High-Redshift Galactic Outflows: Orientation Effects, Kinematics, and Metallicity in TNG50 and SERRA

Context: Recently, JWST/NIRSpec observations have provided the first detections of warm ionised outflows in low-mass galaxies at high redshifts (z>3), revealing an occurrence rate of 25-40% depending on the intensity of the emission lines. This fraction is lower than predicted by simulations, which suggest that fast outflowing gas should be a common feature of all star-forming galaxies in the early Universe. Aims: In order to better understand the discrepancies between simulations and observations, we identify and characterize outflows in high-redshift galaxies using the TNG50 cosmological and SERRA zoom-in simulations. Our study examines how outflow detectability depends on the line of sight, explores the properties of the fast gas, and investigates its relationship with key galactic properties. Methods: We analyse approximately 60000 galaxies from TNG50 and 3000 galaxies from SERRA over the redshift ranges z=3-5 and z=4-5, respectively, spanning stellar masses of Mstar=10^7.5-10^11Msun. Outflows in the immediate vicinity of each galaxy are identified using a Gaussian mixture model algorithm that uses the gas velocity, star-formation-rate, and location as input parameters. We subsequently compare the simulated outflows to those observed in the JWST/JADES NIRSpec survey. Results: Outflow masses in both TNG50 and SERRA broadly reproduce the JWST/JADES measurements within roughly 0.5dex, though simulations tend to predict slightly higher values, suggesting that optical emission lines capture only a fraction of the multiphase outflow. However, simulated outflow velocities are typically an order of magnitude lower than those inferred from observations. TNG50 indicates a clear orientation dependence as outflows in face-on galaxies are approximately 15% more likely to be detected than in edge-on systems, with this difference increasing to nearly 40% for more massive, disc-shaped galaxies.

💡 Research Summary

This research presents a rigorous comparative analysis between high-redshift galactic outflow observations from JWST/NIRSpec and large-scale cosmological simulations, specifically TNG50 and SERRA. The study addresses a critical tension in modern astrophysics: the discrepancy between the high frequency of outflows predicted by theoretical models and the significantly lower occurrence rate (25-40%) observed in low-mass galaxies at redshifts z > 3.

To investigate this gap, the authors analyzed a massive dataset comprising approximately 60,000 galaxies from TNG50 and 3,000 from SERRA, covering stellar masses from $10^{7.5}$ to $10^{11} M_{\odot}$. Utilizing a Gaussian Mixture Model (GMM) that integrates gas velocity, star-formation rate (SFR), and spatial location, the researchers identified outflows in the immediate galactic vicinity and compared these simulated properties with the JWST/JADES survey data.

The findings reveal three pivotal insights. First, regarding outflow mass, the simulations broadly align with JWST/JADES measurements within a 0.5 dex margin. However, the simulations tend to predict slightly higher masses, suggesting that current optical emission line observations likely capture only a specific subset of the multiphase outflow, potentially missing cooler or denser gas components.

Second, a profound discrepancy was identified in outflow kinematics. The simulated outflow velocities are approximately an order of magnitude lower than those inferred from JWST observations. This significant gap implies that current feedback prescriptions in cosmological simulations—which govern how energy from supernovae or AGN is coupled to the interstellar medium—may be underestimating the kinetic energy injection into the circumgalactic medium.

Third, the study highlights the critical role of orientation effects in outflow detectability. The research demonstrates that outflows in face-on galaxies are approximately 15% more likely to be detected than in edge-on systems. This orientation-dependent bias becomes even more pronounced in more massive, disk-shaped galaxies, where the difference in detection probability increases to nearly 40%. This finding underscores the necessity of accounting for geometric viewing angles and galaxy morphology when interpreting high-redshift observational surveys.

In conclusion, this paper provides essential constraints for future galaxy formation models, emphasizing that resolving the tension between theory and observation requires both a refinement of the physical feedback mechanisms and a sophisticated treatment of observational selection effects related to galaxy orientation.