INAF - Astronomical Observatory of Padova

INAF - Astronomical Observatory of Padova is one of the main structures of the National Institute of Astrophysics (INAF). The main activity of the Observatory is to perform scientific research in various  fields of Astrophysics. The Observatory also contributes to  advanced education, assists in the dissemination of knowledge and implements project of education and outreach of Astronomy.

Vista su INAF - Osservatorio Astronomico di Padova.jpg
Vista su INAF - Osservatorio Astronomico di Padova

In evidenza

 Automatic galaxy morphology

Plot
Figure 1:Comparison between visual and automatic (MORPHOT), broad morphological classification (E:S0:SpE=Sp-Early:SpL=Sp-Late:Irr) of ~500 galaxies in the SDSS survey. The percentages of both classifications in each 2D-bin are reported.

 There are basically two alternative approaches to the automatic estimate of galaxy morphology:

  1. the first one assumes that an objective morphological classification can be derived by parameterization of the radial light profiles of galaxies (Saintonge et al. 2005, Grogin et al. 2003, Trujillo et al. 2001). In this case, the most common morphological indicators are the bulge to disk ratio (B/D, with R**1/4 and exponential profiles for the bulge and disk components, respectively) and the Sersic's index (n). This parametric approach presents two serious problems: (i) often the R**1/4 and exponential components derived from the formal best-fitting of galaxy light profiles do not actually correspond to any physical component, like disk and/or bulge; (ii) the correlations between model parameters and visual morphological type are weak and show an high degree of degeneracy for early-type galaxies (Sanchez-Portal et al. 2004, Pignatelli et al. 2005);
  2. the alternative, non-parametric approach relies upon empiric indicators which turn out to correlate with the visual morphological type. The most widely used among them are those defining the so-called CAS parameter set (Concentration / Asymmetry / clumpinesS; Conselice et al. 2000, Conselice 2003). Although promising, up to now this approach is far from providing morphological classifications with resolution and reliability comparable to those obtained from visual inspection.

Our tool MORPHOT provides a generalization of the non-parametric approach, by adding to the CAS parameter set a number suitably devised morphological indicators, easily derived from digital imaging of galaxies. A control sample of a few hundreds of visually classified galaxies has been used to calibrate the whole set of morphological indicators. These have been combined using two independent methods. The first one is based on a Maximum Likelihood analysis (ML). It identifies the best sub-set of meaningful, orthogonal parameters and analyses how each of them depends on the S/N ratio, as well as on the galaxy size (expressed in units of image resolution) and flattening. The second method relies on a purposely devices Neural Network (NN), trained onto the control sample of visually classified galaxies. The final, automatic indicator exploits both methods (ML and NN) and its robustness and reliability turns out to be comparable with that obtained by human experienced classifiers (see Figure 1).

MORPHOT runs on UNIX-based OS and makes use of the F77 compiler, as well as of the widely used software SExtractor, IRAF and SuperMongo. In order to safely run MORPHOT, a Random Access Memory of at least 1Gb is strongly advisable.

In its present version, MORPHOT is not yet completely stable. We plan to release a fully tested version of the tool as soon as possible.

 

People: G. Fasano

Collaboration: E. Vanzella (INAF OA Trieste), E. Pignatelli (external)

Publications: Fasano et al. (2006) MORPHOT: a tool for automatic galaxy morphology in wide and/or deep fields (see the poster)

Go to top