The asymmetries of chromospheric line profiles (Balmer series of hydrogen, CaII lines, ...) as well as soft X-ray lines (Ca XIX, Fe XXV, ...) are related to mass motions in the flare atmosphere. The most extended observational data are available for the line that is regarded as the most important tool for analysis. The observations reveal -line asymmetries of various kind. While the red asymmetry is present in most flares, the blue one was detected only in some of them and only at the very onset of the flare (Švestka 1976; Wülser 1987). The red asymmetry of appears either as a shift of the peak to longer wavelength (Ichimoto & Kurokawa 1984) or as a stronger red wing of the line (Švestka 1976). It is observed throughout the impulsive phase of the flare, reaches its maximum before the maximum of the line intensity and is temporally correlated with X-ray radiation (Ichimoto & Kurokawa 1984; Canfield et al. 1990). The blue asymmetry of is almost every time connected with central reversal (Canfield et al. 1990; Heinzel et al. 1994).
The velocities of the chromospheric plasma responsible for observed asymmetries are usually obtained using bisector technique. So Ichimoto & Kurokawa (1984) found downward motion velocities between and they report drop from 100 to within 30 s. Wülser & Marti (1989) get drop from 90 to within 12 s. Although the bisector technique is widely used, because of its straightforwardness, from the radiative transfer point of view it cannot be generally applied and as was shown by Heinzel et al. (1994) it may lead to misleading results.
The numerical simulations show that the heating of the chromosphere during a flare produces upward as well as downward motions of the chromospheric plasma (Somov et al. 1982; Fisher et al. 1985; Karlický & Hénoux 1992). These simulations predict an explosive evaporation with upward velocities up to , accompanied by a chromospheric condensation - a thin dense and cool layer propagating downward with velocity up to . Various observations based on bisector method seem to confirm this result (Zarro et al. 1988; Zarro & Canfield 1989; Wülser & Marti 1989; Canfield et al. 1990; Wülser et al. 1993).
Falchi et al. (1992) observed a flare in the , , , , CaII K and lines and their results show that the chromospheric condensation probably has a velocity gradient. From the observation of two flares in the , Ca II K, He , Na and other metallic lines Shoji & Kurokawa (1995) conclude that flare emission comes from two layers - one thin, fast-downward-moving and very turbulent, and another stationary and optically thick.
Fang et al. (1992) analysed red asymmetry of CaII K line for 12 solar flares using semi-empirical NLTE calculations. They found the observed asymmetry can be explained by downward motion above temperature minimum region with velocity . Using profiles for two flares observed by Wülser & Marti (1989) and Graeter (1990) Gan et al. (1993) constructed semi-empirical NLTE models with chromospheric condensation. They showed that the chromospheric condensation is responsible not only for red asymmetry of but also for blue asymmetry with central reversal.
A realistic determination of the velocity structure in the flaring atmosphere is extremely important for better understanding of the flare energetic processes. It is highly desirable to replace the simple bisector method by the NLTE dynamic model calculations and compare the synthetic profiles of a number of various lines with the observed ones.
In this paper we would like to present NLTE simulations of the flare atmosphere with prescribed run of the temperature and the velocity field, in order to show the influence of velocities on emergent hydrogen -line intensities. The results are compared with the bisector technique.
The NLTE method we used is described in Sect. 2 (click here), the models we performed the calculations for are described in Sect. 3 (click here) and in Sect. 4 (click here) we discuss the results.