Filling the Cosmos with (Virtual) Stars. Mark Krumholz, UC Santa Cruz HiPACC Computational Astronomy Journalism Boot Camp, June 26, 2012

Transcription

1 Filling the Cosmos with (Virtual) Stars Mark Krumholz, UC Santa Cruz HiPACC Computational Astronomy Journalism Boot Camp, June 26, 2012

2 Outline What is the interstellar medium, and how is it related to star formation? Why is this a problem for supercomputers? What are the big questions, and how do simulations help answer them?

3 What you get if you google image search interstellar medium The black light of the interstellar medium of deep space illumines all that is not, to be seen and not to Be. To Be and not seen this is the path of the Panther woman. From the hydrogen oceans of Deep Space she condenses rain. Water precipitates from Her Vast Womb, falling onto planets open to Her Life giving moisture.

4 Hier ist wahrhaftig ein loch im Himmel!* * Here truly is a hole in Heaven! - - William Herschel, as remembered by his sister Caroline, 1784

5 Postsdam, 1904

6 Observations from Lick Observatory!

7 A Modern View of the ISM ISM made up of many phases Most H is atomic, n ~ 1 cm 3, T ~ 1,000 10,000 K Densest gas is molecular (H 2 ), n > 100 cm 3, T ~ 10 K

8 Milky Way Census Visible Stars: ~10 11 M H Atomic hydrogen: ~10 10 M CO Molecular hydrogen: ~ M

9 Evidence for Star Formation

10 Leroy Maps of HI, H 2, and SFR

11 Why Supercomputers? Clouds filled with hypersonic turbulent motions, so analytic treatment impossible Many physical processes important Gas flows, gravity, magnetic fields, radiation Large range of scales: typical GMC density ~100 cm 3, stellar density ~10 24 cm 3 Timescales too long for direct observation

12 What Does a Simulation Do? Step 1: start with the underlying equations. Example: thermal diffusion Time T t = F F = κ T = T t = κ 2 T Temperature Heat flux Thermal conductivity

13 Step 2: Discretize y L x x j = 3 (3,0) (3,1) (3,2) (3,3) (3,4) (3,5) j = 2 (2,0) (2,1) (2,2) (2,3) (2,4) (2,5) y L y j = 1 (1,0) (1,1) (1,2) (1,3) (1,4) (1,5) j = 0 (0,0) (0,1) (0,2) (0,3) (0,4) (0.5) i = 0 Discretize in space: T ij t i = 1 i = 2 i = 3 i = 4 i = 5 = κ x 2 (T i 1,j + T i+1,j + T i,j 1, +T i,j+1 4T i,j ) x

14 Step 2: Discretize y L x x j = 3 (3,0) (3,1) (3,2) (3,3) (3,4) (3,5) j = 2 (2,0) (2,1) (2,2) (2,3) (2,4) (2,5) y L y j = 1 (1,0) (1,1) (1,2) (1,3) (1,4) (1,5) j = 0 (0,0) (0,1) (0,2) (0,3) (0,4) (0.5) i = 0 Discretize in space and time: T (n+1) ij = T (n) ij i = 1 i = 2 i = 3 i = 4 i = 5 + t x 2 κ T (n) i 1,j + T (n) i+1,j + T (n) (n) i,j 1, +T i,j+1 x (n) 4T i,j

15 Step 3: Turn the Crank

16 Star Formation: The Big Questions What determines stellar masses? Does stellar mass have an upper limit? What determines the rate at which stars form?

17 The IMF Fig. 18. The measured IMF for the ONC, fitted with a log-normal dis top and bottom IMF as measured panels represent in the Orion thenebula mass distributions Cluster (da Rio+ obtained 2012) assumin respectively, whereas the center panels show the result assuming Baraffe mass has been extrapolated from their T eff (see text). The shaded areas represent the IMF of Chabrier (2003) (left panel) and Kroupa (2001) (righ

18 Challenges of the IMF IMF is universal or nearly so, yet star- forming environments vary greatly. Why? Why are stars at the right mass scale to allow nuclear fusion?

19 Simple Collapsing Cluster Simulation Column density Temperature Krumholz, Klein, & McKee (2011) What s included: hydrodynamics, gravity, radiation Numerical method: AMR

20 Doesn t Work

21 Adding More Physics Krumholz, Klein, & McKee (2012) What s included: hydrodynamics, gravity, radiation, protostellar jets Numerical method: AMR

22 Works Much Better

23 Do Stars Have a Mass Limit? Krumholz What s included: hydrodynamics, gravity, radiation Numerical method: AMR

24 Observational Test: The Disk

25 The Instrument Atacama Large Millimeter Array

26 The Prediction Simulated ALMA observation in CH 3 CN GHz (Krumholz+ 2007) Numerical method: long characteristics

27 Star Formation Rates: the Simplest Approach What s included: hydrodynamics, gravity Numerical method: SPH

28 SFR: The Problem in a Nutshell Simulation Bigiel Too fast by a factor of ~100 Too efficient: simulation forms a gravitationally bound cluster, but most stars not in clusters

29 Adding More Physics Simulation: Wang et al. (2009) What s included: MHD, gravity, protostellar jets Numerical method: AMR

30 Concluding Thoughts Simulations are the indispensible tool for understanding star formation Many big problems not fully solved: origin of stellar masses, star formation rates This is probably because we don t have all the necessary physics in the codes yet