Ram pressure stripping occurs when galaxies speeding through dense galaxy clusters lose their gas to the surrounding hot intracluster medium, much like a strong wind stripping leaves from a tree. A new preprint (arXiv:2603.26899, not yet peer-reviewed) simulates this process at the scale of individual gas clouds to study whether and how stars form as the gas is accelerated.

The study methodology involves high-resolution hydrodynamical simulations of isolated gas clouds with masses of roughly one million solar masses and radii spanning a few to several hundred parsecs. These clouds are exposed to a hot wind representing the intracluster medium, with a subgrid model for star formation included. Rather than simulating entire galaxies, the team focuses on single clouds across multiple runs with varying parameters. This approach allows detailed tracking of cloud compression, fragmentation, and stellar age distributions but has clear limitations: it uses idealized initial conditions, omits magnetic fields and cosmic rays, and does not capture the continuous stripping from a live galactic disk. Sample size is modest—several representative clouds rather than a statistical ensemble.

The simulations confirm an overall age gradient, with younger stars tending to appear farther along the wind direction and at higher velocities. However, the gradients are strongly non-monotonic, unlike the smooth progression predicted by the earlier 'fireball' toy model. Two physical effects drive this complexity: rapid compression and gravitational collapse within the cloud produce stars simultaneously at varied locations, while shredding of the cloud into filamentary structures scatters newly formed stars across a range of heights and speeds. Self-gravity from the gas and stars further modifies stellar velocities after birth.

This preprint reveals mechanisms the original abstract and toy model under-emphasized: the messy, filamentary nature of real cloud acceleration creates irregular stellar populations rather than orderly sequences. Synthesizing these results with observations from the GASP survey (Poggianti et al. 2017, arXiv:1705.00012), which used MUSE on the VLT to map star-forming clumps and complex velocity fields in jellyfish galaxy tails, shows striking consistency. The simulated non-monotonic gradients may explain observed age spreads that puzzled earlier studies. Additionally, work from the FIRE-2 simulation suite (Hopkins et al. 2020) on multiphase galactic winds demonstrates that similar cloud-wind interactions occur in feedback-driven outflows, not just ram pressure, suggesting these star-formation pathways are widespread.

These findings connect to larger galactic evolution patterns. Ram pressure is a primary quenching mechanism that stops star formation in cluster galaxies, yet the survival of dense cloud fragments allows star formation to persist in extreme environments. Some of these stars may become unbound, contributing to intracluster light. The research also implies that interpreting tail observations requires more sophisticated models than simple acceleration toys; ignoring shredding and gravity leads to incorrect age-dating of stellar populations.

While limited by its zoom-in approach, the work uncovers novel stellar birth channels in hot winds that enrich our understanding of environmental feedback, galaxy transformation, and the multiphase physics shaping cosmic structure.