If you want to know what the past of our universe looks like, you have to talk to Anne Hutter. The astrophysicist at the University of Vienna can look back millions of years thanks to the latest telescope technology and supercomputing. In this interview, she explains why astrophysics is “back at the starting line” with the James Webb Space Telescope and how she uses Austrian HPC systems to model the early galaxies.
Professor Hutter, you study the early universe, a topic that may feel rather distant to many people. What are some of the most interesting facts about the universe that are not widely known?
Anne Hutter: One important fact is that we can look back into the past at all, but that there is a final “wall” we cannot see beyond. It lies about 380 million years after the big bang. So we cannot see the big bang itself. As a result, many questions are very difficult to answer when it comes to what happened before or shortly after the big bang.
Another fascinating point is how much information we can extract from the light of galaxies. We can model when and how quickly a galaxy formed stars. In some cases we can even narrow down which types of stars formed and how many stars it contains. We can see whether, for example, the galaxy contains a lot of gas or a lot of dust, consisting of carbon rich compounds and silicates. These absorb UV radiation, and from this we can infer which substances a galaxy is made of. We can often read all this from its light.
Most people probably know that the universe is expanding. Do we know how big it is?
Anne Hutter: Unfortunately, we cannot see to the end of the universe, but we do have models for the expansion rate. On that basis we can calculate how large the observable universe is today. Its current diameter is about 28,300 megaparsecs (Mpc). A megaparsec is a unit of length commonly used in astronomy and corresponds to about 3.26 million light years.
For comparison: if we shrank the observable universe to the size of Austria, the Earth would be about as large as an atomic nucleus within a molecule sized solar system sitting in a Milky Way the size of a dinner plate. And that only covers the observable universe. We do not know how big the entire universe actually is.
"The new images from James Webb Telescope have pushed us to question all the standard assumptions. It feels a bit like being back at square one."
Anne Hutter
University of Vienna
You work with the James Webb Space Telescope (JWST) and have gained remarkable insights into the formation of early galaxies. What is special about this telescope?
Anne Hutter: It measures exactly the wavelength ranges we need to observe galaxies far back in time in an expanding universe. As the universe grows, the frequency of their light changes. You can picture it as waves painted onto a balloon that is not yet fully inflated. If you blow it up further, the waves also stretch; the troughs and peaks become longer. That is what happens to the light from these galaxies as the universe expands. The James Webb Telescope is one of the first to be sensitive in precisely this wavelength range.
With one of its predecessors, the Hubble Space Telescope, we could see back to about 500 million years after the big bang. With the James Webb Telescope, we can see galaxies only 200 to 300 million years after the big bang. It allows us to look much further into the past of galaxies and takes us right to the very first ones.
How does that change our view of the universe in concrete terms?
Anne Hutter: JWST has shaken up the previous galaxy models. Those models predicted fewer bright galaxies than JWST is now observing. That naturally sparked a debate about why this is the case. The new images have pushed us to question all the standard assumptions. I have the feeling that our field has become very creative since then. At the moment it is not at all clear which idea is correct. It feels a bit like being back at square one.
"It is precisely these bright galaxies, that give us crucial clues about how the first galaxies worked: which stars formed, how much mass and which chemical elements were present in these young systems."
Anne Hutter
University of Vienna
What's the story with those bright galaxies?
Anne Hutter: Bright galaxies emit a lot of UV radiation. They are exciting because JWST has observed more of them in the early universe than our previous models predicted, and because we can see them at all only because they are so bright. At the extreme distances where the James Webb Telescope operates, faint galaxies practically fall out of view. Only the particularly bright systems emit enough light for us to detect them.
It is precisely these bright galaxies, however, that give us crucial clues about how the first galaxies worked: which stars formed, how efficiently new stars formed from gas, how much mass and which chemical elements were present in these young systems. By studying the particularly bright, observable galaxies in detail and comparing them with predictions for the entire galaxy population, we learn something about the physical conditions in the young universe.
And your project “Early Universe Simulations”, which you run among other places at ASC, aims to shed light on this?
Anne Hutter: In my current project I am trying to solve the problem that we observe many more bright galaxies in the early universe than our existing models predict. I assume that the conditions in the early universe were different – the gas was denser and hotter – and that more massive stars therefore formed. I have built exactly this into my model, and it already explains the observed bright galaxies better.
In the next step, I add further ideas for how these bright galaxies could have formed and link the different physical processes instead of treating them in isolation. The goal is to develop a more self consistent model in which these processes interact in a realistic way.
Massive star clusters 460 million years after the big bang
In 2024, an international team of astronomers used the James Webb Space Telescope to discover gravitationally bound star clusters in the Cosmic Gems arc, a galaxy whose light left it when the universe was 460 million years old – a glimpse back across 97% of cosmic time. This is the first discovery of star clusters in an infant galaxy less than 500 million years after the big bang.
This image shows two panels. On the right is a field of many galaxies on the black background of space, known as the galaxy cluster SPT-CL J0615−5746. On the left is a callout image from a portion of this galaxy cluster showing two distinct lensed galaxies. © ESA/Webb, NASA & CSA, L. Bradley (STScI), A. Adamo (Stockholm University) and the Cosmic Spring collaboration
What are your findings so far?
Anne Hutter: So far, we have mainly found that we can explain the observations better if we assume more massive stars in the early universe than in the present day local universe, especially in massive galaxies. When I adjust the stellar mass distribution in my model accordingly – that is, when I produce more of these heavy stars – the number of bright galaxies seen by JWST matches the simulation much more closely.
Physically, the mechanism is quite simple: massive stars are, in terms of their total mass, much more luminous in the ultraviolet than many small stars. This changes the mass to light ratio, and the galaxies appear much brighter in UV light. This offers at least one possible explanation for the large number of very bright galaxies in the early universe – one of several ideas we are currently testing.
What impact does your research have on society?
Apart from the gain in knowledge, our work makes a major contribution to teaching. In essence, I train students in the modelling of complex systems and in how to translate these processes into computations. It is about questions such as: how do I model the system physically, and how do I implement that model? I believe these are skills that are needed in many other disciplines as well.
Our research also provides a certain push for industry to develop ever better detection systems. Our constant question is: how can we do this better? This has repeatedly had an impact on technology in society; I am thinking, for example, of GPS, which was originally developed, among other things, for space applications.
Webb’s record-breaker: the farthest galaxy ever detected
The galaxy designated MoM-z14 is currently the farthest galaxy ever detected. Through the Webb Telescope, we are seeing this galaxy as it appeared in the distant past, only 280 million years after the universe began in the big bang. Its light has traveled through space for more than 13 billion years to reach us.
This image shows a wide field of view showing deep space, dotted with many small galaxies. One is highlighted with a magnified image in a graphic pull-out box in the lower right corner. The galaxy is labeled MoM-z14 and appears as a blurry yellow blob with a small red area at its top. © NASA, ESA, CSA, STScI, R. Naidu (MIT), Image Processing: J. DePasquale (STScI)
What does your work on the high performance computing (HPC) systems look like in practice?
Anne Hutter: In principle, I mainly run my own models, which I have been developing continuously for almost ten years. I first think about which quantities I need to describe – for example, how much gas or how many stars are in the galaxy. I might want to know at what rate it is currently forming stars.
Then I consider which physical processes are necessary to convert one into the other, or to move something to a different region in space, and I describe these processes with the appropriate equations.
At some point I realise that, yes, I have now described this, and I could programme it that way, but it would take forever to run. So I think about which simplifying assumptions I can make. The model will not be perfectly accurate, but it will come reasonably close. That is the biggest challenge in model development: knowing which assumptions are good and which are not.
Once the model is set up to the point where I think “yes, this might work now”, I implement it on a local machine and spend some time debugging, because of course it does not all run smoothly at first. Most of the time you start with a toy model that runs on a single processor, and you first check whether it does what it is supposed to do.
The next step in my case was to parallelise the simulation code, because we want to simulate many galaxies, which produces enormous amounts of data that definitely do not fit on my laptop. I simulate a section of the universe with all its galaxies. This section has to be large enough to be statistically representative.
Once the model is parallelised, there is another round of debugging, because parallelisation introduces all sorts of new errors. When everything finally works, I run the model on the HPC system. For one simulation run I need about 700 gigabytes of memory (editor’s note: roughly the computing power of 40 to 80 standard laptops combined). That is why I work on a supercomputer, specifically on the Austrian cluster VSC-5.
In principle, the process of building the model never really ends. It is almost impossible to reach the point where you can say “now I have everything”, because new findings and questions are constantly being added, or the computers become faster so that you can describe the physical processes in the model in more detail.
One last question: you run your simulations on VSC-5 at Austrian Scientific Computing (ASC). If this supercomputer were a galaxy, what would it be called?
Anne Hutter: I would call it “42”. That is the answer to everything.
About Anne Hutter
Dr Anne Hutter has been an assistant professor at the Institute of Astrophysics at the University of Vienna since 2025. She obtained her PhD in 2015 at the Leibniz Institute for Astrophysics in Potsdam. Anne Hutter specialises in the formation and evolution of the first galaxies and their influence on the matter between galaxies. She studies their properties and evolution and their contribution to the reionisation of the early universe, and develops theoretical and semi numerical simulation models for this purpose.
At the University of Vienna she is investigating, among other things, why the James Webb Space Telescope observes more bright galaxies in the early universe than models had predicted before its first observations.