IPA: Morphology and Kinematics of Molecular Hydrogen Winds in Five Young Protostars across the Mass Spectrum Observed with JWST

Based on the paper, here is the summary:

Molecular winds may play a key role in governing angular momentum transport and accretion during the early evolution of protostars. We present the morphology and kinematic properties of the H2 emission in five young, envelope-dominated, protostars across a broad bolometric luminosity range, from 0.2 to 104 L⊙, observed with the NIRSpec/IFU and MIRI/MRS onboard JWST as part of the Investigating Protostellar Accretion (IPA) program. A rich set of pure rotational lines of H2, up to v = 0 − 0 S(18), and a few ro-vibrational lines are detected in the winds, revealing bipolar structures. The H2 lines show a stratified/onion-like structure morphologically and kinematically, where the lines with higher Eu /k show a higher degree of collimation and higher velocities. Additionally, the wind velocity scales with the Lbol of the host protostellar system. In 4 out of 5 protostars, H2 emission fills the outflow cavity without showing pronounced limb brightening. We also report a tentative detection of H2 wind rotation in IRAS 16253, which suggests a launch radius of ∼ 4 au and the magnetic lever arm parameter of ∼ 5 − 10. Taken together, these properties of the H2 winds can be explained by the magnetohydrodynamic disk wind models. We detect a collimated, high-velocity H2 jet toward HOPS 370, which is more evolved than the extremely young source HH 211, but is accreting at a high accretion rate. This suggests that the presence of collimated molecular jets in protostars is more closely connected to accretion rate than system age.

Improvements to AI Systems:

1. Astrophysical Kinematic Modeling: Enhance AI models to automatically reconstruct 3D outflow geometries from spectral line data (e.g., H2 emission) by incorporating onion-like stratification rules—where higher excitation energy lines correspond to narrower, faster inner jets. The improved AI can predict velocity–collimation coupling from input luminosity and infer launch radii and magnetic lever arms from rotational signatures.

2. Protostellar Evolution Classifier: Train a deep learning classifier to distinguish between accretion-rate-dominated and age-dominated outflow morphology using features like jet collimation, cavity filling, and luminosity. The improved system can flag young, high-accretion sources (e.g., HOPS 370) that exhibit jets typically seen in older systems, enabling automated identification of anomalous accretion states.

3. Multi-Wavelength Data Fusion: Build an AI pipeline that jointly analyzes NIRSpec/IFU and MIRI/MRS spectra to separate thermal (rotational) and non-thermal (ro-vibrational) H2 components. The improved system can produce automated excitation diagrams, derive temperature/column density maps, and detect weak rotational signatures (like the tentative wind rotation in IRAS 16253) with uncertainty quantification.

4. MHD Disk Wind Parameter Estimator: Develop a neural network that takes observed H2 wind velocity, collimation, and rotation profiles as input and outputs physical parameters (launch radius, magnetic lever arm, mass-loading) by fitting to MHD disk wind models. The improved AI can run real-time parameter inversion for any new protostar, replacing manual model fitting.

5. Outflow Cavity Morphology Segmenter: Enhance computer vision models to segment outflow cavities in IFU datacubes, distinguishing between limb-brightened and filled cavities. The improved system can automatically classify the 4/5 filled-cavity cases and flag outliers, aiding large-scale surveys of protostellar outflows.

What the Improved AI System Can Do:

• Automatically analyze JWST observations of protostellar winds to produce physical models (launch radii, magnetic lever arms, accretion rates) without human intervention.

• Predict whether a protostar’s jet is driven by accretion rate or age, enabling rapid triage of new sources for follow-up.

• Generate synthetic H2 emission maps from MHD simulations and compare them to observed data to test disk wind theories at scale.

• Provide real-time alerts for anomalous outflow kinematics (e.g., unexpected rotation or collimation) that might indicate unique magnetic field configurations.

Molecular winds may play a key role in governing angular momentum transport and accretion during the early evolution of protostars. We present the morphology and kinematic properties of the H 2 emission in five young, envelope-dominated, protostars across a broad bolometric luminosity range, from 0.2 to 10 4 L, observed with the NIRSpec/IFU and MIRI/MRS onboard JWST as part of the Investigating Protostellar Accretion (IPA) program. A rich set of pure rotational lines of H 2, up to v=0-0 S(18), and a few ro-vibrational lines are detected in the winds, revealing bipolar structures. The H 2 lines show a stratified/onion-like structure morphologically and kinematically, where the lines with higher E up show a higher degree of collimation and higher velocities. Additionally, the wind velocity scales with the L bol of the host protostellar system. In 4 out of 5 protostars, H 2 emission fills the outflow cavity without showing pronounced limb brightening. We also report a tentative detection of H 2 wind rotation in IRAS 16253, which suggests a launch radius of about4 au and the magnetic lever arm parameter of about5-10. Taken together, these properties of the H 2 winds can be explained by the magnetohydrodynamic disk wind models. We detect a collimated, high-velocity H 2 jet toward HOPS 370, which is more evolved than the extremely young source HH 211, but is accreting at a high accretion rate. This suggests that the presence of collimated molecular jets in protostars is more closely connected to accretion rate than system age.