[论文解读] GALAXIES AT Z 6: THE REST-FRAME UV LUMINOSITY FUNCTION AND LUMINOSITY DENSITY FROM 506 UDF, UDF-PS, AND GOODS I-DROPOUTS

M. Franx|arXiv (Cornell University)|Sep 21, 2005

Galaxies: Formation, Evolution, Phenomena被引用 1

一句话总结

本研究利用哈勃空间望远镜在HUDF、GOODS和UDF-Parallel区域的深度数据，首次对红移 z ≈ 6 处的506个 i- dropout 星系进行了全面分析。研究得出一个稳健的星等框紫外波段光度函数，其Schechter参数为 M∗ = 20.20 ± 0.35，φ∗ = 1.76⁺¹.¹⁶₋₀.⁷¹ × 10⁻³ Mpc⁻³，α = 1.74 ± 0.24，显示从 z ≈ 3 开始显著演化，其中 M∗ 明亮化了 0.7 ± 0.4 星等，且 z ≈ 6 时的光度密度仅为 z ≈ 3 时的68%，表明恒星形成率较低，对宇宙再电离过程具有重要启示。

ABSTRACT

A very large sample of 506 i-dropouts (z � 6 galaxies) have been obtained from all the deep, wide- area HST ACS fields: HUDF, GOODS (enhanced by the extensive SNe search data - to be the v2.0 GOODS release), and UDF-Parallel ACS fields (UDF-Ps). With our selection criteria, we show that the contamination levels are . 8% (i.e., & 92% or � 465 are at z � 6). This is the first comprehensive quantitative analysis of such a large sample at z � 6, only 0.9 Gyr from recombination, and is used to establish optimal measures of both the luminosity function and the luminosity density at z � 6, and their evolution relative to the z � 3 dropouts. A total of 122 i-dropouts ((i775 z850)AB > 1.3, (V606 z850)AB > 2.8) are detected in the HUDF to z850,AB � 29.5 (the 8� limit). Fits to the optical- infrared (ACS/NICMOS) colors of the HUDF dropout galaxies give mean UV slopes of � = 1.8. This is bluer than the dropout population at z � 3, suggesting lower dust content at z � 6. The sizes of the i-dropouts in our samples are small, with anL ∗ galaxy having rhl � 0.8 kpc or � 0.14 '' . There is clear size evolution, with a redshift dependence between z � 6 and z � 2.5 of (1+z) −1.1±0.3 for fixed luminosity. Significant differences were found in the surface density of i-dropouts between the fields, as expected from cosmic variance. Great care was taken to normalize out these differences. Completeness, flux, and contamination corrections are performed in a very systematic and quantitative way for the three data sets. After establishing and performing these corrections, we combine the data to derive a rest-frame continuum UV (� 1350u) luminosity function at z � 6. The best-fit Schechter parameters are M ∗ 1350,AB = 20.20±0.35, � = 1.74±0.24, and � ∗ = 1.76 +1.16 −0.71 ×10 −3 Mpc −3 . This luminosity function extends to M1350,AB � 17.5 (0.04 L ∗=3 ), the faintest limit yet reached for galaxies at high redshifts z > 2. We find strong evidence for evolution of the luminosity function between z � 6 and z � 3. The most likely way of accommodating this evolution is through a brightening (0.7 ± 0.4 mag) of M ∗ (at 82% confidence) though less brightening is required if the faint-end slopechanges to � � 1.9 (from 1.6 at z � 3). Scenarios, such as density evolution (� ∗ ), which do not include this evolution in M ∗ orare ruled out at 99.9999% confidence - demonstrating quite significantly that galaxies at z � 6 are lower in luminosity than galaxies at z � 3. Integrated down to 0.04L ∗=3, the rest-frame continuum UV luminosity density (unextincted star formation rate) at z � 6 is a very well-determined 1.46±0.14×10 26 ergs/s/Hz/Mpc 3 . Recent determinations of the luminosity density at z � 6 have shown large variations - we identify some of the causes and find that other results are consistent when appropriately corrected. The substantial evolution in the luminosity function is not accompanied by a large change in the luminosity density. The luminosity density at z � 6 (to 0.04L ∗=3 ) is 0.68 ± 0.08× that at z � 3. Accounting for the suggested evolution in dust content over this range indicates that the true evolution is substantially larger than for the unextincted star formation rate, so that the total star formation rate density at z � 6 is just � 0.3× the z � 3 value. Despite large uncertainties in the escape fraction and other assumed quantities, our best-fit UV luminosity function is consistent with z � 6 galaxies providing the necessary UV flux to reionize the universe. Lower luminosity galaxies, in particular, appear to be important for this process. Subject headings: galaxies: evolution — galaxies: high-redshift

研究动机与目标

• 以高统计显著性测量 z ≈ 6 处的星等框紫外光度函数与光度密度。
• 量化在深度HST场中i- dropout选星方法的污染与完整度。
• 评估高红移星系的大小演化与紫外谱指数趋势。
• 通过将 z ≈ 6 星系的光度函数与星暴率与 z ≈ 3 的结果进行比较，评估其在宇宙再电离中的作用。

提出的方法

• 利用HUDF、GOODS和UDF-Parallel区域的多波段ACS与NICMOS数据，识别出506个i- dropout星系。
• 应用严格的颜色与星等选择标准（i775−z850 > 1.3，V606−z850 > 2.8），将污染控制在<8%以内，最大化 z ≈ 6 的可靠性。
• 对三个独立数据集系统性地校正完整度、流量与污染，以实现稳健组合。
• 通过拟合观测的光学-红外颜色，推导出平均紫外谱指数（⟨β⟩ = −1.8），表明尘埃含量较低。
• 使用Schechter函数拟合星等框紫外光度函数：ϕ(L)dL = ϕ∗(L/L∗)α exp(−L/L∗)d(L/L∗)，其中 M∗、α 和 φ∗ 为自由参数。
• 通过归一化场间差异并校正尘埃与完整度，考虑宇宙方差的影响，推导出光度密度。

实验结果

研究问题

• RQ1 z ≈ 6 处的星等框紫外光度函数是什么？与 z ≈ 3 的光度函数相比有何不同？
• RQ2 z ≈ 6 星系的光度与大小相对于 z ≈ 3 有何演化？
• RQ3 z ≈ 6 星系对总紫外光度密度及星际介质再电离的贡献如何？
• RQ4 z ≈ 6 星系的尘埃含量与紫外谱指数与 z ≈ 3 相比如何？这对尘埃消光与恒星形成率有何启示？
• RQ5 z ≈ 6 处观测到的光度函数是否足以提供再电离宇宙所需的紫外辐射通量？

主要发现

• z ≈ 6 处的星等框紫外光度函数最佳拟合Schechter参数为 M∗ = 20.20 ± 0.35，α = 1.74 ± 0.24，φ∗ = 1.76⁺¹.¹⁶₋₀.⁷¹ × 10⁻³ Mpc⁻³。
• 光度函数延伸至 M1350,AB ≈ 17.5（0.04 L∗=3），为目前 z > 2 星系所能达到的最暗极限。
• 有充分证据表明光度函数发生演化，M∗ 从 z ≈ 3 到 z ≈ 6 明亮化了 0.7 ± 0.4 星等（置信度82%）。
• z ≈ 6 处的光度密度（至 0.04L∗=3）为 1.46 ± 0.14 × 10²⁶ erg/s/Hz/Mpc³，即 z ≈ 3 时值的 68% ± 8%。
• 在考虑尘埃演化校正后，z ≈ 6 处的光度密度仅约为 z ≈ 3 时的30%，表明恒星形成率密度显著下降。
• 最佳拟合的紫外光度函数表明，z ≈ 6 星系提供的紫外通量足以支持宇宙再电离，尤其在包含暗淡星系时更为成立。