Periodic Reporting for period 1 - DarkGalaxies (Dark galaxies: Probing the ΛCDM cosmological model on unprecedented scales)
Okres sprawozdawczy: 2022-09-01 do 2024-08-31
A fundamental challenge in modern physics is identifying the nature of dark matter (DM). Nearly a century after this particle's existence was first inferred from its gravitational influence on large cosmic structures, it has yet to be directly detected on Earth.
The leading model for interpreting observations is Λ-Cold Dark Matter (ΛCDM), which has been instrumental for over 40 years in explaining how structures and galaxies form. In this model, "dark energy" (Λ) drives the universe's accelerated expansion, while "cold" dark matter (CDM) governs the gravitational collapse of structures. This model is testable, as it predicts the distribution, structure, and abundance of collapsed halos, the sites where galaxies form. However, these predictions are largely influenced by the complex physics governing galaxy formation and evolution. These issues pose a significant challenge and hamper the steady progress that has defined cosmology over the past decades.
While the comparison between simulation results and observations continues to face scrutiny, it is clear that simulations allow a high degree of freedom to "accommodate" their outcomes to observations. Given these issues, a natural question arises: can competing DM models be tested, constrained, or ruled out through comparisons between observations and simulations of galaxies? The flexibility in galaxy formation models makes it difficult to argue in favour of this.
Our project explored observational probes at scales where galaxies do not form. On these scales, DM-dominated halos’ properties are robust. The existence of these systems is well justified. Observationally, reconciling ΛCDM with the abundance of galaxies requires galaxies to form predominantly in halos above a characteristic mass of about 5 billion solar masses. Theoretically, efficient gas cooling and cosmic reionization dictate a similar mass scale value. ΛCDM predicts myriads of halos below this mass, which, while devoid of stars, should contain neutral hydrogen. The gas in these systems should be in hydrostatic equilibrium with their underlying DM halo and in thermal equilibrium with the external ultraviolet background radiation field. By studying the gas distribution of these systems one can probe the clustering properties of DM, and thus constrain its nature. We shall refer to these systems as REionization-Limited HI Clouds (RELHICs).
The leading model for interpreting observations is Λ-Cold Dark Matter (ΛCDM), which has been instrumental for over 40 years in explaining how structures and galaxies form. In this model, "dark energy" (Λ) drives the universe's accelerated expansion, while "cold" dark matter (CDM) governs the gravitational collapse of structures. This model is testable, as it predicts the distribution, structure, and abundance of collapsed halos, the sites where galaxies form. However, these predictions are largely influenced by the complex physics governing galaxy formation and evolution. These issues pose a significant challenge and hamper the steady progress that has defined cosmology over the past decades.
While the comparison between simulation results and observations continues to face scrutiny, it is clear that simulations allow a high degree of freedom to "accommodate" their outcomes to observations. Given these issues, a natural question arises: can competing DM models be tested, constrained, or ruled out through comparisons between observations and simulations of galaxies? The flexibility in galaxy formation models makes it difficult to argue in favour of this.
Our project explored observational probes at scales where galaxies do not form. On these scales, DM-dominated halos’ properties are robust. The existence of these systems is well justified. Observationally, reconciling ΛCDM with the abundance of galaxies requires galaxies to form predominantly in halos above a characteristic mass of about 5 billion solar masses. Theoretically, efficient gas cooling and cosmic reionization dictate a similar mass scale value. ΛCDM predicts myriads of halos below this mass, which, while devoid of stars, should contain neutral hydrogen. The gas in these systems should be in hydrostatic equilibrium with their underlying DM halo and in thermal equilibrium with the external ultraviolet background radiation field. By studying the gas distribution of these systems one can probe the clustering properties of DM, and thus constrain its nature. We shall refer to these systems as REionization-Limited HI Clouds (RELHICs).
We made progress towards achieving the ultimate goal of our project: providing a novel and independent test of the nature of the DM. We successfully identified the first RELHIC candidate, Cloud-9, using radio observations with the Five Hundred-Meter Spherical Telescope (FAST), followed by Very Large Array (VLA) observations. Further Hubble Space Telescope (HST) optical observations will confirm or rule out the RELHIC hypothesis.
Cloud-9’s properties broadly match those expected for RELHICs. Its morphology, velocity, and emission pattern suggest it is a RELHIC requiring a significant amount of DM, with an estimated total and neutral hydrogen mass of ∼ 5 billion and 6 hundred thousand solar masses, respectively.
To achieve this milestone, we tackled the problem from three fronts: (1) We developed cosmological numerical simulations at unprecedented high resolution. These allowed us to understand the clustering of RELHICs for the first time, and understand that RELHICs should be homogeneously distributed in the sky, with a slight enhancement towards bright galaxies. We also developed simulations with alternative DM models, specifically warm DM and self-interacting DM. These simulations are essential for examining RELHIC properties under varying assumptions about the nature of the DM particle. (2) We created a novel GPU-based radiative transfer code achieving a 100x speed-up over similar CPU counterparts, enabling efficient modelling of ionization effects on RELHICs. (3) We conducted targeted observations of Cloud-9 to elucidate its nature.
To ensure the robustness of results, we tackled the theoretical aspects from two fronts: (1) We contributed to developing the novel COLIBRE galaxy formation model. We ran simulations to study the population of RELHICs emerging in this model and compared our results with results arising from previous models. (2) We employed a moments method scheme designed to solve the radiation transport equation coupled with the novel SWIFT code to study the formation of RELHICs at early times when the first ionization fronts travelled through the Universe.
Our radiative transfer tools, RELHIC analysis pipeline, and high-resolution numerical simulations will likely have a lasting impact in the field and help carry out research complementing ours. The work done during this project was published in leading peer-reviewed journals with open-access policy, and are all publicly available in ArXiv (https://arxiv.org(odnośnik otworzy się w nowym oknie)).
Cloud-9’s properties broadly match those expected for RELHICs. Its morphology, velocity, and emission pattern suggest it is a RELHIC requiring a significant amount of DM, with an estimated total and neutral hydrogen mass of ∼ 5 billion and 6 hundred thousand solar masses, respectively.
To achieve this milestone, we tackled the problem from three fronts: (1) We developed cosmological numerical simulations at unprecedented high resolution. These allowed us to understand the clustering of RELHICs for the first time, and understand that RELHICs should be homogeneously distributed in the sky, with a slight enhancement towards bright galaxies. We also developed simulations with alternative DM models, specifically warm DM and self-interacting DM. These simulations are essential for examining RELHIC properties under varying assumptions about the nature of the DM particle. (2) We created a novel GPU-based radiative transfer code achieving a 100x speed-up over similar CPU counterparts, enabling efficient modelling of ionization effects on RELHICs. (3) We conducted targeted observations of Cloud-9 to elucidate its nature.
To ensure the robustness of results, we tackled the theoretical aspects from two fronts: (1) We contributed to developing the novel COLIBRE galaxy formation model. We ran simulations to study the population of RELHICs emerging in this model and compared our results with results arising from previous models. (2) We employed a moments method scheme designed to solve the radiation transport equation coupled with the novel SWIFT code to study the formation of RELHICs at early times when the first ionization fronts travelled through the Universe.
Our radiative transfer tools, RELHIC analysis pipeline, and high-resolution numerical simulations will likely have a lasting impact in the field and help carry out research complementing ours. The work done during this project was published in leading peer-reviewed journals with open-access policy, and are all publicly available in ArXiv (https://arxiv.org(odnośnik otworzy się w nowym oknie)).
We have been able to pursue a high-risk/high-reward research idea: the existence of RELHICs. While the model's prediction is clear, the low abundance and faint emission of these systems makes it hard for existing radio facilities to detect them.
To maximise the probability of identifying RELHICs–and to characterise them among the sea of galaxies that populate the Universe—we developed advanced simulations and numerical tools. We succeeded in identifying and characterising the most compelling RELHIC candidate to date, detected using the largest radio telescope on Earth. This achievement surpasses our initial expectations and underscores the groundbreaking impact of our work.
In parallel, the development of a GPU-based radiative transfer code addresses key challenges in our research and extends its applicability to future questions beyond the original scope of the project. To materialise this code, we also developed algorithms that enabled us to create, as a by-product, software for real-time visualisation of large datasets using Virtual Reality headsets. These tools were showcased in science outreach events, effectively communicating complex cosmological simulations to a wide audience ranging from 6-year-old kids to adults. Also, our solutions achieved a 2 to 10x increase in energy efficiency over traditional CPU computations, advancing computational astrophysics and the EU's sustainability goals.
The tools and methodologies developed during this project will have a lasting impact on the astronomical community. Before this project, the existence of RELHICs was well-supported, but their low abundance made their detection elusive with existing radio facilities. By leveraging FAST's capabilities and our advanced simulations, we overcame these limitations. Our transition from optical to radio wavelengths opens up a transformative avenue to constrain the properties of DM particles on the smallest cosmological scales.
Our work aligns with the capabilities of next-generation mega-projects, such as the Square Kilometre Array (SKA). SKA's unprecedented resolution and sensitivity will unravel additional RELHIC candidates akin to Cloud-9. Our simulations and tools will undoubtedly shape and guide future explorations in this field.
To maximise the probability of identifying RELHICs–and to characterise them among the sea of galaxies that populate the Universe—we developed advanced simulations and numerical tools. We succeeded in identifying and characterising the most compelling RELHIC candidate to date, detected using the largest radio telescope on Earth. This achievement surpasses our initial expectations and underscores the groundbreaking impact of our work.
In parallel, the development of a GPU-based radiative transfer code addresses key challenges in our research and extends its applicability to future questions beyond the original scope of the project. To materialise this code, we also developed algorithms that enabled us to create, as a by-product, software for real-time visualisation of large datasets using Virtual Reality headsets. These tools were showcased in science outreach events, effectively communicating complex cosmological simulations to a wide audience ranging from 6-year-old kids to adults. Also, our solutions achieved a 2 to 10x increase in energy efficiency over traditional CPU computations, advancing computational astrophysics and the EU's sustainability goals.
The tools and methodologies developed during this project will have a lasting impact on the astronomical community. Before this project, the existence of RELHICs was well-supported, but their low abundance made their detection elusive with existing radio facilities. By leveraging FAST's capabilities and our advanced simulations, we overcame these limitations. Our transition from optical to radio wavelengths opens up a transformative avenue to constrain the properties of DM particles on the smallest cosmological scales.
Our work aligns with the capabilities of next-generation mega-projects, such as the Square Kilometre Array (SKA). SKA's unprecedented resolution and sensitivity will unravel additional RELHIC candidates akin to Cloud-9. Our simulations and tools will undoubtedly shape and guide future explorations in this field.