A study of galaxy evolution: stacking emission lines from distant galaxies

Licentiate thesis, 2019

To draw up a thorough description of galaxy evolution exhaustive observations are needed, of distant but mainly of faint galaxies. Describing low mass galaxies is important to move the focus from the tip of the iceberg. While advances in telescopes capabilities have allowed to reach further galaxy, high-redshift studies are especially biased toward the brightest galaxies. In addition, well resolved spectral observations are needed to fully characterise the studied galaxies. Permitting the detection of emission or absorption lines, giving a direct access to galaxy composition. But spectral observations imply reducing further the signal to noise ratio (SNR), making line observations in faint high-redshift galaxies a difficult task.
Line stacking allows to get around intrinsic limitations in spectral observations, by averaging large galaxy samples, leading to drastic SNR improvement. However stacking, and especially line stacking, is not straightforward and requires a good handle of the population distribution, along with the output stack. To facilitate the use of such method, and to make data analysis tightly linked to stacking accessible, we produced an open source/open access tool, Line-Stacker.
After introducing the current state of knowledge on galaxy evolution, I will present Line-Stacker, describing both its main algorithm as well as the embedded data analysis tools. I will then introduce the two included papers and present their results. Paper I consists in a thorough description of Line-Stacker, and its test on numerous different simulated data-sets. Paper II is the first application of Line-Stacker on real data, in which we studied outflows from z ∼ 6 quasars.