Future of asteroseismology II

Jørgen Christensen-Dalsgaard

Institut for Fysik og Astronomi, Aarhus Universitet

Dansk AsteroSeismologisk Center

We need

• Better data

• Better models

Better data

• Better frequency precision (() < 0.1 Hz)

• Lower noise level to reach more modes

• Data on a broader variety of stars

• Identification of the modes (l, m)

• Better ‘classical’ observables (M, R, L, Teff, X, Z)

• g modes in the Sun to study the solar core

Frequency precision

Simply observe for longer

• Easy for heat-engine modes (() / tobs-1)

• Harder for stochastically excited modes (() / tobs

-1/2 for t > tlife)

Longer observations also improve detection of lower-amplitude modes

Observational strategies

• For very extended observations (weeks or months) we need dedicated instrumentation.

• Space observations in intensity? Discussed by HK.

• Helioseismology has shown the way: dedicated networks (BiSON, IRIS, TON) and

• GONG (Global Oscillation Network Group)

Hence we need ……

SONG: Stellar Oscillation Network Group

SONG proposal (the Aarhus dream):

• Network of small telescopes (60 cm equivalent)

• Very efficient and highly stabilized spectrograph

Science goals:

• Solar-like oscillations in relatively bright stars

• Search for low-mass extrasolar planets in close orbits

Possible distribution of sites

?

Asteroseismic capabilities

Planet-search capabilities

Better data

• Better frequency precision (() < 0.1 Hz)

• Lower noise level to reach more modes

• Data on a broader variety of stars

• Identification of the modes (l, m)

• Better ‘classical’ observables (M, R, L, Teff, X, Z)

• g modes in the Sun to study the solar core

Data on a broader variety of stars

• Multi-object spectrographs (but hard to ensure radial-velocity precision)

• Intensity observations of multiple stars from space (HK lecture)

Better data

• Better frequency precision (() < 0.1 Hz)

• Lower noise level to reach more modes

• Data on a broader variety of stars

• Identification of the modes (l, m)

• Better ‘classical’ observables (M, R, L, Teff, X, Z)

• g modes in the Sun to study the solar core

Mode identification

• For stochastically excited oscillators, use nearly complete spectrum, regular structure of frequencies

• For heat-engine oscillators, in general need independent information about mode geometry:

• Combine amplitudes and phases of observations with different techniques (intensity in different colours, intensity and radial velocity, etc.)

Doppler imaging

Tau Peg (Kennelly et al. 1998; ApJ 495, 440)

Doppler imaging

Tau Peg (Kennelly et al. 1998)

Major difficulty: Modelling of structure and oscillations of rapidly rotating star

Better data

• Better frequency precision (() < 0.1 Hz)

• Lower noise level to reach more modes

• Data on a broader variety of stars

• Identification of the modes (l, m)

• Better ‘classical’ observables (M, R, L, Teff, X, Z)

• g modes in the Sun to study the solar core

Better ‘classical’ observablesDirect observations:

• Magnitude

• Colours

• Spectra

With calibrations:

• Luminosity (needs distance, bolometric correction)

• Effective temperature (needs calibration)

• Composition (needs model atmosphere)

Solar abundance revisions are a reminder of the uncertainties in these analyses

Better data

• Better frequency precision (() < 0.1 Hz)

• Lower noise level to reach more modes

• Data on a broader variety of stars

• Identification of the modes (l, m)

• Better ‘classical’ observables (M, R, L, Teff, X, Z)

• g modes in the Sun to study the solar core

Well, not yet, after 30 years of intensive efforts

Better models of stellar evolution and oscillations

• Better numerical reliability, accuracy

• Better microphysics (equation of state, opacity, …)

• Better treatment of convection

• Better (i.e., some) treatment of energetics of oscillations

• Inclusion of effects of rotation, on structure and oscillations

• What about magnetic fields???

Use analysis of oscillation results to inspire improvements to the physics

Numerical treatment

• Are the evolution codes correct???? (Probably not)

• Is the numerical precision adequate? (Compared with the observational precision)

• How do we find out?

Detailed comparisons of results of independent codes.

Better microphysics

• Extremely complex problems in many-body atomic physics

• Coulomb interactions, excluded-volume effects, partial degeneracy, interaction with radiation ….

Some detailed testing using the Sun as a laboratory.

No relativistic effectsIncluding relativistic effects

Example: relativistic electrons in the Sun

Elliot & Kosovichev (1998; ApJ 500, L199)

Modelling stellar convection

• Mixing-length treatment (calibrated against the Sun)

• Detailed hydrodynamical simulations (for a range of stellar parameters)

• Simpler treatments, but calibrated against simulations

Note: treatment of convection and hydrodynamics of stellar atmospheres crucial for the abundance determinations, calibrations of photometric indices.

Simulation of convection in the Sun

Nordlund et al.

Effects of rotation on stellar structure

• Spherically symmetric component of centrifugal force in hydrostatic equilibrium: fairly simple

• Effects of circulation and instabilities: extremely hard

• Evolution of internal angular momentum: worse

Recall uniform slow rotation of solar interior

Meridional circulation

20 Msol on the ZAMS

Meynet

Circulation and associated instabilities lead to

• transport of elements

• transport of angular momentum

Effect of rapid rotation on oscillations

Analysis by Soufi et al. (1998; Astron. Astrophys. 334, 911)

1st order

2nd order

3rd order

Development of analysis techniques

• Fits to determine global parameters

• Must worry about possible multiple maxima in likelihood function: use Monte-Carlo techniques (e.g. genetic algorithm)

• Inversion based on just low-degree modes.

Examples of potential analyses

Tests based on artificial data with realistic (we hope) properties

• Properties of convective overshoot

• Structure of the stellar core

Base of convective envelope

Monteiro et al. (2000; MNRAS 316, 165)

Effect of He ionization

Signal from base of

convective envelope

Monteiro et al. (2000)

Inversion for core structure

Models: 1 M¯

(Mixed core) – (normal)

Degree l = 0 - 3

(Basu et al. 2002; ESA-SP 485, 249)

The future: stellar tachoclines??

NASA vision study. Launch 20??