Research:

Quantitative galaxy morphology

In recent years, observational extragalactic astronomy has become more and more dominated by extremely deep images of selected portions of the sky, like the Hubble Ultra Deep Field or the large-area COSMOS survey. These data reveal a large number of galaxies at intermediate and high redshifts that can hardly be described by the standard Hubble classification scheme: many objects consist of several bright clumps, are highly distorted, or appear to be galaxy pairs in a phase of heavy interaction.

Due to both this variety of galaxy shapes and the large number of objects – e.g. several thousand galaxies in HST/GOODS – structure parameters have been introduced that are supposed to reduce a galaxy's appearance to a set of numbers: its central light Concentration, its degree of Asymmetry, or its degree of Smoothness. This CAS-system, or more general, "non-parametric morphology", has been advertised to be able to replace the standard Hubble classification scheme, and moreover, to be computable for thousands of galaxies faster and more objectively than any visual classification could be done.

However, non-parametric morphology is far from being successful beyond any doubt. While various diagnostic diagrams can be found to crudely separate early-type (i.e., bulge-dominated) from late-type (i.e., disk-dominated and irregular) galaxies, things get much more complicated when close-to-complete samples of a certain galaxy type are desired (cf. Ferreras et al. 2005): it needs to be understood, depending on the desired study, how a certain contaminating fraction of "unwanted" galaxies can affect the results, and how the parameters used depend statistically and systematically on image quality, i.e., signal-to-noise ratio and resolution.

We therefore aim at performing detailed analyses of the statistical and systematic behaviour of "non-parametric" structure parameters with respect to image depth and galaxy type, and at comparing this approach to other methods of quantitative morphological analysis, like, e.g., shapelet decomposition. Understanding the advantages and shortcomings of automated galaxy classification is of particular importance in the light of the ever increasing number of deep and large imaging surveys.

Related refereed papers:

T. Lisker, V. P. Debattista, I. Ferreras, and P. Erwin.
Double barred galaxies at intermediate redshifts: A feasibility study.
2006, Monthly Notices of the R.A.S, 370: 477-487
( PDF )

I. Ferreras, T. Lisker, C. M. Carollo, S. J. Lilly, and B. Mobasher.
Evolution of field early-type galaxies: the view from GOODS/CDFS.
2005, Astrophysical Journal, 635: 243-259
( URL )