ABSTRACT
We have searched for optical coronal lines in the 3100–3700 Å region of eight M dwarf stars with rather low levels of activity. This brief survey supplements a similar search in 15 active stars published in 1991. No coronal lines could be identified. However, the emission spectra including lines of H i, He i, Ca ii, Ca i, Si i, and Fe i are described and illustrated. Radial velocities of the emission lines show no systematic differences from the stellar absorption lines. Coronae with temperatures similar to those in the solar corona seem to be rare among the M dwarfs, although at least one example has been found by Schmitt & Wichmann.
Export citation and abstract BibTeX RIS
1. INTRODUCTION
Hydrogen emission lines in the spectra of M dwarf (henceforth dM) stars have been known for many years (Joy & Humanson 1941). Joy & Humason (1949) also found emission lines of He i in the flare star L726‐8. Since solar flares are known to be one of the heating sources of the solar corona, a search for coronal lines in the ground‐based accessible region of dM stars was initiated about 15 years ago (Wallerstein et al. 1991, henceforth WBO). That search was concentrated on known active stars, but no coronal lines were found. Soft X‐ray observations were later found to show that active dM stars usually have a two‐temperature corona, with most of the flux originating at temperatures that are too high to excite the lines that might have been visible between 3100 and 3700 Å. In the meantime, Schmitt & Wichmann (2001) found the [Fe xiii] line at λ3388 in the M5.5 V star CN Leo. This demonstrated that coronal lines in dM stars could indeed be found from the ground. Shortly before that, one of us (G. W.) took advantage of an opportunity to observe eight additional dM stars with the Cerro Tololo Inter‐American Observatory (CTIO) 4 m echelle spectrograph. The short camera with the 79 lines mm−1 echelle grating, and the KPGL1 cross‐disperser in the first order, whose resolving power was about 20,000, were used to record the spectra. This arrangement provided usable spectra from 3100 to 4950 Å. The S/N varied from about 30 at the short wavelength end, increasing rapidly to about 200 at 3600 Å, and several hundred at Hβ. The spectra were reduced with standard IRAF programs by S. T. to yield the intensity as a function of wavelength over the entire interval. The wavelength interval was fully covered, with significant overlap at the ends of each order. The individual orders were measured separately using the splotroutine. By measuring them separately rather than first combining the orders, spurious lines could be recognized by their appearance in one order only.
The observed stars are described in Table 1 and are largely from the SIMBAD database. In Table 2 we show the ROSAT all‐sky counts for both the 2001 targets and the stars observed in 1988. Our measured velocities of both emission and absorption lines are shown in Table 3.
 |
 |
 |
2. THE EMISSION LINES
No coronal lines were recognized. However, our spectra show a number of emission lines of lower ionization presumably formed in the stellar chromospheres. Absorption velocities were derived from the measurement of about 12 lines of neutral Fe, Mn, Cr, and Ca. For two stars, our measurements differ from catalog values in SIMBAD, for reasons that are unclear. It is known that the CTIO Cassegrain echelle suffers from flexure, so frequent ThAr spectra were obtained during the observing session. However, it is still possible that some spectra were not properly calibrated for wavelength.
Emission in the Ca ii H and K lines was seen in every star. Our resolution was not sufficient to measure their intrinsic widths. We show two Ca ii line profiles in Figure 1. Note the presence of emission by H
in V577 Mon. Emission by Fe ii multiplets 1, 6, and 7, which is often seen in M giants and supergiants, was not seen in any of our stars (Bidelman & Pyper 1963). None of our stars showed the Ti ii emission lines reported by Fuhrmeister et al. (2004)
Fig. 1.— Spectrum of V577 Mon. Note the presence of Ca ii H and K lines in emission as well as H in emission.
Four of our stars show H i emission, and V577 Mon shows weak emission by He i at λλ4026 and 4471. V1216 Sgr also shows emission at 4026 Å. He ii emission was searched for unsuccessfully at both the λλ3203 and 4686 lines. The spectrum of V1216 Sgr includes emission by Fe i, Si i, and Ca i. Such a cool chromosphere in a dM star is probably rare. It was first seen in the well‐known flare star EV Lac (Wilson 1961). We show some line profiles in V1216 Sgr in Figure 2. In Figure 3 we show emission by one of several Fe i lines, as well as the H i line at λ3889. The He i line at λ3888.64 does not seem to be present. Figure 4 shows the resonance line of Ca i at λ4226 in V1216 Sgr. As shown in Table 3, the velocities of the emission lines in our target stars show very small and uncertain differences from the measured absorption velocities. Hence, we have no evidence that supports an expanding chromosphere around these stars.
Fig. 2.— Spectrum of V1216 Sgr. A weak emission line of He i is present at λ4026 along with a number of absorption lines.
Fig. 3.— Emission lines of Fe i and H i are shown in V1216 Sgr.
Fig. 4.— Region around the resonance line of Ca i at λ4226 in V1216 Sgr. Note the likely presence of the Fe i line at λ4216.19, and the possible emission at λ4232.72 also of Fe i.
3. DISCUSSION
As discussed by WBO, coronal lines might be expected to be present in dM stars if their coronas were of similar temperatures as the solar corona, and they should be visible in the near‐UV region. The discovery by Schmitt & Wichmann (2001) of one star with a coronal line shows that coronal lines in dM stars can be found from the ground. However, our unsuccessful surveys of 1988 and 2001 show that the phenomenon is rare. The reason for this is likely to be the temperature structure of dM coronas. Most of the optical coronal lines in the Sun are formed in regions with temperatures of 2–3 × 106 K. If the coronas of dM stars are either hotter or cooler than this, they will show different lines. In fact, there is some evidence that dM stars have two temperature peaks in their coronas (Redfield et al. 2003; Sanz‐Forcada et al. 2003). Detailed high‐resolution X‐ray spectra covering a sufficient energy range will be very useful in restricting model coronas of dM stars.
S. T. is pleased to acknowledge the Washington NASA Space Grant Consortium for support during the summer of 2002. Chris Laws provided useful help on the use of IRAF to reduce echelle spectra, Suzanne Hawley made helpful suggestions on target selection, and Marcel Agueros looked up the ROSAT data that appear in Table 2. We have made frequent use of the SIMBAD database and the ADS Abstracts.