Solar models Composition, neutrinos & accretion Aldo Serenelli (MPA)

1. Solar models Composition, neutrinos & accretion Aldo Serenelli (MPA)

2. Outline Abundances from solar photosphere 3D-atmosphere model line formation Effects on helioseismology and neutrinos can n experiments tell something about Z? Accretion onto the Sun? Surface/core composition difference

3. Revision of solar abundances Solar atmosphere: convection  3-D models Improved atomic and molecular trans. prob. Relaxation of LTE assumption

4. Credit: N. Brummell Solar atmosphere 1D models do not capture basic structure of atmosphere Energy transported by convection, radiated away at the top Flow is turbulent Up- and downward flows asymmetric Not unique relation T vs. depth Magnetic fields

5. Credit: Bob Stein 3D models “Box in a star”: simultaneous solution of radiative transfer (RT) and hydrodynamic equations RT is “rudimentary”: frequency dependent opacities combined in a few (4 to 12) bins (10^5 in 1D) LTE  not good opacities (e.g Mg I, Ca I, Si I, Fe I) 1D models used as benchmark effects on 3D may be different However… it looks good

6. Credit: Bob Stein 3D models – testing the model Good handle on granulation topology timescales convective velocities intensity brightness contrast

7. 3D models – testing the model Line shapes: bisectors Asplund et al. 2000

8. Asplund et al. 2009 3D models – testing the model Limb darkening Credit: Matt Carlsson

9. Ludwig et al. 2010 3D models – testing the model Other models CO5BOLD (aka Paris group) Max Planck for Solar System Research (coming from solar MHD) Comparison of structure between models undergoing (slowly)

10. Asplund et al. 2009 3D models – line formation 3D background model atmosphere  detailed radiation transfer for line formation  determination of abundances

11. 3D models – line formation Formal requirement: 3D background model + 3D-RT and 3D-NLTE In practice 3D-RT + 3D-NLTE only for O Different combinations 3D-RT + 1D-NLTE (e.g. C) Multi 1D + 1D-NLTE (e.g. Fe) 1D + 1D-NLTE 1D + LTE Warning: e- and H collision rates (crucial for NLTE) missing for almost all elements, including C & N

12. 3D models – line formation Pros: identification of blends Asplund et al. 2009 But be aware of inconsistent treatment of lines in blends e.g. [OI] and Ni

13. 3D models – line formation Pros: consistency (after some massage), e.g. atomic and molecular (very T sensitive) indicators Asplund et al. 2009

14. Asplund et al. 2009 3D models – line formation Agreement with meteoritic abundances Si used as reference D=0.00 +- 0.05 dex

15. 3D models – last two words of warning Hydrogen (T sensitive) lines poorly reproduced Regardless of central values, small uncertainties (too optimistic?) Caffau et al. give systematically larger CNO abundances & uncertainties, x2 for C

16. Effect on solar models: Helioseismology Reduction in CNONe (30-40%) boundary in RCZ Reduction in Ne + Si, S, Fe (10%)  YS Sound speed Density Z/X= 0.0229 (GS98), 0.0178 (AGSS09)

17. Effect on solar models: Helioseismology Estimation of uncertainties in sound speed and density from MC simulations (5000 models) AGS05 Density profile: excellent example of correlated differences

18. Effect on solar models: Helioseismology Low degree modes (l=0, 1, 2, 3) from +4700 days BiSON Separation ratios Enhance effects in the core

19. Effect on solar models: Helioseismology

20. GS98 models AGS05 models Effect on solar models: Helioseismology me averaged over R < 0.2R8 Both compositions me = 0.723±0.003 Chaplin et al. (2007)

21. Effect on solar models: Helioseismology Christensen-Dalsgaard (2009)

22. Effect on solar models: neutrino fluxes Direct measurements of 7Be and 8B from Borexino and SNO

23. Effect on solar models: neutrino fluxes Gonzalez-Garcia et al. arxiv: 0910.4584 Global analysis of solar & terrestrial n data 3 flavor-mixing framework Basic constraints from pp-chains and CNO cicles Luminosity constraint (optional) Exhaustive discussion of importance of Borexino

24. Luminosity constraint No luminosity constraint Effect on solar models: neutrino fluxes No Borexino Borexino

25. GS98 AGS05 Effect on solar models: neutrino fluxes Gonzalez-Garcia et al. arxiv: 0910.4584 P(GS98) = 43% P(AGS05)= 20%

26. Effect on solar models: neutrino fluxes GS98AGS05AGS09 2= 5.2 (74%) 5.7 (68%) 5.05 (76%) Is comparison fair?

27. Additional motivation to measure CN fluxes Solar vs. solar twins (Melendez et al. 2009, Ramirez et al. 2009) same Teff, same gravity, same Fe abundance  same systematics  differential study

28. Volatiles Refractories Additional motivation to measure CN fluxes (V/R)8 ~ 0.05-0.08 dex > (V/R)twins IFevidence for accretion after planet formation solar interior  solar twins interior ≠ surface (eg AGS09)

29. “Solar comp.” Transition Twins composition Additional motivation to measure CN fluxes Accretion: schematic composition stratification (V/R) 0.05-0.10dex contrast between interior and surface in addition to overall Z contrast CN fluxes affected linearly

30. Conclusions Abundances 3D atmosphere models big step forward line formation still mostly 1D NLTE effects: sometimes, not in 3D atm. model inconsistent treatment of different elements and lines systematics not well understood different groups – different results (abundances and errors) Solar models: helioseismology nothing fits in low-Z models different constraints sensitive to different abundances Solar models: neutrinos current experiments do not discriminate between Z CN measurement needed test of accretion?