The population of SNe/SNRs in the starburst galaxy Arp 220. A self-consistent analysis of 20 years of VLBI monitoring

Context. The nearby ultra-luminous infrared galaxy (ULIRG) Arp 220 is an excellent laboratory for studies of extreme astrophysical environments. For 20 years, Very Long Baseline Interferometry (VLBI) has been used to monitor a population of compact sources thought to be supernovae (SNe), supernova remnants (SNRs), and possibly active galactic nuclei (AGNs). SNe and SNRs are thought to be the sites of relativistic particle acceleration powering star formation induced radio emission in galaxies, and are hence important for studies of for example the origin of the FIR-radio correlation.
Aims: In this work we aim for a self-consistent analysis of a large collection of Arp 220 continuum VLBI data sets. With more data and improved consistency in calibration and imaging, we aim to detect more sources and improve source classifications with respect to previous studies. Furthermore, we aim to increase the number of sources with robust size estimates, to analyse the compact source luminosity function (LF), and to search for a luminosity-diameter (LD) relation within Arp 220.
Methods: Using new and archival VLBI data spanning 20 years, we obtained 23 high-resolution radio images of Arp 220 at wavelengths from 18 cm to 2 cm. From model-fitting to the images we obtained estimates of flux densities and sizes of detected sources. The sources were classified in groups according to their observed lightcurves, spectra and sizes. We fitted a multi-frequency supernova light-curve model to the object brightest at 6 cm to estimate explosion properties for this object.
Results: We detect radio continuum emission from 97 compact sources and present flux densities and sizes for all analysed observation epochs. The positions of the sources trace the star forming disks of the two nuclei known from lower-resolution studies. We find evidence for a LD-relation within Arp 220, with larger sources being less luminous. We find a compact source LF n(L)∝L^β with β = -2.19 ± 0.15, similar to SNRs in normal galaxies, and we argue that there are many relatively large and weak sources below our detection threshold. The brightest (at 6 cm) object 0.2195+0.492 is modelled as a radio SN with an unusually long 6 cm rise time of 17 years.
Conclusions: The observations can be explained by a mixed population of SNe and SNRs, where the former expand in a dense circumstellar medium (CSM) and the latter interact with the surrounding interstellar medium (ISM). Nine sources are likely luminous SNe, for example type IIn, and correspond to few percent of the total number of SNe in Arp 220. Assuming all IIns reach these luminosities, and no confusion with other SNe types, our data are consistent with a total SN-rate of 4 yr^-1 as inferred from the total radio emission given a normal stellar initial mass function (IMF). Based on the fitted luminosity function, we argue that emission from all compact sources, also below our detection threshold, make up at most 20% of the total radio emission at GHz frequencies. However, colliding SN shocks and the production of secondary electrons through cosmic ray (CR) protons colliding with the dense ISM may cause weak sources to radiate much longer than assumed in this work. This could potentially explain the remaining fraction of the smooth synchrotron component. Future, deeper observations of Arp 220 will probe the sources with lower luminosities and larger sizes. This will further constrain the evolution of SNe/SNRs in extreme environments and the presence of AGN activity.

Data, images, and analysis scripts presented in this paper (see Appendix A) are available at the CDS via anonymous ftp to http://cdsarc.u-strasbg.fr (ftp://130.79.128.5) or via http://cdsarc.u-strasbg.fr/viz-bin/qcat?J/A+A/623/A173