The 21cm signal at epoch of reionization (EoR) should be observed within next decade. We expect that cosmic 21cm signal at the EoR provides us both cosmological and astrophysical information. In order to extract fruitful information from observation data, we need to develop inversion method. For such a method, we introduce artificial neural network (ANN) which is one of the machine learning techniques. We apply the ANN to inversion problem to constrain astrophysical parameters from 21cm power spectrum. We train the architecture of the neural network with 70 training datasets and apply it to 54 test datasets with different value of parameters. We find that the quality of the parameter reconstruction depends on the sensitivity of the power spectrum to the different parameter sets at a given redshift and also find that the accuracy of reconstruction is improved by increasing the number of given redshifts. We conclude that the ANN is viable inversion method whose main strength is that they require a sparse extrapolation of the parameter space and thus should be usable with full simulation.

The fraction (fesc) of ionizing photons that escape their host galaxy and so are able to ionize hydrogen in the intergalactic medium is a critical parameter in studies of the reionization era. In this paper we constrain fesc at high redshift by modelling the observed Lyα transmission. Using an N-body simulation of the cosmic web, we first model an ionizing background due to star-bursting galaxies. We then extract an ensemble of mock absorption spectra from the simulation and compute the mean transmission. We adjust the free parameter, fesc in our model of the ionizing background to reproduce the observed transmission. We find that fesc increases with the minimum galaxy mass considered to contribute to the ionizing background, and that at high redshift fesc must be greater than 5% irrespective of galaxy mass. Based on our numerical results we calibrate a semi-analytic model which shows that the ionizing background observed at z ∼ 5.5–6 implies a sufficient number of ionizing photons to have reionized the Universe by z ∼ 6. However, if the minimum mass for star-formation were ≳109 M⊙, the ionizing background would be over-produced at z ≲ 5. In summary, our results indicate that the Universe was reionized by low mass galaxies.

Bright point sources associated with extragalactic active galactic nuclei and radio galaxies are an important foreground for low-frequency radio experiments aimed at detecting the redshifted 21-cm emission from neutral hydrogen during the epoch of reionization. The frequency dependence of the synthesized beam implies that the sidelobes of these sources will move across the field of view as a function of observing frequency, hence frustrating line-of-sight foreground subtraction techniques. We describe a method for subtracting these point sources from dirty maps produced by an instrument such as the MWA. This technique combines matched filters with an iterative centroiding scheme to locate and characterize point sources in the presence of a diffuse background. Simulations show that this technique can improve the dynamic range of epoch-of-reionization maps by 2—3 orders of magnitude.

The Extragalactic Background Light (EBL) at UV, optical and NIR wavelengths consists of the integrated light of all unresolved galaxies along the line of sight plus any contributions by intergalactic matter including hypothetical decaying relic particles. The measurement of the EBL has turned out to be a tedious problem. This is because of the foreground components of the night sky brightness, much larger than the EBL itself: the Zodiacal Light (ZL), Integrated Starlight (ISL), Diffuse Galactic Light (DGL) and, for ground-based observations, the Airglow (AGL) and the tropospheric scattered light. We have been developing a method for the EBL measurement which utilises the screening effect of a dark nebula on the EBL. A differential measurement in the direction of a high-latitude dark nebula and its surrounding area provides a signal that is due to two components only, i.e. the EBL and the diffusely scattered ISL from the cloud. We present a progress report of this method where we are now utilising intermediate resolution spectroscopy with ESO's VLT telescope. We detect and remove the scattered ISL component by using its characteristic Fraunhofer line spectral signature. In contrast to the ISL, in the EBL spectrum all spectral lines are washed out. We present a high quality spectrum representing the difference between an opaque position within our target cloud and several clear OFF positions around the cloud. We derive a preliminary EBL value at 400 nm and an upper limit to the EBL at 520 nm. These values are in the same range as the EBL lower limits derived from galaxy counts.

Unit: We will use in this paper the abbreviation 1 cgs = 10−9erg s−1cm−2sr−1Å−1

Very high energy (VHE ≳0.1 TeV) gamma-rays from extragalactic sources, interacting by γ-γ collisions with diffuse intergalactic radiation fields, provide an alternative way to constrain the diffuse background light, completely independent of direct measurements. The limits depend however on our knowledge of the physics of the gamma-ray sources. After clarifying the interplay between background light and VHE spectra, I summarize the extent and validity of the obtainable limits, and where future improvements can be expected.

We present the detection of escaping Lyman Continuum (LyC) radiation from two local starburst galaxies, Tol 1247-232 and Tol 0440-381, using archival data from the Far Ultraviolet Spectroscopic Explorer (FUSE). From profile fitting of metal lines, high densities of neutral gas were derived for both galaxies, implying that the LyC radiation is escaping from holes in the interstellar medium. The first results for one of the galaxies, Tol 1247-232, give an escape fraction (fesc) of 17–35%. These new detections, together with that of the blue compact galaxy Haro 11 in Bergvall et al. (2006) where fesc between 4–10% was found, will have a feedback on cosmic reionization models.