All Figures

Figure 1: The temporally and spatially averaged temperature structure of the 3D solar surface convection simulation (solid line) used for the 3D spectral line formation. The spatial averaging has been performed over surfaces of the same continuum optical depths at 500 nm. Note that the actual 3D simulation extends to much greater optical depths than shown here. Also shown are the temperature structures for the 1D Holweger-Müller (1974) semi-empirical solar atmosphere (dashed line) and the 1D MARCS (Asplund et al. 1997) theoretical solar atmosphere (dashed-dotted line).

Open with DEXTER


In the text

Figure 2: Grotrian term-diagram of the employed 23-level O model atom with 22 bound states of O  I and the ground state of O  II. The forbidden [O  I] 630.0 and 636.3 nm lines originate from the 2p4 3P ground state, while the O  I 777 nm triplet is between the high-excitation 3s 5So and 3p 5P levels. The 43 bound-bound transitions are drawn. All O  I levels are connected by bound-free transitions with the ground state of O  II but are not shown for clarity.

Open with DEXTER


In the text

Figure 3: The granulation pattern in one of the solar snapshots seen in disk-center continuum intensity at 777 nm ( upper left panel) and equivalent width of the O  I 777.41 nm line in LTE ( upper right panel) and non-LTE ( lower left panel); the equivalent width images have the same relative intensity scale to emphasize the overall difference in line strengths. Also shown is the ratio of the non-LTE and LTE equivalent widths ( lower right panel).

Open with DEXTER


In the text

Figure 4: Predicted intensity () equivalent widths of the [O  I] 630.03 nm ( left panel) and O  I 777.41 nm ( right panel) lines across the granulation pattern in LTE ( upper panel) and non-LTE ( middle panel). Both cases are computed for log  for the two lines. The mean intensity equivalent widths are denoted by horizontal lines. Also shown are the ratios of the non-LTE and LTE equivalent widths ( lower panel), with the horizontal lines representing the average ratio for the two lines. As expected, the [O  I] line is in perfect LTE while the O  I line shows significant departures from LTE. Note that for clarity only a selection of the vertical atmospheric columns are shown.

Open with DEXTER


In the text

Figure 5: Upper panel: the temporally and spatially averaged LTE flux profile from the 3D model atmosphere (+) together with the observed profile (solid line) for the O  I 777.53 nm line. The theoretical profile has been computed with log  and convolved with a Gaussian corresponding to the instrumental resolution. Clearly the 3D LTE profile lacks effects of departures from LTE in the line formation. Middle panel: the 3D non-LTE (solid line) and the 3D LTE (dashed line) line profiles computed for one snapshot with MULTI3D (Botnen 1997; Asplund et al. 2003). The 3D non-LTE profile has been computed with log  while for the LTE case an abundance of log  was adopted. In terms of equivalent widths the two profiles have the same line strengths. Lower panel: the temporally and spatially averaged 3D profile taking into account departures from LTE (+) shown together with the observed profile (solid line). The 3D non-LTE profile has been obtained by multiplying the 3D LTE profile in the upper panel with the ratio of non-LTE and LTE profiles presented in the middle panel. While not a perfect substitute for the temporally averaged 3D non-LTE profile it nevertheless shows a very encouraging agreement with observations. The observed line profiles have been corrected for the solar gravitational redshift of 633 m s-1.

Open with DEXTER


In the text

Figure 6: The observed center-to-limb variations of the equivalent widths of the O  I 777 nm triplet are shown as bands with estimated internal errors. The solid lines denote the average of two snapshots of the theoretical 3D non-LTE results for the three lines, while the dashed lines represent the corresponding 3D LTE case. The here shown 3D LTE curves are very similar to the average of all 100 snapshots computed with the same 3D LTE line formation code used for the calculations of the [O  I] and OH lines. As two snapshots are insufficient to yield an accurate abundance estimate, the 3D non-LTE results have been interpolated to log  to overlap with the observational estimates to high-light the excellent agreement in the center-to-limb variation in 3D non-LTE. In all cases, the LTE curves correspond to a 0.2 dex higher abundance than the non-LTE cases. Clearly, LTE fails to reproduce the observational evidence, while the non-LTE calculations are in good agreement.

Open with DEXTER


In the text

	Figure 7: The derived solar oxygen abundance (filled circles) from OH vibration-rotation lines using the 3D hydrodynamical time-dependent simulation of the solar atmosphere (Asplund et al. 2000b) as a function of wavelength, lower level excitation potential and line strength (in pm). The solid lines denote least-square-fits giving equal weights to all lines.
Open with DEXTER


In the text

	Figure 8: Same as Fig. 7 but using the Holweger-Müller (1974) semi-empirical model atmosphere and a microturbulence of  km s-1 (filled circles with solid lines denoting least-square-fits). The open circles and dashed lines correspond to a larger microturbulence (  km s-1).
Open with DEXTER


In the text

	Figure 9: Same as Fig. 7 but using a theoretical MARCS model atmosphere (Asplund et al. 1997) and a microturbulence of  km s-1 (filled circles with solid lines denoting least-square-fits). The open circles and dashed lines correspond to a larger microturbulence (  km s-1).
Open with DEXTER


In the text

Figure 10: The derived solar oxygen abundance from OH pure rotational lines using a 3D hydrodynamical solar model atmosphere as a function of wavelength, lower level excitation potential and line strength (in pm). The solid lines denote the least-square-fits. The rotational lines are more sensitive to the adopted microturbulence than the vibration-rotation lines since the stronger lines are partly saturated. The increased scatter for the weakest lines (which correspond to the highest excitation lines) is most likely due to increased observational difficulties in measuring the equivalent widths accurately.

Open with DEXTER


In the text

	Figure 11: Same as Fig. 10 but using the Holweger-Müller model atmosphere. The filled circles correspond to a microturbulence of  km s-1 while the open circles are for  km s-1 with the solid and dashed lines, respectively, denoting the least-square-fits.
Open with DEXTER


In the text

	Figure 12: Same as Fig. 10 but using a MARCS model atmosphere. The filled circles correspond to a microturbulence of  km s-1 while the open circles are for  km s-1 with the solid and dashed lines, respectively, denoting the least-square-fits.
Open with DEXTER


In the text