Day 1044

Friday.


1601.07559
Binary stars can provide the "missing photons" needed for reionizaton
Ma, Hopkins, Kasen, Quataert, Faucher-Giguere, Keres, Murray

Empirical constraints on reionization require galactic ionizing photon escape fractions fesc>20%, but recent high-res radiation-hydrodynamic calculations have consistently found much lower values ~1-5%.  While these models have included strong stellar feedback and additional processes such as runaway stars, they have almost exclusively considered stellar evolution models based on single (isolated) stars, despite the fact that most massive stars are in binaries.  Re-visit these calculations, combining radiative transfer and high-res cosmo sims of galaxies with detailed models for stellar feedback from the Feedback in Realistic Environments (FIRE) project.  For the first time, use a stellar evolution model that includes a physically and observationally motivated treatment of binaries (the BPASS model).  Binary mass transfer and mergers enhance the population of massive stars at late times (>3 Myr) after SF, which in turn strongly enhances the late-time ionizing photon production (especially at low metallicities).  These photons are produced after feedback from massive stars has carved escape channels in the ISM, and so efficiently leak out of galaxies.  As a result, the time-averaged "effective" escape fraction (ratio of escaped ionizing photons to observed 1500 A photons) increases by factors 4-10, sufficient to explain reionization.  While important uncertainties remain, conclude that binary evolution may be critical for understanding the ionization of the Universe.


1601.07694
Searching for filaments are large-scale structure around DAFT/FADA clusters
Durret, et al

Search for extensions and filaments around the 30 clusters of DAFT/FADA (0.4<z<0.9).  Previous detections have been limited until now to relatively low redshifts (z<0.3).  Based on a color-magnitude diagram, select red-sequence galaxies, and hence at the cluster redshift, and build density maps, then draw density contours (3 sigma).  Find clear elongations in 12 clusters, with sizes reaching up to 7.6 Mpc.  11 other clusters have neighboring structures, but the zones linking them are not detected in the density maps at 3 sigma.  3 clusters show no extended structure and no neighbors, and 4 clusters are of too low contrast to be clearly visible on the density map.  Plan to apply to a larger photometric survey such as CFHTLS and SDSS-Stripe82.


1601.07857
Overconfidence in photometric redshift estimation
Wittman, Bhaskar, Tobin

Describe a new test of photo-z performance given a spectro-z sample.  This test complements the traditional comparison of z differences by testing whether the probability density functions p(z) have the correct width.  Test two photo-z codes, BPZ and EAZY, on each of 2 data sets and find that BPZ is consistently overconfident (the p(z) are too narrow) while EAZY produces approximately the correct level of confidence.  Show that this is because EAZY models the uncertainty in its spectral energy distribution templates, and that post-hoc smoothing of the BPZ p(z) provides a reasonable substitute for detailed modeling of template uncertainties.  Either remedy still leaves a small surplus of galaxies with spectroscopic redshift very far from the peaks.  Thus, better modeling of low-probability tails will be needed for high-precision work such as DE constraints with the LSST and other large surveys.