ABSTRACT

We present a computationally efficient and fast semi-numerical technique for simulating the Lyman α (Ly α) absorption optical depth in presence of neutral hydrogen ‘islands’ left over from reionization at redshifts 5 ≲ z ≲ 6. The main inputs to the analysis are (i) a semi-numerical photon-conserving model of ionized regions during reionization (named script) along with a prescription for simulating the shadowing by neutral islands and (ii) the fluctuating Gunn–Peterson approximation to model the Ly α absorption. Our model is then used for simulating the large-scale fluctuations in the effective optical depth as observed along sightlines towards high-z quasars. Our model is fully described by five parameters. By setting two of them to default values and varying the other three, we obtain the constraints on reionization history at 5 ≲ z ≲ 6 as allowed by the data. We confirm that reionization is not complete before z ∼ 5.6 at ≳2σ confidence, with the exact confidence limits depending on how the non-detections of the flux in the data are treated. We also confirm that the completion of reionization can be as late as z ∼ 5.2. With further improvements in the model and with more sightlines at z ∼ 6, we can take advantage of the computational efficiency of our analysis to obtain more stringent constraints on the ionization fraction at the tail end of reionization.

中文翻译：

---

使用快速半数值方法研究z〜5.5处的Lymanα光学深度波动

摘要

我们提出了一个计算上有效的和快速的半数值技术用于模拟莱曼 α（Ly的 α在红移5个≲在从再电离遗留中性氢“孤岛”的存在下）的吸收光深度Ž ≲6.主输入到分析是（i）与用于通过中性岛屿和（ii）所述波动耿氏彼得森近似模拟遮蔽到Ly的建模处方沿再电离（名为脚本）中电离区域的半数值光子节约模型 α吸收。然后，我们的模型用于模拟沿视线向高Z方向观察到的有效光学深度的大规模波动。类星体。我们的模型由五个参数充分描述。通过设置他们两个到默认值和改变其他三个，我们在5≲获得再电离历史的约束ž ≲6所允许的数据。我们确认，再电离是不完整的前ž〜5.6和≳2 σ信心，与这取决于如何在数据通量的未查出被视为确切的置信区间。公司谨确认，再电离的完成可迟至Z ^〜5.2。随着模型的进一步改进和z处更多视线 〜6，我们可以利用分析的计算效率来获得对电离尾部电离分数的更严格的约束。