http://www.skyandtelescope.com.au 33~10–^5 parsec
Stars, protostellar disks~10–^2 -10^0 parsec
Cores, clusters,
supernova blastwaves~10^1 -10^2 parsecs
Molecular clouds,
star-forming regionsLEAH TISCIONE /
S&T
, SOURCE PHIL HOPKINS (2)
scales — how they form, how they die, and how they affect
their environs,” outlines Faucher-Giguère, joint principal
investigator (PI) on FIRE. Working in concert, these small-
scale processes from billions of stars affect the larger-scale
properties of the galaxy as a whole.
“With the FIRE approach, we can directly test physics on
the smallest resolved scales and see if additional processes
are needed to explain galaxy formation,” says Dušan Kereš
(University of California, San Diego), another FIRE PI. This
tack is highly novel, as simulations traditionally start from
a large cosmic volume, resolve everything that is possible
to resolve, and then approximate what happens on smaller,
unresolved scales.
Meanwhile, CLUES focuses on the Milky Way’s Local
Group of galaxies — a galaxy cluster 10 million light-years
across consisting of our own Milky Way, the neighbouring
Andromeda Galaxy, and the Triangulum Galaxy, as well
as dozens of smaller galaxies — and ‘constrains’ the large-
scale structures around it. “What this means is that we
ensure, through a kind of manipulation, that the galaxies
we simulate are in the correct place,” clarifies CLUES joint-PI
Noam Libeskind (Leibniz Institute for Astrophysics Potsdam,
Germany). “What we are constraining is the initial conditions
of the problem. We are essentially artificially selecting initial
conditions that we know, when run to conclusion, will result
in simulations that mock the local cosmography.”
This enables researchers to conduct various experiments
more easily, because large galaxies will all be in the right
places regardless of the assumptions on dark matter, gas
physics, and star formation used, while the amounts and
distributions of dark matter, gas and stars will be different.
They can then compare these alternatives with observations
to see which matches reality best.
Although wildly different in approach, both FIRE and
CLUES have already offered new insights into a specific issue
that has confounded astrophysicists for nearly two decades:
the missing satellites problem. Dark-matter-only experiments
have shown that a huge number of small satellite galaxies
should have formed in our cosmic neighbourhood, yet only ahandful of these galaxies are seen orbiting the Milky Way.
However, in 2013 the CLUES collaboration analysed its
simulations and discovered that some of the dwarf galaxies in
the Local Group move with such high velocities with respect
to the cosmic web that most of their gas can be stripped as
they pass through, essentially destroying their ability to make
stars. Named ‘cosmic web stripping,’ the mechanism can help
explain the observed paucity of dwarf galaxies compared with
that predicted.
The bottom-up, FIRE-based Latte simulations also suggest
normal matter solves the missing satellites problem. Wetzel,
who leads the project, used Latte — the highest-resolution
cosmological simulation of Milky Way–mass galaxies to
date — to look at the dwarfs that form as the main galaxy
coalesces and evolves. Compared to dark-matter-only
simulations, only about a thousandth as many starry satellite
galaxies formed in Latte. The team reasons this discrepancy
is caused by two processes: The dwarf galaxies have their gas
ripped away from them as they pass through the big galaxy’s
surrounding gas cloud, preventing star formation; and the
main galaxy’s central stellar disk gravitationally destroys the
invisible dark matter clumps in which the satellite galaxies
would otherwise have been born.
FIRE also naturally reproduces the ‘low’ density of dark
matter in galactic cores, the highly inefficient rate of star
formation in galaxies (roughly 30% of the Milky Way’s
baryons are in stars, instead of the nearly 100% predicted
without feedback), and several other observations.A bigger slice
If the zoom-in simulations — sacrificing scale for detail
— equate to the serene pleasure of home baking, larger
simulations that incorporate baryons, such as BlueTides,
EAGLE, and Illustris, can be likened to wedding catering,
constraining the number of ingredients and simplifying the
recipe in order to feed the crowds in a reasonable amount of
time. As a result, compromises have to be made.
Illustris and EAGLE (Evolution and Assembly of Galaxies
and Their Environments), for example, run from early after