Preview

PDF, English Download (4MB) | Terms of use

Abstract

The small-scale structure of the intergalactic medium (IGM) is fundamental to our understanding of cosmology and structure formation. Although the baryons trace dark matter fluctuations on megaparsec scales, on small scales ($\sim 100$ kpc), gas perturbations are regulated by hydrodynamics and they are thought to be suppressed by pressure below a characteristic \emph{filtering scale} $\lambda_J$, analogous to the classic Jeans scale. The value of this Jeans filtering scale is set by the interplay between pressure support and gravity across the cosmic history, and has fundamental cosmological implications. First it provides a thermal record of heat injected by ultraviolet photons during cosmic reionization events, and thus constraints the thermal and reionization history of the universe. Second, it determines the clumpiness of the IGM and the minimum mass for gravitational collapse from the IGM, playing a pivotal role in galaxy formation and reionization. In principle, the sign of Jeans smoothing could be probed by the redshifted \mlya\ absorption lines in the spectra of high-redshift quasars (The \mlya\ forest). Unfortunately, this is extremely challenging to do because the thermal Doppler broadening of \mlya\ lines along the observing direction is highly degenerate with pressure smoothing. In this work, I explicitly show what degeneracies hold among the thermal parameters of the IGM when only line-of-sight observations are possible. For this purpose, I devised a rigorous statistical algorithm based on Gaussian processes and Markov-Chain Monte Carlo methods, trained on a grid of semianalytical models of the IGM. I then introduce a novel method able to measure the Jeans scale by estimating the transverse coherence in the spectra of close quasar pairs (transverse separation $r_{\perp}< 1$ Mpc). This method is based on the phase differences of homologous Fourier modes in the \mlya\ forests of quasar pairs, and I prove that it is maximally sensitive to $\lambda_J$ and only weakly dependent on the other considered parameters. The available sample of quasar pairs is analyzed, after careful calibration of noise, resolution, and other possible systematics. Our new method applied to this dataset provides the first measurement of the filtering scale of the intergalactic medium. A first comparison of our findings with hydrodynamical simulations suggests that the filtering scale predicted by the standard thermal models of the IGM is significantly higher than what we observe, motivating further theoretical studies to understand this discrepancy.