Improving Accretion Diagnostics for Young Stellar Objects with Mid-infrared Hydrogen lines from JWST/MIRI

TL;DR

This study leverages JWST/MIRI spectroscopy to extend accretion diagnostics into mid-infrared neutral hydrogen lines for 79 nearby young stellar objects. By detecting 22 H I transitions and meticulously removing H2O contamination with LTE slab modelling, the authors derive updated L_acc–L_line relations for six MIR HI lines and identify robust, contamination-resistant tracers at 7.502 μm (8–6) and 8.760 μm (10–7). MIR line ratios analyzed with Kwan2011 models constrain the hydrogen density in accretion columns to $n_{ m H} obreak= obreak 10^{10.6}$–$10^{11.2}$ cm$^{-3}$ for most sources, with some approaching optically thin conditions. The results demonstrate the power of JWST to provide clean, high-sensitivity accretion diagnostics in disks, including embedded systems, and highlight the value of combining multi-wavelength HI lines to break degeneracies between temperature and density.

Abstract

We present a comprehensive study of mid-infrared neutral hydrogen (H~\textsc{i}) emission lines in 79 nearby (d $<$ 200 $pc$) young stars using JWST/MIRI. We aim to identify mid-infrared H~\textsc{i} transitions that can serve as reliable accretion diagnostics in young stars, and evaluate their utility in deriving physical conditions of the accreting gas. We identify and measure 22 H~\textsc{i} transitions in the MIRI wavelength regime (5-28 $μm$) and perform LTE slab modelling to remove the H\textsubscript{2}O contribution from selected H~\textsc{i} transitions. We find that mid-IR H~\textsc{i} line emission is spatially compact, even for sources with spatially extended [Ne~\textsc{ii}] and [Fe~\textsc{ii}] jets, suggesting minimal contamination from extended jet. Although Pfund~$α$ (H~\textsc{i}~6--5) and Humphreys~$α$ (H~\textsc{i}~7--6) are the strongest lines, they are blended with H$_2$O transitions. This blending necessitates additional processing to remove molecular contamination, thereby limiting their use as accretion diagnostics. Instead, we identify the H~\textsc{i}~(8--6) at 7.502 $μm$ and H~\textsc{i}~(10--7) at 8.760 $μm$ transitions as better alternatives, as they are largely unaffected by molecular contamination and offer a more reliable means of measuring accretion rates from MIRI spectra. We provide updated empirical relations for converting mid-IR H~\textsc{i} line luminosities into accretion luminosity for 6 different H~\textsc{i} lines in the MIRI wavelength range. Moreover, comparison of observed line ratios with theoretical models shows that MIR H~\textsc{i} lines offer robust constraints on the hydrogen gas density in accretion columns, $n_\mathrm{H} = $10$^{10.6}$ to 10$^{11.2}$ cm$^{-3}$ in most stars, with some stars exhibiting lower densities ($<10^{10}$~cm$^{-3}$), approaching the optically thin regime.

Figures (13)

• Figure 1: Continuum-subtracted JWST/MIRI spectrum of FT Tau. The main panel shows the full MIRI spectrum from 5 to 24 $\mu m$, with the positions of 22 H i transitions marked. Sub-panels (a) through (u) show Gaussian fits (colored curves) to individual H i line profiles used in the analysis. The green dashed horizontal line indicates the $10 \times \mathrm{RMS}$ threshold used for line identification.
• Figure 2: H i line detection statistics. Vertical stacked bar plot showing the detection statistics of H i emission lines in our sample. Dark-green bars denote lines detected in more than 40 sources, which are analysed in detail in this work. Light-green bars correspond to lines detected in fewer than 40 sources, and red-hatched bars indicate the number of non-detections for each line.
• Figure 3: Heatmap showing the overlap of molecular transitions with the H i. Each cell represents the number of molecular lines (E$_{\rm up} < 5000$ K) from species such as H$_2$O, HCN, CH$_4$, C$_2$H$_2$, CO, and CO$_2$ that fall within $\pm1\,\Delta\lambda$ of the rest wavelength of a given H i transition. Darker colors indicate a higher number of overlapping molecular features, with H i (13--9) exhibiting the strongest contamination, primarily from CH$_4$ and H$_2$O. Based on our analysis, lines such as H i (9--6), H i (10--6), and H i (10--7) do not show significant molecular contamination.
• Figure 4: Representative best-fit LTE spectra for three Class II disks. Sz 114 (top row), FZ Tau (middle row), and FT Tau (bottom row), showing varying levels of water contribution to H i (7--6), H i (6--5), and H i (12--7) lines. In each panel, the black curve represents the continuum-subtracted JWST/MIRI spectrum, while the coloured curve shows the best-fit LTE H$_2$O emission model. Inset boxes indicate the fitted physical parameters: temperature ($T$), column density ($\log(N)$), and emitting radius ($R$) for the water model. Other H i lines are annotated in each of the sub-panels.
• Figure 5: Contribution of H2O to 3 H i lines. Left panels: Estimated percentage contribution of H$_2$O emission to the observed H i transitions—H i (12--7), H i (7--6), and H i (6--5)—across the Class II disk sample. Blue stars indicate the fractional H$_2$O contribution for each source. Green upward arrows denote sources where the entire flux is attributable to H$_2$O, while red downward arrows mark sources for which H$_2$O modelling was not applied due to either dominant H i flux or negligible H$_2$O emission. Right panels: Violin plots showing the distributions of best-fit physical parameters from LTE H$_2$O models for each wavelength region. The temperature and effective emitting radius exhibit an anti-correlation, consistent with compact hot inner disk emission dominating at shorter wavelengths. The H$_2$O column densities remain relatively constant across the different spectral regions.