IfA Home IGM temperature

Simulations of the Lyman-α forest at ECCA

Related Links:

IGM simulations
CGM simulations
Baryonic power spectrum
Metagalactic UV background
IGM transmissivity
IGM metallicity
21cm signature


ECCA Research

Structure of the Intergalactic Medium

Numerical simulations of the structure of the Intergalactic Medium in the context of Cold Dark Matter (CDM) dominated models show that the most of the Universe is filled with a weblike distribution of dark matter and baryons of low to moderate overdensities. While the highest density regions form virialised haloes that become the sites of galaxy formation, moderately overdense structures form filaments connected at the haloes. Suspended between the filaments are sheets with densities comparable to the mean density of the Universe. The sheets in turn enclose low density regions.

The Lyman-α Forest
The structures seen in simulations are not directly observable, since they produce no light. Instead they are detected through the formation of absorption features in the spectra of luminous radiation sources like Quasi-Stellar Objects (QSOs). The neutral Hydrogen scatters out radiation at wavelengths corresponding to the Lyman resonance transitions. The most prominent of these is the Lyman-α transition. Because the Universe is expanding, the observed wavelength at which the light is scattered out from the background object is redshifted, depending on the distance to the hydrogen in the IGM doing the scattering. The result is a choppy series of absorption troughs in the QSO spectrum, known as the Lyman-α Forest. The absorption features are characterised by their widths and neutral hydrogen column densities, as well as by the light fluctuation in any given pixel of the QSO spectrum. The statistical properties of these quantities are used to test cosmological structure formation models.

Numerical simulations have been broadly successful at reproducing the observed properties of the IGM as measured in QSO spectra (e.g., Zhang et al. 1997, 1998; Meiksin et al. 2001). Most striking is the agreement between the predicted and measured flux per pixel distributions, with the models recovering the measured cumulative distributions to a precision of 3-5% (Figure 1) . This is an important success of the CDM model for structure formation, especially in that no assumptions relating the scattered light to the dark matter are required, except for the intensity of the UV background, which may be determined by fitting to the mean transmission of the QSO light through the IGM. This contrasts with galaxy simulations, all of which may be compared with observations only through extensive additional modelling of the amount of light produced by stars inside the collapsed dark matter haloes. IGM simulations instead are more of a non-linear extension of predictions for Cosmic Microwave Background (CMB) fluctuations, bridging the linear density fluctuation regime of the CMB with the highly nonlinear fluctuation regime of galaxies. The IGM therefore provides an independent and complementary means of testing cosmological structure formation scenarios and constraining cosmological parameters.

The best-fitting flat universe models impose the dark matter density constraint 0.26 < ΩM < 0.43 (Meiksin et al. 2001). The measured distributions of absorption line parameters are moderately well recovered by the simulations, but statistically significant differences remain. The measured widths of the absorption features are significantly broader than the models predict (Figure 2) . The disagreement in linewidths is particularly pronounced for the weakest Ly&alpha absorbers, those optically thin at their line centre (Figure 3) , suggesting that the discrepancy is indicative of gas that is warmer than the model predictions throughout the IGM, as these more rarefied systems occupy the underdense regions between the galaxies (Zhang et al. 1998). A likely explanation is a combination of additional heat input during the reionisation process itself, as well as the possibility that HeII was re-ionized late (at z < 3.5), resulting in a boost to the IGM temperature through the additional photoelectric heating. Estimates show this brings the distributions into much better agreement (Figure 4) . Ultimately, precise modelling of of the widths of the absorption features requires the inclusion of radiative transfer into the reionisation calculation.


Meiksin A., Bryan G., Machacek M., 2001, MNRAS, 327, 296

Zhang Y., Anninos P., Norman M.L., Meiksin A., 1997, ApJ, 485, 496

Zhang Y., Meiksin A., Anninos P., Norman M.L., 1998, ApJ, 495, 63

Return to Astrophysical fluids