2002ApJ...579..800S

In deeply convective stars, the nonthermal energy required to heat the chromosphere ultimately is supplied by turbulent magnetoconvection. Because the early and middle A stars have very shallow convective layers, they are not expected to produce enough magnetoconvective power to sustain luminous chromospheres or hot coronae. Here we describe a search for chromospheric emission in the far-ultraviolet (905-1185 Å) spectra of seven main-sequence A stars, based on observations from the Far Ultraviolet Spectroscopic Explorer (FUSE) telescope. Our survey spans the interval in effective temperature along the main sequence over which powerful subsurface convection zones and hence chromospheric emission are expected to vanish. The presence or absence of high-temperature emissions in our FUSE spectra therefore can be used to identify the locus for the transition from convective to radiative envelopes–a change in stellar structure that is difficult to assess by other means. We present our observations and analysis of the subcoronal emission lines of C III λλ977, 1175 and O VI λλ1032, 1037, which bracket a range in formation temperatures from 50,000 to 300,000 K. To supplement our FUSE observations, we also report Goddard High Resolution Spectrograph measurements of Si III λ1206 and H I Lyα λ1215, obtained from archival observations of the Hubble Space Telescope, as well as X-ray measurements from previous ROSAT survey and pointed observations. We detected C III and O VI emission features in the FUSE spectra of the coolest stars of our sample, at T_eff≲8200 K. When normalized to the bolometric luminosities, the detected emission-line fluxes are comparable to solar values. We detected none of the hotter stars in our survey at T_eff≥8300 K. Upper limits on the normalized flux in some instances approach 40 times less than solar. Within an uncertainty in the effective temperature scale of up to several hundred kelvins, our FUSE observations indicate that the transition between convective and radiative stellar envelopes takes place at, or very near, the point along the main sequence where stellar structure models predict and, moreover, that the changeover occurs very abruptly, over a temperature interval no greater than ∼100 K in width. Our FUSE sample also includes two binary stars. In both cases, the narrow UV line profiles we have observed suggest that the high-temperature emission is most likely associated with the late-type companions rather than the A stars themselves.