The response of an isothermal atmosphere to small disturbances in entropy is studied taking compressible effects fully into account. The method of Green's functions is applied to solve the linearized hydrodynamic equations by Fourier transformation. A bubble may be created by perturbing the entropy within a finite volume. At first Lamb waves will be then emitted radially and the bubble undergoes a series of Brunt-Väisälä oscillations. We find that horizontally propagating waves are generated only by large bubbles “exceeding a radius of about ten pressure scale heights, whereas smaller bubbles lead to motions propagating principally in the vertical direction.

The numerically generated dynamical evolution of an RR Lyrae model from different initial conditions is subjected to a time-dependent Fourier analysis, which yields the temporal behavior of the amplitudes and phases of the few longlived transient modes in addition to the ultimate winner. It is shown that the amplitude equation formalism of Buchler and Goupil gives an almost perfect fit to the observed transient behavior of the amplitudes and phases of the excited modes. Prospects and applications are discussed.

It is shown that in a certain range of physical parameters a proper choice of the artificial viscosity parameters can lead to double-mode pulsation of RR Lyrae models. Here we present one such model and exhibit its double-mode behavior both through the stability analysis of the periodic limit cycles and through straightforward numerical integrations.

The linear spectrum of a non-homogeneous, compressible and thermally conducting planar plasma with a vertical gravitational field and a sheared horizontal magnetic field is studied. It is shown that the spectrum has two continuous parts. The Alfvén continuum of linear ideal MHD is unaffected by radiative conduction, but the slow continuum is removed and replaced by a new continuous part which has been called isothermic continuum. Purely exponentially growing and overstable normal modes have been determined numerically and the distribution of their eigenfrequencies in the complex plane has been studied. The eigenfrequencies lie on specific curves in the complex plane which are partially determined by the isothermic continuum.

The linear stability of massive extreme Pop I stars at the main sequence is investigated with a full nonadiabatic treatment. Contrary to earlier beliefs these stars become vibrationally unstable only beyond a critical mass as high as 440 M⊙ at the main sequence.

I present a review of the art of using acoustic and gravity waves as a tool to probe the structure of the solar atmosphere. The principle ideas behind the technique are discussed and a small bouquet of results is presented. The current discrepancies are pointed out, and in the last section a few thoughts are let loose about the inversion of data to deduce the physical state of the atmosphere.

The aim of the present study is to examine whether seismologic diagnostic techniques may be used to improve the models of the solar chromosphere. We are mainly concerned by oscillations observed in the cores of strong lines such as the resonance lines of Ca II or Mg II. It is intuitive, and confirmed by numerical simulations (Gouttebroze and Leibacher, 1980), that the displacements of the cores of these lines are strongly correlated with the vertical velocity of the atmospheric layer where the line center optical depth is equal to 1. With classical atmospheric models such as those of Vernazza et al. (1981) (hereafter VAL), these line cores are formed in the chromospheric plateau, around 1500 km for the Ca II lines and 1900 km for the Mg II ones.

We have observed solar oscillations using a new instrumental technique in a relatively unexplored region of the solar spectrum. We obtained a 2-day sequence of line profiles, at 30 second intervals, for a pure rotation line of OH at 11.065 μm, using a laser heterodyne spectrometer to view a 2 arc-sec portion of the quiet Sun at disk center. The continuous opacity of the solar atmosphere increases with wavelength longward of 1.6 μm, so 11 μm lines are formed in the upper photosphere, near h = 250 km. In this region the OH rotational transitions have δJ=1 collisional rates which are two orders of magnitude larger than their radiative rates. Hence the OH lines have source functions which are equal to the Planck function, and the high spectral purity provided by the laser heterodyne technique makes their line profiles especially appropriate for investigating the dynamics of the solar atmosphere. We have recently reported (Deming et al. 1986) that oscillations in this OH line show evidence of a resonance due to a cavity in the solar chromosphere.

The aim of this work is to compute the amplitude of the limb darkening fluctuations due to oscillations in order to compare with the observations made with the heliometer at the Pic du Midi (Rösch and Yerle 1983–1984). It is shown that 5 min oscillations lead to variation in the brightness gradient of less than 2%.

We show the evidence at chromospheric level (Call K line) of mesostructures, “mesocells”, reminiscent of the mesogranulation by their spatial size (8 Mm). These cells present very regular oscillations in intensity, preferably in the 3–5 min. period range, and we show that the phase of the sustaining wave extends smoothly (coherently) over the mesocell area.

The amount of energy carried in the solar atmosphere by short period acoustic waves is particularly uncertain in the range of frequencies (ν ≥ 10mHz) which is potentially relevant for acoustic heating.

Calculations of the frequency of solar oscillation are sensitive to the upper boundary conditions of the model. Our investigations of the velocity fields of the overshoot layers (photosphere) have shown that there is a minimum of velocity at about 200 km above τ = 1: We suggest that this minimum provides a natural upper boundary condition for the calculations of solar oscillation. The propagation of sound in these layers has to be regarded as a propagation in a turbulent medium.

The most accurate measurements of red shifts in the solar spectrum, made in different epochs by different astronomers, show a systematic difference, in the average 2 mA° (0.2 pm), that seems quite independent on the wavelengths. Such a result can be explained, at least from a qualitative point of view, by the Compton effect. Indeed a variation of the normalized width U of the spectral lines causes a consequent variation of the Compton red shift: increasing U of the 50%, also the red shift increases of about 0.1 pm. Besides that a variation of the average depth h for the formation of the spectral lines in the reversing layer, may cause, in some model atmospheres a change of the red shift: an increase of about 90 Km for h may induce an increase of about 0.4 pm in the Compton red shift.

There is currently a wide range of methods for observing the properties of solar oscillations. It is generally true, however, that the means used to analyze the data are just as important to the final result as the instrumentation. I discuss first the instrumental techniques used to observe solar p-modes, with attention given both to low- and no-resolution systems, and to systems with spatial resolution. Then I describe the reduction techniques that are used to convert the raw observations into useful form.

Modulation of the magnetic field in optical resonance spectrometers in order to calibrate the slope of the observable intensity ratios in terms of differential velocities is described.

We present here the first full-disk solar Dopplergram obtained with the new 1024 × 1024-pixel CCD camera which has recently been installed at the 60-Foot Tower Telescope of the Mt. Wilson Observatory. This Dopplergram has a spatial resolution of 2.2 arcseconds and was obtained in less than one minute of time. The Dopplergram was obtained with a magneto-optical filter which was designed to obtain images in the two Na D lines. The filter and the camera were operated together as part of the development of a Solar Oscillations Imager (SOI) esperiment which is currently being designed at JPL for the joint NASA/ESA Solar and Heliospheric Observatory (SOHO) mission.

A unique solar lineshift analyzer described by Rust, Burton and Leistner (1986) has been used at the Sacramento Peak Observatory to study solar oscillations. Operation of this “Stablized Solar Analyzer” (SSA) depends on the electro-optic effect in crystalline lithium niobate, the substrate of the solid Fabry-Perot etalon. Voltage on the etalon shifts the passband by ∼ 4.5 × 10−4 Å/V. The etalon has a passband of 0.175 Å at 6102.7 Å. The stabilization system uses a tunable diode laser to relate the Fabry-Perot passband to the D2 line of atomic Cs 133, at 8521.46 Å. This system reduced instrumental noise to less than Δλ/λ = 10−9 over a six-hour interval.

Heterodyne spectroscopy at infrared (IR) wavelengths is a technique well suited for measuring small velocities in the solar atmosphere. An IR heterodyne spectrometer for solar oscillation measurements has been located at the McMath Solar Telescope of the Kitt Peak National Solar Observatory. It is now being used for single point Doppler shift measurements of 11 micron OH absorption features formed in the upper photosphere, with sampling rates as high as 33 mHz. The instrument employs a stabilized CO2. laser permitting absolute velocity measurements with an uncertainty of < 10 ms−1.

The acoustic oscillation modes of the Sun cluster along ridges of power in the ω-k plane. Fitting curves to these ridges provides input for methods that reveal information about the Sun's interior. This curve-fitting task is difficult due to noise in the data, close spacing between ridges at low k, and heuristic approaches to the fitting problem. The procedure we are investigating employs a simple but powerful rule from Bayesian decision theory in an effort to minimize the impact of such problems. This Bayesian approach allows one to make systematic use of prior physical and phenomenological information to assign a prior probability that a candidate curve gives the best fit to a ridge. Bayes' rule then permits one to update this probability using the new ridge power data. The maximally probable candidate curve given both new and prior information is chosen as the best fit.

The task of interpreting p-mode spectra is complicated by the presence of a very large number of oscillation modes, each of which may appear (because of aliasing) in the power spectra corresponding to several values of l and m. Identifying peaks in a power spectrum with particular modes in an interactive fashion thus quickly becomes impractical. Here I describe an automated method for doing this identification. The method is based on an application of Bayes' theorem, which provides a simple way to use prior knowledge about the oscillation spectrum. The method takes as input the observed power spectra, and a model of the amplitudes and frequencies one expects to see.