Nathan Adams
30-40 years ago, galaxies of such a distance that their light has taken over 12 billion years to reach us were the record holders for the most distant astronomical objects known. These objects presented Astronomers with an opportunity to study galaxies as they were when the universe was 10% its current age. Today, modern telescope surveys can cover larger areas of the sky to ever greater depths, pushing the record further to study objects that existed when the universe was a mere 3.5% its current age. In the process, these new datasets now unveil tens of thousands of galaxies around the epoch where the universe was 10% its current age. This has allowed for studies of these galaxies to expand from analysing individual cases to exploring the population of galaxies as a whole in the early Universe. I will present the first measurement of the Luminosity Function (galaxy number density vs intrinsic luminosity) of these galaxies that is capable of probing 5 orders of magnitude in number density (a factor 100,000). Using both optical and near-infrared data from three telescopes allows us to solve issues experienced with selecting reliable samples of objects in the early Universe, revealing there to be more galaxies with ‘faint’ active galactic nuclei (AGN) at this time than found with conventional selection methods which used galaxy colours and morphology.
The first page displays two pictures of galaxies taken by the Hubble Space Telescope. A traditional spiral galaxy (like our Milky Way) and the 2013 record holder for the most distant known galaxy (which looks like a tiny red spec). We are trying to answer the questions regarding how the population of galaxies has changed over time, from these tiny, messy red objects to the more massive galaxies that we see today.
To conduct research on the matter, we need to be able to robustly select a large sample of galaxies in the early Universe. Thankfully, there are a couple physical concepts to assist us in searching for these kinds of galaxies.
The first concept is known as the “Lyman Break”. It is a consequence of the Universe being mostly made of Hydrogen. Light with a wavelength of 91.2 nanometres or shorter is capable of ionising Hydrogen. This is where the electron of the atom absorbs the photon of light and the resultant energy gain is enough to knock it out of the atom. This results in a large absorption feature in a galaxy spectrum that can be detected by telescopes.
The second concept is that of “redshift” and is a consequence of us living in a Universe that is expanding. In such a Universe, objects that are more distant from you move increasingly faster away from you. These speeds are so high that the Doppler Effect can take over. This causes the Lyman Break to shift to longer/redder wavelengths the faster a galaxy moves away from us. We can thus use the measurement of the wavelength of the Lyman Break as a proxy measurement for the speed of the galaxy and, through our knowledge of Cosmology, convert this speed into a distance measurement.
We observe an area of sky with an area of 30 full moons (6 square degrees) using 3 telescopes. This provides us with measurements of 1.8 million objects in 10 different coloured filters. We use these to search for galaxies with Lyman Breaks around ‘blue/green’ wavelengths of 400-500nm. Converting this into a distance reveals these galaxies to have existed when the Universe was 10% of its current age. In our data we find 47,700 galaxies located at this epoch.
Using our knowledge of how bright these objects are and how distant they are, we can calculate the total power output of the galaxy at ultraviolet wavelengths, just before the Lyman Break takes effect. We then measure how common galaxies are of various power outputs (a Luminosity Function). Our dataset is the first one to be capable of probing a large dynamical range of the Luminosity Function (factors of 100,000 in number density). We find some very rare, ultra-luminous objects known as Active Galactic Nuclei (AGN) and a large number of more typical galaxies. Fitting models to the Luminosity Function shows a sharp transition between where AGN dominate number counts to when normal galaxies begin to dominate. We also predict there to be more “faint” AGN than previously thought at this time.