The Degeneracy of Galaxy Formation Models

TL;DR

This paper introduces a flexible galaxy formation modeling formalism that captures degeneracies among different scenarios and identifies observational data, like star-formation rate distributions, to constrain models effectively.

Contribution

A new formalism for galaxy formation modeling that uses efficiency parameters dependent only on halo mass and redshift, allowing easy exploration of diverse scenarios.

Findings

Models can reproduce observed galaxy populations with different efficiencies.

Full star-formation rate distributions help break model degeneracies.

The formalism closely matches existing semi-analytical models.

Abstract

We develop a new formalism for modeling the formation and evolution of galaxies within a hierarchical universe. Similarly to standard semi-analytical models we trace galaxies inside dark-matter merger-trees. The formalism includes treatment of feedback, star-formation, cooling, smooth accretion, gas stripping in satellite galaxies, and merger-induced star bursts. However, unlike in other models, each process is assumed to have an efficiency which depends only on the host halo mass and redshift. This allows us to describe the various components of the model in a simple and transparent way. By allowing the efficiencies to have any value for a given halo mass and redshift, we can easily encompass a large range of scenarios. To demonstrate this point, we examine several different galaxy formation models, which are all consistent with the observational data. Each model is characterized by a different unique feature: cold accretion in low mass haloes, zero feedback, stars formed only in merger-induced bursts, and shutdown of star-formation after mergers. Using these models we are able to examine the degeneracy inherent in galaxy formation models, and look for observational data that will help to break this degeneracy. We show that the full distribution of star-formation rates in a given stellar mass bin is promising in constraining the models. We compare our approach in detail to the semi-analytical model of De Lucia & Blaizot. It is shown that our formalism is able to produce a very similar population of galaxies once the same median efficiencies per halo mass and redshift are being used. We provide a public version of the model galaxies on our web-page, along with a tool for running models with user-defined parameters. Our model is able to provide results for a 62.5 h^{-1} Mpc box within just a few seconds.