Supermassive black holes are, as you might expect, rather large – millions and sometimes billions of times as massive as the sun. They lurk at the centre of all large galaxies, including our Milky Way, shaping the growth of these cosmic structures. And yet we can say precious little for certain about how they form and why they grow so big.
These mysteries have come into sharper focus in recent years thanks to the James Webb Space Telescope (JWST), which has peered back in deep time to spot a surprising abundance of supermassive black holes in the early universe. Intriguingly, it seems that just a few hundred million years after the big bang brought our universe into being, the cosmos already contained black holes that were far too hefty to make sense under our current models of how the cosmos evolved. There simply hadn’t been enough time for anything that enormous to form.
Advertisement
, an astrophysicist at the University of Cambridge, is among those trying to solve this conundrum. She uses supercomputer simulations to model galaxies and supermassive black holes in the early universe, testing ideas about their origins and growth and even predicting what we should be looking for in future observations.
Koudmani spoke to Â鶹´«Ă˝ about why supermassive black holes are so fascinating, the joy of discovering surprises in the early universe that throw up new questions, and how ambitious computer simulations can help us finally make sense of them.
Daniel Cossins: What is so interesting about supermassive black holes?
Sophie Koudmani: What I really like about astrophysics is that it brings together all the different disciplines of physics but under extreme conditions, and that is especially true of supermassive black holes. We’re talking about extremely dense objects here, where a huge amount of mass – at least 100,000 times the mass of the sun – is concentrated in a relatively small space at the centre of every large galaxy. So there’s an incredibly rich set of physics associated with supermassive black holes, which are really the engines of galaxies.
How exactly do supermassive black holes shape galaxies?
As gas spirals towards a supermassive black hole, it forms a flattened, disc-like structure called an accretion disc. The immense gravitational forces and friction cause the gas in the disc to heat up to millions of degrees, making it glow brightly – imagine the intense orange-white glow of molten metal, but far hotter and more luminous. But this energy doesn’t just stay near the black hole. It is propelled outward in powerful jets and winds, travelling to the galaxy’s outer edges.
Normally, stars form when galactic gas clumps together and collapses into dense spheres. But the fierce radiation and outflows from the supermassive black hole blow apart these gas clouds, preventing them from collapsing into stars. It is this interplay between the supermassive black hole and its host galaxy that drives the evolution of galaxies, shaping their gas structure and stellar populations. And this is something we see in simulations. The only way to get galaxies in the computer that look like the ones we can see is to include a supermassive black hole.
What are the big mysteries about these cosmic giants?
The two big questions are how they formed and how they became so big so quickly in the early universe. We can start with the second question, which has really come to the fore thanks to discoveries made by the James Webb Space Telescope. JWST has opened a whole new window into the early universe in the past few years, because it is able to capture light from the very distant reaches of the observable universe, which has travelled to us from the very early universe. And we can now see supermassive black holes as they were when the universe was just 500 million years old – just a fraction of its current age.
The discovery that these early supermassive black holes can get so big so soon has been nothing short of mesmerising. It is an exciting moment because these early-universe supermassive black holes are appearing in much lower-mass galaxies than expected. What is particularly exciting is these discoveries open up a host of new questions. How did these black holes grow so large so quickly? What conditions could have led to their formation in such small, early galaxies?
The problem is that we would expect smaller black holes to be the seeds of supermassive black holes, but it’s very, very difficult for them to grow so big in the timescales needed. So this really shows that we don’t understand supermassive black hole growth very well: how the gas actually gets funnelled towards the centre and allows the black hole to grow so rapidly, especially in small galaxies.
You say that we expect smaller black holes to be the seeds for the creation of supermassive black holes, but is that the only way for them to form?
That remains a major open question, and we have several possibilities. Some people say that because we see these supermassive black holes so early on, they must have formed from the direct collapse of massive gas clouds, which would allow for them to start off quite massive.
Others say you could start with a different formation channel that produces a lower-mass black hole that then grows very rapidly, even if we’re not sure how. Here, the seed for the supermassive black hole could be the remnants of the first generation of stars, which were a lot more massive than today’s stars and a lot more short-lived. Or it could be the extremely dense star clusters found at the cores of many galaxies, including our Milky Way, where runaway collisions create a supermassive star that collapses to become the seed of a supermassive black hole.
The ongoing controversy is whether the discovery of numerous large supermassive black holes in the early universe favours the heavy seed idea or whether there is still scope for light seeds, even though these would have had to grow extremely rapidly. The good thing is that these surprisingly large early black holes should help us to tease it all apart. It’s an exciting time for us, because we have all these new questions and we are developing new simulations that we hope can tackle them.
Why are computer simulations so important?
If you have an observation of a galaxy, you only see a snapshot, whereas with computer models we can trace it from its birth to its present state. We get the full picture. Importantly, we can also use our simulations to test theories that cannot be tested with current observations and make predictions that we can then go out and test with future observations.
We have a few scenarios in mind for how supermassive black holes can form, and these will have different signatures in the early universe. For example, we can say that if supermassive black holes formed in a certain way, then an X-ray telescope would expect to see a certain thing. In astrophysics, it’s very hard to do experiments because you can’t rearrange the galaxies in the sky. But you can rearrange them in the computer and then see what impact different assumptions about supermassive black holes have.
The discovery that these early supermassive black holes can get so big so soon has been nothing short of mesmerising.
How do these simulations work?
You can think of it like a galaxy in a box, where we can simulate the galaxy from its birth not long after the big bang to the present day. You start with a few basic ingredients in the early universe – gas and dark matter and so on – and then you add hydrodynamics to model the gas evolution, and you have gravity. This is how you model basic structure formation, and then you plug in some assumptions about processes we are unsure about and run it forward.
Then there are processes that happen below the scale that you can see in the simulation, including most of black hole physics. That is what makes it so difficult, because the supermassive black holes are so tiny compared with the scale of the galaxy you are simulating. So you need to assume models for how supermassive black hole growth happens and then you feed that back onto the larger scales that it’s impacting.
How can we use simulations to unravel the mysteries of supermassive black holes?
You can run lots of different models for how they form and how they grow in your simulations and compare the results to the data we have from JWST. So you can say, “OK, if this model is correct then we should see this in JWST data”, or conversely you can say, “OK, JWST has seen this, what do we need to tweak in our models of supermassive black holes to replicate those observations in the results of our simulations”.
There is one idea, for example, that says supermassive black holes actually form from the collapse of a very early, very massive star. This would probably be the formation channel that would form the lowest-mass supermassive black holes. In that case, you could use the simulation to try to see, given the observation constraints, whether it is at all possible to reach these black hole masses very, very early on. If so, when would the star have needed to form and then collapse? You can narrow it down more and more in your simulations until you can start ruling things out.
Given the amount that we don’t know about supermassive black holes, how accurate can these simulations be?
Obviously, the computing capabilities have just reached incredible levels, so we can do much more accurate simulations and across many more scales than we ever could before. Some groups are now able to run hugely impressive simulations that span the scales from galaxies to the event horizon of a black hole, the boundary inside which gravity is so strong that light cannot escape.
Ultimately, we want to have simulations that encode all the information at every scale, from the cosmic web, which stretches across billions of light years, down to the event horizon. We want to feed the laws of physics in, along with our assumptions about supermassive black holes, and watch the entire universe evolve. That is beyond us at the moment, and that challenge is at the heart of my work.
How has the relationship between simulations and observations changed in recent years?
Simulations have gone from being tools we primarily use to interpret observations to being instrumental in making predictions that drive observational discoveries. It’s a profound change and it’s really pushing the boundaries of our understanding of galaxy formation and evolution, which remains one of these major unsolved challenges in astrophysics.
When JWST starts getting observations from even earlier times in the universe’s history, we will be able to settle this question of how supermassive black holes form and grow in conjunction with simulations that predict the expected population at these times depending on different seeding models.
These questions about supermassive black holes are clearly tough to answer. Why is it so important to get to the bottom of them?
I would say that everything in our universe is connected. The detail of the astrophysics of supermassive black holes and galaxy evolution obviously has an impact on cosmology, which seeks to explain the evolution of the universe as a whole.
Cosmology has now entered this high-precision era, where if we are to reconcile conflicting observations and stress test our best model of how the universe evolved, known as the standard model of cosmology, then we really do need to include supermassive black holes.
I really think that before we develop more exotic answers to some of these puzzles, as tempting as that may be, we need to fully understand galaxies and supermassive black holes. Without that, we cannot properly grasp how our universe has come to be what we see today.
