The effect of spatial correlations in a chaotic velocity field on the D/H measurements from QSO absorption spectra
We describe numerical simulations of hydrogen and deuterium Lyalpha lines formed in turbulent media with a stochastic velocity field. The line broadening is assumed to be the result of both macroscopic and thermal motions, and the line profile function Phi(Deltalambda) is a random function of coordinate along the line of sight. This is in contrast to the standard assumption of the microturbulence, where Phi(Deltalambda) is a given quantity and any bulk motions are treated as completely uncorrelated. The randomization of Phi(Deltalambda) stems from the chaotic character of the velocity field, which results from the Doppler shifts in the absorption coefficient. Numerical results are presented for models with N_HI=10^16, 10^17 and 10^18 cm^-2 D/H=2x10^-5, 4x10^-5, 10^-4 and 3x10^-4 rms turbulent velocities sigma=8, 12 and 16 km s^-1 kinetic temperature T_kin=5000K and different velocity correlation lengths. A finite correlation length is shown to affect strongly the intensity ratio of DI to HI Lyalpha lines. In particular, uncertainties of the D/H ratio determination may be as large as a few orders of magnitude. It is important that the actual D/H ratios may appear to be higher or lower than the values obtained from the standard Voigt-fitting procedure applied to our simulated mesoturbulent H+D Lyalpha spectra. We suggest that deuterium absorption lines should be observable in the Lyalpha forest systems with the apparent hydrogen column density, which is obtained in the conventional microturbulent model fitting, N_HI,micro~10^14-10^15 > cm^-2. The actual HI densities in these systems may be 10-100 times higher if the clouds of the Lyalpha forest have correlated large-scale motions. An example is presented based on high-resolution observations of Q1100-264 by Carswell et al.: two Lyalpha forest lines at z=2.1029 and 2.1037 can be re-identified as DI and HI Lyalpha, respectively, at z=2.1037. We conclude that the precise measurements of the D abundance are actually impossible without thorough investigation of the turbulent characteristics of the gas for each absorption system.