Simulating galaxies on computers:
a story of the Universe
across space and time

Dr. Corentin Cadiou

University College London

@cphyc

@corentin-cadiou

Credits: NASA & ESA

Sombrero galaxy

M83

"Penguin" galaxy

Magellanic clouds

M87 galaxy

What is a galaxy?

Credits: A. Russell/ESO

What is a galaxy?

Credits: A. Russell/ESO

  • Hundreds of billions of stars*
  • Dust
  • Gas
  • Dark matter
  • Supermassive black hole

* only true for Milky Way, may vary by order of magnitudes

Dust in the horsehead nebula, credits: HST

Star-forming region (Westerlund 2), credits: HST

SMBH in galaxy M87, credit: EHT

DM (in blue) in a simulation

Credits: Adams Evans

Observations in astronomy

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax
  2. Times — seemingly no evolution

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax
  2. Times — seemingly no evolution

Andromeda in 1899 by  Isaac Roberts

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax
  2. Times — seemingly no evolution
  3. Scale coupling

mm

km

10,000 km

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax
  2. Times — seemingly no evolution
  3. Scale coupling

km

10,000 km

\(10^{-3} \ \mathrm{ly} \)

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax
  2. Times — seemingly no evolution
  3. Scale coupling

\(10\ \mathrm{ly}\)

10,000 km

\(10^{-3} \ \mathrm{ly} \)

Credits: Adams Evans

Observations in astronomy

Three key issues

  1. Distances — no parallax
  2. Times — seemingly no evolution
  3. Scale coupling

\(10\ \mathrm{ly}\)

\(100,000 \ \mathrm{ly}\)

\(10^{-3} \ \mathrm{ly} \)

NASA; ESA; and F. Summers

Simulations are proxy for experiments

Merging of the Milky Way (our own galaxy)
with Andromeda [in a few billion years]

NASA; ESA; and F. Summers

Simulations are proxy for experiments

NASA; ESA; and F. Summers

How to simulate the Universe?

... and galaxies therein

Each bright dot is one galaxy

?

HST

  1. Create a box for space and time
  2. Include known physics
  3. Set some initial conditions
  4. Move time forward
  5. ☕*
  6. SCIENCE!
    • Compare to observations
    • Make predictions
    • …

How to simulate the Universe?

... and galaxies therein

* of the order of \( 10^6-10^7\ \mathrm{hr} \approx 100-1000\ \mathrm{yr} \)

  1. Create a box for space and time
  2. Include known physics
  3. Set some initial conditions
  4. Move time forward
  5. ☕*
  6. SCIENCE!
    • Compare to observations
    • Make predictions
    • …

How to simulate the Universe?

... and galaxies therein

* of the order of \( 10^6-10^7\ \mathrm{hr} \approx 100-1000\ \mathrm{yr} \)

  1. Create a box for space and time
  2. Include known physics
  3. Set some initial conditions
  4. Move time forward
  5. ☕
  6. SCIENCE !
    • Compare to observations
    • Make predictions
    • …

How to simulate the Universe?

... and galaxies therein

Forming galaxies...

New Horizon simulation — IAP, CNRS

50 million light year

50 million light year

Galaxies

  1. form in a "cosmic web"
  2. grow by accreting
  3. merge together

1 million light year

100,000 light year

We have a scenario for galaxy formation

We have a scenario for galaxy formation

  1. Formation of dark matter "halos"...
  2. ... and of the cosmic web
  3. Gas falls in, galaxies merge
  4. Gas condenses to form stars & black holes
  5. Stars & black holes expel gas

3C 348 seen by HST & VLA

Crab nebula seen by HST

We have a scenario for galaxy formation

Can we test it?

Effect of physical parameters

Formation of stars

Formation of black holes

Feedback effects

[...]

Effect of physical parameters

Dubois+16

Black holes prevent the formation

of large spiral galaxies

Effect of physical parameters

Dubois+16

The initial conditions of the Universe

Our whole universe was in a hot, dense state*
Then nearly fourteen billion years ago expansion started, wait […]

*(and homogenous)

The initial conditions of the Universe

Planck satellite. Credits: ESA/NASA/JPL-Caltech

The initial conditions of the Universe

The initial conditions of the Universe

Initial conditions

Evolved Universe (+ galaxies)

The initial conditions of the Universe

Initial conditions

Evolved Universe (+ galaxies)

The Universe is determined

by its initial conditions

Initial conditions = DNA of galaxies
Genetically modified galaxies?
 

\(14\ \mathrm{Gyr}\)

The initial conditions of the Universe

Initial conditions:

  • Tiny density fluctuations \(\sim  0.001\% \)
  • Very well described mathematically*
    → easy to generate

*By a Gaussian random field with known spectrum

The initial conditions of the Universe

Initial conditions:

  • Tiny density fluctuations \(\sim  0.001\% \)
  • Very well described mathematically*
    → easy to generate & constrain

"Splicing" method, Cadiou+ in prep.

The initial conditions of the Universe

"Splicing" method, Cadiou+ in prep.

\(14\ \mathrm{Gyr}\)

The initial conditions of the Universe

Galaxies are influenced by a region at least \( 100\times\) larger

"Splicing" method, Cadiou+ in prep.

Galaxies are influenced by a region at least \( 100\times\) larger

"Splicing" method, Cadiou+ in prep.

\(100\ \mathrm{kly}\)

\(10 \ \mathrm{Mly}\)

We can simulate galaxies on (super)computers

→ insight on actual galaxies

→ better understanding of physics

→ testbench of astrophysics

Questions?

Forming an elliptical galaxy...

Forming an elliptical galaxy...

Adaptive Mesh Refinment

Domain decomposition