Australian Sky & Telescope — July 2017

36 AUSTRALIAN SKY & TELESCOPE July 2017

EAGLE GALAXIES: J. TRAYFORD AND THE EAGLE COLLABORATION (2); GRAPH: LEAH TISCIONE /

S&T

, SOURCE PHIL HOPKINS

data on small and large scales in our universe. “Illustris

producesthecorrectmixofellipticalandspiralgalaxies—none

oftheprevioussimulationswasabletoachievethisindetail,”

explains Illustris team member Mark Vogelsberger (MIT).

“Thisproblemwassoseverethatsomeastronomers

even questioned the underlying cosmological model,” adds

teammate Debora Sijacki (University of Cambridge, UK).

“Illustris’successstemsfromitsmoresophisticatednumerical

code,morerealisticphysicalprocesses,andhighresolution.”

Theresultisasimulationthattheteamcanusetolookbackin

time and understand how different galaxies formed.

Reuniting the cooks in the kitchen

Rapidprogressinthezoom-inandlarge-scalecommunities

isbeginningtobringthetwofieldstogether,sothathigh-

resolution, large-volume cosmological simulations that include

baryons are almost in reach. “What will be exciting, over the

nextseveralyearsandascomputationaltoolsimprove,ishow

wewillgraduallyseeaconvergenceofthesetwofrontiers,”

says Faucher-Giguère. “On the one hand, we will increase

ourvolumes,andconversely,large-volumesimulationswill

improve their resolutions and physical realism.”

Butthereisstillalongwaytogo.Intruth,mostbaryon-

inclusivesimulationsusealotofapproximations.EAGLE

principal investigator, Joop Shaye (Leiden University, The

Netherlands), notes that although hydrodynamics is explicitly

treated, radiation emission, star formation and evolution, and

blackholeprocessesareallapproximatedintheirsimulations.

This is essentially because current simulations are not

granularenoughtomodeltherealphysicsofprocesseslikestar

formationortheinfluenceofblackholesongalaxies,sothey

are put in by hand using a semi-analytic approach. Libeskind

illuminates this best: “Take star formation. This happens on

thescaleofastar—or,let’sbegenerousandsayonthescale

ofasolarsystem.Mostcosmologicalsimulationsdon’tresolve

suchsmallscales.Instead,ourhardlimitcanbesomething

like100parsecs(ifyou’relucky!)andthatcorrespondsto

somethinglike20milliontimesthedistancetoPluto.Sincewe

◗Although there is huge diversity in the cosmological

simulations mentioned here, they all share the same roots.

In the 1980s, Bolshoi co-Principal Investigator Joel Primack

(now at University of California, Santa Cruz) co-authored

the theory dubbedcold dark matter(CDM). This theory

described how the universe went from a smooth initial state

at early times to the large-scale, intricate structure we see

today, assuming that most matter is cold (ie. moves slowly)

and dark (ie. does not emit electromagnetic radiation or

scatter light). With the simple addition of an omnipresent

anti-gravity force calledΛ(lambda, or nowadays ‘dark

energy’), the currentΛCDM paradigm also accounts for the

accelerating expansion of the universe.

Simulations useΛCDM for good reason, as it has done an

exceptionally good job at explaining what telescopes have

revealed of the universe’s history and structure. Yet there is

still an elephant in the room, as EAGLE team member Rob

Crain (Liverpool John Moores University, UK) explains: “We

still don’t know the fundamental nature of theΛor the CDM!”

Many believe there is still scope for a slightly more

complicated theory, in which dark energy evolves, or

where dark matter is not perfectly cold and collisionless.

“But the impact on structure formation is very indirect

and very difficult to distinguish fromΛCDM,” says Illustris

team member Volker Springel (Max Planck Institute for

Astrophysics, Germany).

And there is still an even more extreme possibility. “One

of the most uncomfortable aspects regardingΛCDM is that

most of the universe is composed of a particle we have yet

to discover — dark matter — and a type of energy which

we understand only phenomenologically: dark energy,”

remarks Libeskind. “It could very well be that neither of these

proposed dark entities exists and that alternatives to the

laws of gravity are more likely.”

SEAGLE GALAXIES The EAGLE simulation produces a population

of galaxies similar to the one we see in the real universe, including

elliptical (left) and bluer disk (right) galaxies.

The dark universe

don’t resolve the scales on which stars are born, we simply put

them in ‘by hand’ when certain conditions are met. Needless

to say, it would be better if we could actually resolve the

formation of a protostar and follow its evolution in real detail.”

Striving for an ever more complete description of the

physics, researchers are also adding new ingredients to the mix,

such as magnetic fields, cosmic rays, dust physics and radiative

transfer. Doing this efficiently on upcoming generations of

supercomputers adds significantly to the challenge.

But it’s a challenge worth rising to, because cosmological

simulations are our one and only laboratory in which to test

and expand upon theories of the origin and evolution of the

cosmos. If these simulations capture enough of the physics of

the universe, they might not only tell us why galaxies form

— and therefore in essence how the conditions for life were

first constructed — but they may also offer our first glimpse of

what the dark universe is really made of.

„BENJAMIN SKUSE is a science writer based in Bristol, UK.

UNIVERSE 2.0