SN 2006oz: rise of a super-luminous supernova observed by the SDSS-II SN Survey

Context. A new class of super-luminous transients has recently been identified. These objects reach absolute luminosities of M-u < -21, lack hydrogen in their spectra, and are exclusively discovered by non-targeted surveys because they are associated with very faint galaxies. Aims. We aim to contribute to a better understanding of these objects by studying SN 2006oz, a newly-recognized member of this class. Methods. We present multi-color light curves of SN 2006oz from the SDSS-II SN Survey that cover its rise time, as well as an optical spectrum that shows that the explosion occurred at z similar to 0.376. We fitted black-body functions to estimate the temperature and radius evolution of the photosphere and used the parametrized code SYNOW to model the spectrum. We constructed a bolometric light curve and compared it with explosion models. In addition, we conducted a deep search for the host galaxy with the 10 m GTC telescope. Results. The very early light curves show a dip in the g-and r-bands and a possible initial cooling phase in the u-band before rising to maximum light. The bolometric light curve shows a precursor plateau with a duration of 6-10 days in the rest-frame. A lower limit of M-u < -21.5 can be placed on the absolute peak luminosity of the SN, while the rise time is constrained to be at least 29 days. During our observations, the emitting sphere doubled its radius to similar to 2 x 10(15) cm, while the temperature remained hot at similar to 15 000 K. As for other similar SNe, the spectrum is best modeled with elements including O II and Mg II, while we tentatively suggest that Fe III might be present. The host galaxy is detected in gri with 25.74 +/- 0.19, 24.43 +/- 0.06, and 24.14 +/- 0.12, respectively. It is a faint dwarf galaxy with M-g = -16.9. Conclusions. We suggest that the precursor plateau might be related to a recombination wave in a circumstellar medium (CSM) and discuss whether this is a common property of all similar explosions. The subsequent rise can be equally well described by input from a magnetar or by ejecta-CSM interaction, but the models are not well constrained owing to the lack of post-maximum observations, and CSM interaction has difficulties accounting for the precursor plateau self-consistently. Radioactive decay is less likely to be the mechanism that powers the luminosity. The host is a moderately young and star-forming, but not a starburst, galaxy.