Astron. Astrophys. 318, 791-796 (1997)

4. Discussion

The ROSAT PSPC spectra of Gem show two temperature components, at 2 MK and 12 MK. Previously another component at 40 MK (Singh et al. 1987) was obtained from EXOSAT observations (ROSAT is rather insensitive for gas at this temperature). The corona of Gem therefore seems to accommodate plasma at three characteristic temperatures. This situation has a Solar analogy: quiet corona (low temperature), active regions (medium temperature), and regions connected with magnetic field disruptions (highest temperature). This is not to be interpreted so as to imply that there is only coronal plasma at two or three discrete temperatures. The coronal plasma is likely to have a wide range of temperatures.

It is noteworthy that the emission measure for the medium temperature (12 MK) component is 4-5 times larger than that of the lower temperature (2MK) component. Furthermore the emission measure of the 12 MK component is about 3 times larger than that of the 40 MK component (see Tables 1 and 3). This implies that the emission from active regions is very important. The results are in good accordance with the results of Schrijver et al. (1995), who found that their EUVE observations of Gem might be explained by a 3-5 MK component, a stronger 15 MK component and a relatively strong component with temperatures exceeding 20-30 MK, when Solar abundance values were assumed. But when the iron abundance was halved, the 20 - 30 MK component disappeared. Contemporaneous observations with EUVE and ASCA are desirable in order to settle the question of the coronal Fe abundance.

Further evidence in favor of the suggestion that the 12 MK component is associated with emission from active regions comes from the fact that the emission measure of this component shows rapid variations (see Fig. 2d). There is a significant increase of the emission measure of the 12 MK component from March 31 through April 1, whereas changes in 2 MK emission are small. Rapid changes in X-ray emission is likely to occur if the emission is confined to (relatively few) active regions on Gem. The quiet (2 MK) component should be rather stable, as is actually found.

Fig. 2. Emission measures at low temperature a and high temperature b versus total X-ray fluxes. The symbols diamond, triangle, square, and cross represent, respectively, results for March 30 to April 1 1991, April 22 1992, April 27 1992, and Oct. 9 1992. The lower panels c and d show the variations of emission measures with the orbital phase. The error bars represent the 90% confidence limits.

The increase in the EM of the 12 MK component leads to an increase in the total X-ray emission, and again suggests that the active regions dominate the X-ray emission on Gem. One does not see any strong variations in the temperature (Fig. 3).

Fig. 3. Best fit to the two temperatures a and b versus total X-ray fluxes. The symbols diamond, triangle, square, and cross represent, respectively, results for March 30 to April 1 1991, April 22 1992, April 27 1992, and Oct. 9 1992. The lower panels c and d show the variations of temperatures with the orbital phase. The error bars represent the 90% confidence limits.

The energy relations between the 2 MK, 12 MK, and the 40 MK plasma components may be estimated when the respective emission measures and electron densities are known. One has

hence

where E is the energy of the gas, is the electron density, k is Boltzmann's constant and T the temperature. Average emission measures for the 2 MK and 12 MK components may be found from Table 3. The emission measure for the 40 MK component is given in Table 1. The following relations are then found:

where are the electron densities associated with, respectively, the 2 MK, 12 MK, and the 40 MK components.

The electron density is likely to be higher in the active region corona than in the quiet regions. An important aspect of the above relation is that the two higher temperature components are dominating over the low temperature component which we associate with quiet regions. If the inclination of Gem is small (35 degrees according to Dümmler, 1995) and the spotted regions are at high latitudes, the above result appears quite reasonable. The result is in accordance with the results of Schrijver et al. (1995) who found, from EUVE observations of Gem, that the emission measure below 2 MK was relatively low.

One may conclude from the above discussion that Gem shows a weak plasma component in the one MK range, but the bulk of the plasma is at temperatures in the range between a few MK and 15 MK. There may also be plasma at higher temperatures in the range of some tens of MK. If the latter component is associated with field disruptions, it may be rather variable. More observations to this point are highly desirable.

The UV, X-ray and radio emission of Gem is quite variable. A long term activity cycle of 8.5 years has been detected by Henry et al. (1995). From Table 2 one notes a systematic difference between the count rates in the 1991 and 1992 observations. This may be an outcome of the activity cycle, but far more observations are needed in order to reach a conclusion.

It was found by Elgaroy et al. (1995) that a possible rotational modulation was masked by stronger short time fluctuations in the UV range. Our ROSAT observations show that short time fluctuations also occur in the X-ray range. From Table 2 and 3 together with Fig. 2 it may be concluded that variations in the emission occur on a time scale corresponding to the time between subsequent observing sequences, i.e. about 1.5 hours.

In order to search for variations on the shortest possible time scale, we selected three of the longest observations and binned the data in 402 s bins (according to the wobbling period of ROSAT). Fig. 4 shows that there are variations on a time scale of the order of some minutes. Taking into account our whole set of X-ray observations one may conclude that there is good evidence for intensity variations on time scales ranging from years to minutes. Existing observations are far too scanty to give a good insight into the time variation of the chromospheric and coronal emissions of Gem and their underlying physical causes, but present evidence has at least shown that further observations may give interesting results.

Fig. 4. Time variation of the count rate for the observations from March 31 to April 1, 1991. The telescope wobbling has been removed by rebinning the data in 400 s bins.

Radio observations of several RSCVn systems (Lefévre et al. 1994) with high time resolution showed that the emission was time variable both during flares as well as during a significant fraction of the "quiescent" phases, on time scales ranging from minutes to some hours. It was concluded that low level flaring must be a permanent feature of these stellar atmospheres. It is in good accordance with our findings concerning UV and X-ray emission from Gem.

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de