Tjitske Starkenburg

Galaxy Evolution

The galaxies in our universe are the evidence we see of how our Universe and its components formed and evolved. At the same time galaxies are the centers of their halos' gravitational potential and host the detailed gas physics and chemistry in the interstellar matter and the formation and evolution of stars and black holes. Galaxy properties are governed by how, when and where they formed through the detailed physics taking place inside of them, and through interactions with their larger scale environment and neighboring systems.
I am fascinated by the combination of smaller and larger scale physics, and internal and external effects, that are all relevant for galaxy formation and evolution. I aim to dientangle the physical processes that shape the galaxies we see and which observational properties can serve as tracers of their evolution.

Galactic Dynamics

Stellar halos are records of galaxies’ pasts: both major and minor galaxy mergers can leave unique traces in the outer reaches of galaxies. Additionally, stellar halos host tidal features that are relics of accreted and disrupted lower-mass galaxies such as satellites, stellar streams and shells. The coming decade will open up this discovery space for galaxies in the Nearby Universe with the Euclid mission, the Nancy Grace Roman Space Telescope, and the Vera Rubin Observatory.
I explore what this new observational data can tell us about and hierarchical galaxy formation physics based on these upcoming facilities. With this work I bridge the gaps between galactic dynamics and extragalactic astronomy, and between observation and theory.

Synthetic Observations, Inference and Machine Learning

I am leading the Isolated and Quiescent galaxies (IQ) Collaboratory, a ~30 astrophysicists strong collaboration started during a workshop at the Center for Computational Astrophysics of the Flatiron Institute. The IQ-Collaboratory endeavors to learn about galaxy quenching through careful, in depth comparisons of large simulated and observational galaxy samples. We aim to get as close to a one-to-one comparison as possible to be able to connect the different quenching methods to observational galaxy properties. Using mock galaxy spectra from a very large set of simulated galaxies we compare observational star forming, quiescent, and green valley definitions to enable comparison between different observational datasets and empower the community. We employ data-driven analysis techniques to optimally explore the large and diverse datasets in our sample.
(Hahn, Starkenburg et al. 2019;
Dickey, Starkenburg et al. 2021;
Starkenburg et. al. in prep.;
(Hahn, Starkenburg et al. 2021;
Choi, Starkenburg et al. in prep.;
Maller, Starkenburg et al. in prep.)

Together with my co-coordinators Francisco Villaescusa Navarro, John Wu, and Peter Behroozi and scientific advisors Romeel Dave, Shirly Ho, and Josh Peek, I am organizing a 10-week collaborative program at the Kavli Institute for Theoretical Physics and the Center for Computational Astrophysics to advance the field of galaxy formation and evolution in the data science era. I initiated and lead the proposal for this program so it is great to see this program taking form! I think there is much more that can be done with data-driven approaches in the field of galaxy formation, especially considering the data deluge that we are facing. Developing new techniques to combine information from multi-wavelegth and multi-resolution observations (e.g. photometry, IFU observations, and spectroscopy) as well as improving techniques to compare observations and theory would greatly advance the field. However, connecting statistics and machine learning results to an improved understanding of key astrophysical processes remains a major challenge, and is a topic that I hope this program will provide novel insights in.
(KITP program - Winter 2023 and Associated Conference - March 2023)

Together with Martin Rey I looked at how small changes in a galaxy's merger history influenced the distribution of stars in the galaxy halo, the so-called stellar halo. We found that the effects from small changes in the merger history could have major and observable effects, and that by varying the timing or size of speicifc mergers could results in stellar halos that span the broad observed range of halo masses. To do this study we combined genetic modifications of the halo's Lagrangian region in the early universe with an empirical galaxy formation model and particle tagging. Because of this flexible setup we could also identify that the diversity of galaxy stellar halos is connected to galaxy formation at the low-mass end, which suggests an exiting new way to probe the unique galaxy formation physics at the lowest masses. Some of our simulated halos have now been followed-up with high-resolution hydrodynamic simulations as well, allowing us to study the influence of merger histories on a much wider range of properties of the stellar halos as well as the central galaxies themselves.
(Rey & Starkenburg 2021)

Stellar streams and shells from accreted and disrupted dwarf galaxies or globular clusters that are discovered in our own Milky Way have provided crucial insights in hierarchical galaxy formation. Moreover, in addition to surviving satellites, these features are the closest objects to provide constraints on the formation and evolution of galaxies at the low-mass end. With current and upcoming observational efforts we will be able to observe such features around external galaxies. For example, the Roman Telescope will be able to observe, and resolve into individual stars, the cold, thin, globular cluster stellar streams in the stellar halo of the Andromeda galaxy. Dwarf galaxy stellar streams may be observable around galaxies in the Local Universe. Most of the nearby galaxies are dwarf galaxies themselves, but we predict that the majority of LMC-sized galaxies do host stellar streams and shells, and that a fraction of those will be observable with upcoming surveys as the Roman Telescope and the Rubin Observatory.
(Pearson, Starkenburg et al. 2019;
Starkenburg, Pearson et al. in prep.; Pearson et al., incl. Starkenburg, 2021)

This Summer (2021) I was a lecturer and mentor at the Applied Dynamics Summer School at the CCA in New York. It was wonderful to be able to discuss galactic dynamics with so many great students and mentors, and especially to be able to do so in-person. Being on the SOC it was also very gratifying to see the program take off.
Students watched pre-recorded lectures and did workshops, but also worked on individual research projects for 5 weeks and presented these at the end. I had the honor to work with Sachi Weerasooriya and other mentors on the evolution of stellar streams during major mergers, and to be involved in other projects like that of Christian Aganze on the observability of gaps in streams in external galaxies with the upcoming Nancy Grace Roman Space Telescope. All projects I've been involved in at the school are continuing, so keep an eye out for results!
(Applied Galactic Dynamics Summer School 2021)

The large surveys utilizing Integral Field Units such as SAMI, MANGA, CALIFA, ATLAS3D and more, now allow for the study of spatially resolved structure and kinematics of a large sample of galaxies. Additionally, on the theory side galaxy simulations are now reaching resolutions and galaxy samples that make spatially resolved studies and comparisons possible. These detailed studies can provide the next stage of insights in all physical processes that govern galaxy evolution. A first step is to compare radial distributions. We show that while simulated star-forming galaxies approximately match those observed, this is not the case for green valley galaxies in simulations. On the other hand, observed galaxies that show centrally concentrated star formation, dubbed breakBRD galaxies, can also be found in simulations. These galaxies have reduced gas and star formation in their outskirts and are slowly quenching their star formation. Careful tracing these galaxies back in time indicates a likely cause in slow an subtle environmental effects instead of a single origin.
(Starkenburg et al. 2019;
Kopenhafer, Starkenburg et al. 2020)

Galaxies with uncommon properties are exceptionally interesting to study as these may provide key insights into the physics of galaxy evolution. One example of such are counterrotating galaxies. We find and analyze systems in the Illustris simulation where the stellar and gas disk rotate in the opposite sense, providing the first statistical prediction of the occurence and formation history of these galaxies. This work is particularly timely as observational IFU surveys such as CALIFA, MANGA, and SAMI measure stellar and gas kinematics for a large number of galaxies. We find that star-gas counterrotating low-mass galaxies have experienced significant episodes of gas removal in the past, either through strong feedback or gas stripping, and subsequently accreted gas with misaligned angular momentum. Therefore counterrotation may be more strongly correlated with the environment and properties of a galaxy at earlier times instead of those at the present-day.
(Starkenburg et al. 2019;
Duckworth, Starkenburg et al. 2020)
Additionally, I was involved in looking for retrograde moving structures within our own Milky Way. It was exciting to look for these effects in our own neighborhood, and with observational data on individual stars!
(Koppelman et al. incl. Starkenburg 2019)

The SMAUG consortium (Simulating Multiscale Astrophysics to Understand Galaxies) endeavors to improve our understanding of baryonic processes in galaxy formation by connecting large-scale cosmological galaxy simulations to high-resolution small-scale simulations describing the details of ISM structure, star formation feedback, and AGN accretion and feedback. I am a member of the consortium and a (co)-leader of the working group for synthetic observations.
(Pandya et al. incl. Starkenburg 2020)
My interest in connecting results from different scale simulations and in using high-resolution simulations to produce mock observations and provide insights into observational trends is also represented by two projects I advised on. One started at the Kavli Summer School at the CCA in 2017 and combined information from many simulations at different scales an resolutions. The other I got involved in when moving to Northwestern University and highlights to what degree observational star formation rate measurements at different wavelengths can constrain the timescales of star formation.
(Iyer et al. incl. Starkenburg 2020;
Flores Velazquez, Gurvich et al. incl. Starkenburg 2021)

Merging dwarf galaxies can show many of the features of their more massive counterparts. However, as the galaxy halo mass-stellar mass relation is predicted to be steep at the low-mass end apparantly minor mergers in stellar mass ratio can be much more significant in halo mass ratio and cause major merger-like effects such as creating dwarf collisional ring galaxies and significantly increasing central supermassive black hole accretion.
(Arabsalmani, Roychowdhury, Starkenburg et al. 2019)
During the Summer of 2017 I worked with the amazing Jonathan Mercedes Feliz (then undergraduate student at CUNY Lehman College, now a graduate student at UConn). He showed that low-mass black holes in dwarf galaxies will accrete more gas during dwarf galaxy mergers, leading to increased AGN brightness. Interacting dwarf galaxies may thus be ideal candidates to find the lowest mass super-massive black holes in galaxies!
(Jonathan has presented this work at multiple meetings, e.g.: Mercedes-Feliz, Starkenburg and Bellovary 2018)

Within the LCDM cosmological framework dark matter halo mass function is almost completely scale-free. The formation of galaxies however is not scale-free. There should therefore be a large number of low-mass halos that have not been able to form stars or a galaxy within them, and that are thus not observable in light. These so-called dark halos can however have a strong gravitational effect, for example on low-mass structures like dwarf galaxies. In this series of papers we study the interaction of dwarf galaxies with their dark satellites. Such an interection can produce effects similar to major mergers for larger galaxies, creating large tidal arms and tails, and central or offset starbursts and long-term increases in star formation in the dwarf galaxy. Studying candidates of these mergers in the universe may provide an additional route to understanding the properties of dark matter.
(Helmi, Sales, Starkenburg, E., Starkenburg et al. 2012;
Starkenburg et al. 2015; 2016a; 2016b)