Astron. Astrophys. 351, 368-372 (1999)

2. The flare atmospheric model

So far, several atmospheric models for solar WLFs have been proposed. They can be divided into two groups. The first group includes models with a hot and condensed chromosphere (Avrett et al. 1986) or a chromospheric condensation (Gan et al. 1992). The continuum emission from these models originates mainly from the recombination of hydrogen atoms in the chromosphere and thus shows a distinguishable Balmer or Paschen jump. In comparison, models in the second group exhibit a heated upper photosphere (e.g., Avrett et al. 1986; Mauas et al. 1990; Ding et al. 1994). The continuum emission is then mainly due to the negative hydrogen emission in the upper photosphere. No Balmer or Paschen jump is present in this second case. The above two groups of models correspond to Types I and II WLFs respectively (Machado et al. 1986; Fang & Ding 1995).

For the event of 1974 October 11 discussed here, Fang et al. (1995) stated that it might belong to a Type I WLF, since its radio burst coincided with the maximum of continuum emission and higher Balmer lines appeared. However, the fact that the Balmer jump was very weak implies that it was a complex event showing also some features of a Type II WLF. Simply, we cannot propose a very hot and condensed chromosphere or a very strong chromospheric condensation to account for the continuum enhancement because it produces a significant Balmer jump that is inconsistent with observations. Therefore, to obtain an atmospheric model for this event, we put our effort mainly in adjusting the temperature structure around the TMR and the upper photosphere, where the K₁ and visible continuum are formed.

Adopting a non-LTE code similar to that used in Ding et al. (1994), we have computed various atmospheric models and obtained by trial and error a model responsible for the flare emission at 03:29:17 UT, the time when the K₁ and the visible continuum reached their maximum, and a model corresponding to the time of 03:34:21 UT when the K₁ and continuum intensities had decreased substantially (see Figs. 2 and 3 in Fang et al. 1995). Fig. 1 shows the temperature structures for these two models, labeled as A and B, together with the F1 and F2 flare models of Machado et al. (1980). The most interesting thing is, as expected, a very hot TMR in model A. The minimum temperature reaches a value as high as K, in comparison with the highest value of K in previous flare models (the F3 model in Avrett et al. 1986). Besides, model A also shows an increased temperature in the upper photosphere relative to model B, which results in a continuum enhancement. Model A shows no more strange behavior than model F1 except for this very hot temperature structure in the TMR.

Fig. 1. Temperature distributions for the flare atmospheric models A (solid line) and B (dashed line), corresponding to the times of 03:29:17 and 03:34:21 UT, respectively. The flare models F1 and F2 (dotted lines) of Machado et al. (1980) are also plotted for reference. The coronal column mass densities, , are 3.46 10-3 g cm-2 for model F2 and 3.14 10-4 g cm-2 for other models

The Ca II K line profiles computed from models A and B are plotted in Fig. 2, together with the continuum intensities at Å. They match the observations fairly well. Since we have not computed the higher Balmer lines (only four bound levels are considered for the hydrogen atom in the present code), there is probably an ununiqueness in the chromospheric structure of these models. However, this does not affect the following discussions on the energy balance in lower layers, provided that the chromosphere is not extremely perturbed, as evidenced in this event.

Fig. 2. Ca II K line profiles computed from models A (solid line) and B (dashed line). The profile from model F2 (dotted line) is also given for reference. A Gaussian macro-velocity of 20 km s-1 has been used to convolve the profiles. The horizontal lines represent the corresponding continuum intensities at Å

© European Southern Observatory (ESO) 1999

Online publication: November 2, 1999
helpdesk.link@springer.de