A new map of dark matter in the regional universe exposes numerous formerly unexplored filamentary formations linking galaxies. The map built applying machine learning by a global team involving a Penn State astrophysicist. It might facilitate studies regarding the nature of dark matter as well as concerning the past and future of our regional universe.
Dark matter is an obscure substance that covers 80% of the universe. It also extends the skeleton for what cosmologists call the cosmic web. It is the large-scale composition of the universe that, owing to its gravitational impact, affects the movement of galaxies and other cosmic substances. However, the allocation of regional dark matter is presently unspecified. This is because it cannot be quantified clearly. Scientists must in its place infer its allotment on the basis of its gravitational impact on other objects in the universe, such as galaxies.
“Ironically, it is simpler to research the dissemination of dark matter much further away. This is because it signifies the very faint past, which is much less intricate,” says Donghui Jeong. He is the associate professor of astronomy and astrophysics at Penn State and a consequent author of the research. “Over time, as the large-scale composition of the universe has increased, the intricacy of the universe has arisen. Thus, it is fundamentally tougher to make measurements regarding dark matter locally.”
Earlier efforts to map the cosmic web began with a model of the initial universe and then simulates the evolution of the model across billions of years. Though, this approach is computationally exhaustive and yet has not been capable to deliver results. In the latest study, the scientists took an entirely different method, applying machine learning. This is in order to create a model that utilizes data about the dissemination and movement of galaxies to foresee the distribution of dark matter.
The scientists developed and trained their model applying a large set of galaxy simulations, known as Illustris-TNGwhich includes:
- Other visible matter
- Dark matter
The squad specially chooses simulated galaxies in comparison to those in the Milky Way. This finally identifies which features of galaxies are essential in order to foresee dark matter dissemination.
In addition to Jeong, the research team includes:
- Sungwook Hong at the University of Seoul/Korea Astronomy and Space Science Institute in Korea
- Ho Seong Hwang at the Seoul National University in Korea
- Juhan Kim at the Korea Institute for Advanced Study
This study was supported in share by the National Research Foundation of Korea subsidized by the following institutes:
- Korean Ministry of Education
- Korean Ministry of Science
- S. National Science Foundation
- S. National Aeronautics and Space Administration Astrophysics Theory program