Machine Studying Pins Down Cosmological Parameters

The historical past of the Universe is ruled by a handful of cosmological parameters, such because the present-day matter density and amplitude of density fluctuations ( ${\Omega }_m$ and ${𝜎}_8$, respectively). Astronomers have measured these parameters by observing the large-scale construction of the Universe, however the measurement precision has been restricted by a number of parameter degeneracies. For instance, a rise in ${\Omega }_m$ can counteract a lower in ${𝜎}_8$ and vice versa. Now Tomasz Kacprzak and Janis Fluri on the Swiss Federal Institute of Expertise (ETH), Zurich, have discovered a strategy to break such degeneracies, strengthening the constraints on these parameters [1].

The Universe’s large-scale construction is often probed by learning galaxy clustering or through a statistical gravitational-lensing signature known as weak gravitational lensing (see Viewpoint: Weak Lensing Turns into a Excessive-Precision Survey Science). Every of those methods has related parameter degeneracies: for galaxy clustering, ${\Omega }_m$ and ${𝜎}_8$ can tackle any variety of values relying on the assumed galaxy bias, which quantifies how nicely seen matter traces invisible darkish matter; for weak lensing, ${\Omega }_m$ and ${𝜎}_8$ are equally degenerate with the intrinsic alignment amplitude, which measures the extent to which galaxies are aligned with each other.

Kacprzak and Fluri discovered that these degeneracies may very well be lifted through the use of deep studying—a kind of machine studying impressed by the human mind—to carry out a collective evaluation of galaxy-clustering and weak-lensing knowledge. The researchers present that, for future large-scale-structure surveys, this method might tighten constraints on the galaxy bias and on the intrinsic alignment amplitude by elements of about 1.5 and eight, respectively. In flip, the tactic might considerably strengthen constraints on each ${\Omega }_m$ and ${𝜎}_8$. Nonetheless, the group says that an improved knowledge evaluation and understanding of survey systematic results is required earlier than the method can be utilized in observe.

–Ryan Wilkinson

Ryan Wilkinson is a contract science editor and author based mostly in Durham, UK.

Associated Articles

Extra Articles