Probing clumpy star formation using Euclid observations and cosmic simulations
Lead Research Organisation:
The Open University
Abstract
The observable Universe is filled with trillions of galaxies with a bewildering array of different shapes, sizes and colours. Using observations and theoretical models, we can try to decipher and understand the intricate mechanisms by which these galaxies formed and this task is one of the primary challenges for modern astrophysics and cosmology.
This project focusses on a specific population of star-forming galaxies that have a visibly clumpy appearance. These "clumpy galaxies" get their appearance from dense knots of newly-forming stars which can drive stellar winds and - when the most massive stars die - supernova shocks out into their host galaxies and even intergalactic space. It is thought that most galaxies in our Universe (including our own Milky Way) pass through one or more phases of clumpy star formation during their evolution. The clump-driven winds and shocks deliver energy and momentum that has a powerful (and perhaps dominant) effect in building and shaping galaxies, so by studying galaxies that are clumpy today, we gain a critical understanding of how all galaxies today evolved to their present state.
Astrophysics is on the brink of a revolution. Next-generation survey telescopes like the Euclid space telescope, the Vera Rubin Observatory, and the Square kilometre array promise enormous, petabyte-scale datasets containing exquisite measurements of the Universe. The first of these instruments - Euclid - is already in operation. Its mission is to collect high-resolution images and spectra for 35% of the night sky. In the process, it will capture images of 1 billion galaxies, of which at least 250 million will be spatially resolved so we can examine their morphology (i.e. the details of their shapes) and substructure.
At the same time, modern cosmological simulations have recently achieved levels of sophistication where they can model enormous volumes of the Universe, containing thousands of simulated galaxies, while also attempting to accurately model the small-scale physical processes of star formation that drive and regulate those galaxies' growth. What is needed now is a way to check whether the star formation models used by these simulations produce simulated galaxies with properties that match those in the real Universe.
This project will provide that critically-needed validation by comparing the detailed substructure of real galaxies that are observed by Euclid with the detailed substructure of galaxies that is predicted by a next-generation cosmological simulation called COLIBRE. We will use a deep-learning model that we have developed to search for clumps in Euclid images of galaxies and in galaxies from the COLIBRE simulation and compare the detailed substructure of the real and simulated clumps. If we find observed clumps that have no counterparts in COLIBRE (or vice-versa) this tells us that the COLIBRE star formation models have shortcomings and we can refine them in future simulations.
On the other hand, if the observed clumps look very similar to simulated ones, then we can use the simulation to infer the physical processes that are happening in the real clump (because we know exactly what physics is being modelled in the simulation) and we will have gained new understanding of the effects of clumpy star formation on galaxies throughout the Universe.
Either way, the unprecedented dataset provided by Euclid and the ground-breaking sophistication of COLIBRE mean that this project is guaranteed to deliver a step-change in our understanding of galaxy formation.
This project focusses on a specific population of star-forming galaxies that have a visibly clumpy appearance. These "clumpy galaxies" get their appearance from dense knots of newly-forming stars which can drive stellar winds and - when the most massive stars die - supernova shocks out into their host galaxies and even intergalactic space. It is thought that most galaxies in our Universe (including our own Milky Way) pass through one or more phases of clumpy star formation during their evolution. The clump-driven winds and shocks deliver energy and momentum that has a powerful (and perhaps dominant) effect in building and shaping galaxies, so by studying galaxies that are clumpy today, we gain a critical understanding of how all galaxies today evolved to their present state.
Astrophysics is on the brink of a revolution. Next-generation survey telescopes like the Euclid space telescope, the Vera Rubin Observatory, and the Square kilometre array promise enormous, petabyte-scale datasets containing exquisite measurements of the Universe. The first of these instruments - Euclid - is already in operation. Its mission is to collect high-resolution images and spectra for 35% of the night sky. In the process, it will capture images of 1 billion galaxies, of which at least 250 million will be spatially resolved so we can examine their morphology (i.e. the details of their shapes) and substructure.
At the same time, modern cosmological simulations have recently achieved levels of sophistication where they can model enormous volumes of the Universe, containing thousands of simulated galaxies, while also attempting to accurately model the small-scale physical processes of star formation that drive and regulate those galaxies' growth. What is needed now is a way to check whether the star formation models used by these simulations produce simulated galaxies with properties that match those in the real Universe.
This project will provide that critically-needed validation by comparing the detailed substructure of real galaxies that are observed by Euclid with the detailed substructure of galaxies that is predicted by a next-generation cosmological simulation called COLIBRE. We will use a deep-learning model that we have developed to search for clumps in Euclid images of galaxies and in galaxies from the COLIBRE simulation and compare the detailed substructure of the real and simulated clumps. If we find observed clumps that have no counterparts in COLIBRE (or vice-versa) this tells us that the COLIBRE star formation models have shortcomings and we can refine them in future simulations.
On the other hand, if the observed clumps look very similar to simulated ones, then we can use the simulation to infer the physical processes that are happening in the real clump (because we know exactly what physics is being modelled in the simulation) and we will have gained new understanding of the effects of clumpy star formation on galaxies throughout the Universe.
Either way, the unprecedented dataset provided by Euclid and the ground-breaking sophistication of COLIBRE mean that this project is guaranteed to deliver a step-change in our understanding of galaxy formation.
People |
ORCID iD |
| Hugh Dickinson (Principal Investigator) |