We have used multiband imaging to investigate the nature of an extreme starburst environment in the nearby Lyman break galaxy analogue Haro 11 (ESO 350−IG038) by means of its stellar cluster population. The central starburst region has been observed in eight different high‐resolution Hubble Space Telescope (HST) wavebands, sampling the stellar and gas components from UV to near‐infrared. Photometric imaging of the galaxy was also carried out at 2.16 μm by NaCo AO instrument at the ESO Very Large Telescope. We constructed integrated spectral energy distributions (SEDs) for about 200 star clusters located in the active star‐forming regions and compared them with single stellar population models (suitable for physical properties of very young cluster population) in order to derive ages, masses and extinctions of the star clusters. The cluster age distribution we recover confirms that the present starburst has lasted for 40 Myr, and shows a peak of cluster formation only 3.5 Myr old. With such an extremely young cluster population, Haro 11 represents a unique opportunity to investigate the youngest phase of the cluster formation process and evolution in starburst systems. We looked for possible relations between cluster ages, extinctions and masses. Extinction tends to diminish as a function of the cluster age, but the spread is large and reaches the highest dispersion for clusters in partial embedded phases (<5 Myr). A fraction of low‐mass (below 104  M⊙), very young (1–3 Myr) clusters is missing, either because they are embedded in the parental molecular cloud and heavily extinguished, or because of blending with neighbouring clusters. The range of the cluster masses is wide; we observe that more than 30 per cent of the clusters have masses above 105  M⊙, qualifying them as super star clusters. Almost half of the cluster sample is affected by flux excesses at wavelengths >8000 Å which cannot be explained by simple stellar evolutionary models. Fitting SED models over all wavebands leads to systematic overestimates of cluster ages and incorrect masses for the stellar population supplying the light in these clusters. We show that the red excess affects also the HST F814W filter, which is typically used to constrain cluster physical properties. The clusters which show the red excess are younger than 40 Myr; we discuss possible physical explanations for the phenomenon. Finally, we estimate that Haro 11 has produced bound clusters at a rate almost a factor of 10 higher than the massive and regular spirals, like the Milky Way. The present cluster formation efficiency is ∼38 per cent of the galactic star formation rate.

The origins of Lyman continuum (LyC) photons responsible for the reionization of the universe are as of yet unknown and highly contested. Detecting LyC photons from the Epoch of Reionization is not possible due to absorption by the intergalactic medium, which has prompted the development of several indirect diagnostics to infer the rate at which galaxies contribute LyC photons to reionize the universe by studying lower-redshift analogs. We present the Low-redshift Lyman Continuum Survey (LzLCS) comprising measurements made with the Hubble Space Telescope Cosmic Origins Spectrograph for a z = 0.2–0.4 sample of 66 galaxies. After careful processing of the far-UV spectra, we obtain a total of 35 Lyman continuum emitters (LCEs) detected with 97.725% confidence, nearly tripling the number of known local LCEs. We estimate escape fractions from the detected LyC flux and upper limits on the undetected LyC flux, finding a range of LyC escape fractions up to 50%. Of the 35 LzLCS LCEs, 12 have LyC escape fractions greater than 5%, more than doubling the number of known local LCEs with cosmologically relevant LyC escape.

The Lyman continuum (LyC) cannot be observed at the epoch of reionization (z ≳ 6) owing to intergalactic H i absorption. To identify LyC emitters (LCEs) and infer the fraction of escaping LyC, astronomers have developed various indirect diagnostics of LyC escape. Using measurements of the LyC from the Low-redshift Lyman Continuum Survey (LzLCS), we present the first statistical test of these diagnostics. While optical depth indicators based on Lyα, such as peak velocity separation and equivalent width, perform well, we also find that other diagnostics, such as the [O iii]/[O ii] flux ratio and star formation rate surface density, predict whether a galaxy is an LCE. The relationship between these galaxy properties and the fraction of escaping LyC flux suggests that LyC escape depends strongly on H i column density, ionization parameter, and stellar feedback. We find that LCEs occupy a range of stellar masses, metallicities, star formation histories, and ionization parameters, which may indicate episodic and/or different physical causes of LyC escape.

Using four different suites of cosmological simulations, we generate synthetic spectra for galaxies with different Lyman continuum escape fractions (f esc ) at redshifts z ≈ 7-9, in the rest-frame wavelength range relevant for the James Webb Space Telescope (JWST) NIRSpec instrument. By investigating the effects of realistic star formation histories and metallicity distributions on the EW(Hβ)-β diagram (previously proposed as a tool for identifying galaxies with very high f esc ), we find that neither of these effects are likely to jeopardize the identification of galaxies with extreme Lyman continuum leakage. Based on our models, we expect that essentially all z ≈ 7-9 galaxies that exhibit restframe EW(Hβ) 30Å to have f esc > 0.5. Incorrect assumptions concerning the ionizing fluxes of stellar populations or the dust properties of z > 6 galaxies can in principle bias the selection, but substantial model deficiencies of this type should at the same time be evident from offsets in the observed distribution of z > 6 galaxies in the EW(Hβ)-β diagram compared to the simulated one. Such offsets would thereby allow JWST/NIRSpec measurements of these observables to serve as input for further model refinement.