loader image


Latest results

The metallicity gradients of star-forming regions store information of the assembly history of galaxies

ERIS adopts the hypothesis that chemical abundances store information on the main events that form galaxies. In this paper our master student, Francisco Jara,  together with members of our team, studied galaxies from the EAGLE project.  By studying the metal content and spatial patterns within these galaxies, scientists can uncover clues about their history and formation. Recent research has focused on comparing observations of real galaxies with simulations to better understand these processes.

One important discovery is the relationship between a galaxy’s mass and its metal content, known as the mass-metallicity relation (MZR). Researchers found that low-mass galaxies with strong negative metallicity gradients tend to be more enriched in metals compared to the MZR median. These galaxies also show high levels of star formation activity. However, there are discrepancies between observations and simulations, particularly in massive galaxies. Simulations suggest that massive galaxies should have smaller central black holes in galaxies with strong negative metallicity gradients, but this trend is not seen in observed data.nThese findings highlight the complexity of galaxy evolution and the need for further research to bridge the gap between observations and simulations. Understanding these processes can lead to a deeper understanding of the universe and its formation.

Relevant links:

Francisco’s paper: https://arxiv.org/pdf/2403.06836.pdf

Research Lines: L2

Understanding the mass-assembly histories of bulges and discs of spiral galaxies

One of the goals for ERIS in the near future is to connect  stellar evolution and phylogenies within the Milky Way with unresolved stellar populations observed in nearby galaxies, thus linking galactic and extragalactic astronomy and contributing to a deeper understanding of how galaxies form and evolve. One approach towards this goal is to study the stellar populations of different components within galaxies, such as the bulge and disc, to better understand how they assembled their mass.

In this paper we use a novel technique to apply bulge-disc decomposition, a procedure long applied to imaging data to model the light of different galaxy components, to IFU datacubes of galaxies from the SDSS MaNGA survey. This technique allowed us to cleanly separate the spectra of the bulges and discs in around 950 spiral galaxies in order to study their stellar populations with minimal contamination. This analysis gave us details of the mass assembly histories, indicating how much mass was created within that component at a series of lookback times throughout the lifetime of the galaxy. We find evidence of a clear downsizing effect in the bulges, such that more massive bulges formed earlier and more rapidly while lower mass bulges formed more recently over longer timescales. The discs show a similar but weaker trend, with longer formation timescales than the bulges. In general we find that the bulges are older than the discs, indicating that the bulges formed first in these galaxies over a short timescale while the discs formed later and in many cases are still assembling their masses through star formation and mergers. This work will help build on many other research projects within ERIS by providing a deeper understanding of a statistical sample of other spiral galaxies like the Milky Way.

Relevant links:

Keerthana’s paper: https://arxiv.org/abs/2402.00959

Research lines: L2, L4

On the evolutionary history of a simulated disc galaxy as seen by phylogenetic trees

 “The paper seeks to better understand the limits of a novel approach by using phylogenetic methods to dissect the evolutionary history of a simulated disc galaxy. The application of phylogenetics in this realm is promising, it represents an ambitious interdisciplinary journey that merges the concepts of phylogenetic methods from biology and the evolutionary history of galaxies” (taken from referee report) 

Our recent paper, which consists of a substantial part of our ERIS ambassador Danielle de Brito Silva’s PhD thesis, is the first study of phylogenetic trees in numerical simulations. On this occasion, we considered a simulation of a galaxy that evolved in isolation, because to know how to reconstruct the complex histories of galaxies with phylogenies, we first need to know how trees of simple histories look like. They look very imbalanced, with star particles ranked by their ages e.g. one lineage evolves over time. Their lengths, however, depend on the star formation efficiency. The more stars formed, the longer the trees. This work marks a highlight, as it represents a solid PhD chapter supervised by the ERIS directors and multiple discussions in person and online over the past two years with the entire collaboration. This paper is the seed for many following works for more masters and PhD theses. The natural next questions are how can we use the tree imbalance and length to study different star formation and merging histories. 

Relevant links: 

Danielle’s paper: https://arxiv.org/pdf/2310.12235.pdf

Instagram reel: https://www.instagram.com/nucleomilenioeris/reel/C1F3V3KO3F8/

Research lines: L2, L4

This paper started as a research project to be done for the Mathematical Modelling course in mathematical engineering at University of Chile. In 2021 Holly Jackson published our latest paper on solar twins and we were curious to try such analysis using state-of-the-art spectroscopic survey data. We chose the GALAH survey, because of the large number of solar twins with several chemical abundances measured. Our preliminary told us that the precision of the abundances were not enough to resolve the trees, so Kurt applied a machine learning algorithm based on The Cannon to improve the precision of the stellar abundances and generate a catalog of 40,000 solar twins with high precision abundances, the largest one to-date. Kurt has stayed in ERIS working on his master thesis in mathematics, building and analyzing phylogenetic trees from that catalog. As expected, the robustness of the trees improve with the Cannon catalog compared to the GALAH DR3 abundances. But the tree signal is not as strong as we wished, which could be due to the fact that many solar twins in the solar neighborhood are probably siblings, and other large numbers are migrants, making the tree signal quite noisy. 

Relevant links:

Kurt’s paper:  https://arxiv.org/pdf/2310.15107.pdf

Holly’s paper: https://academic.oup.com/mnras/article/502/1/32/6081048

Research lines: L1, L4

Exploring the dependence of chemical traits on metallicity: chemical trends for red giant stars with asteroseismic ages

This recent paper is the result of an ESO student’s internship which is a competitive opportunity for PhD students working in Chile. Sara worked during the first semester of 2023 at ESO on the spectral analysis of giant stars for which very high resolution UVES spectra and information from astroseismology is available. This allowed her to determine accurate ages and very high precision abundances and so study the evolution of chemical abundances with time and studied this dependency with metallicity. She found that there is in general a strong dependency with metallicity, which was attributed to a differential production rate of the various chemicals with metallicity, but also to different star formation histories across the Milky Way disk. This paper builds up from another paper published by our group which focused on the solar twin data used by Holly Jackson (paper above). 

Relevant links:

Sara’s paper: https://ui.adsabs.harvard.edu/abs/2024arXiv240102328V/abstract

Paula’s paper: https://ui.adsabs.harvard.edu/abs/2020A%26A…633L…9J/abstract

Research Lines: L1, L3

The study introduces the “Galactic Archaeology Neural Network” (GANN), a method based on neural network models (NNMs), designed to identify accreted stars in galactic discs using their chemical fingerprint and age. The network is trained on simulated galaxies from the Auriga Project, focusing on the target galaxy’s local environment defined by the stellar halo and surviving satellites. GANN effectively detects accreted stars spatially mixed into the disc, with performance measured by recovery fraction and the probability of true-positive results. Flexible threshold values based on the NNM output enable refined selection. The results show high accuracy in identifying accreted disc stars within simulated galaxies and successful recovery of stars in Gaia-Enceladus-Sausage mass systems. A conglomerated NNM, trained on halo and satellite stars from all Auriga galaxies, yields consistent results, suggesting a promising future approach as observational capabilities advance.

Relevant links:

Published paper: Machine learning for galactic archaeology: a chemistry-based neural network method for identification of accreted disc stars

Research lines: L2

The study investigates the influence of the environment on galaxies falling into and orbiting within the potential well of a Local Group  analogue. Utilizing high-cadence observations of eight disc satellite galaxies from the CIELO hydrodynamical simulations, the analysis includes measuring tidal torques, ram pressure, and specific star formation rates over time. These factors are correlated with the amount of gas lost by satellites along their orbits. Notably, stronger gas removal episodes occur when the disc plane is perpendicular to the direction of motion, and multiple peripassages are often required to significantly alter disc orientations. Gas removal during interactions with central galaxies may occur opposite to the direction of motion. Despite reaching their first peripassage, satellites are not completely quenched, continuing to form approximately 10 percent of the final stellar mass. The fraction of removed gas is influenced by the joint action of tidal torque and ram pressure, potentially triggering new star formation and subsequent supernova feedback.

Relevant links:

Published paper: https://ui.adsabs.harvard.edu/abs/2022MNRAS.514.6157R/abstract

Research lines: L2

The study delves into the chemodynamical properties of a selection of very metal-poor (VMP) stars from the Hamburg/ESO survey, cross-referenced with Gaia EDR3, in the phase space defined by three integrals of motion (Lz, E, I3). Employing criteria by Carollo & Chiba to distinguish disk and halo orbits, 26 stars with [Fe/H] ≤ -2.5 exhibiting disk kinematics are identified, with 13 being extremely metal-poor. Notably, at these low metallicities, the number of stars with disk kinematics surpasses its retrograde counterpart by threefold. Additionally, 37 halo stars, most tightly bound to the progenitor halo’s gravitational potential, are identified in the same metallicity range. The study investigates the origins of these stars by comparing observational results with simulated galaxies from the Aquarius Project and the IllustrisTNG simulations. Two mechanisms of formation for VMP stars with disk kinematics are identified: dominant accretion from early satellites and in situ formation. These stars are notably old, with ages exceeding 12.5 Gyr (z > 5), and display α-enrichment. Accretion and in situ formation are also observed for retrograde counterparts, with accretion prevailing. The contributing accreted satellites exhibit stellar masses in the range of 10^6-10^9 M ☉ and are notably gas-rich. The most bound halo stars detected are the oldest, with a median age of approximately 13.3 Gyr (z ~ 11), and also display α-enrichment. The findings underscore the importance of very old, VMP stars in preserving crucial information about the initial stages of our Galaxy’s assembly and its halo.

Relevant links:

Published paper: https://ui.adsabs.harvard.edu/abs/2023ApJ…946…99C/abstract

Research lines: L1, L2

The impact of binary stars on the dust and metal evolution of galaxies

This recent paper builds on the L-Galaxies 2020 semi-analytic simulation of galaxy evolution by including (a) binary stellar evolution (via the binary_c BSE code) and (b) dust production and destruction into the code. This allows us to study the effect of binary stellar evolution and dust production on the evolution of galaxies, and also includes all 118 chemical elements (487 isotopes) from 16 distinct enrichment channels. We  find that binary effects, such as common envelope ejection and novae, affect carbon and nitrogen enrichment in galaxies; however, heavier alpha elements are more affected by the choice of supernova and wind yields. On sub-galactic scales, there is very good agreement between this new L-Galaxies and observed dust and metal radial profiles in nearby systems.

Relevant links:

Rob Y’s paper: https://ui.adsabs.harvard.edu/abs/2024MNRAS.527.6292Y/abstract

Research Lines: L2, L3