Observations and multiwavelength modelling of active galactic nuclei and starburst galaxies

This thesis contains observations and multiwavelength modelling of Active Galactic Nuclei (AGN) and Starburst (SB) galaxies. The drive behind this work is the identification of the powering mechanism of high luminosity galaxies, where obscuration due to gas and dust hide the central regions from direct view. I present near infrared (NIR) observations of Ultraluminous Infrared galaxies (ULIRGs) and X-ray luminous starburst galaxies (XLSBs). Observations in the NIR are particularly useful in the case of dust obscuration since much less attenuation is found at these wavelengths than in the optical (by a factor of around 10). ULIRGs are defined as galaxies with extremely high infrared luminosities, attributed variously to a hidden AGN, SB region, or combination of both. We examine the dominant power source in the sample observed, finding evidence for both star formation and active nuclei. XLSBs are a group of objects, with X-ray luminosities much higher than those of normal galaxies, thought to originate only from star formation. We investigate the properties of the SB regions and rule out hidden AGN as potential power sources, providing more evidence for the existence of this group. In addition to NIR based work, I detail the building and application of a computer model which aims to deconvolve multiwavelength spectral energy distributions (SEDs), encompassing radio to X-ray frequencies, in order to quantify the relative strength of the principal components, and hence better understand the physical processes involved. Emission components used include synchrotron emission, thermal emission at various temperatures from dust, stellar emission, recombination emission and accretion disk emission (when relevant). Preliminary results indicating that reasonable fits can be achieved, highlighting distinctions between stellar and non-stellar activity.