We present observations and analysis of PS1-10bzj, a superluminous supernova (SLSN) discovered in the Pan-STARRS Medium Deep Survey at a redshift z = 0.650. Spectroscopically, PS1-10bzj is similar to the hydrogen-poor SLSNe 2005ap and SCP 06F6, thoug
h with a steeper rise and lower peak luminosity (M_bol = -21.4 mag) than previous events. We construct a bolometric light curve, and show that while PS1-10bzjs energetics were less extreme than previous events, its luminosity still cannot be explained by radioactive nickel decay alone. We explore both a magnetar spin-down and circumstellar interaction scenario and find that either can fit the data. PS1-10bzj is located in the Extended Chandra Deep Field South and the host galaxy is imaged in a number of surveys, including with the Hubble Space Telescope. The host is a compact dwarf galaxy (M_B ~ -18 mag, diameter < 800 pc), with a low stellar mass (M_* ~ 2.4 * 10^7 M_sun), young stellar population (tau_* ~ 5 Myr), and a star formation rate of ~ 2-3 M_sun/yr. The specific star formation rate is the highest seen in a SLSN host so far (~ 100 Gyr^{-1}). We detect the [O III]lambda 4363 line, and find a low metallicity: 12+(O/H) = 7.8 +/- 0.2 (~ 0.1 Z_sun). Together, this indicates that at least some of the progenitors of SLSNe come from young, low-metallicity populations.
We combine the high-resolution Aquarius simulations with three-dimensional models of reionization based on the initial density field of the Aquarius parent simulation, Millennium-II, to study the impact of patchy reionization on the faint satellite p
opulation of Milky Way halos. Because the Aquarius suite consists of zoom-in simulations of halos in the Millennium-II volume, we follow the formation of substructure and the growth of reionization bubbles due to the larger environment simultaneously, and thereby determine the reionization redshifts of satellite candidates. We do this for four different reionization models, and also compare results to instantaneous reionization. Using a simple procedure for selecting satellites and assigning luminosities in the simulations, we compare the resulting satellite populations. We find that the overall number of satellites depends sensitively on the reionization model, with a factor of 3-4 variation between the four models for a given host halo, although the difference is entirely in the population of faint satellites (M_V > -10). In addition, we find that for a given reionization model the total number of satellites differs by 10%-20% between the patchy and homogeneous scenarios, provided that the redshift is chosen appropriately for the instantaneous case. However, the halo-halo scatter from the six Aquarius halos is large, up to a factor of 2-3, and so is comparable to the difference between reionization scenarios. In order to use the population of faint dwarf galaxies around the Milky Way as a probe of the local reionization history, then, it is necessary to first better understand the general distribution of substructure around Milky Way-mass halos.
