Formation of the Circumgalactic Medium in Cosmological Zoom-in Simulations

Abstract

The gaseous halo around galaxies, known as the circumgalactic medium (CGM), has been inferred from observations and simulations to transition from cold chaos to hot stability (virialisation), but the road it takes is uncertain. One of the possible routes is via the ejections of energy and matter from stars, called stellar feedback, within the galaxy, providing the necessary thermal support to stabilise the CGM. In this thesis, the evolution of the CGM was analysed and compared between multiple state-of-the-art simulations provided by the AGORA Collaboration. These simulations varied primarily in their numerical implementation and stellar feedback model. By matching their gas properties, each simulation could be distinguished in their CGM by their unique set of assumptions, and an improved understanding of their respective models and implementations could be gleaned. Each code predicted a CGM baryon mass fraction relative to the cosmic baryon fraction in the range of ∼ 20 − 50%, implying all could regulate and maintain a significant fraction of baryons from their central galaxy in the CGM, albeit with different efficiency. The thermal phases of these baryons varied significantly, where certain codes with strong preventive feedback developed a hot gas dominated CGM early at z ≳ 4, before predicted accretion shocks, while some with instead strong ejective feedback predicted a much later transition at z ≲ 1.5. It was shown that including kinetic feedback, together with thermal, also greatly improved the ability to eject baryons, improving the regulation of the total baryon fraction. An odd commonality between most codes, independent of their numerical implementation or stellar feedback model, was a shared tendency towards halo wide virialisation, at similar times in the outer regions. This does not align with the current standard theories that predict a stronger radial dependence for the virialisation, either stabilising from the inside-out or from the outside in, and therefore has interesting implications for the evolution of the CGM.

Popular Abstract

If the Universe is the Earth, then galaxies are its biomes. Like the rainforests and the savannas, galaxies also have ecosystems, although not primarily made from organisms but rather a long series of complicated and extreme interactions covering almost every field of physics. A critical component of any ecosystem is the atmosphere and in the galactic context, the atmosphere is represented by a vast gaseous halo called the circumgalactic medium, henceforth the CGM. The CGM acts as the interface between the central galaxy, called the interstellar medium (ISM), and the outside environment, named the intergalactic medium (IGM). The primary purpose of the CGM is to facilitate and mediate the circulation of gas, called the baryon cycle, via the inflows from the IGM and outflows from the ISM. Gas is the primary fuel for galaxies because from it, given the right conditions, stars can be born. But its ability to allow gas to pass is determined by its stability. First it will be unstable, meaning the gas that flows through it can pass with little hindrance. In this scenario, the galaxy will consume as much of the surrounding gas as it can, but will also easily lose gas from e.g. its stars ejecting energy, matter, and momentum as they evolve via e.g. supernovae. You can think of this as a typical teenager who just got their first job, a lot of money coming in but also no sense of saving. This is a chaotic time and not sustainable, and, for galaxies, does not promote long lived structures, such as disks. If the CGM is instead heated enough, matured, it can provide a sufficient and stable thermal support, acting as a pressure, that will impede gas flows. In this state, it would be considered virialised, and gas can be stored inside the CGM for a longer time and instead slowly fed into the galaxy in an inner region of Semi-stability.

How the CGM reaches this hot, virialised state is currently up for debate. Some models predict that it naturally occurs when the galaxy grows massive enough for the infalling gas to be gravitationally accelerated to supersonic velocities. When this gas encounters the medium inside the CGM, like a jet plane through the sky, a sonic boom, or in this case just a shock wave will form. Along these waves’ journey, they transfer heat into the surrounding gas, which in turn supports them from behind with a thermal pressure, and as they travel throughout the CGM they provide the necessary temperature for stability. But recent observations reveal that this approach might be too slow. This work explored the alternative, or possibly complimentary, route where the stars from within the galaxy might be able to provide the needed thermal energy for the CGM to stabilise much earlier. As stars age, they eventually evolve depending on their mass, and the result of this evolution is often violent which ejects significant energy and gas into their surroundings, called stellar feedback, or in this work just feedback. But due to current limited observational capabilities, especially for galaxies like our own Milky Way, we must test our theories via simulations. In order to do this, we must ensure that the simulations we construct and the physics they include are robust, which is the aim of this project and the involved AGORA Collaboration. A lot of detail goes into these simulations, and due to the very fickle nature of simulations (think butterfly effect), we must first run them before we can see the effects of our chosen assumptions. By comparing the CGM evolution in simulated galaxies like our own Milky Way from using different implementations and different models for the stellar feedback, this work showed that we can reverse engineer galaxies like our own and learn more about which assumptions result in what. Specifically, by analysing and matching the distributions of gas properties in both space and time inside the CGM, this work managed to link the resulting evolution and formation of the galaxies to their respective assumptions. This is one of the many stepping stones on a long journey towards understanding how galaxies, like our own, form and evolve.

• author: Lindh, Isak
• supervisor: Oscar Agertz; Santi Roca Fàbrega
• organization: Astrophysics; Department of Physics
• course: ASTM32 20261
• year: 2026
• type: H2 - Master's Degree (Two Years)
• subject: Physics and Astronomy
• keywords: analytical, numerical, galaxy simulations, galaxy evolution, galaxy formation, galaxy haloes
• report number: EXA264
• language: English
• additional info: This work is a paper thesis and the authors of the paper within were not only the author of the thesis itself and the two listed supervisors, but also Víctor Rufo-Pastor, a PhD student in the Astronomy Divison. For an overview of the contribution from each author, please see the final section of the thesis.
• id: 9246690
• date added to LUP: 2026-08-10 12:29:13
• date last changed: 2026-08-10 12:29:13