Researchers mapping dark matter in the local universe stunned t previously undiscovered filamentary structures bridging galaxies. The map, constructed with the help of machine learning by scientists from Penn State, could further studies regarding the nature of dark matter, its history, and the future of the local universe.
Understanding Dark Matter
Dark matter, according to Space.com has eluded scientists ever since its conceptualization. The substance is thought to consist of up to 80% of the known universe. In addition, it provides a skeletal frame for the cosmic web, a large-scale structure that because of its gravitational influence, is able to dictate the motion of celestial bodies, galaxies, and various cosmic materials.
On the other hand, researchers are unsure of the local distribution of dark matter and which restricts them from accurately measuring the force. Instead, scientists must infer the distribution of dark matter based on factors such as gravitational influences on various objects in the known universe, such as galaxies.
Donghui Jeong, co-author and an associate professor at Penn state explain that today it is much easier to study dark matter distribution much further away due to its reflection of the distant universe which is simpler than what we have locally.
He adds that over time, as the universe grows in scale and structure, its complexity has also increased making it harder for researchers to measure dark matter locally.
Mapping the Cosmos
Previously, mapping the cosmic web began with models of the early universe which researchers then simulated to evolve over billions of years. On the other hand, this method is computationally intensive and has only produced results regarding the local universe.
In a study slated to be published in the upcoming weeks in the journal Astrophysical Journal, researchers used a different approach, by using machine learning, to design a model that uses vital information regarding the distribution and motion of galaxies in the known universe to more precisely predict the distribution of dark matter.
The team built and trained the model via large set galaxy simulations known as Illustris-TNG that includes gasses, galaxies, visible matter, and dark matter. Specifically, the team simulated galaxies that are comparable in characteristics in the Milky Way and identified which properties of the known galaxy were needed in order to accurately predict the distribution of dark matter.
Phys.org reports that when given vital information, the model can theoretically fill in the gaps based on what has been looked at before. Jeong clarifies that maps generated from the models aren't perfectly fit with data simulations, however, this allows researchers to construct detailed structures.
The team discovered that the motion of galaxies, radial peculiar velocities, and their distribution enhances the quality of the map generated allowing researchers to see in great detail.
Researchers are ecstatic that the map was successfully able to reproduce prominent structures in the known universe. Jeong explains that a local map of the cosmic web broadens cosmological studies. Now it is possible to study dark matter distributions and how it relates to emissions of data, which in the long run will be vital in understanding the nature of dark matter. This will also allow researchers to study filamentary structures connecting the galaxies.
Check out more news and information on Space on Science Times.
