The reanalysis of spectra of GRB 080913 to estimate the neutral fraction of the IGM at a redshift of 6.7

TL;DR

This study reanalyzed spectra of GRB 080913 at redshift 6.7 with improved data quality to better estimate the neutral hydrogen fraction of the early universe's IGM, but found limited constraints due to spectral complexities.

Contribution

It provides a refined spectral analysis with higher S/N and explores the impact of ionized regions on measuring the IGM's neutral fraction at high redshift.

Findings

Identified a significant absorption line at z=6.733.

Established an upper limit of x(HI) < 0.73 for the IGM neutral fraction.

Determined the ionized region size r < 2 proper Mpc.

Abstract

Aims: We reanalyse optical spectra of the z=6.7 gamma-ray burst GRB 080913, adding hitherto unpublished spectra, in order to reassess the measurement of the neutral fraction of the IGM at high redshifts. Methods: In the data reduction we take particular care to minimise systematic errors in the sky subtraction, which are evident in the published spectrum, and compromise the analysis. The final combined spectrum has a higher S/N than the previously published spectrum by a factor of 1.3. Results: We find a single significant absorption line redward of the Ly-alpha continuum break, which we identify with the SII+SiII blend (rest wavelength of 0.1260 micron), at z=6.733. The sharp spectral break at Ly-alpha implies a comparatively low total column density of neutral hydrogen along the line of sight, log [N(HI)/cm^-2] < 20. We model the absorption with a host-galaxy DLA, surrounded by an ionised region of unknown size r, within the IGM of neutral fraction, x(HI). Despite knowing the source redshift, and the improved S/N of the spectrum, when fitting only over wavelengths redward of Ly-alpha, no useful constraints on x(HI) can be obtained. We consider the possibility of including the ionised region, blueward of Ly-alpha, in constraining the fit. For the optimistic assumption that the ionised region is transparent, tau_{GP} << 1, we find that the region is of small size r < 2 proper Mpc, and we obtain an upper limit to the neutral fraction of the IGM at z=6.7 of x(HI) < 0.73.