Early spectra of AT2017gfo from scaled 3D merger ejecta: mass--velocity distribution, 3D structure, and treatment of line opacity
来自缩放三维并合抛射物的 AT2017gfo 早期光谱:质量–速度分布、三维结构与线不透明度处理
We present ARTIS radiative transfer calculations for neutron star merger ejecta with line-by-line opacities, wavelength-calibrated lanthanide data, and time-dependent r-process compositions and thermalisation, testing how the early kilonova depends on the mass-velocity distribution, dimensionality, and line treatment. We rescale 3D dynamical ejecta from a merger simulation in density and velocity to approximate AT2017gfo. Mass scaling alone provides good agreement with the bolometric light curve but inconsistent spectra. With velocities additionally halved, the model also reproduces the observed spectra at 1.4, 2.4, and (albeit less accurately) 3.4 days. The Sr II triplet contributes most to the 10500 $\mathring{\mathrm{A}}$ peak, and La III and Gd III to the 13500 $\mathring{\mathrm{A}}$ shoulder. Although hydrodynamically not fully self-consistent, this model is to our knowledge the first kilonova model based on calibrated atomic data that reproduces the light curve and main spectral features of the early AT2017gfo, suggesting that total (dynamical plus post-merger) ejecta with higher mass, lower velocity, and similar composition compared to our dynamical ejecta could explain AT2017gfo. A 2D azimuthal average mixes lanthanide-poor with lanthanide-rich regions, smooths the density, and makes the averaged model three times fainter at 1 day and redder throughout than the 3D model, highlighting the importance of 3D modelling. We compare the line-by-line macroatom treatment with expansion-opacity and line-binned schemes with thermal redistribution, using identical atomic data. Luminosities agree to $\lesssim 10$ per cent after 1 day, but the expansion-opacity scheme, unlike converged line-binned schemes, overestimates the spectral flux by up to 20 per cent, which can bias inferred abundances.
展开 ▾首个基于校准原子数据、同时大致复现 AT2017gfo 早期光度和主要光谱特征的多维千新星模型;发现展宽不透明度对光谱通量有最高 20% 系统高估,可能偏差元素丰度推断。
该工作延续了用 ARTIS 对中子星并合动力学抛射物做 3D 辐射转移的路线:Collins+ 2023 建立了原始 3D 模型,但在 AT2017gfo 早期偏暗、偏蓝且演化过快。作者借质量标度和速度减半来模仿缺失的后并合抛射物,并采用 Flörs+ 2026 的波长校准镧系原子数据,首次在 1.4–2.4 天同时接近观测光变曲线和主要光谱特征。为厘清数值处理影响,他们与 Fontes+ 2020 线分档方案及展宽不透明度对比,发现展宽方案使通量最多偏高 20%,而收敛的线分档与 macroatom 一致。未来需要与 Fernández+ 2013 等自洽的后并合外流模型及更多非 LTE、锕系校准数据结合,打破质量–速度–几何简并。
预印本 2026-11-18 · 收录 2026-11-19