Â鶹ÒùÔº

Pictures are the key to new insights in the field of astrophysics. Such images include simulations of cosmic events, which astrophysicists at UZH use to investigate how stars, planets and galaxies came into existence.

Lucio Mayer says, "The field of astrophysics today is on the cusp of a golden age. Never before have such vast amounts of data from so many different sectors of the universe been at our disposal," he says. The James Webb Space Telescope, for example, allows very distant galaxies to be viewed in an unprecedented image quality.

"The telescope delivers new, astonishing discoveries almost every week," the UZH professor of astrophysics says. The Square Kilometer Array Observatory (SKAO) being built in Australia and South Africa is still under construction. That enormous telescope array will collect more data than any scientific project has before.

Understanding the universe

The new information is bringing astrophysics ever closer to its ultimate goal of understanding the universe as a whole and in detail—from the genesis of stars to the structure of the cosmos. Astrophysicists at UZH like Lucio Mayer and Ravit Helled are right in the thick of this frenzy of new discoveries.

Mayer conducts research into how stars and galaxies come into being, and Helled investigates how planets form. Both researchers work with computer simulations that make it possible to model and analyze how planets, stars and entire galaxies formed and evolved over time.

The simulations require massive amounts of computing power. In an international competition, Mayer's team succeeded in snagging a time slot to use the LUMI computer, Europe's most powerful supercomputer, in Finland.

The key to Mayer's success is a new code for computing cosmic events. It took nearly seven years to develop it, the researcher recounts. A team from Zurich, Basel and the Swiss National Supercomputing Center (CSCS) in Lugano staffed with specialists in informatics, computational science, astrophysics and cosmology worked on it.

"With the aid of the supercomputer, we are able to model how planets, stars, galaxies—indeed, the entire cosmos—came into existence," Mayer says. Those simulations are so computationally intensive that they would take years if they were done using customary methods and conventional computing power. "Now we can perform them in a matter of a few days."

The new graphics processing units (GPUs) make it all possible. They run the simulations up to 1,000 times faster than conventional computers equipped with classical processors.

Vast clouds full of stars

Thanks to the greater performance capacity, much more complex processes such as those that occur inside giant molecular clouds where millions of stars form can be modeled. "That heretofore wasn't possible," Mayer says, "but we will be able to do it with the new code. At the moment we're working on enlarging the scale of the models."

Giant molecular clouds are particularly important because that's where stars originate. "If we can simulate giant molecular clouds, we can reconstruct how star formation takes place throughout an entire galaxy," Mayer says. The stars that come into being and form stellar clusters in specific locations in a galaxy play a crucial role in the life of that galaxy. They release energy that affects the future evolution of the galaxy.

The simulation stage with the LUMI computer has just recently been completed. "Now we're starting to take a look at the properties of the clouds in which the stars are located," Mayer explains. The simulations enable astrophysicists to look at the different stages of stellar genesis and to determine their precise chronological order.

"We see what happened and when it happened," Mayer says. The findings are then compared with the data from large telescopes to check if the results of the simulations match what's visible in the universe. If they don't match, the models and calculations have to be adjusted.

Mayer describes this as a "dialogue" aimed at explaining as accurately as possible how stars and galaxies originated and evolved. One spectacular simulation performed by Mayer's research group depicts the first galaxies and star clusters that formed in the universe. The simulation reconstructs how those galaxies, which were just recently discovered by the James Webb Space Telescope, came into existence.

Simulations of that kind used to take months and sometimes years to perform. But afterwards, Mayer says, it was pretty easy to analyze the resulting data because the data sets were small. "We could do it on our laptops," he recounts. Today it's entirely different: the simulations run much faster and deliver vast amounts of data to analyze. Artificial intelligence is specifically trained and employed to do that.

"The Swiss Data Science Center has expertise in AI and machine learning and is collaborating with our team to develop new methods to analyze the results of the simulations," Mayer says.

Simulations of that kind are astrophysical laboratory experiments of sorts that attempt to replicate and reconstruct cosmic processes because when one peers at the universe, one sees only a snapshot in time but doesn't know how it came about.

"We try to work backwards to gain an understanding of how stellar clusters, galaxies and indeed the entire universe came into existence and why the universe is the way it is today," Mayer explains. "When I compare my simulations with the real universe, I'd like to be able to say, 'okay, the galaxy that I can create actually looks like that in reality.'"

Lucio Mayer researches how galaxies and stars came into being. The energy of stars and stellar clusters affects the evolution of galaxies. And planets form around stars. Ravit Helled conducts research into how planets originate. The UZH professor of astrophysics likewise works with models and simulations. Like in Mayer's research work, they serve to close gaps in knowledge left open by telescopic observation.

"We see protoplanetary disks and we see today's planets, but we don't see what happens in between," Helled says.

Helled points out that some of the simulations generated by the models developed by her are visually less spectacular than those created by Mayer. But they are especially well suited for illustrating the basic physical processes that take place during planetary formation.

And Helled is no less enthusiastic than Mayer when she talks about her research. She wants to understand how planets form and why they differ from each other. Planets are born from the gas and dust in rotating protoplanetary disks that surround young stars. Their diversity is explained by the varying initial conditions—such as the temperature, density, composition and distribution of matter—that prevail in protoplanetary disks.

Small changes with a big effect

Slight variations in formation conditions give rise to very different planets, as we can observe in our own solar system, where there are terrestrial planets like Earth, Mercury, Venus and Mars that are comparatively small and have a solid, rocky surface. There are also gas and ice giants—Jupiter, Saturn, Uranus and Neptune—composed mainly of hydrogen and helium. They do not have a solid surface and are much larger than the terrestrial planets. "With our models, we are able to show how even the tiniest changes affect planetary formation," Helled explains.

In her research, the astrophysicist comes to some remarkable conclusions that call existing theories into question. For example, the ice giants Uranus and Neptune differ from one another significantly despite having many similarities. Uranus has an extreme axial tilt and has regular satellites, no internal heat source and a greater density, while Neptune has irregular moons and does possess an inner heat source. Neptune also has a different density distribution.

How did those differences come about? Through simulations, Helled discovered a possible explanation: she modeled what happens when a large celestial body collides with a young planet. Disruptive occurrences of that kind might explain why Uranus and Neptune are so different from each other today.

In the Uranus simulation, a glancing collision is the cause leading to the planet's axial tilt, satellite disk and internal structure. In the case of Neptune, a head-on collision may have affected that planet's inner structure and energy profile.

Solving a puzzle

Colossal collisions of that kind would also explain why Uranus and Neptune formed the way they did in their place in the solar system, because according to classical theory, it is improbable that they came into being in this form in their present position.

"This demonstrates to us how certain events can alter the evolution of planets and knock them out of their expected orbit," Helled explains. And it shows just how rewardingly simulations can be employed. "Explaining how Uranus and Neptune came into existence in their present form was a puzzle and a challenge for science for years."

That statement also applies to Jupiter. That planet possesses a fuzzy core with no sharp boundary. Helled's simulations suggest that here, too, a tremendous collision with another celestial body during Jupiter's early formation may explain this phenomenon. The impact partially ruptures the planet's core, but it then reaccretes relatively quickly.

Modeling simulations of that kind is a challenging endeavor. Helled says, "The cosmic processes are diverse and complex, and there are so many parameters that have to be taken into account." That's why numerous iterations are run with different variables, she adds. "Then you see the outcome, and it's sometimes very surprising." Surprises of that kind are what make the work fascinating to Helled.

The research conducted by Ravit Helled and Lucio Mayer exemplifies how new, spectacular insights can be gained through simulations. "It's truly an exciting time," Mayer says, "especially for students. I tell them that they picked a great moment to do research in this field."