Simulating the Carbon Footprint of Galactic Halos

Simeon Bird, Kate H. R. Rubin, Joshua Suresh, Lars Hernquist

Published 2015-12-07Version 1

We use observations of CIV and CII absorption in background quasars to constrain the parameters of supernova feedback models based on the Illustris cosmological simulation. We compare our simulations to two CIV absorber surveys at z=2-4, spanning a column density range $10^{12} - 10^{15}$ cm$^{-2}$, and an equivalent width 0.1 - 2 \AA, respectively. We find that reproducing results from the first survey requires that the energy per unit mass of the supernova feedback be increased by a factor of two over the Illustris feedback model. We suggest that winds which deposit a fraction of their energy into heating, rather than accelerating, the surrounding gas can achieve this without altering the star formation rate. However, even our most energetic wind models do not produce enough absorbers with a CIV equivalent width greater than 0.6 Angstrom to match the results of the second survey. We connect these absorbers to the most massive haloes present in our simulations, and suggest possible ways to alleviate the discrepancy, either by further increasing the wind energy per unit mass, or by modifying the AGN feedback model. We also compare to the covering fractions and equivalent widths of CIV and CII absorbers around Damped Lyman-alpha absorbers, showing generally good agreement. Finally, we show that the CIV in our simulations is predominantly photoionized.