Hidden information in maps of galaxies spread across the universe could soon come forth, thanks to a new way of interrogating the data that preserves the three-dimensional nature of these maps.
The hidden information could be vital in telling us whether the standard model of cosmology is correct, or whether there are deviations from it that could affect our understanding of the “dark universe,” which comprises dark matter and dark energy.
The research, led by astronomer Minh Nguyen of the University of Tokyo, utilizes powerful computer algorithms that are able to compare the relative positions of galaxies in a 3D map of the universe with detailed simulations that depict the growth and behavior of galaxies and haloes of dark matter.
Back in the old days, astronomers would conduct galaxy surveys by taking deep space images on photographic plates and then measuring directly on the plates, in two-dimensions, the spatial distribution of the galaxies. They’d try answering questions like “How close are these galaxies to their neighbors?” and “How well-aligned are they with one another?”
In modern times, a third dimension can be added to these surveys. It’s all thanks to multi-object spectroscopy, which measures the redshift of these galaxies, and hence the distance to them in an expanding universe, in an observed volume of space. With such galactic distance measurements, it’s actually possible to create a three-dimensional map of the universe.
However, the calculating power required to statistically analyze this three-dimensional galaxy data is fiendish, and so, for efficiency, the 3D data has traditionally been compressed down into what are called “n-point correlation functions,” the “n” referring to a number (usually two or three points of data as mentioned above).
That’s all well and good in most cases, but there has also been a nagging suspicion that compressing and analyzing the data this way results in information being missed — or hidden. And now, using a technique called “field-level inference” (FLI) in combination with a suite of algorithms in a framework called “LEFTfield” that models galaxy growth and clustering from the early universe to the present day, Nguyen’s team has shown that vital information is indeed being suppressed by the compression. The team won third place in the Buchalter Cosmology Prize for this result.
“In field-level inference, we work directly…
Click Here to Read the Full Original Article at Latest from Space…