4 Chromospheric activity indicators

The several chromospheric activity indicators in our observations are formed at different heights in the chromosphere, Na I D₁, D₂ are formed in the upper photosphere and lower chromosphere, Ca II IRT lines in the lower chromosphere, H$_\alpha$ in the middle chromosphere, and He I D₃ in the upper chromosphere. Through detailed investigation, some information about the chromosphere structure can be inferred.

In all of the activity indicators analysed in our observations, we have found that the emission features in the spectrum come from the cooler component. The main spectroscopic features of the excess emissions in the H$\alpha$ and Ca II IRT lines have been measured in the subtracted spectra and listed in Table 2, which includes orbital phase, full width at half maximum FWHM of the excess emissions, equivalent width EW of the excess emissions, the intensity of the excess emission peak, and the absolute flux (logarithm) at the stellar surface $\log F_{\rm S}$ derived using the calibration of Hall (1996) as the relation of color index (V-R). The equivalent width EWs of the excess emissions are corrected for the contribution of the cooler component to the total continuum before they are converted to the absolute flux $\log F_{\rm S}$ at the stellar surface. The measured excess emission EWs have been plotted vs. the orbital phase in Fig. 8.

In order to analyse the short time-scale chromospheric variability of UX Ari, we have divided our three observing campaigns into two different periods: February 2000 when the observation almost covers a complete orbital period, and September-November 2000 when the observation covers about half the orbital period. Figures 2-8 show that He I D₃, Na I D₁, D₂, H$\alpha$, and Ca II IRT indicators except the Ca II IRT (8542) line during the second period exhibit similar variation patterns in the two periods, and their behaviours are coincident with each other. In the second period, the spectra at orbital phase 0.781 show stronger excess emission in all of the activity indicators. The H$\alpha$ and Ca II IRT emission lines show a dramatic increase, the He I D₃ line changes to emission, and the Na I D₁, D₂ lines show a stronger fill-in. According to Zirin (1988), if the He I D3 line is in emission, it suggests a flare-like event has happened. Therefore, these results suggest that a flare-like event was detected.

When this flare happened, the excess emission EW of H$\alpha$ line increased by about a factor of 1.5 in one day. This factor is smaller than in the previous flare of UX Ari detected by Montes et al. (1996) during which the excess emission EW of the H$\alpha$ line increased by about a factor 2.2 in one day. Compared with the typical increase factor of 2 of the chromospheric lines during flares, the flare detected here is somewhat weak. Converting the excess emission EW of the H$\alpha$ line into luminosity with the radius of the cooler component given by Duemmler & Aarum (2001), we have derived the energy released in the flare, $2.1\times 10^{31}$ erg s^-1, which is of the same order as the H$\alpha$ flares in other RS-CVn systems such as HK Lac (Catalano & Frasca 1994) and HR 1099 (Foing et al. 1994). In addition, the flare we detected took place at the orbital phase 0.781, which is very similar to the orbital phase (0.74) of the flare detected by Montes et al. (1996), and also to the orbital phase where the excess emission EW of H$\alpha$ line reached a maximum in our Feb. 2000 observations. Therefore, the activity longitude region of UX Ari should lie around here.