellar evolution plays an essential role in the accuracy and precision of star formation history measurements for nearby galaxies. However, fainter sources are more contaminated by unresolved background galaxies. In the second project, we investigated the contamination effect of background galaxies in star formation history measurements by simulating stellar populations of nearby galaxies and a population of unresolved background galaxies. We found that deeper photometry helps to reduce the contamination only for distance moduli larger than (m-M)=23 or 0.4 Mpc. Most of the contamination effect comes from galaxies less than 2 magnitudes brighter than the photometry limit, and the contamination affects the older stellar populations more than younger populations. We also showed that including a model of background galaxies in the fitting process can result in a more accurate and precise measured star formation history compared to removing the contaminated part of the colour-magnitude diagram.