Constraining the cosmic UV background at z>3 with MUSE Lyman-{\alpha} emission observations

TL;DR

This study introduces a new method using 3D stacking of MUSE Lyman-alpha observations to independently constrain the cosmic UV background at z>3, revealing potential non-monotonic evolution and low covering fractions of Lyman limit systems.

Contribution

It presents a novel approach combining deep MUSE data and simulations to measure the UVB and LLS covering fractions at high redshift, providing independent constraints.

Findings

UVB intensity at z=3 is relatively low and possibly non-monotonic between z=3 and 5.

Covering fraction of LLSs within 150kpc of LAEs at 3<z<4.5 is less than 25%.

Results suggest tension with some existing UVB models, indicating the need for deeper observations.

Abstract

The intensity of the Cosmic UV background (UVB), coming from all sources of ionising photons such as star-forming galaxies and quasars, determines the thermal evolution and ionization state of the intergalactic medium (IGM) and is, therefore, a critical ingredient for models of cosmic structure formation. Most of the previous estimates are based on the comparison between observed and simulated Lyman-$\alpha$ forest. We present the results of an independent method to constrain the product of the UVB photoionisation rate and the covering fraction of Lyman limit systems (LLSs) by searching for the fluorescent Lyman-$\alpha$ emission produced by self-shielded clouds. Because the expected surface brightness is well below current sensitivity limits for direct imaging, we developed a new method based on three-dimensional stacking of the IGM around Lyman-$\alpha$ emitting galaxies (LAEs) between 2.9<z<6.6 using deep MUSE observations. Combining our results with covering fractions of LLSs obtained from mock cubes extracted from the EAGLE simulation, we obtain new and independent constraints on the UVB at z>3 that are consistent with previous measurements, with a preference for relatively low UVB intensities at z=3, and which suggest a non-monotonic decrease of $\Gamma$HI with increasing redshift between 3<z<5. This could suggest a possible tension between some UVB models and current observations which however require deeper and wider observations in Lyman-$\alpha$ emission and absorption to be confirmed. Assuming instead a value of UVB from current models, our results constrain the covering fraction of LLSs at 3<z<4.5 to be less than 25% within 150kpc from LAEs.