Context. Asymptotic giant branch (AGB) stars are one of the major sources of dust in the universe. The formation of molecules and dust grains and their subsequent expulsion into the interstellar medium via strong stellar winds is under intense investigation. This is in particular true for oxygen-rich stars, for which the path of dust formation has remained unclear. Aims. We conducted spatially and spectrally resolved mid-infrared multi-epoch interferometric observations to investigate the dust formation process in the extended atmospheres of oxygen-rich AGB stars. Methods. We observed the Mira variable AGB stars S Ori, GX Mon, and R Cnc between February 2006 and March 2009 with the MIDI instrument at the VLT interferometer. We compared the data to radiative transfer models of the dust shells, where the central stellar intensity profiles were described by dust-free dynamic model atmospheres. We used Al 2 O 3 and warm silicate grains, following earlier studies in the literature. Results. Our S Ori and R Cnc data could be well described by an Al 2 O 3 dust shell alone, and our GX Mon data by a mix of an Al 2 O 3 and a silicate shell. The best-fit parameters for S Ori and R Cnc included photospheric angular diameters Θ Phot of 9.7 ± 1.0 mas and 12.3 ± 1.0 mas, optical depths τ V (Al 2 O 3 ) of 1.5 ± 0.5 and 1.35 ± 0.2, and inner radii R in of 1.9 ± 0.3 R Phot and 2.2 ± 0.3 R Phot , respectively. Best-fit parameters for GX Mon were Θ Phot = 8.7 ± 1.3 mas, τ V (Al 2 O 3 ) = 1.9 ± 0.6, R in (Al 2 O 3 ) = 2.1 ± 0.3 R Phot , τ V (silicate) = 3.2 ± 0.5, and R in (silicate) = 4.6 ± 0.2 R Phot . Our data did not show evidence of intra-cycle and cycle-to-cycle variability or of asymmetries within the error-bars and within the limits of our baseline and phase coverage. Conclusions. Our model fits constrain the chemical composition and the inner boundary radii of the dust shells, as well as the photospheric angular diameters. Our interferometric results are consistent with Al 2 O 3 grains condensing close to the stellar surface at about 2 stellar radii, co-located with the extended atmosphere and SiO maser emission, and warm silicate grains at larger distances of about 4-5 stellar radii. We verified that the number densities of aluminum can match that of the best-fit Al 2 O 3 dust shell near the inner dust radius in sufficiently extended atmospheres, confirming that Al 2 O 3 grains can be seed particles for the further dust condensation. Together with literature data of the mass-loss rates, our sample is consistent with a hypothesis that stars with low mass-loss rates form primarily dust that preserves the spectral properties of Al 2 O 3 , and stars with higher mass-loss rate form dust with properties of warm silicates.

Aims. We investigate the structure and shape of the photospheric and molecular layers of the atmospheres of four Mira variables. Methods. We obtained near-infrared K-band spectro-interferometric observations of the Mira variables R Cnc, X Hya, W Vel, and RW Vel with a spectral resolution of about 1500 using the AMBER instrument at the VLTI. We obtained concurrent JHKL photometry using the the Mk II instrument at the SAAO. Results. The Mira stars in our sample are found to have wavelength-dependent visibility values that are consistent with earlier low-resolution AMBER observations of S Ori and with the predictions of dynamic model atmosphere series based on self-excited pulsation models. The corresponding wavelength-dependent uniform disk (UD) diameters show a minimum near the near-continuum bandpass at 2.25 μm. They then increase by up to 30% toward the H 2 O band at 2.0 μm and by up to 70% at the CO bandheads between 2.29 μm and 2.48 μm. The dynamic model atmosphere series show a consistent wavelength-dependence, and their parameters such as the visual phase, effective temperature, and distances are consistent with independent estimates. The closure phases have significantly wavelength-dependent and non-zero values at all wavelengths indicating deviations from point symmetry. For example, the R Cnc closure phase is 110 • ± 4 • in the 2.0 μm H 2 O band, corresponding for instance to an additional unresolved spot contributing 3% of the total flux at a separation of ∼4 mas. Conclusions. Our observations are consistent with the predictions of the latest dynamic model atmosphere series based on self-excited pulsation models. The wavelength-dependent radius variations are interpreted as the effect of molecular layers lying above the photosphere. The wavelength-dependent closure phase values are indicative of deviations from point symmetry at all wavelengths, thus a complex non-spherical stratification of the extended atmosphere. In particular, the significant deviation from point symmetry in the H 2 O band is interpreted as a signature on large scales (there being a few across the stellar disk) of inhomogeneities or clumps in the water vapor layer. The observed inhomogeneities might possibly be caused by pulsation-and shock-induced chaotic motion in the extended atmosphere.

We report the first asteroseismic results obtained with the Hertzsprung Stellar Observations Network Group Telescope from an extensive high-precision radial-velocity observing campaign of the subgiant μ Herculis. The data set was collected during 215 nights in 2014 and 2015. We detected a total of 49 oscillation modes with l values from zeroto three, including some l=1 mixed modes. Based on the rotational splitting observed in l=1 modes, we determine a rotational period of 52 days and a stellar inclination angle of 63°. The parameters obtained through modeling of the observed oscillation frequencies agree very well with independent observations and imply a stellar mass between 1.11 and 1.15 M e and an age of -+ 7.8 0.4 0.3 Gyr. Furthermore, the high-quality data allowed us to determine the acoustic depths of the He II ionization layer and the base of the convection zone.

Aims. We aim at comparing spectro-interferometric observations of Mira variable asymptotic giant branch (AGB) stars with the latest 1D dynamic model atmospheres based on self-excited pulsation models (CODEX models) and with 3D dynamic model atmospheres including pulsation and convection (CO5BOLD models) to better understand the processes that extend the molecular atmosphere to radii where dust can form. Methods. We obtained a total of 20 near-infrared K-band spectro-interferometric snapshot observations of the Mira variables o Cet, R Leo, R Aqr, X Hya, W Vel, and R Cnc with a spectral resolution of about 1500. We compared observed flux and visibility spectra with predictions by CODEX 1D dynamic model atmospheres and with azimuthally averaged intensities based on CO5BOLD 3D dynamic model atmospheres. Results. Our visibility data confirm the presence of spatially extended molecular atmospheres located above the continuum radii with large-scale inhomogeneities or clumps that contribute a few percent of the total flux. The detailed structure of the inhomogeneities or clumps show a variability on time scales of 3 months and above. Both modeling attempts provided satisfactory fits to our data. In particular, they are both consistent with the observed decrease in the visibility function at molecular bands of water vapor and CO, indicating a spatially extended molecular atmosphere. Observational variability phases are mostly consistent with those of the best-fit CODEX models, except for near-maximum phases, where data are better described by near-minimum models. Rosseland angular diameters derived from the model fits are broadly consistent between those based on the 1D and the 3D models and with earlier observations. We derived fundamental parameters including absolute radii, effective temperatures, and luminosities for our sources. Conclusions. Our results provide a first observational support for theoretical results that shocks induced by convection and pulsation in the 3D CO5BOLD models of AGB stars are roughly spherically expanding and of similar nature to those of self-excited pulsations in 1D CODEX models. Unlike for red supergiants, the pulsation-and shock-induced dynamics can levitate the molecular atmospheres of Mira variables to extensions that are consistent with observations.

Context. Benchmark stars are crucial as validating standards for current as well as future large stellar surveys of the Milky Way. However, the number of suitable metal-poor benchmark stars is currently limited, owing to the difficulty in determining reliable effective temperatures (Teff) in this regime. Aims. We aim to construct a new set of metal-poor benchmark stars based on reliable interferometric effective temperature determinations and a homogeneous analysis. The aim is to reach a precision of 1% in Teff, as is crucial for sufficiently accurate determinations of the full set of fundamental parameters and abundances for the survey sources. Methods. We observed ten late-type metal-poor dwarfs and giants: HD 2665, HD 6755, HD 6833, HD 103095, HD 122563, HD 127243, HD 140283, HD 175305, HD 221170, and HD 224930. Only three of them (HD 103095, HD 122563, and HD 140283) have previously been used as benchmark stars. For the observations, we used the high-angular-resolution optical interferometric instrument PAVO at the CHARA array. We modelled angular diameters using 3D limb-darkening models and determined effective temperatures directly from the Stefan-Boltzmann relation, with an iterative procedure to interpolate over tables of bolometric corrections. Surface gravities (log(g)) were estimated from comparisons to Dartmouth stellar evolution model tracks. We collected spectroscopic observations from the ELODIE and FIES spectrographs and estimated metallicities ([Fe/H]) from a 1D non-local thermodynamic equilibrium (NLTE) abundance analysis of unblended lines of neutral and singly ionised iron. Results. We inferred Teff to better than 1% for five of the stars (HD 103095, HD 122563, HD 127243, HD 140283, and HD 224930). The effective temperatures of the other five stars are reliable to between 2 and 3%; the higher uncertainty on the Teff for those stars is mainly due to their having a larger uncertainty in the bolometric fluxes. We also determined log(g) and [Fe/H] with median uncertainties of 0.03 dex and 0.09 dex, respectively. Conclusions. This study presents reliable and homogeneous fundamental stellar parameters for ten metal-poor stars that can be adopted as a new set of benchmarks. The parameters are based on our consistent approach of combining interferometric observations, 3D limb-darkening-modelling and spectroscopic observations. The next paper in this series will extend this approach to dwarfs and giants in the metal-rich regime.