🔭 astrophysics

TRINITY: A coupled model of winds, radiation, and photoionised gas in molecular clouds. I. Methods and validation

This paper introduces TRINITY, an efficient 1D thin-shell code that couples winds, radiation, and photoionized gas pressure to model molecular cloud dispersal, demonstrating that both photoionized gas pressure and cloud density structure critically determine whether clouds expand, collapse, or leak photons.

Original authors: Jia Wei Teh, Ralf S. Klessen, Simon C. O. Glover, Kathryn Kreckel

Published 2026-05-28

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Jia Wei Teh, Ralf S. Klessen, Simon C. O. Glover, Kathryn Kreckel

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine a giant, fluffy cloud of cosmic gas and dust floating in space. This is a "molecular cloud," and it's the nursery where new stars are born. For a long time, astronomers have wondered: What happens when the first massive stars wake up inside this cloud? Do they gently push the cloud away, or do they blow it apart? And how long does it take?

This paper introduces a new computer program called Trinity to answer these questions. Think of Trinity as a high-speed, one-dimensional video game that simulates the life of a star cluster inside a cloud, tracking how the gas moves, heats up, and eventually disperses.

Here is the story of what the paper found, explained with everyday analogies:

1. The Setup: A Balloon in a Room

Imagine the molecular cloud is a large, heavy room filled with fog. In the center, a group of massive stars (the "cluster") suddenly turns on. These stars are like powerful blowtorches and firehoses combined. They shoot out:

Stellar Winds: Fast streams of particles (like a firehose).
Radiation: Intense light and heat (like a blowtorch).
Supernovae: Later, some stars explode like bombs (though this paper focuses on what happens before the explosions).

As these forces push out, they sweep up the surrounding fog into a thick, moving wall (a "shell"). Trinity tracks how this wall expands.

2. The New Twist: The "Ghost" Pressure

Previous models (like the one Trinity replaces, called warpfield) mostly focused on the heat from the stellar winds. But Trinity adds a crucial new ingredient: Photoionized Gas Pressure (PH ii).

The Analogy:
Imagine you are blowing up a balloon inside a room.

Old Model: It only counted the air inside the balloon pushing out.
Trinity: It realizes that the air outside the balloon (but still inside the room) is getting heated by the light from the stars. This heated air also pushes back against the balloon.
The Result: Trinity found that this "heated outside air" (PH ii) is a significant helper. In their test runs, including this pressure made the bubble expand about 17% larger by the time the simulation ended. It's like realizing your balloon is being pushed from the outside as well as the inside.

3. The Three Acts of the Story

The paper describes the life of the bubble in three distinct phases, like a movie with three acts:

Act 1: The Energy-Driven Phase (The Hot Bubble):
At the start, the stellar winds create a super-hot, high-pressure bubble in the center. It's like a pressure cooker pushing the shell outward. The bubble expands rapidly, growing like $t^{3/5}$ (a specific mathematical speed).
Act 2: The Transition (The Leak):
Eventually, the hot gas inside the bubble starts to cool down and lose its heat (like a radiator losing steam). The "pressure cooker" starts to fail.
Act 3: The Momentum-Driven Phase (The Snowplow):
Once the heat is gone, the bubble stops being a pressure cooker and becomes a snowplow. It keeps moving forward not because of internal heat, but because it has built up speed (momentum) from the wind and the direct push of the stars. It coasts along, pushing the shell until gravity or the edge of the room stops it.

4. The Shape of the Cloud Matters

One of the most surprising findings is that the shape of the cloud changes the ending of the story, even if the stars are exactly the same.

The Uniform Cloud (Flat Fog): If the cloud has the same density everywhere, the shell hits a lot of heavy gas as it moves out. It gets tired, slows down, and eventually collapses back in (re-collapses).
The Steep Cloud (Heavy Center, Light Edges): If the cloud is very dense in the middle but gets very thin and light toward the edges (like a steep slope), the shell hits less and less gas as it moves out. It speeds up and flies away, dispersing the cloud completely.

The Takeaway: You can't just look at how many stars are born to predict if a cloud will survive. You have to know how the gas is arranged. A "steep" cloud is much easier to blow apart than a "flat" one.

5. The "Escape" Clock

Trinity also tracks Lyman-continuum (LyC) photons. These are high-energy light particles that can escape the cloud and travel across the universe.

The Insight: Trinity separates two clocks. One clock measures when the cloud physically blows apart (dispersal). The other measures when the light starts leaking out.
Why it matters: Sometimes the light leaks out before the cloud is fully gone. This helps astronomers understand why we see light escaping from some regions even if the gas is still there.

6. The Bottom Line

The paper concludes that Trinity is a fast, efficient tool that helps astronomers map out exactly how stars clear their nurseries.

It's not just about the stars: The outcome depends on how the feedback (winds, light, gas pressure) couples to the cloud's shape.
Timing is key: The cloud usually disperses before the first supernova explosions even happen. The stars themselves (via winds and light) do the heavy lifting.
The "55% Rule": If you have a cloud with a specific "Bonnor-Ebert" shape (a realistic, rounded shape), it takes about 55% longer to clear than a simple, uniform cloud, even if they have the same mass and stars.

In short, Trinity is a new lens that shows us that the architecture of the gas cloud is just as important as the power of the stars in deciding whether a star-forming region survives or gets blown apart.

Technical Summary: TRINITY: A coupled model of winds, radiation, and photoionised gas in molecular clouds – I. Methods and validation

Problem and Context
Giant molecular clouds (GMCs) disperse within 1–5 Myr of massive star emergence, typically before the first core-collapse supernovae arrive. Observational surveys (e.g., PHANGS, LEGUS) indicate that pre-supernova feedback mechanisms—stellar winds, radiation pressure, and photoionised gas pressure ( $P_{\text{H II}}$ )—are the primary drivers of cloud clearing. However, the relative dominance of these channels, their timing, and their dependence on cloud properties (mass, density profile, star-formation efficiency) remain uncertain. Existing semi-analytic tools often treat these feedback mechanisms in isolation or rely on simplified density profiles, while 3D simulations are computationally expensive for statistical surveys. There is a need for an efficient, self-consistent forward model that couples all pre-supernova feedback channels to diverse cloud structures to map feedback dominance across parameter space.

Methodology
The authors introduce Trinity, a one-dimensional, spherically symmetric thin-shell code that evolves a feedback-driven bubble and swept-up shell within a molecular cloud. Trinity succeeds the warpfield framework but introduces two critical architectural changes:

Explicit Photoionised Gas Pressure: $P_{\text{H II}}$ is included explicitly in the shell equation of motion, rather than being treated solely as a boundary condition or ignored.
Phase-Aware Driving Prescription: The code dynamically selects the driving pressure term based on the evolutionary phase to avoid double-counting forces:
- Energy-driven phase: The driving pressure is the maximum of the hot-bubble thermal pressure ( $P_b$ ) and photoionised gas pressure ( $P_{\text{H II}}$ ).
- Transition phase: The driving pressure is the maximum of the remaining $P_b$ and the sum of $P_{\text{H II}}$ plus the ram pressure of newly injected wind and supernova ejecta ( $P_{\text{ram}}$ ).
- Momentum-driven phase: Once the hot bubble cools ( $E_b \to 0$ ), the driving pressure is the sum $P_{\text{H II}} + P_{\text{ram}}$ .

The model solves a coupled system of ordinary differential equations for the bubble radius ( $R_b$ ), velocity ( $v_b$ ), and thermal energy ( $E_b$ ). It incorporates:

Initial Conditions: Supports uniform, piecewise power-law ( $\rho \propto r^\alpha$ ), and Bonnor–Ebert (BE) sphere density profiles.
Physics Modules: Includes stellar winds, supernovae, direct and dust-reprocessed radiation pressure, self-gravity of the shell and cluster, and conductive evaporation from the shell into the hot bubble.
Radiation-Shell Coupling: A hydrostatic shell-structure calculation determines the ionisation-front position ( $R_{\text{if}}$ ), absorbed radiation fractions, and infrared optical depth self-consistently.
Cooling: Treats the transition from collisional ionisation equilibrium (CIE) to non-CIE regimes and calculates cooling losses to determine the energy-to-momentum transition.

Key Results
The paper validates the code against analytic limits (Weaver wind, Spitzer expansion, snowplow) and warpfield runs, then presents results for clouds with $M_{\text{cloud}} = 10^5–10^{6.5} M_\odot$ , core densities $n_{\text{core}} = 10^3–10^4 \text{ cm}^{-3}$ , and star-formation efficiencies $\epsilon = 0.01–0.30$ .

Impact of Photoionised Gas: Including $P_{\text{H II}}$ measurably alters dynamics. In a fiducial validation run ( $M_{\text{cloud}}=10^6 M_\odot, \epsilon=0.01$ ), the inclusion of $P_{\text{H II}}$ increases the shell radius by $\sim 17\%$ at 10 Myr compared to a warpfield-equivalent run where $P_{\text{H II}}$ is disabled. $P_{\text{H II}}$ remains dynamically important even after the transition to momentum-dominated evolution.
Density Profile Dependence: The cloud structure dictates the outcome even when stellar populations are fixed. For a fixed $M_{\text{cloud}}$ $M_{cloud}$ , $n_{\text{core}}$ $n_{core}$ , and $\epsilon$ $ϵ$ :
- Homogeneous ( $\rho \propto r^0$ ) and shallow ( $\rho \propto r^{-1}$ ) profiles re-collapse.
- A steep profile ( $\rho \propto r^{-2}$ ) allows the shell to accelerate down the density gradient and continue expanding.
- Bonnor–Ebert spheres re-collapse in cases where homogeneous clouds might disperse, or disperse significantly later.
Dispersal Timescales: In a grid of models where dispersal is guaranteed, Bonnor–Ebert clouds disperse roughly 55% later than homogeneous clouds. This systematic offset arises because BE structures distribute the same mass over a larger physical radius, requiring the shell to travel further.
Pre-Supernova Dominance: In the tested parameter space, cloud dispersal or re-collapse is determined by pre-supernova feedback (winds, radiation, $P_{\text{H II}}$ ) before the first core-collapse supernova arrives (typically $>3$ Myr).
LyC Escape: The code tracks the Lyman-continuum (LyC) escape fraction ( $f_{\text{LyC}}^{\text{esc}}$ ) as a separate "clock" from dynamical clearing, distinguishing the time feedback alters shell dynamics from the time photons begin leaking into the diffuse ISM.

Significance and Claims
The authors claim that Trinity provides a highly efficient, interpretable framework for mapping feedback dominance across cloud parameter space. The primary conclusion is that feedback dominance cannot be inferred from the stellar population alone. The outcome depends critically on:

How feedback channels couple to the shell (specifically the role of $P_{\text{H II}}$ ).
The timing of the hot-bubble cooling transition.
The density structure of the natal cloud.

The paper asserts that assuming a canonical homogeneous cloud or a single power-law slope introduces systematic errors in inferred emergence times and dispersal thresholds. By offering a tool that self-consistently couples winds, radiation, and photoionisation with flexible density profiles, Trinity enables the study of resolved H II regions and provides sub-grid prescriptions for galaxy-scale simulations regarding ionising photon leakage and cloud dispersal.

Limitations
The authors explicitly note that the 1D spherical symmetry averages over angular structures (e.g., champagne flows, blister breakouts) and cannot resolve instabilities like Kelvin–Helmholtz or Rayleigh–Taylor that may enhance cooling. The treatment of $P_{\text{H II}}$ uses a single radius for the bubble-ionisation front system rather than evolving $R_b$ and $R_{\text{if}}$ as separate dynamical variables. Additionally, the model currently neglects magnetic support, cosmic-ray pressure gradients, and the evolution of the covering fraction (leaky shells). These are identified as targets for future development.