Hydrodynamical simulations of the Lya forest: Data comparisons

Abstract

Numerical hydrodynamical simulations are used to predict the expected absorption properties of the Lyα forest for a variety of cold dark matter (CDM) dominated cosmological scenarios: CHDM, OCDM, ACDM, SCDM, and tCDM. Synthetic spectra are constructed duplicating the resolution, signal-to-noise ratio and wavelength coverage of several published high resolution spectra, and their statistical properties are compared on the basis of the flux distribution of the spectra, the distribution of coefficients in a wavelet decomposition of the spectra and the distributions of absorption line profile parameters. Agreement between the measured and predicted cumulative distributions is found at the few to several per cent level. The best-fitting models to the flux distribution correspond to normalizations on the scale of the cosmological Jeans length of 1.3 < σj < 1.7 at z = 3. No single model provides a statistically acceptable match to all the distributions. Significantly larger median Doppler parameters are found in the measured spectra than predicted by all but the lowest normalization models (CHDM and tCDM), which provide poor fits to the flux distributions. The discrepancy in Doppler parameters is particularly large for absorption systems optically thin at the Lyα line centre. This may indicate a need to introduce additional energy injection throughout the intergalactic medium, as may be provided by late He n reionization (ZHe II ≃ 3.5) or supernovae-driven winds from young galaxies, and/or a larger baryon fraction than given by recent determinations of the deuterium abundance within the context of standard Big Bang nucleosynthesis. The models require a hydrogen ionization rate at redshifts 1.7 < z < 3.8 within a factor of 2 of that predicted from quasi-stellar objects (QSOs) alone as the sources of the UV photoionization background, although with a slower rate of decline with redshift at z > 3.5 than predicted from current QSO counts. Principal systematic uncertainties in comparing the models with the observations are the setting of the continuum level of the QSO spectra and the prevalence of metal absorption lines, particularly at z < 3.