Results

Figure 47.1 shows the resulting star formation and metallicity distribution for the thick (left panel) and thin (right panel) disks. In the case of the thick disk we show several possible models computed with different assumptions regarding (a) its mass with respect to that of the thin disk, (b) the late evolution of the star formation (truncated, threshold, continuous), and (c) the infall timescale (0.4 or 0.8 Gyr). Given the uncertainty in all these quantities, we check the range of model parameters that could still explain the observed thick-disk metallicity distribution. In the next figures we will show the predictions corresponding to the model shown by a thick solid line in Figure 47.1 (which has a timescale of 0.8 Gyr and a normalization with respect to the thin disk of ~10% - see details in Chiappini et al. (2008, in preparation). In what follows we will see that such a model naturally accounts for the observed abundance patterns in the thick disk. The chemical-evolution models for the thick and thin disks were computed assuming the same initial mass function and stellar yields for both components. For the stellar yields we adopted the Woosley & Weaver (1995) prescriptions for SNII, model W7 from Nomoto et al. (1997) for SN Ia, and the yields of van den Hoek & Groenewegen (1997) for low- and intermediate-mass stars. The yields for massive stars and SN Ia were modified according to the prescriptions given in Francois et al. (2004) (model A). Moreover, we also show the results obtained from models computed with the same prescriptions as detailed above, except for the massive stars for which we adopted the new stellar yields of Nomoto et al. (2006) (model B).

2.1 The a-elements

Figure 47.2 shows the predicted patterns for the a-elements for the thick (right panel) and thin disks (left panel) for models A (solid line) and B (dotted line) described above. Figure 47.3 (left panel) shows data for the thick- and thin-disk stars plotted together in an [X / O] versus [O/H] diagram, on which the differences, if they are indeed caused by different formation timescales, should be clearer. The solid curves correspond to the thick-disk predictions; the dashed ones to the predictions for the thin disk (for model A). It is clearly seen that, despite some absolute shifts (which will depend on the stellar yields adopted), the relative behavior of the two curves is in very good agreement with the data, namely (a) Mg/O relationships in the thick and thin disks trace each other; (b) larger differences are seen in the Fe/O ratio, as expected; (c) the a-elements do nothave the same overabundances, because of the different amounts coming from massive stars but also due to the contribution of SN Ia for S, Ca, and Si; and (d) the variations of the shifts between the thick-and thin-disk patterns with metallicity are in good agreement with those predicted by the models. A detailed discussion of these results and their implications for the stellar yields of particular elements can be found in Chiappini et al. (2008, in preparation).

2.2 The iron-peak elements

Figure 47.3 (right panel) shows a comparison between our predictions (model A) for the variation of [X/Fe] versus metallicity, for X = Sc, Co, V, Ni, Cr, Cu, Mn,

e]

0.5

/F

ro

0

S

-0.5

e]

0.5

/F

0

Ö

-0.5

e]

0.5

/F

ro

0

O

-0.5

e]

0.5

/F

0

ä

-0.5

e]

0.5

/F

0

21

-0.5

e]

0.5

/F

0

F

-0.5

+

e]

0.5

/F

0

-0.5

e]

0.5

/F

0

f)

-0.5

e]

0.5

/F

m

0

O

-0.5

e]

0.5

/F

0

(/I

-0.5

e]

0.5

/F

0

U)

-0.5

e]

0.5

/F

0

H—I—I—I—I—I—I—I—I—I—I—I—I—H

H—I—I—I—I—I—I—I—I—I—I—I—I—H

H—I—I—I—I—I—I—I—I—I—I—I—I—H H—I—I—I—|—I—I—I—I—|—I—I—I—I-

Figure 47.2. Plots of [a/Fe] versus [Fe/H] for a selected dataset for which it is possible to assign the stars to either the thick or the thin disk with more than 70% probability. See Chiappini et al. (2008 in preparation for the choice of the datasets included in this figure. The data are from Brewer & Carney (2006), Bensby et al. (2003, 2004, 2005), Caffau et al. (2006), Prochaska et al. (2000), Reddy et al. (2003), Fuhrman (2004), Melendez et al. (2002), Nissen et al. (2004), and Reddy et al. (2006.) Left panel: thin disk. Right panel: thick disk. For explanation of models A (solid line) and B (large-dotted line) see the text. For color figures, see online version.

and Zn, for the thick (solid line) and thin (dashed line) disks. It can be seen that the two curves overlap for Co, V, Ni, Cr, and Zn. A small shift between the curves is obtained for Mn, Cu, and Sc. The data seem to behave in the same way, there being only little shifts for Sc, Mn, and, perhaps, Cu. The elements K, Sc, and V are odd-Z elements produced mainly by oxygen-burning (K in hydrostatic burning; Sc and V in explosive burning). The nucleosynthesis prescriptions for the last two elements are thus more uncertain. Here we adopted the prescriptions given in Francois et al. (2004), which assume that the stellar yields of the above elements are not metallicity-dependent. Although this seems to be the case for Cr and Ni, a certain dependence on metallicity is expected for Mn, Co, Cu, and Zn. For models including the metallicity dependencies of the latter elements, see Chiappini et al. (2008, in preparation).

Was this article helpful?

0 0

Post a comment