5 Discussion and conclusions

Using spectral scans taken co-temporarily and co-spatially with the 9-channel MSDP instrument in the H$\alpha$ line, we document four hours of the evolution of the NOAA 7912 active region on 19 October 1995. Utilizing Yohkoh/SXT images co-aligned carefully with the H$\alpha$ observations we relate the chromospheric features to the coronal ones.

In the quiet state of this reversed polarity AR the brightest, highly sheared X-ray loops (they were almost parallel to the filament) appeared above a system of strong dark chromospheric fibrils connecting the penumbra of the big leading spot and a bright chromospheric area. These fibrils did not show the characteristic flow pattern of arch filament systems, so we conclude that they did not belong to emerging flux. It is interesting to note that all the brightenings, except for the B2, occurred at the base of anchorage of the feet of the corresponding fibrils systems. The brightest X-ray loop was rooted in the brightest part of the chromospheric plage, in the vicinity of a local (positive) magnetic field maximum. Apart from the time of flare and jet event, this X-ray loop remained the brightest X-ray feature in the AR for several hours.

The studied B3.2 flare associated with the X-ray jet affected the entire AR. Strong up and downflows appeared at both sides of the big leading spot, far away from the flare. The remote fibrils appear to be connected by long sheared coronal loops. The important downflow of the material (V2) was associated with a minor brightening in the H$\alpha$ line centre, thus the impact of the matter very probably created some chromospheric heating. The flare footpoints were observed at both sides of a filament. This filament changed the most considerably, as expected, due to the flare. Inside the AR the filament separated into two parts, the one close to the flare footpoints appeared temporarily at a higher altitude after the flare. Similar behaviour of a filament after an X-flare was observed by [4, Dezso et al. (1980).] It is remarkable that even a minor B3.2 flare can have such effect. There were important velocities observed along the filament during and even half an hour after the flare. The two parts appeared dynamically different: one, located close to the big spot, was dominated by downflow, another - located close to the flare footpoints, by upflow. The south continuation of the filament disappeared, probably erupted, at the time of the flare, and re-formed only about three hours later. The fact, that the filament did not erupt inside the AR can be attributed to the stabilizing effect of the strong overlying loops, while outside of the AR such loops did not prevent the eruption. Such duality in the reaction of a filament to a flare was reported earlier by [10, Raadu et al. (1988)] and [16, van Driel-Gesztelyi et al. (1998b).] [10, Raadu et al. (1988)] related the activity of the filament to photospheric motion of pores corresponding to change in the small scale magnetic polarity pattern. Such behaviour could be explained with the new MHD models of prominences which show the direct relationship between prominence foot and the presence of small parasitic polarities in the filament channel [2, (Aulanier & Démoulin 1998;] [2, Aulanier et al. 1998 and 1999).] Any disturbance in the magnetic field pattern (cancellation, emerging flux) close to a foot could lead to eruption of the foot and more generally of the filament itself.

The presented series of high-resolution spectroscopic H$\alpha$ observations demonstrated how dynamical is the chromosphere and even a minor flare can lead to important morphological and kinematic changes.

Acknowledgements

This work was supported by the French-Polish grant of "Le Conseiller pour la Science et la Technologie" and KBN, No. 6039 (1996). MSDP observations were obtained in the framework of the International time offered by the CCI of the Canarian Observatories and supported by the European Commission through the Access to the Large-Scale Facility "Activity of the Human Capital and Mobility Programme". Thanks are due to J. Staiger, C. Coutard, R. Hellier, the technical team of the VTT Telescope and the Teide Observatory, for their efficient assistance during the campaign. We thank the Yohkoh Team and the YDAC at Mullard Space Science Lab. for the SXT data. The NSO/Kitt Peak magnetogram data used in the paper were kindly provided us by Dr. Karen L. Harvey, the data are produced co-operatively by NSF/NOAO, NASA, GSFC, and NOAA/SEL. LvDG acknowledges the research grant OTKA T026165 and AKP 97-58 2,2. BR and PR were supported by KBN grants no. P03D.025.09 and 2P03D.005.15.