IUE observations of cool stars

A summary of IUE results concerning late-type stars is presented. Observations show that high-temperature outer atmospheres, as indicated by N V, C IV emission at T ≈ 10⁵K, are generally present only in high-gravity (log g ? 2) stars. Objects with high-temperature emission tend not to exhibit cool circumstellar shells, and vice versa, although there are several transition objects, the hybrid atmosphere stars, which combine C IV emission with cool winds. Ultraviolet emission from stellar transition regions correlates well with chromospheric and X-ray emission. Transition-region line ratios indicate that many stars have differential emission measure distributions similar to the Sun's. Ultraviolet observations also give indications of important dynamical effects in low-gravity stars. Density diagnostics indicate extended chromospheres for some red giants and supergiants. In addition, the large widths of lines of high temperature ions in several luminous stars indicate supersonic motions.     

Cool luminous stars

A broad theme emerging from $\text{IUE}$ and $\text{Einstein}$ observations of cool stars is that magnetic fields control the structure and energy balance of the outer atmospheres of these stars. I summarize the phenomena associated with magnetic fields in the Sun and show that similar phenomena occur in cool luminous stars. High dispersion spectra are providing unique information concerning densities, atmospheric extension, and emission line widths. A recent unanticipated discovery is that the transition lines are redshifted (an antiwind) in β Dra (G2 Ib) and perhaps other stars, which I interpret as indicating downflows in closed magnetic flux tubes as are seen in the solar flux tubes above sunspots. Finally, I classify the G and K giants and supergiants into three groups — active stars, quiet stars, and hybrid stars — depending on whether their atmospheres are dominated by closed magnetic flux tubes, open field geometries, or a predominately open geometry with a few closed flux tubes embedded.     

Coronal activity in F-, G-, and K-type stars; relations between parameters characterizing stellar structures and X-ray emission

A sample of 52 stars containing dwarfs and giants is subjected to a multidimensional factor analysis. The parameters used are the soft X-ray flux at the stellar surface F_x, the Ca II H+K line-core flux F_H+K, the stellar radius and mass. We find a high correlation between F_x and the Ca II H+K excess flux ΔF_H+K obtained by subtracting an observational lower-limit flux from F_H+K. We conclude that the lower-limit Ca II flux is uncorrelated with the stellar X-ray emission. The common-factor analysis shows that, for the present sample, F_x depends only on ΔF_H+K, and not on the stellar radius or mass. All stars included in our analysis follow the relation F_x ∝ Δ^1.4_H+K over almost four decades in F_x.     

IUE observations of globular clusters and blue horizontal branch stars

This paper presents a review of ultraviolet photometry and spectroscopy of globular clusters and blue horizontal branch stars. Current observations with IUE are placed within the framework of earlier photometric studies carried out by OAO and ANS. In order to provide basic information on the nature of the horizontal branch star, whose light dominates in the ultraviolet, observations of field horizontal branch stars as well as individual stars in globular clusters are discussed. Finally the IUE observations of globular clusters in the Magellanic Clouds are examined, since they provide a sequence spanning the age range from 7 × 10⁶ years to 10¹⁰ years.     

Upper limits to ultraviolet line emission in fully convective M-dwarfs

Theoretical stellar structure models find that stars later than ∼dM3 are fully convective. It is widely believed that this should result in reduced non-radiative heating and activity levels, however, the observational evidence is sparse and somewhat contradictory. In order to expand the number of atmospheric diagnostics for these stars, we have obtained deep short wavelength (SWP) IUE spectra of 5 M dwarfs later than spectral type dM5. Exposure times ranged from 4 to 7 hours. Only upper limits were measured for chromospheric and transition region lines, thus appearing to rule out enhanced line emission activity. The fractional luminosities of C IV (λ1550), C II (λ1335), and C I (λ1657) relative to L_bol indicate that emission lines in the temperature range 6000 K to ∼1 × 10⁵K are up to an order of magnitude weaker than in earlier M dwarfs having radiative cores. The lower limits on line fluxes for Barnard's star are significantly below those of the other stars in the program.     

Coronal heating mechanisms

As a result of the large body of data available from solar and stellar coronae, our understanding of the mechanisms responsible for the heating of coronal plasmas to temperatures of the order of ～ 10⁸ K has changed. The solar corona is highly structured by magnetic fields and the acoustic shocks which, according to early theories, should have acted as the coronal energy source have not been observed. Einstein Observatory data show moreover that coronae are present in most regions of the H-R diagram. The observed relationship between X-ray luminosity and rotational velocity in dwarf stars from spectral types F to M again suggests an active role for the magnetic fields.The basic picture which is emerging is that coronae in stellar types from F to M are produced because of the interaction of the magnetic field with the convective velocity fields generated in the photosphere resulting in MHD waves or currents which dissipate in the corona. X-ray emission in early type stars cannot be explained with this mechanism and the models which have been proposed for these stars are not yet completely satisfactory.     

EXOSAT observations of two gamma-ray burst sources

EXOSAT observations of the 19 Nov 1978 and 25 Mar 1979b gamma bursters are presented, for the 0.02–2.5 keV energy range. No source was detected in either case, leading to temperature upper limits for the neutron stars assumed to be present of 10⁵–10⁶ °K. Polar cap accretion rates are constrained to values of 10⁻¹²–10⁻¹⁸ M_⊙/y km². The EXOSAT non-detection of the source discovered by the Einstein satellite in the 19 Nov 1978 error box may be explained if the burster is more distant than about 1.5 kpc.     

The galactic center black hole: Clues for the evolution of black holes in Galactic nuclei

The evidence for a black hole located at the dynamical center of the Milky Way and identified with the unusual radio source, Sgr A¹, is now very compelling. Proper motion and radial velocity surveys of stars clearly demonstrate the presence of a non-luminous concentration of 2.6 × 10⁶ M_⊙ within a volume of radius ∼0.01 pc centered on Sgr A¹. At present, the accretion rate onto this object is rather small, leading to a total accretion luminosity at radio through far-IR wavelengths < 10³ L_⊙. The accreted material apparently originates in the winds of nearby massive stars. However, neither the stellar nor the gaseous environments are static. The surrounding cluster of massive stars, most lying well within a parsec, is only a few million years old, and is destined to fade substantially within another 10⁷ years. How did such a cluster form in the immediate and tidally stressed vicinity of a supermassive black hole? The circumnuclear disk of gas, which presently has an inner radius of 1 pc, seems destined to migrate inwards and eventually cause a much higher accretion rate onto Sgr A¹, with a consequent flurry of new activity. Because the young stars and gas in the vicinity of the black hole interact with each other, the episodes of recurrent activity there can be described in terms of a limit cycle, which effectively controls the growth of the central black hole. In addition to describing the steps of this cycle, we identify several key observations which serve as potential clues to the past activity not only of our Galactic center, but to the activity of gas-rich nuclei in general.     

Late stages of stellar evolution – Herschel’s contributions

Cool objects glow in the infrared. The gas and solid-state species that escape the stellar gravitational attraction of evolved late-type stars in the form of a stellar wind are cool, with temperatures typically ?1500 K, and can be ideally studied in the infrared. These stellar winds create huge extended circumstellar envelopes with extents approaching 10¹⁹${10}^{19}$ cm. In these envelopes, a complex kinematical, thermodynamical and chemical interplay determines the global and local structural parameters. Unraveling the wind acceleration mechanisms and deriving the complicated structure of the envelopes is important to understand the late stages of evolution of ∼97% of stars in galaxies as our own Milky Way. That way, we can also assess the significant chemical enrichment of the interstellar medium by the mass loss of these evolved stars. The Herschel Space Observatory is uniquely placed to study evolved stars thanks to the excellent capabilities of the three infrared and sub-millimeter instruments on board: PACS, SPIRE and HIFI. In this review, I give an overview of a few important results obtained during the first two years of Herschel observations in the field of evolved low and intermediate mass stars, and I will show how the Herschel observations can solve some historical questions on these late stages of stellar evolution, but also add some new ones.     

Highly variable X-ray emitting objects in the Rho Oph dark cloud

The Rho Ophiuchi dark cloud region has been the subject of an extensive guest investigation using the $\text{Einstein}$ Observatory. The set of observations comprise 14 IPC fields and 3 HRI fields. The densest part of the cloud has been observed 6 times. Forty seven sources were detected at a level > 3.5 σ and twenty more above 2 σ. The majority of these sources have optical, IR, or even radio continuum counterparts; nine are identified with known T Tauri stars, while several others are identified with stars showing H α in emission. All show a high degree of time variability; flux variations reach factors of 5 in a few hours, or 25 in a day. Apparent luminosities are in the range 10(30) – 10(31)(1) erg.s^?1. The possibility that the X-ray variability is due to flares is examined. If this interpretation is correct, one source has been the seat of the largest stellar flare ever recorded in X rays [L_x = 10(32) erg.s^?1, E_x ?10(36) ergs^-].     

Radio wavelength observations of magnetic fields on active dwarf M,RS CVn and magnetic stars

The dwarf M stars YZ Canis Minoris and AD Leonis exhibit narrow-band, slowly varying (hours) microwave emission that cannot be explained by conventional thermal radiation mechanisms. The dwarf M stars AD Leonis and Wolf 424 emit rapid spikes whose high brightness temperatures similarly require a nonthermal radiation process. We attribute them to coherent mechanisms such as an electron-cyclotron maser or coherent plasma radiation. If the electron-cyclotron maser emits at the second or third harmonic of the gyrofrequency, the coronal magnetic field strength H = 250 G or 167 G and constraints on the plasma frequency imply an electron density of N_e = 6 × 10⁹cm⁻³. Coherent plasma radiation requires similar values of electron density but much weaker magnetic fields. Radio spikes from AD Leonis and Wolf 424 have rise times τ_R ⩽ 5 ms, indicating a linear size of L ⩽ 1.5 × 10⁸cm, or less than 0.005 of the stellar radius. Although Ap magnetic stars have strong dipole magnetic fields, they exhibit no detectable gyroresonant radiation, suggesting that these stars do not have hot, dense coronae. The binary RS CVn star UX Arietis exhibits variable emission at 6 cm wavelength on time scales ranging from 30 s to more than one hour. The shortest variation implies a linear size much less than that of the halo observed by VLBI techniques, and most probably sizes smaller than those of the component stars. The observed variations might be due to absorption by a thermal plasma located between the stars.     

Supernova remnant 1987A: The latest report from the Chandra X-ray Observatory

We continue monitoring supernova remnant (SNR) 1987A with the Chandra X-ray Observatory. As of 2004 January, bright X-ray spots in the northwest and the southwest are now evident in addition to the bright eastern ring. The overall X-ray spectrum, since 2002 December, can be described by a planar shock with an electron temperature of ∼2.1 keV. The soft X-ray flux is now 8 × 10⁻¹³ ergs cm⁻² s⁻¹, which is about five times higher than four years ago. This flux increase rate is consistent with our prediction based on an exponential density distribution along the radius of the SNR between the HII region and the inner ring. We still have no direct evidence of a central point source, and place an upper limit of L_X = 1.3 × 10³⁴ ergs s⁻¹ on the 3–10 keV band X-ray luminosity.     

Multiple wavelength observations of flaring active regions

We present observations of flaring active regions with the Very Large Array (V.L.A. at 6 cm and 20 cm wavelengths) and the Westerbork Synthesis Radio Telescope (W.S.R.T. at 6 cm wavelength). These are compared with photospheric magnetograms (Meudon) and with Hα and offband Hα photographs (Big Bear and Ottawa River Solar Observatories). The 6 cm radiation of these active regions marks the legs of dipolar loops which have their footpoints in lower-lying sunspots. The intense, million degree radiation at 6 cm lies above sunspot umbrae in coronal regions where the longitudinal magnetic field strength H_? = 600 Gauss and the height above the sunspot umbrae h = 3.5±0.5 × 10⁹ cm. Circularly polarized horseshoe structures at 6 cm ring the sunspot umbrae. The high degree of circular polarization (?_c = 95%) of the horseshoes is attributed to gyroresonant emission above sunspot^? penumbrae. The 20 cm radiation of these active regions exhibits looplike coronal structures which extend across regions of opposite magnetic polarity in the underlying photosphere. The 20 cm loops are the radio wavelength counterparts of the X-ray coronal loops. We infer semilengths L = 5 × 10⁹ cm, maximum electron temperatures T_e(max) = 3 × 10⁶ K, emission measures ∫N_e²d$\text{l}$ = 10²⁸ cm^?5, and electron densities N_e = 10⁹ cm^?3 (or pressures p = 1 dyn cm^?2) for the 20 cm bremsstrahlung. A total of eight solar bursts were observed at 6 cm or 20 cm wavelength with second-of-arc angular resolution. The regions of burst energy were all resolved with angular sizes between 5″ and 30″, brightness temperatures between 2 × 10⁷ K and 2 × 10⁸ K, and degrees of circular polarization between 10% and 90%. The impulsive phase of the radio bursts are located near the magnetic neutral lines of the active regions, and between the flaring Hα kernels which mark the footpoints of magnetic loops. In one case there was preburst heating in the coronal loop in which a burst occurred. Snapshot maps at 10 s intervals reveal interesting burst evolution including rapid changes of circular polarization and an impulsive burst which was physically separated from both the preburst radio emission and the gradual decay phase of the burst.     

Einstein observations of hot stars

Surveys with instruments on the $\text{Einstein Observatory}$ have shown that essentially all 0 and B main sequence stars are X-ray sources as are many, if not all, 0B supergiants and Wolf-Rayet stars. The X-ray luminosities are sufficient to explain broad lines from the superionization stages seen in the UV spectra of the stars. High energy resolution spectra from the Solid State Spectrometer are shown to place severe constraints on various models for the location of the X-ray sources in the outer atmospheres of the stars. Coronal and embedded shock models for the X-ray emission are discussed and each is found to have some problems in explaining the X-ray emission of 0B stars. X-ray line emission of Si XIII and S XV in ? Ori is discussed and interpreted as arising from magnetically confined loops.     

Detection of thermal X-ray emission in the halo of the plerionic supernova remnant G21.5-0.9

The detection of a soft thermal X-ray component in the spectrum of a bright knot in the halo of the plerion G21.5-0.9 is reported. Using a collisional ionization equilibrium model for an hot optically thin plasma, a temperature of kT 0.12–0.24 keV, a mass of 0.3–1.0 M_⊙ and a density of 1.6–6 cm⁻³ is derived. The spectral analysis suggests a possible overabundance of Silicon with respect to the solar value in the knot; if this will be confirmed this object may be a clump of shocked ejecta.     

Solar and late-type dwarfs

The Einstein Observatory and the IUE satellite have provided the observational basis for a major restructuring in theories of coronal formation for late-type stars. For the first time, coronal and transition region emission from a large sample of low mass (1 M_o) dwarf stars has been directly observed, with the unexpected result that essentially all such stars are x-ray emitters. The Sun, which was previously assumed to be typical, is now known to be at the low end of the x-ray luminosity function for solar-type stars. K- and M-dwarfs are observed to have nearly the same luminosity distributions as G-dwarfs and all of these spectral types have a large spread in x-ray luminosity.Observationally, there is a strong correlation between the strength of coronal emission in stars with outer convective zones and the rotation rates of these stars. At the present time we have only the beginnings of a satisfactory theoretical explanation for this correlation; although we are beginning to understand the connection between coronal emission strength and the magnetic field, we do not yet understand the stellar dynamo which generates the magnetic field. Studies of the coronal emission of stars may lead to a better understanding of stellar dynamos.     

Microwave observations of red dwarf flare stars

We discuss some recent observations of red dwarf flare stars. When observed over periods of about 8 hours, each of 4 flare star systems displayed at least one major flare at 20 cm. Quiescent emission at 6 cm was seen from UV Ceti and EQ Peg A, but flares were much less frequent at 6 cm than at 20 cm. We also summarize earlier observations of quiescent emission from UV Ceti. Observations of highly polarized flares with brightness temperatures in excess of 10¹⁰ K appear to be common on red dwarf stars. We have also found narrowband flares which strengthen the argument that a coheren emission mechanism is involved in these flares. One of those narrowband flares allows us to place severe constraints on conditions in the flare source, and if the flare is cyclotron maser emission it seems unlikely that magnetic reconnection is involved in the flare.     

Thermal emission from isolated neutron stars and their surface magnetic field: Going quadrupolar?

In the last few years considerable observational resources have been devoted to study the thermal emission from isolated neutron stars. Detailed XMM and Chandra observations revealed a number of features in the X-ray pulse profile, like asymmetry, energy dependence, and possible evolution of the pulse profile over a time scale of months or years. Here we show that these characteristics may be explained by a patchy surface temperature distribution, which is expected if the magnetic field has a complex structure in which higher order multipoles contribute together with the dipole. We reconsider these effects from a theoretical point of view, and discuss their implications to the observational properties of thermally emitting neutron stars.     

The plerionic supernova remnant G21.5-0.9: In and out

The absence of a supernova remnant (SNR) shell surrounding the Crab and other plerions (pulsar wind nebulae) has been a mystery for three decades. G21.5-0.9 is a particularly intriguing plerionic SNR in which the central powering engine is not yet detected. Early CHANDRA observations revealed a faint extended X-ray halo which was suggested to be associated with the SNR shell; however its spectrum was non-thermal, unlike what is expected from an SNR shell. On the other hand, a plerionic origin to the halo is problematic since the X-ray plerion would be larger than the radio plerion. We present here our analysis of an integrated 245 ks of archival CHANDRA data acquired with the High-Resolution Camera (HRC) and 520 ks acquired with the Advanced CCD Imaging Spectrometer (ACIS). This study provides the deepest and highest resolution images obtained to date. The resulting images reveal for the first time: (1) a limb-brightened morphology in the eastern section of the halo, and (2) a rich structure in the inner (40″-radius) bright plerion including wisps and a double-lobed morphology with an axis of symmetry running in the northwest–southeast direction. Our spatially resolved spectroscopic study of the ACIS-I data indicates that the photon index steepens with increasing distance from the central point source out to a radius of 40″ then becomes constant at ∼2.4 in the X-ray halo (for a column density N_H = 2.2 × 10²² cm⁻²). No line emission was found from the eastern limb; however marginal evidence for line emission in the halo’s northern knots was found. This study illustrates the need for deep CHANDRA observations to reveal the missing SNR material in Crab-like plerions.     

Possible Suzaku detection of non-thermal X-ray signals from a rotating magnetized white dwarf

Although rotating neutron stars (NSs) have been regarded as being textbook examples of astrophysical particle acceleration sites for decades, details of the acceleration mechanism remain a mystery; for example, we cannot yet observationally distinguish “polar cap” models from “outer gap” models. To solve the model degeneracy, it is useful to study similar systems with much different physical parameters. Strongly magnetized white dwarfs (WDs) are ideal for this purpose, because they have essentially the same system geometry as NSs, but differ largely from NSs in the system parameters, including the size, magnetic field, and the rotation velocity, with the induced electric field expected to reach 10¹³–10¹⁴ eV. Based on this idea, the best candidate among WDs, AE Aquarii, was observed with the fifth Japaneses X-ray satellite, Suzaku. The hard X-ray detector (HXD) on-board Suzaku has the highest sensitivity in the hard X-ray band over 10 keV. A marginal detection in the hard X-ray band was achieved with the HXD, and was separated from the thermal emission. The flux corresponds to about 0.02% of its spin-down energy. If the signal is real, this observation must be a first case of the detection of non-thermal emission from WDs.