1997A&A...319..747A

The traditional method of morphological classification, by visual inspection of images of uniform quality and by reference to standards for each type, is critically examined. The rate of agreement among traditional morphologists on the morphological type of galaxies is estimated from published classification works, and is estimated at about 20%, when galaxies are classified into three bins (E, S0, S+Irr). The advantages of the quantitative method of structural classification for classifying galaxies in clusters are outlined. This method is based on the isophotal analysis of galaxy images, and on the examination of quantitative structural parameters derived from this analysis, such as the profiles of luminosity, ellipticity and deviations from ellipticity of the galaxy. The structural and traditional methods are compared on a complete sample of 190 galaxies in the Coma cluster. The morphological types derived by both methods agree to within 15 or 20%, the same rate as among traditional morphologists alone, thus showing that our morphological classes do coincide with the traditional ones. The galaxies with discrepant types are mostly faint (mag_B>16.0), or have features typical of spirals, but which have not been detected, noticed or taken into account by traditional morphologists. The rate of agreement is also good for galaxies in a distant cluster (z∼0.4). The structural method, which requires an image quality adapted to the difficulty of classifying a given galaxy, gives highly reproducible results, never reached by traditional estimations of the morphological type. Thus, the morphological types obtained with this method should be preferred, even if their determination is more time and telescope consuming, because they are less subjective, therefore more reproducible, and based on images of adequate resolution. The advantages of the method are further demonstrated by new results on the properties of galaxies in clusters.