Astron. Astrophys. 357, 767-776 (2000)

4. Model predictions

In September 1996 a SUMER observing sequence called OPAC was run which consisted of around 20 west-east scans of the east solar limb. Each scan involved 18 steps using the 1x120 arc sec slit. The C II and C III data used in the analysis described below are from this sequence. The theoretical atomic data used for the modelling of C II and C III populations follow from Paper I and are drawn from the Atomic Data and Analysis Structure (ADAS, see Summers 1993, 1999).

4.1. Fluxes

The observed spectral line fluxes for the C II  2s²2p ²P - 2s2p^2 2S line at 1037.020 Å and the C III  2s2p ³P₂ - 2p^2 3P₂ line at 1175.711 Å are shown in Fig. 5a and b respectively with the predicted fluxes (calculated via Eq.  27) overlayed. Model 1 fails completely for both C II and C III as expected as it does not recognise the extension of the transition region into the corona due to spicules and other structures. Models 2, 3 and 4 track the trend in both cases in broad terms but not in detail, failing most markedly at the limb and well off limb. On disk despite the averaging over the slit, the fluxes display a sensitivity to structure which is evident in the surface plots of Fig. 6. This sensitivity to column density is implied by Eq.  27 since is insensitive to optical depth for optical depths greater than  1.

Fig. 5a and b. Observed and model fluxes for the a C II  2s22p 2P - 2s2p2 2S line at 1037.020 Å and b the C III  2s2p 3P2 - 2p2 3P2 line at 1175.711 Å. The stars correspond to the observed fluxes and the solid lines are the model predictions numbered as above. Note that the visible limb is at 959.6 arc sec.

At heights of  970 arc sec and above the observed signal is orders of magnitude greater than that predicted. Since the scattered light calculations are expected to be in error by a maximum factor of this signal is possibly of solar origin. If this is so then the fact that the ratios at these heights indicate optical depths similar to that on disk, implies that the emitting structures here must be unresolved.

4.2. Ratios

The observed and model flux ratios of the (3/2-1/2)/(1/2-1/2) components of the C II  2s²2p ²P - 2s2p^2 2S multiplet and the (2-2)/(1-2) components of the C III  2s2p ³P - 2p^2 3P multiplet are shown in Fig. 7a and b respectively. The models here are much more effective since the linear dependence on column density is factored out (see Eq.  15). In the C III case the dip in the ratios at the limb is underestimated yet optical depths at this point are such that the ratio, in the absence of blending, is insensitive to optical depth. When blending is included, absorption of 2-2 line photons by the 1-1 line leads to a decrease in the ratio. The underestimation of the ratios at the limb suggests an underestimate in lower level population density of either the 2-2 line or the 1-1 line yet at this point the emergent intensities are overestimated suggesting an overestimate in the upper level column densities. This discrepancy is possibly due to a dependence of the absorption factor on homogeneity. The general insensitivity of the ratios themselves, and the relative success of the stratified model in predicting them suggests, however, that this dependence is weak. This is further supported by the insensitivity of to optical depth for optical depths greater than  1 (see Fig. 2b).

Fig. 6a and b. Surface plots of the a C II  2s22p 2P - 2s2p2 2P and b C III  2s2p 3P - 2p2 3P multiplets at  904 Å and 1175 Å respectively, showing total flux in the slit/raster plane. The slit dimension runs from bottom right to top left and the raster dimension runs from top right to bottom left.

Fig. 7a and b. Observed and model flux ratios of the a (3/2-1/2)/(1/2-1/2) components of the C II  2s22p 2P - 2s2p2 2S multiplet and b the (2-2)/(1-2) components of the C III  2s2p 3P - 2p2 3P multiplet. Note that in a the results of models 3 and 4 are almost identical.

© European Southern Observatory (ESO) 2000

Online publication: June 5, 2000
helpdesk.link@springer.de