Comments by the editor:The main message of the following paper is to report the detection of nonradial pulsation in a K giant. This is an important and controversial topic.A key issue for this investigation is the reliability of the extracted frequencies. Your editor has spent his life with Fourier amplitude spectra of high-precision ground-based photometry. From that experience, Figure 3 of the paper represents the natural result of instrumental instabilities and atmospheric transparency changes and does not deserve a second look. However, these are satellite observations without the typical ground-based difficulties -but there may be different problems.The statistical analyses can only help us in a limited way. It does not matter much whether one applies my favorite criterion of amplitude signal/noise ratio ≥ 4, the SigSpec values, or another method. All these statistical analyses rely on assumptions of the nature of noise present in the data. Again, in ground-based photometry, we have found that the statistics are not reliable for low-frequency peaks. The obvious method to verify that certain peaks are not noise is to look at a comparison sample: the authors provide the results for another star with a simpler light curve. In my opinion, the issue is not yet solved.T. Kallinger, D. B. Guenther, W. W. Weiss, et al.
85
AbstractWe found evidence for radial p-modes and nonradial mixed modes in the oscillation spectrum of the K giant HD 20884 based on 20.6 days of nearly continuous high-precision photometry obtained by the Canadian microsatellite MOST 1 . Oscillation frequencies range from 5-31 µHz (periods of about 2.3 d -9 hr) with luminosity amplitudes between about 300 and 950 ppm and mode lifetimes exceeding 10 days are indicated. The mode identifications are based on searches of a large grid of models for a best fit to the frequencies and temperature of HD 20884. The latter is better constrained now by spectroscopy obtained at the David Dunlap Observatory as part of this work.