The unprecedentedly bright optical afterglow of GRB 130606A located by Swift at a redshift close to the reionization era (z = 5.913) provides a new opportunity to probe the ionization status of intergalactic medium (IGM). Here we present an analysis
of the red Ly alpha damping wing of the afterglow spectrum taken by Subaru/FOCAS during 10.4-13.2 hr after the burst. We find that the minimal model including only the baseline power-law and HI absorption in the host galaxy does not give a good fit, leaving residuals showing concave curvature in 8400-8900 A with an amplitude of about 0.6% of the flux. Such a curvature in the short wavelength range cannot be explained either by extinction at the host with standard extinction curves, intrinsic curvature of afterglow spectra, or by the known systematic uncertainties in the observed spectrum. The red damping wing by intervening HI gas outside the host can reduce the residual by about 3 sigma statistical significance. We find that a damped Ly alpha system is not favored as the origin of this intervening HI absorption, from the observed Ly beta and metal absorption features. Therefore absorption by diffuse IGM remains as a plausible explanation. A fit by a simple uniform IGM model requires HI neutral fraction of f_HI ~ 0.1-0.5 depending on the distance to the GRB host, implying high f_HI IGM associated with the observed dark Gunn-Peterson (GP) troughs. This gives a new evidence that the reionization is not yet complete at z ~ 6.
We examine variability of absorption line strength of intervening systems along the line of sight to GRB 060206 at $z = 4.05$, by the low-resolution optical spectra obtained by the Subaru telescope from six to ten hours after the burst. Strong variab
ilities of Feemissiontype{II} and Mgemissiontype{II} lines at $z=1.48$ during $t=$5--8 hours have been reported for this GRB citep{Hao07}, and this has been used to support the idea of clumpy Mgemissiontype{II} cloudlets that was originally proposed to explain the anomalously high incidence of Mgemissiontype{II} absorbers in GRB spectra compared with quasars . However, our spectra with higher signal-to-noise ratio do not show any evidence for variability in $t=$6-10 hours. There is a clear discrepancy between our data and Hao et al. data in the overlapping time interval. Furthermore, the line strengths in our data are in good agreement with those observed at $t sim$ 2 hours by citet{Tho08}. We also detected Feemissiontype{II} and Mgemissiontype{II} absorption lines for a system at $z = 2.26$, and these lines do not show evidence for variability either. Therefore we conclude that there is no strong evidence for variability of intervening absorption lines toward GRB 060206, significantly weakening the support to the Mgemissiontype{II} cloudlet hypothe sis by the GRB 060206 data.
