![Page 1: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/1.jpg)
Computational Astrophysics versus the Big Questions: An Assessment
Michael L NormanPhysics Department and San Diego
Supercomputer CenterSupercomputer CenterUCSD
![Page 2: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/2.jpg)
Computational Science: h d ll fThe 3rd Pillar of Science
“The purpose of computingis insight, not numbers”
--Hamming
“The purpose of computingis prediction”
--Worlton
simulation
theory observationtheory observation
“Computer simulations are the most“Simulation is a bridge between theory and observation”
Computer simulations are the mostcomplete descriptions of complexphenomena we have”
![Page 3: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/3.jpg)
60 years of supercomputer f k ’ lperformance tracks Moore’s law
sequential parallel multicore
Sourcebook, Ch. 1
![Page 4: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/4.jpg)
Importance of Numerical Algorithms
Springel et al. (2005)
![Page 5: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/5.jpg)
Mature Multiscale MethodsMature Multiscale Methods
• N‐body/SPH tree codes • AMR hydro/MHDN body/SPH tree codes AMR hydro/MHD
DM substructure in Milky WayDiemand et al. (2008)
Dense molecular cloud cores Collins (2009)
![Page 6: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/6.jpg)
Where’s the Beef?Where s the Beef?
• What Grand Challenge problems hasWhat Grand Challenge problems has computational astrophysics solved?– “For every problem solved 10 new problems are– For every problem solved, 10 new problems are identified”
• If not solved then what progress has been• If not solved, then what progress has been achieved, and how?
Wh t l l h l d b t• What general lessons have we learned about what is needed for genuine progress?
![Page 7: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/7.jpg)
Some Grand Challenge ProblemsSome Grand Challenge Problems
• Formation of stars and planetary systemsFormation of stars and planetary systems
• Type Ia and II supernovae mechanisms
i f l i d l l• Formation of galaxies and large scale structure
• Formation of supermassive black holes
• Origin of cosmic magnetic fields
• Origin of highest energy cosmic raysOrigin of highest energy cosmic rays
• Nature of the solar dynamo
![Page 8: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/8.jpg)
Why Grand Challenge Problems are ff lDifficult
• Phenomena arePhenomena are– Complex
Dynamical
M101
– Dynamical
– Multidimensional
M lti l– Multiscale
– Inter‐related
b bl• Direct observations sometimes not possible or yield meager information (e.g, supernovae)
![Page 9: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/9.jpg)
Galactic Star FormationGalactic Star Formation
• Driving questionsDriving questions– Origin of mass scale?
– Origin of IMF?
– Why star formation efficiency is so low?
i i f bi i ?– Origin of binarity?
– Role of feedback (outflows, radiation) in(outflows, radiation) in setting final mass?
– Properties of Young Stellar Objects (YSOs)
NCG 602 in LMC
![Page 10: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/10.jpg)
Molecular Cloud Complex in Perseusp
![Page 11: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/11.jpg)
Molecular Clouds, Clumps, and CoresMolecular Clouds, Clumps, and Cores
Highly complex structure:g y pHierarchical, fractal
![Page 12: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/12.jpg)
![Page 13: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/13.jpg)
Tale of 2 ReviewsSh t l (1987) M K & O t ik (2007)Shu et al. (1987) McKee & Ostriker (2007)
FOCUS
o Low mass star formationf
o Stars of all massesl l l b l fo How dense cores form stars o How molecular cloud turbulence forms
dense cores
PARADIGM
o “Magnetic star formation”o Ambipolar diffusion creates dense cores quasi‐statically
o “Turbulent star formation”o Molecular cloud turbulence dynamicallycompresses gas to beyond stability limit
MAIN PREDICTIONSMAIN PREDICTIONS
o Subcritical clouds: isolated low mass stars form at low efficiencyo Supercritical clouds: high mass stars and
o Density and velocity statisticso Core IMFo Star formation efficiencyo Supercritical clouds: high mass stars and
clusters form at high efficiencyo Star formation efficiency
TYPICAL SIMULATIONS
o 1D 2D cloud collapse models o 3D turbulence in a boxo 1D, 2D cloud collapse modelso Synthetic spectra of YSOs
o 3D turbulence in a boxo Synthetic molecular cloud maps
![Page 14: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/14.jpg)
Formation of Low Mass StarsShu, Adams & Lizano (1987), ARAA 25
• Stage 1– Dense cores form via ambipolar diffusion
• Stage 2• Stage 2– Inside‐out collapse to form protostar/diskp /
• Stage 3– Inflow + outflow triggered by deuterium burning
St 4• Stage 4– Isolated star/disk system
![Page 15: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/15.jpg)
Magnetically Supported CloudsMouschovias (1976a,b)
J i t bilit
2
⎟⎞
⎜⎛⎟⎞
⎜⎛Φ RB
• Jeans instability requires M > Mcr
3
2301013.0 ⎟⎟
⎠
⎞⎜⎜⎝
⎛⎟⎟⎠
⎞⎜⎜⎝
⎛≈
Φ=
pcR
GB
GM cr μ
Subcritical: M< M– Subcritical: M< Mcr
– Supercritical: M> Mcr
• M/Φ increases due toM/Φ increases due to ambipolar diffusion, inevitably leading to y gcollapse
![Page 16: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/16.jpg)
Collapse of Singular Isothermal Sphere Shu (1977)Shu (1977)
• SIS: no characteristicSIS: no characteristic mass scale
==)(2
mkTaarρ
∞⇒∞⇒∞⇒⇒
==
)()( ,0
,2
)( 2
rMrrr
mkTaGr
r
ρπ
ρ
• SIS: characteristic mass
∞⇒∞⇒ )( , rMr
SIS: characteristic mass accretion rate
am3
970=&G
mSIS 97.0=
![Page 17: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/17.jpg)
Critiqueh l’ ll f h• Shu et al’s 4 stages are essentially a summary of what is observed, not our theoretical understanding
N th f i i f t IMF• No theory for origin of core mass spectrum or IMF
• Intermediate and high mass stars not addressed
f f f• Mass scale for low mass stars cannot fall out of SIS theory since it is scale free
Either mass scale is set by– Either mass scale is set by: • core mass, for which no theory was presented
• protostellar feedback (Shu), for which no theory presented
– Or • Magnetic or turbulent support of envelope
N i l i l i i l f d• Numerical simulations not prominently featured
![Page 18: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/18.jpg)
However…However…
• Shu et alShu et al. summarize good agreement
Stellar “birthline”
agreement between Phase 4 models andmodels and observations of YSOsYSOs
Reprocessed radiation inReprocessed radiation in Star-dusty disk system
![Page 19: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/19.jpg)
Star Formation and Turbulence:h dThe New Paradigm
Star s:
MMass distribution functionuniversal
Turbulent fragmentation
Molecular clouds:
Velocity distribution functioni luniversal
![Page 20: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/20.jpg)
Universal Linewidth – Size Relation in l l l d ( ’ )Molecular Clouds (Larson’s Law)
Sound speed
Molecular cloud turbulence is supersonic
![Page 21: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/21.jpg)
Universal Stellar Mass FunctionUniversal Stellar Mass Function
![Page 22: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/22.jpg)
Turbulent Fragmentation Paradigm ( d dl d )(Padoan & Nordlund 2002)
• Supersonic turbulence • Young stars in Taurusinduces large compressions in the gas– Origin of core mass
function• Regions of high density
collapse to form stars• Hypothesis: statistics of
supersonic turbulence governg– statistics of star masses
and – Star formation rate
![Page 23: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/23.jpg)
Core Mass DistributionCore Mass Distribution
Enoch et al. (2007)
![Page 24: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/24.jpg)
Turbulence in a Box: Dissipation RatesLemaster & Stone (2008)
10243 gas dynamics 10243 MHD
Mach 7 Mach 6.9
Turbulence decays on a crossing time unless drivenDissipation rate converges by 643 for HD, but not until 5123 for MHDVery high resolution needed to measure inertial range slopes
![Page 25: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/25.jpg)
Column Density MapsLemaster & Stone (2008)
10243 gas dynamics 10243 MHD
![Page 26: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/26.jpg)
Turbulent Cascade a la h d lRichardson‐Kolmogorov
![Page 27: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/27.jpg)
![Page 28: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/28.jpg)
Turbulence in a Box: Scaling RelationsKritsuk et al. 2006, 2007, 2008, 2009
20483 gas dynamics 10243 MHDg y
Supersonic - Ms=6 Supersonic - Ms=10, super-Alfvenic MA=3
![Page 29: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/29.jpg)
![Page 30: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/30.jpg)
![Page 31: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/31.jpg)
![Page 32: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/32.jpg)
Forms starsForms stars
![Page 33: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/33.jpg)
David Collins PhD thesis (UCSD, 2009)first self‐gravitating AMR MHD sim of turbulent fragmentation
1283 root grid4 l l f fi t
ENZO-MHD code
4 levels of refinement
![Page 34: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/34.jpg)
Movie without AMR grids
Movie with AMR grids
![Page 35: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/35.jpg)
Effect of Self‐Gravity on PDFEffect of Self Gravity on PDF
![Page 36: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/36.jpg)
Core Mass Function:Comparison with Data
![Page 37: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/37.jpg)
CNZeeman
OHZeeman
![Page 38: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/38.jpg)
CNCNZeeman
OHZeeman
![Page 39: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/39.jpg)
Star Formation EfficiencyStar Formation EfficiencyFreefall time in units of the depletion time, measured by various tracers
Klessen, Krumholz & Heitsch (2009)
![Page 40: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/40.jpg)
Krumholz & McKee (2005) TheoryKrumholz & McKee (2005) Theory
• Assume onlyAssume only gravitationally bound regions of turbulent 25 2
1 <=RDσα
flow collapse to form stars 3/
2
2/131 ∝=
<=
RGM
DD
vir
σσ
α
• Assume turbulence obeys Larson’s law
f h l d scalewithdecreases
231
⇒∝∴ Rvir
DD
α
• Fraction of the cloud at or near the sonic scale will form stars
scalewith decreases⇒
will form stars
![Page 41: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/41.jpg)
Comparison with KM05Comparison with KM05
Simulations(1283 base grid)
KM05 predictions
![Page 42: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/42.jpg)
Assessment: Galactic Star Formation• Turbulent star formation has displaced magnetic star formation paradigm because– Zeeman measurements which show cores are mildly supercritical
Provides a natural explanation for origin of cloud cores– Provides a natural explanation for origin of cloud cores that agrees with observations
– Provides a natural explanation for low star formation efficiencies
• Progress simulating TSF has been paced by growth in computing power and availability of stable super‐Alfvenic MHD algorithms
P li i AMR l l k i i b h• Preliminary AMR results look promising, but much more work is required to critically test predictions
![Page 43: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/43.jpg)
Formation and Evolution of Disk lGalaxies
• Stellar structureStellar structure– Bulge, disk, halo
• Kinematics– stars, gas
• Tully‐Fisher relationy
• Gas content
• Stellar agesStellar ages
• Role of mergers on disk formation and
M101
destruction
![Page 44: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/44.jpg)
Theoretical NotionsTheoretical Notions• Bottom‐up structure formation (Davis et al. 1985)
id l i i f l ( ll &• Tidal torque origin of angular momentum (Fall & Efstathiou 1980; Fall 1983)
Di i ti l ll f b d t ll di k• Dissipational collapse of baryons and stellar disk formation via fragmentation (White and Rees 1978)
• Destruction of disks by major mergers (Toomre &• Destruction of disks by major mergers (Toomre & Toomre 1972; Barnes & Hernquist 1996)
• Secular processes (gas accretion galactic dynamics)• Secular processes (gas accretion, galactic dynamics) reshape galaxy at late times (e.g., Valenzuela & Klypin 2003)yp )
Galaxy formation is continuous, ongoing process and history dependent
![Page 45: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/45.jpg)
Formation of Disk Galaxies: i l i d ( hi & )Conventional Wisdom (White & Rees 1978)
Insert cartoon here
Baugh (2006)
![Page 46: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/46.jpg)
Early Numerical Experiments:Abject Failure
• Poor force resolution Catastrophic loss ofPoor force resolution Catastrophic loss of baryonic angular momentum tiny disks (Navarro & White 1994)(Navarro & White 1994)
• Lack of SN feedback star formation rate too high (White & Frenk 1991 Balogh et al 2001)high (White & Frenk 1991, Balogh et al. 2001)
• Combined effects yielded compact disk l i hi h di d i h T ll Fi hgalaxies which disagreed with Tully‐Fisher
relation (Navarro & Steinmetz 2000; Eke et al. 2001)2001)
![Page 47: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/47.jpg)
Angular Momentum LossNavarro, Frenk & White (1995)
Cooling but no feedback
![Page 48: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/48.jpg)
Possible Reasons for Angular Moment Loss in Disk Galaxies
• Dynamical friction on clumpy gas distributionDynamical friction on clumpy gas distribution (Navarro & White 1996)
• Gravitational torques in gaseous spiral arms• Gravitational torques in gaseous spiral arms (Lynden‐Bell & Kalnajs 1971)
A ifi i l i i h / ld SPH i f• Artificial viscosity at hot/cold SPH interfaces (Okamoto 2006)
• Torques from “grainy” dark matter halos (Kaufmann 2007)
![Page 49: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/49.jpg)
The Overcooling ProblemWhit & F k 1991 B l h t l 2001White & Frenk 1991, Balogh et al. 2001
• Simulations with• Simulations with radiative cooling but no no star formation and feedback produce too much cool gas relative to observations
• This problem led to lmany mostly
unsuccessful attempts to model SF+FBto model SF+FB
![Page 50: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/50.jpg)
I‐band Tully‐Fisher RelationNavarro & Steinmetz (2000)
• N‐body/SPH simulationsN body/SPH simulations of GF with SF/FB– N=32,000 particles
– ε=1 kpc
• I‐band Tully‐Fisher too many
relation sloperecovered, but not
li ti
too manystars
AM loss
normalization
• Due to excessively compact DM halos andcompact DM halos and high mass/light ratio
![Page 51: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/51.jpg)
Other Challenges to CDM: The “Missing Satellite” Problem
1414Cluster 5 10 M×
2 Mpc
12Galaxy 2 10 M×
300 kpc
B. Moore et al. (1999)
![Page 52: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/52.jpg)
Yet Another Challenges to CDM: “Galaxy Downsizing”
• Galaxies with lessGalaxies with less massive stellar component have younger stellar populations (Cowie et l 1996 M A th tal. 1996; MacArthur et al. 2004)
• Contrary to naieve• Contrary to naieve interpretation of hierarchical modelhierarchical model
MacArthur et al. (2004)
![Page 53: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/53.jpg)
Slide courtesy F. Governato
![Page 54: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/54.jpg)
Achieving Agreement with Observations ( )(Governato et al. 2007, 2008; Zavala et al. 2008)
• Improved star formation + FB recipe– More astrophysically motivated
– Calibrated with data
• Substantially better mass and force resolution– Nvir > 106Nvir 10
– εsoft << disk scale length, scale height (~300 pc)
![Page 55: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/55.jpg)
Star formation/feedback recipesC & O t ik (1992) K t W i b H i t (1996) Y t l (1997)Cen & Ostriker (1992); Katz, Weinberg, Hernquist (1996), Yepes et al. (1997),
Springel & Hernquist (2003), Kravtsov (2003), Stinson et al. (2006)
d it• deposit_energy– Locally as thermal energy radiated away
forall particles) SPHor (cells energy radiated away
– Locally as kinetic energy escape galaxythen
if
i l
.true.} criteria ofset { =
– In neighborhood region as thermal energy still radiated away2MEd it
body-Nevolve_as_ r_particlecreate_sta
&& radiated away
– radiative cooling suppressed in region for
endifc2
SFMEergy deposit_en ∝
some time Δt Sedov blast wave
![Page 56: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/56.jpg)
Slide courtesy F. Governato
![Page 57: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/57.jpg)
Slide courtesy F. Governato
![Page 58: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/58.jpg)
Slide courtesy F. Governato
![Page 59: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/59.jpg)
Slide courtesy F. Governato
![Page 60: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/60.jpg)
Calibrating Star Formation/Feedback RecipesS i l (2000)Springel (2000)
• Build isolated galaxy d l himodel matching
observations (B/D/H)
• Calibrate SFE andCalibrate SFE and feedback parameters to Kennicutt law
![Page 61: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/61.jpg)
Disk Galaxies from CosmoSimsGovernato et al. (2007)
DWF1
MW1mass
GAL1
![Page 62: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/62.jpg)
Observable PropertiesObservable Properties
![Page 63: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/63.jpg)
Missing Satellite ProblemMissing Satellite Problem
Effect of feedback Effect of resolution
M31
MWM31
MWM31
![Page 64: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/64.jpg)
Q: Do Major Mergers Destroy Galaxy Disks f All Ti ?for All Time?
A: Not Necessarily
Governato et al. (2008)
![Page 65: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/65.jpg)
![Page 66: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/66.jpg)
![Page 67: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/67.jpg)
Assessment: Disk Galaxy Formation• Tremendous progress in last 5 years
• Conventional Wisdom is wrong: stellar disks gform and reform even after major mergers– Primarily from cold flow accretionPrimarily from cold flow accretion
– Secondarily from hot flow accretion
• Models agree quite well now with observations• Models agree quite well now with observations (structure, kinematics, populations)
Require quite high resolution and SN feedback– Require quite high resolution and SN feedback implemented in a way that suppresses SF for SNR cooling timecooling time
– Missing satellite problem largely goes away
![Page 68: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/68.jpg)
Open IssuesOpen Issues
• LF of satellite galaxiesLF of satellite galaxies
• origin of Morphology‐Density relation (Dressler)(Dressler)
• resolving bulge formation/evolution
• dynamical erasure of DM cusps
![Page 69: Computational Astrophysics versus the Big Questions.ppt · Computational Astrophysics versus the Big Questions: An Assessment ... (White & Frenk 1991, Balogh et al. 2001) • Combined](https://reader033.vdocument.in/reader033/viewer/2022042020/5e776812973076421615d74f/html5/thumbnails/69.jpg)
Lessons Learned from these Two Examples
• “the role of simulation is insight notthe role of simulation is insight, not numbers” –Hamming
• “there is no free lunch at the table of• there is no free lunch at the table of computational physics” –Norman age 25
“ b i h h i d• “…..but, with correct physics, adequate algorithms, and sufficient computer power to
l h l l l hresolve the relevant scales, only then may we be in a position to obtain the insights we seek, d l hi ” N 55and learn something new” –Norman age 55