During the four years of my PhD studies in radio astronomy I have concentrated my efforts on the following two questions: 1) Is there an evolutionary connection between nuclear starburst and AGN activity? 2) Are the starbursts observed in other galaxies scaled up versions of what we observe in local (Galactic) star forming regions?
These questions were tackled with two complementary approaches which I refer as the ``statistical'' and the ``detailed'' approaches. In the statistical approach I have conducted radio interferometric observations of the northern half of a sample of 217 infrared selected galaxies known as COLA (Corbett et al. 2002; Parra et al. 2005a) aiming to detect weak or buried AGN and correlate its presence (or absence) with global galactic properties. The results of this survey will be published in a forthcoming paper (Parra et al. in prep.). The main result of this work is that we found 20/90 detections at mJy levels on transatlantic baselines at 6 cm wavelength. All 20 sources have a small but significant radio excess on the far-infrared/radio correlation however this excess cannot be ascribed to the detected weak VLBI cores but instead to larger scale structures. This result is thus evidence in favor of an association between weak AGN and compact starbursts.
In the detailed approach, I have carried out multiwavelength/multiepoch VLBI observations of the prototypical ultra-luminous infrared galaxy (ULIRG) Arp 220. In these observations we detected for the first time, at 13, 6 and 3.6 cm wavelength, 16 of the 49 known 18 cm compact sources in Arp 220 (Lonsdale et al. 2006) and 2 new sources (see Figure 1). We conclude that these sources are a mixed population of supernovae and supernova remnants (Parra et al. 20061). This conclusion offers a consistent explanation to the observed stability of the 18 cm light curves (Rovilos et al. 2003) without invoking supernova events many times more luminous than the unusual SN1986J. If the rate of appearance of new radio sources is identified with core-collapse radio-supernovae, then the required number of massive stars is much larger than predicted by the star formation rate implying a top-heavy initial mass function. This suggests that the initial mass function in starbursts is radically different than in normal galaxies.
Also within the detailed approach and in a more theoretical aspect we have modelled the observed OH megamaser emission from the luminous infrared galaxy (LIRG) IIIZw35 (Pihlström et al. 2001) as arising from a narrow rotating starburst ring of radius 22 pc (Parra et al. 2005b). We show how both the compact and apparently diffuse maser emission from this ring arise from a single phase of unsaturated maser clouds amplifying background radio continuum. From the observed maser properties, these clouds are estimated to have sizes of less than 0.7 pc and internal velocity dispersions in the order of 20 km/s. Our model applied to these observations give one of the few direct estimates of molecular cloud sizes and velocity widths in luminous infrared galaxies.
|
|
|
|
|