It all began roughly 13.8 billion years ago with a massive, cosmic “bang” that brought the cosmos into being instantly and dramatically. Soon later, the baby cosmos began to cool significantly and turned utterly black.
The cosmos then awoke a few hundred million years after the Big Bang, when gravity collected stuff into the first stars and galaxies. The light from these earliest stars converted the surrounding gas into hot, ionized plasma, a critical event known as cosmic reionization that drove the cosmos into the intricate structure we see today.
With Thesan, a novel simulation constructed by scientists at MIT, Harvard University, and the Max Planck Institute for Astrophysics, scientists may now receive a precise perspective of how the universe may have evolved during this key epoch.
Thesan, named for the Etruscan goddess of the morning, is intended to recreate the “cosmic dawn,” and especially cosmic reionization, a phase that has been difficult to reconstruct due to enormously intricate, chaotic interactions, such as those involving gravity, gas, and radiation.
Thesan’s simulation resolves these interactions in more depth and across a larger volume than any prior simulation. It does this by combining a realistic model of galaxy formation with a novel algorithm that records how light interacts with gas, as well as a cosmic dust model.
Thesan allows the researchers to replicate a cubic chunk of the cosmos covering 300 million light years. They advance the simulation forward in time to see the birth and development of hundreds of millions of galaxies inside this space, starting roughly 400,000 years after the Big Bang and continuing for the first billion years.
So far, the simulations match the scant observations of the early cosmos that astronomers have. As further observations of this epoch are produced, like as with the recently deployed James Webb Space Telescope, Thesan may aid in placing such data in cosmic perspective.
For the time being, the models are beginning to offer information on processes such as how far light can travel in the early cosmos and which galaxies were responsible for reionization.
Aaron Smith, a NASA Einstein Fellow at MIT’s Kavli Institute for Astrophysics and Space Research, describes Thesan as “a gateway to the early cosmos.” “It is designed to be a perfect simulation equivalent for emerging observational facilities that are set to dramatically transform our knowledge of the universe.”
Smith and Mark Vogelsberger of MIT, Rahul Kannan of the Harvard-Smithsonian Center for Astrophysics, and Enrico Garaldi of Max Planck have presented the Thesan simulation in three publications, the third of which was published today in the Monthly Notices of the Royal Astronomical Society.
Pay attention to the light
The cosmos was a black and homogeneous realm during the early phases of cosmic reionization. The cosmic development during these early “dark ages” is extremely straightforward to compute for physicists.
“In theory, you could do this using pen and paper,” Smith explains. “However, at some point, gravity begins to pull and compress things together, first slowly, but later so fast that calculations become too intricate, and we must do a complete simulation.”
The researchers attempted to integrate as many important elements of the early cosmos as possible in order to completely mimic cosmic reionization. They began with a successful model of galaxy formation established earlier by their organizations, known as Illustris-TNG, which has been demonstrated to effectively reproduce the features and populations of developing galaxies. They then created a new algorithm to account for how light from galaxies and stars interacts with and reionizes the surrounding gas, an exceedingly complicated process that previous models have not been able to mimic correctly on a wide scale.
“Thesan studies how the light from these initial galaxies interacts with the gas during the first billion years, transforming the cosmos from neutral to ionized,” Kannan explains. “In this manner, we can automatically track the reionization process as it proceeds.”
Finally, the researchers incorporated a preliminary model of cosmic dust, which is another element unique to early universe simulations. This preliminary model seeks to explain how small grains of material impact the creation of galaxies in the early, sparse cosmos.
The cosmic bridge
Based on accurate observations of relic light from the Big Bang, the scientists set the simulation’s beginning circumstances for about 400,000 years after the Big Bang. They then used the SuperMUC-NG machine, one of the world’s biggest supercomputers, to replicate a slice of the cosmos by concurrently harnessing 60,000 processing cores to carry out Thesan’s calculations over an equivalent of 30 million CPU hours (an effort that would have taken 3,500 years to run on a single desktop).
The simulations gave the most comprehensive depiction of cosmic reionization over the biggest extent of space of any previous simulation. While some simulations simulate across long distances, they do so at a low resolution, and other, more comprehensive simulations do not.
“We’re connecting these two approaches: we have both big volume and high resolution,” Vogelsberger says.
Early simulation evaluations indicate that towards the conclusion of cosmic reionization, the distance light could travel grew more substantially than scientists had previously anticipated.
“Thesan discovered that light does not travel long distances in the early cosmos,” Kannan explains. “In actuality, this distance is relatively tiny and only becomes significant towards the conclusion of reionization, expanding by a factor of ten in only a few hundred million years.”
The researchers also notice clues of the sort of galaxies that are responsible for reionization. The mass of a galaxy seems to impact reionization, but the team believes that further data from James Webb and other telescopes will assist to narrow down these prominent galaxies.
“[Modeling cosmic reionization] has a lot of moving elements,” Vogelsberger concludes. “When we can put this all together in some type of machinery and get it running and it generates a dynamic cosmos, that’s a fairly satisfying moment for all of us.”
NASA, the National Science Foundation, and the Gauss Center for Supercomputing all contributed to this study.