Key points are not available for this paper at this time.
We present Lya luminosity function (LF), clustering measurements, and Lya line profiles based on the largest sample, to date, of 207 Lya emitters (LAEs) at z=6. 6 on the 1-deg² sky of Subaru/XMM-Newton Deep Survey (SXDS) field. Our z=6. 6 Lya LF including cosmic variance estimates yields the best-fit Schechter parameters of phi*=8. 5 +3. 0/-2. 2 x10^ (-4) Mpc^ (-3) and L* (Lya) =4. 4 +/-0. 6 x10⁴2 erg s^ (-1) with a fixed alpha=-1. 5, and indicates a decrease from z=5. 7 at the >~90% confidence level. However, this decrease is not large, only =~30% in Lya luminosity, which is too small to be identified in the previous studies. A clustering signal of z=6. 6 LAEs is detected for the first time. We obtain the correlation length of r₀=2-5 h^ (-1) Mpc and bias of b=3-6, and find no significant boost of clustering amplitude by reionization at z=6. 6. The average hosting dark halo mass inferred from clustering is 10¹0-10¹1 Mo, and duty cycle of LAE population is roughly ~1% albeit with large uncertainties. The average of our high-quality Keck/DEIMOS spectra shows an FWHM velocity width of 251 +/-16 km s^ (-1). We find no large evolution of Lya line profile from z=5. 7 to 6. 6, and no anti-correlation between Lya luminosity and line width at z=6. 6. The combination of various reionization models and our observational results about the LF, clustering, and line profile indicates that there would exist a small decrease of IGM's Lya transmission owing to reionization, but that the hydrogen IGM is not highly neutral at z=6. 6. Our neutral-hydrogen fraction constraint implies that the major reionization process took place at z>~7.
Ouchi et al. (Fri,) studied this question.