JWST observations of photodissociation regions: II. Warm molecular Hydrogen spectroscopy in the Horsehead nebula

TL;DR

JWST observations of the Horsehead PDR reveal that H2 emission originates from a thin, edge-adjacent layer with three dissociation fronts, showing a spatial separation between FUV-pumped and collisionally excited lines. Through NIRSpec and MIRI-MRS spectroscopy, the study derives an extinction profile across the PDR, two-component H2 excitation with $T_{obs} \approx 512 \pm 19\ \mathrm{K}$ (DF1) and $478 \pm 12\ \mathrm{K}$ (DF2), and a lower-limit thermal pressure at the edge of $P_{\rm gas} \geq 6\times10^{6}\ \mathrm{K\,cm^{-3}}$, indicating overpressure relative to the adjacent H II region. Template 1D stationary PDR models fail to reproduce the observed temperatures and excitation, pointing to missing heating terms or strong dynamical effects such as photoevaporation-induced mixing at the PDR front. The results emphasize the crucial role of 2D geometry and dynamical processes in PDR physics and motivate advanced dynamical thermochemical modeling to interpret H2 emission in illuminated molecular clouds.

Abstract

H2 is the most abundant molecule in the interstellar medium and is a useful tool to study photodissociation regions, where radiative feedback from massive stars on molecular clouds is dominant. The James Webb Space Telescope, with its high spatial resolution, sensitivity, and wavelength coverage provides unique access to the detection of most of H2 lines and the analysis of its spatial morphology. Our goal is to use H2 line emission detected with the JWST in the Horsehead nebula to constrain the physical parameters (e.g., extinction, gas temperature, thermal pressure) throughout the PDR and its geometry. The study of H2 morphology reveals that FUV-pumped lines peak closer to the edge of the PDR than thermalized lines. From H2 lines, we estimate the value of extinction throughout the PDR. We find that AV is increasing from the edge of the PDR to the second and third H2 filaments. Then, we study the H2 excitation in different regions across the PDR. The temperature profile shows that the observed gas temperature is quite constant throughout the PDR, with a slight decline in each of the dissociation fronts. This study also reveals that the OPR is far from equilibrium. We observe a spatial separation of para and ortho rovibrational levels, indicating that efficient ortho-para conversion and preferential ortho self-shielding are driving the spatial variations of the OPR. Finally, we derive a thermal pressure in the first filament around P > 6x10$^6$ K cm$^{-3}$, about ten times higher than that of the ionized gas. We highlight that template stationary 1D PDR models cannot account for the intrinsic 2D structure and the very high temperature observed in the Horsehead nebula. We argue the highly excited, over-pressurized H2 gas at the edge of the PDR interface could originate from the mixing between the cold and hot phase induced by the photo-evaporation of the cloud.

Figures (20)

• Figure 1: (Top) JWST NIRCam RGB image of the Horsehead nebula, located in the Orion molecular cloud. Red is the 3.35 $\upmu$m emission (F335M NIRCam filter), blue is the emission of Pa$\alpha$ (F187N filter), and green is the emission of the H$_{\rm 2}$ 1--0 S(1) line at 2.12 $\upmu$m (F212N filter). (Left) The field of view of NIRSpec is overlayed in magenta on the image. (Right) The field of view of the different channels of MIRI-MRS is overlaid on the image (channel 1: blue, channel 2: green, channel 3: yellow, channel 4: red). The black boxes correspond to the aperture used to derive the spectra in the dissociation front regions and the black dashed boxes correspond to the aperture in the "molecular" region behind DF1 defined in misselt_jwst_2025. The dashed line corresponds to the position of cut #3 from abergel_jwst_2024. (Bottom) JWST NIRCam and MIRI-MRS composite image of the Horsehead nebula, zoomed on the edge where the faint striated features, attributed to an evaporative flow, are more visible. This image is rotated by 90$\degree$ with respect to the top panel. Credit: ESA/Webb, NASA, CSA, K. Misselt (University of Arizona) and A. Abergel (IAS/University Paris-Saclay, CNRS).
• Figure 2: NIRSpec (top) and MIRI-MRS (bottom) spectrum averaged on the first filament, DF1. Red lines (resp. blue lines) correspond to the detected rotational transitions (resp. rovibrational transitions) of H$_{\rm 2}$. Most of the lines detected in the dissociation front are attributed to H$_{\rm 2}$. The identification of other lines can be found in misselt_jwst_2025.
• Figure 3: Maps of the brightest H$_{\rm 2}$ rotational lines emission and the rovibrational lines 1--0 S(1) and 2--1 S(1) emission obtained with MIRI/MRS and NIRSpec across the PDR front. White contours are from the 0--0 S(1) line emission.
• Figure 4: Normalized intensity (around 0.5 and 3" around the first peak) profiles across the front abergel_jwst_2024 averaged on 0.5" perpendicular to the line cut. The illuminating star is on the right. (Top) Intensities not corrected for extinction. (Middle and bottom) Intensities corrected for extinction using the attenuation profile derived in Sect. \ref{['extinction']} using a parametrized extinction curve of gordon_one_2023 at $R_V = 3.1$ (middle) and $R_V = 5.5$ (bottom). The filled areas are uncertainties. They are particularly important after 15" because the estimation of extinction is very uncertain due to the very low signal-to-noise ratio of the NIR lines used to derive it.
• Figure 5: Profile of $A_V$, the attenuation by the foreground matter, across the PDR front derived from H$_{\rm 2}$ line ratios. The signal was averaged on the width of the maps and over four columns of pixels to compute H$_{\rm 2}$ maps with sufficient SNR. Before 1" and above 14" from the front, H$_{\rm 2}$ emission is too faint to derive $A_V$. $A_V$ is increasing from the edge of the PDR to the second and third H$_{\rm 2}$ filament.