UC Riverside

Star-forming galaxies are widely believed to be responsible for the reionization of the Universe and much of the ionizing background at z>3. Therefore, there has been much interest in quantifying the escape fraction of the Lyman continuum (LyC) radiation of the star-forming galaxies. Yet direct detection of LyC has proven to be exceptionally challenging. Despite numerous efforts only 8 galaxies at z<2 (all but one with escape fractions less than 0.04) and 4 galaxies at z>2 have been robustly confirmed as LyC leakers. To avoid these challenges many studies use indirect methods to infer the LyC escape fraction. We tested these indirect methods by attempting to detect escaping LyC with a 10-orbit Hubble near-UV (F275W) image that is just below the Lyman limit at the redshift of the Cosmic Horseshoe (a lensed galaxy at z=2.4). We concluded that the measured escape fraction is lower, by more than a factor of five, than the expected escape fraction based on the indirect methods. This emphasizes that indirect determinations should only be interpreted as upper-limits.

We also investigated the deepest near-UV Hubble images of the SSA22 field to detect LyC leakage from a large sample of candidate star-forming galaxies at z $\sim3.1$, whose redshift was obtained by deep Keck/LRIS spectroscopy and for which Keck narrow-band imaging was showing possible LyC leakage. The high spatial resolution of Hubble images is crucial to confirm our detections are clean from foreground contaminating galaxies, and also to ascertain the escape fraction of our final candidates. We identify four clean LyC emitting star-forming galaxies, which doubles the sample of confirmed LyC leakers found till now. The follow up investigation of these galaxies will significantly increase our knowledge of the LyC escape fraction and the mechanisms allowing for LyC escape.