Документ взят из кэша поисковой машины. Адрес оригинального документа : http://www.mrao.cam.ac.uk/~bn204/mk2/publications/2004/dusty04-sdss-interacting-poster.pdf Дата изменения: Wed Nov 25 23:11:01 2009 Дата индексирования: Thu Apr 8 14:12:29 2010 Кодировка: Поисковые слова: http www.arcetri.astro.it

Star Formation in Close Pairs selected from the Sloan Digital Sky Survey
B. Nikolic, H. Cullen and P. Alexander
Cavendish Lab., Cambridge CB3 0HE, UK

1. Introduction
The most intensely starforming galaxies in the local Universe, as discovered by the IRAS mission, contain a high fraction of morphologically disturbed or interacting systems (Sanders & Mirabel, 1996); at luminosities greater than 3 1012 L they are universally classified as interacting/merging/postmerger (ibid). Interacting/merging galaxies selected on the basis of peculiar optical morphologies also appear to have enhanced star formation relative to the general population (Larson & Tinsley, 1978). We have begun investigating the external factors which may influence the rate of star formation in galaxies, starting with the effects of tidal interactions. To gauge the magnitude of these effects, we have taken care to select an objective and complete sample with as little bias as possible towards star forming systems. Tidal influence on each of the galaxies in our sample is quantified by identifying and measuring the distance to the nearest companion, using a photometric technique with good completeness characteristics.

2. Methods


Our primary sample is a complete volume (z<0.1) and luminosity limited (Mr<20.45) subsample of the SDSS ` ain Galaxy Sample'MGS). M ( We have calculated starformation rates for all of the members of the primary sample from extinction and aperture corrected H luminosities. For those galaxies detected in the IRAS allsky survey, we have also calculated far infraredderived starformation rates. Massnormalised starformation rates were derived using galaxy stellar masses estimated from zband luminosities Nearest companions were identified using the algorithm illustrated in Figure 1. We use the rband concentration index, C = Petrosian halflight/ Petrosian 90%light radius, as a proxy to galaxy morphology.







3. Results: H








Figure 2: The distribution of starformation rates of galaxies in three morphologically classified subsamples.


The distribution of starformation rates in the volume limited sample is shown in Figure 2. Some of the lowredshift galaxies in our sample with close companions are illustrated in Figure 3. Although generally accurate, it can be seen that our algorithm does not identify very close companions or merging galaxies. We find that there is an anticorrelation between SSFR and projected separation to the nearest companion in the range 0 < rp < 50 kpc in the latetype, mixedtype and early type morphological subsamples (Figure 4). The anticorrelation appears to extend to larger separations in the latetype sub sample. KolmogorovSmirnov test confirms, at a formal confidence level greater than 99.9%, the difference in distributions of SSFR between galaxies with close companions and isolated galaxies. Our data do not indicate that the morphological type or mass (within the factor of six that we probe) of the companion galaxy has an effect on the triggering of star formation.

Figure 1: Schematic representation of the algorithm used to select nearest companions.

Figure 3: Montage of some of the galaxies in our sample at low redshift (z<0.04) and with companions at projected separations closer than 30 kpc. Cross denotes the primary galaxy and the green triangle denotes the companion identified by our algorithm.

Figure 4: Mean specific star formation rate as a function of separation to the nearest companion for three morphologically classified subsamples (from top to bottom): Late ( C > 0.375), mixed ( 0.33< C < 0.375) and early ( C < 0.33 ).

Figure 5: Mean concentration index as a function of separation to the nearest companion.

4. Results: Farinfrared


5. Results: light profiles






We compare in Figure 6 the dependence on companion separation of H and far infrared derived starformation rates for the subset of galaxies detected by IRAS. There is clearly broad agreement between the two very different measures of star formation. In order to account for the possibility that both the companion and primary galaxies contribute to the measured farinfrared flux, we have indicated two normalisations in Figure 6: normalisation by just the mass of the primary galaxy, as was done for all other plots; and, normalisation by the sum of the masses of the primary and companion galaxies. The latter normalisation appears to agree better with the H star formation rates. SDSS imaging of the eight most luminous infrared galaxies in our sample is shown in Figure 7. All but one show signs of interaction or merger; however, as was expected, only half have been correctly identified as such by our algorithm.



The relationship between the concentration index and the separation to the nearest companion is shown in Figures 5. Galaxies with companions closer than 50 kpc have on lightprofiles which are on average significantly more concentrated in the nucleus. Inspection of imaging of some of these galaxies suggests that they have normal disk components and that the observed effect may be due to nuclear starbursts. Some or all of the observed enhancement of star formation in mixed and earlytype galaxies with close companions may in fact be due to this effect.

Figure 6: Mean specific starformation rate as function of pair separation for the subsample detected by IRAS. Open circles: Halpha derived SSFR; open diamonds: FIR derived SSFR, normalising by the mass of the primary galaxy only; filled squares: FIR derived SSFR, normalising by masses of both the primary and companion galaxy if within 90 arcseconds.

6. Discussion & Conclusions








We confirm previous findings (Barton et al 2000, Lambas et al 2003) that galaxies with close companions have moderately enhanced star formation compared to isolated galaxies. At least as far as it' discernible with the concentration index, this enhancement is s present in galaxies across the morphological spectrum. A tight relation between the concentration index and separation to the nearest companion is observed which we interpret as due to triggering of nuclear starbursts. Under this assumption, the observations suggest that 50 kpc is the approximate length scale at which tidal triggering occurs. The observed enhancement of star formation in galaxies with close companions is equally reflected in optical spectra and farinfrared emission. Although the coarse resolution of IRAS doesn' allow us to investigate in detail, our data suggest that t significant star formation occurs in both of the galaxies involved in interaction. We do not find any statistical dependence on the morphology or mass of the companion galaxy.

Figure 7: Images of eight out of nine of the galaxies in our sample with farinfrared derived SFR > 70 M yr1. The ninth is the second galaxy from left in the top row of Figure 3.

References

:

Sanders & Mirabel, ARA&A, 34:749, 1996 Larson & Tinsley, ApJ, 219:46, 1978 Lambas et al, MNRAS, 346:1189, 2003 Barton et al, ApJ, 530:660, 2000

Acknowledgements

:

This research is based on the Sloan Digital Sky Survey. It has made use of the Two Micron All Sky Survey, IRAS/SCANPI service at IPAC, NASA/IPAC extragalactic database and NASA' Astrophysics s Data System.