Title: The history of the simulated and observed star formation in galaxies for the last 13 Billion years of the Universe
Speaker: Antonios Katsianis
Date: November 18 at 14:30
Location: R124, Physics Building
In order to obtain a deeper understanding of our Cosmos we gather observations of galaxies via our telescopes and construct theoretical models to reproduce these observations. A particular focus is given on the rate at which stars form within galaxies (labelled as the SFR). The “observed” star formation rate-stellar mass (SFR-M*) relation and the "observed" cosmic star formation rate density (CSFRD) represent both important canvases for our current knowledge of galaxy formation and are both routinely used to constrain cosmological models/simulations. I have been studying galaxy SFRs and galaxy stellar masses from 2015 (Katsianis et al. 2015, 2016, 2017, 2017b) to today (Katsianis et al. 2019, 2020, 2021, 2021, 2022). This investigation had some exciting and unexpected results. In this seminar I present the history of the SFRs within galaxies for the last 13 billion years. I start by presenting some severe limitations for both the simulated galaxy SFRs obtained from state-of-the-art cosmological simulations (Katsianis et al. 2021a, 2021b, EAGLE, IllustrisTNG, Simba, Semi-analytic models) and observations (Katsianis et al 2020, 2021b, UV, IR, radio, SED, Ha, OII indicators). I employ cosmological simulations combined with radiative transfer and demonstrate that the adopted methodology / indicator (e.g. IR, UV, Ha) to obtain the observed galaxy SFRs is bound to heavily affect the final result. I demonstrate that different authors who used different methods, assumptions and indicators to obtain their SFRs indeed reported results in tension by a factor of 4 for the period of 2007 to 2022. These interesting limitations / systematics between different methodologies were discussed extensively at the European Astronomical Society meeting (EAS) 2021 with numerous Polls pointing out a turmoil of opinions in the Field and conflicting qualitative results between different groups. In this seminar I will present these Polls. In addition, state-of-the-art simulations (EAGLE, TNG) are found to suffer from troubling limitations (Zhao et al. 2020, Katsianis et al. 2021) mostly connected to resolution effects and the adopted feedback prescriptions which have to be reconsidered. After demonstrating these limitations for both observations and simulations, I keep trying to find a way to address them. I demonstrate that the observed star formation rate density can be described by only two parameters (against 4, Madau et al. 2014) and a function that resembles a Γ distribution, like numerous other physical processes in Nature (from Economy to Biology) while it has a plateau from z ~ 1-4 and not a peak at z ~ 2. I demonstrate that a simple formalism relying on dark matter halo growth can emerge using this finding. The later can help significantly our efforts with cosmological simulations and Semi-analytic models which currently seem to lose legitimacy according to the Polls. Last, using up to date data and insights for different SFR indicators I present evidence that the observed CSFRD is dictated by episodic spikes/bursts and does not have a smooth evolution as previously thought.