AI is today’s need and by using it, a new map of dark matter has been made that can help in studying the unknown features and filaments of invisible stuff that connects two galaxies. The resulting map studies the local universe mainly, thus around the Milky Way galaxy. Although the local universe is much closer yet it is very difficult to map as it is full of complex structures that are made from visible matter.
Donghui Jeong, an astrophysicist at Pennsylvania State University and the lead author of the new research, told Live science “We have to reverse engineer to figure out where dark matter is, by looking at galaxies.” Any substance that predominantly interacts through gravity with visible matter (like stars and planets).
The way by which study proceeded
Researchers believed that this invisible matter probably consists of weakly interactive huge particles or WIMPs, which would possess electromagnetic neutrality so that they are not easily linked with anything on the electromagnetic spectrum, thus light, etc.
There can be another perspective with some evidence to accept that dark matter perhaps consists of UV particles called axions. Any dark matter’s effects in the gravitational forces are traceable which are present in the universe.
The mapping of invisible gravitational forces is not easy. Often, researchers proceed with this by running large computer simulations. Initially, they start with a model of the early universe and fast-forward up to billions and millions of years of expansion and growth of dark matter.
Thus by filling the gravitational blanks researchers try to figure out where the dark matter was and where it has to be in the present. Jeong says that “Such execution requires lots of computing power and a very large amount of time.”
Ideas used in this Study
A different perspective has been taken into account in this new study. Researchers initially trained machine learning programs on nearly thousands of computer simulations of dark and visible matter generally present in the local universe.
Machine learning is a part of AI which studies the computer algorithm and automatically improves through experience and by the use of data. It is a technique that can pick up patterns from large datasets.
The model universes in the study came from an advanced set of simulations called Illustris-TNG. Before applying to real-world data, researchers tested the machine learning algorithms on the second set of Illustris-TNG universe simulations for higher accuracy.
Cosmicflows-3 galaxy catalogue has been used to account for the data on the distribution and movement of visible matter within the period of 200 megaparsecs or 6.5 billion light-years of the Milky Way galaxy. That area consists of more than 17000 galaxies.
The resulting map showed the connection between dark matter in the local universe with visible matter. In optimistic research, the machine learning algorithms tell much of what was already known about the Milky Way galaxy’s neighborhood from cosmological simulation. Interestingly it also gives the idea of new features including long filaments of dark matter that link the galaxies around the Milky way with it and with one another. Jeong said, “it is important to understand how galaxies will move with time.”
As a result, there is a new map of dark matter. Thus the Milky Way and the Andromeda galaxies are expected to crash into one another in nearly 4.5 billion years.
By studying the local dark matter’s role in such collisions might help to address more accurately how and when that fusion and others will occur.
“Now that we know the distribution of dark matter we can calculate more precisely about the acceleration that will move the galaxies around us,” Jeong says in a statement.