58
Simulating Solar Convection QuickTime™ and a GIF decompressor are needed to see this picture. b Stein - MSU vid Benson - MSU ke Nordlund - Copenhagen Univ. ts Carlsson - Oslo Univ. mulated Emergent Intensity

Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

  • View
    216

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Simulating Solar Convection

QuickTime™ and aGIF decompressor

are needed to see this picture.

Bob Stein - MSUDavid Benson - MSUAake Nordlund - Copenhagen Univ.Mats Carlsson - Oslo Univ.

Simulated Emergent Intensity

Page 2: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

METHOD• Solve conservation equations for:

mass, momentum, internal energy & induction equation

• LTE non-gray radiation transfer

• Realistic tabular EOS and opacities

No free parameters (except for resolution & diffusion model).

Page 3: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Conservation Equations

∂ρ∂t

= −∇ • ρu

∂ρui

∂t= −

∂x j

ρuiu j + Pδij + ρυ∂ui

∂x j

+∂u j

∂x i

⎝ ⎜ ⎜

⎠ ⎟ ⎟

⎣ ⎢ ⎢

⎦ ⎥ ⎥+ ρgi + J × B( )i

∂ρe∂t

= −∇ • ρeu− P∇ • u+ ρν∂ui

∂x j

+∂u j

∂x i

⎝ ⎜ ⎜

⎠ ⎟ ⎟

2

+ηJ 2 +Qrad

∂B∂t

= −∇ × E, E = −u× B +ηJ, J = ∇ × B /μ0

Mass

Momentum

Energy

Magnetic Flux

Page 4: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Simulation Domain

48 Mm

48 M

m

20 M

m

500 x 500 x 500 -> 2000 x 2000 x 500

Page 5: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Variables

Page 6: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Spatial Derivatives

Spatial differencing– 6th-order finite difference, non-uniform mesh

∂U∂x

⎝ ⎜

⎠ ⎟j−1/ 2

=

a U j( ) −U j −1( )[ ]

+b U j +1( ) −U j − 2( )[ ]

+c U j + 2( ) −U j − 3( )[ ]

⎜ ⎜ ⎜ ⎜

⎟ ⎟ ⎟ ⎟J j( )

c = (-1.+(3.**5-3.)/(3.**3-3.))/(5.**5-5.-5.*(3.**5-3))b = (-1.-120.*c)/24., a = (1.-3.*b-5.*c)

Page 7: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Time Advance

Time advancement– 3rd order Runga-Kutta

∂U∂t

=α i

∂U

∂t,

∂U

∂t=∂U

∂t+ f U( ),

U =U + β i

∂U

∂tdt

α = 0,−0.64,−1.3[ ], β = 0.46,0.92,0.39[ ]

For i=1,3 do

Page 8: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Radiation Heating/Cooling

• LTE• Non-gray, 4 bin multi-group• Formal Solution

Calculate J - B by integrating Feautrier equations along one vertical and 4 slanted rays through each grid point on the surface.

• Produces low entropy plasma whose buoyancy work drives convection

Qrad = 4π κ λλ

∫ (Jλ − Sλ )dλ

Page 9: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Solve Feautrier equations along rays through each grid point at

the surfaced2Pdτλ

2 =Pλ −Bλ

Pλ =12

[I λ(Ω)+Iλ(−Ω)]

Page 10: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Actually solve for q = P - B

qλ =Pλ −Bλ

d2qλdτλ

2 =qλ −d2Bλdτλ

2

Page 11: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Simplifications• Only 5 rays• 4 Multi-group opacity bins• Assume L C

Page 12: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

5 Rays Through Each Surface Grid Point

μ=cosθ=1,1/3, wμ =1/4, 3/4, ϕ rotates15oeachtimestep

Interpolate source function to rays at each height

φ€

Θ

Page 13: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Opacity is binned, according to its magnitude, into 4 bins.

Page 14: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Line opacities are assumed proportional to the continuum opacity

Weight = number of wavelengths in bin

κ i =10i κ0, i=0(continuum),2,3,4(strongestlines)

wi = wλ jj(i )∑ , j(i) =wavelengthsλ j inbini

Bi = Bλ jj(i )∑ wλ j

⎝ ⎜ ⎜

⎠ ⎟ ⎟ wi

Page 15: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Solve Transfer Equation for each bin i

qi = Pi − Bi

d2qi

dτ i2

= qi −d2Bi

dτ i2

Page 16: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Finite Difference Equationqj−1

1τj −τ j−1

2τ j+1 −τj−1

⎝ ⎜ ⎜

⎠ ⎟ ⎟

−qj 1+1

τj −τj−1

+1

τ j+1 −τj

⎝ ⎜ ⎜

⎠ ⎟ ⎟

2τj+1 −τ j−1

⎣ ⎢ ⎢

⎦ ⎥ ⎥

+qj+1

1τj+1 −τ j

2τj+1 −τ j−1

⎝ ⎜ ⎜

⎠ ⎟ ⎟ =Sj−1

1τj −τj−1

2τj+1 −τ j−1

⎝ ⎜ ⎜

⎠ ⎟ ⎟

−Sj1

τj −τj−1

+1

τ j+1 −τj

⎝ ⎜ ⎜

⎠ ⎟ ⎟

2τj+1 −τ j−1

⎣ ⎢ ⎢

⎦ ⎥ ⎥ +Sj+1

1τ j+1 −τj

2τ j+1 −τj−1

⎝ ⎜ ⎜

⎠ ⎟ ⎟

Ajqj−1 +Bjqj +C jqj+1 =Dj

• Problem: at small optical depth the 1 is lost re 1/2 in B

• Solution: store the value -1, (the sum of the elements in a

row) and calculate B = - (1+A+B)

Page 17: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Advantage• Wavelengths with same (z) are grouped together, so

integral over and sum over commute

κλ (J λ −Bλ )dλ∫ λ = κλ j

j (i)∑

i∑ (J λ j −Bλ j )wλ j

= κλ jj (i)∑

i∑ Lτλ j

(Bλ j )wλ j

(J λ −Bλ) =Lτλ (Bλ ) =dμμ0

1

∫ eτλ / μ dte−t /μ

0

∫ Bλ (t)−Bλ

κλ jj (i)∑

i∑ Lτλ j (Bλ j )wλ j ≅ κ i

i∑ Lτi ( Bλ j

j (i)∑ wλ j )

≡ κ ii∑ Lτi (Bi )wi ≡ κ i

i∑ (J i −Bi)wi

Page 18: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Interpolate q=P-B from slanted grid back to Cartesian grid

φ€

Θ

Page 19: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Radiative Heating/Cooling

Qrad =4πρ κ ii∑

Ω∑ qiwiwΩ

Page 20: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Energy Fluxes

ionization energy 3X larger energy than thermal

Page 21: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Equation of State

• Tabular EOS includes ionization, excitationH, He, H2, other abundant elements

Page 22: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Diffusion stabilizes scheme

• Spreads shocks

• Damps small scale wiggles

ν =amax −∇ • u,0( )

ν =b csound + cAlfven( ) + c urms

Page 23: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Boundary Conditions• Current: ghost zones loaded by extrapolation

– Density, top hydrostatic, bottom logarithmic

– Velocity, symmetric

– Energy (per unit mass), top = slowly evolving average

– Magnetic (Electric field), top -> potential, bottom -> fixed value in inflows, damped in outflows

• Future: ghost zones loaded from characteristics normal to boundary(Poinsot & Lele, JCP, 101, 104-129, 1992)modified for real gases

Page 24: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Observables

Page 25: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Gra

nula

tion

Page 26: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Emergent Intensity Distribution

Page 27: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Line Profiles

Line profile without velocities. Line profile with velocities.

simulation

observed

Page 28: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Convection produces line shifts, changes in line widths. No microturbulence, macroturbulence.

Average profile is combination of lines of different shifts & widths.

average profile

Page 29: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Velocity spectrum, (kP(k))1/2

*

* ***

*

MDI doppler (Hathaway) TRACE

correlation tracking (Shine)

MDI correlation tracking (Shine)

3-D simulations (Stein & Nordlund)

Page 30: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Oscillation modes

Simulation MDI Observations

Page 31: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Local Helioseismologyuses wave travel times through the atmosphere

(by former grad. Student Dali Georgobiani)

Dark line is theoretical wave travel time.

Page 32: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

P-Modes Excitedby PdV work

Triangles = simulation, Squares = observations (l=0-3)

Excitation decreases at lowfrequencies because oscillationmode inertia increases andcompressibility (dV) decreases.

Excitation decreases at highfrequencies because convectivepressure fluctuations have longperiods.

(by former grad. studentsDali Georgobiani & Regner Trampedach)

Page 33: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

P-Mode Excitation

Page 34: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Solar Magneto-Convection

Page 35: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Mean AtmosphereTemperature, Density and Pressure

(105 dynes/cm2)

(10-7 gm/cm2)

(K)

Page 36: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Mean AtmosphereIonization of He, He I and He II

Page 37: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Inhomogeneous T (see only cool gas), & Pturb

Raise atmosphere One scale height

3D atmosphere not same as 1D atmosphere

Page 38: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Never See Hot Gas

Page 39: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Granule ~ Fountain

Page 40: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Granules:diverging warm

upflow at center,

converging cool, turbulent downflows at

edges

Red=diverging flowBlue =converging flowGreen=vorticity

Page 41: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Fluid Parcels

reaching the

surface Radiate away their

Energy and

Entropy

Z

SE

Page 42: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

QuickTime™ and aPhoto - JPEG decompressor

are needed to see this picture.

Page 43: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Magnetic Boundary Conditions

Magnetic structure depends on boundary conditions

• Bottom either:1) Inflows advect in horizontal field

or2) Magnetic field vertical

• Top: B tends toward potential

Page 44: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

QuickTime™ and aYUV420 codec decompressor

are needed to see this picture.

B Swept to Cell Boundaries

Page 45: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Magnetic Field Lines - fed horizontally

Page 46: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Flux Emergence & Disappearance1 2

3 4

Emerging flux

Disappearing flux

Page 47: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity
Page 48: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Magnetic Flux Emergence

Magnetic field lines rise up through theatmosphere and open out to space

Page 49: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

G-band image & magnetic

field contours

(-.3,1,2 kG)

Page 50: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

G-band &

Magnetic Field

Contours: .5, 1, 1.5 kG (gray)20 G (red/green)

Page 51: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Magnetic Field & Velocity (@ surface)

Up Down

Page 52: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

G-band Bright Points = large B, but some large B dark

Page 53: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

QuickTime™ and aGIF decompressor

are needed to see this picture.

G-bandimages from simulation

at disk center & towards limb

(by Norwegian collaboratorMats Carlsson)

Notice:Hilly appearance of granulesBright points, where magnetic field is strongStriated bright walls of granules, when looking through magnetic fieldDark micropore, where especially large magnetic flux

Page 54: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Comparison with observationsSimulation, mu=0.6 Observation, mu=0.63

Page 55: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Height where tau=1

Page 56: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Magnetic concentrations:

cool, low ρlow opacity.

Towards limb,radiation

emerges from hot granule

walls behind.

On optical depth scale,

magneticconcentrations

are hot, contrast

increases with opacity

Page 57: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

Magnetic Field &Velocity

High velocity sheets at

edges of flux concentration

Page 58: Simulating Solar Convection Bob Stein - MSU David Benson - MSU Aake Nordlund - Copenhagen Univ. Mats Carlsson - Oslo Univ. Simulated Emergent Intensity

The End