This paper introduces a novel feature extraction technique for the analysis of spectral line Stokes profiles. The procedure is based on the use of an autoassociative artificial neural network containing nonlinear hidden layers. The neural network extracts a small subset of parameters from the profiles (features), from which it is then able to

We study the presence of low intensity high velocity components, which we have termed wing features in the integrated Hα emission line profiles of the H II region populations of the spiral barred galaxies NGC 1530, NGC 3359 and NGC 6951. We find that more than a third of the H II region line profiles in each galaxy show these components. The

We report redshift measurements for 12 BL Lacertae objects from a program aimed at obtaining high signal-to-noise ratio (up to ~500) optical spectroscopy of a mixed sample of objects. The new observations, gathered with the 8 m ESO Very Large Telescope, allowed us to detect weak spectral features down to a line equivalent width as small as ~1 Å

We analyse N-body galaxy merger experiments involving disc galaxies. Mergers of disc-bulge-halo models are compared to those of bulgeless, disc-halo models to quantify the effects of the central bulge on merger dynamics and the structure of the remnant. Our models explore galaxy mass ratios 1:1 through 3:1, and use higher bulge mass fractions than

We report the discovery of a new, non-eclipsing SU UMa-type dwarf nova, HS 2219+1824. Photometry obtained in quiescence (V≈17.5) reveals a double-humped light curve from which we derive an orbital period of ≃86.2 min. Additional photometry obtained during a superoutburst reaching V≃12.0 clearly shows superhumps with a period of ≃89.05 min. The

The observed coronal loop oscillations and their damping are often theoretically described by the use of a very simple coronal loop model, viz. a straight, longitudinally invariant, axi-symmetric, and pressureless flux tube with a different density inside and outside of the loop. In this paper we generalize the model by including longitudinal