Stellar feedback and galactic matter cycles

Feedback has several physical channels

Forming and evolved stars affect their surroundings through radiation, winds, protostellar outflows and later ejecta or explosions. These processes can heat or ionize gas, transfer momentum, drive flows and change where material accumulates. Their effects depend on the conditions and geometry of the region.

Radiation that escapes without interacting does not deposit all its energy locally. Energy deposited in dense gas may be radiated away, while momentum transfer can still influence motion. A hot expanding component can also do work on surrounding material. A single injected energy total does not determine how much gas is removed.

Compression can promote contraction in some places, while heating, dispersal or loss of gas can inhibit it elsewhere. Feedback is therefore not a universal promise of either more or fewer stars in every local setting. A physical explanation must state the scale, interval and process being compared.

Krumholz and collaborators distinguish feedback mechanisms and the difficulty of inferring their net consequences. This lesson follows deliberately supplied material and energy budgets. It does not adopt universal feedback efficiencies, cloud lifetimes or an automatic outcome from the presence of young stars.

Formation and return are separate terms

Choose a region and an arbitrary positive mass unit M0. Initially the region contains 100M0 of gas and no tracked newly formed stars. Over each interval, separately record gas entering, gas converted into stellar birth mass, material returned from the tracked new stellar population and gas crossing outward through the boundary.

All tracked new stars and remnants remain inside this toy region. No other material source or flux is present. Return values are already integrated over their actual delays in the stipulated intervals; no instantaneous-recycling law or universal return fraction is imposed.

For each interval, final gas equals initial gas plus inflow minus stellar birth mass plus returned material minus outflow. The current retained mass in the tracked stars and remnants increases by birth mass minus return. Ordinary classical mass accounting here neglects the small nuclear rest-mass-to-radiation correction.

Outflow means crossing this chosen boundary. It does not automatically mean escaping the gravitational potential of a galaxy. The material could remain bound or later return; the toy specifies no orbit, potential or future boundary flux.

A cumulative birth total is not the retained mass

In interval I, add 20M0 of inflow, form 30M0 of stellar birth mass, return 8M0 and export 12M0. Final gas is 100 + 20 − 30 + 8 − 12 = 86M0. Retained new stars and remnants hold 30 − 8 = 22M0, and cumulative export is 12M0.

In interval II, begin with 86M0 of gas, add no inflow, form 10M0, return 3M0 and export 9M0. Final gas is 86 − 10 + 3 − 9 = 70M0. Cumulative retained stellar and remnant mass becomes 22 + 10 − 3 = 29M0; cumulative export becomes 21M0.

Across both intervals, the initial material plus inflow is 120M0. The final accounting is 70M0 gas + 29M0 retained stars and remnants + 21M0 exported = 120M0. The cumulative birth mass is 40M0, which exceeds the current retained 29M0 because 11M0 has returned.

The table gives integrated masses, not rates; interval durations have not been supplied. Nor is the remaining 70M0 all guaranteed future stellar mass. Its later flow and physical evolution are unspecified.

Loss from a reservoir is not destruction of energy

Now make a separate illustrative energy ledger for one declared interval, in an arbitrary positive energy unit E0. Supply injected energy 50E0, radiated energy 35E0, exported kinetic energy 10E0 and a net increase in stored energy 5E0. The specified balance is 50 = 35 + 10 + 5.

Energy leaving as radiation or through a boundary no longer remains in the chosen reservoir, but it has not disappeared from the wider energy accounting. In a real region, pressure work, gravitational potential changes and other boundary or stored terms must be included consistently.

No relation between E0 and M0 is supplied. The energy row therefore cannot be divided into the material rows to establish an outflow speed or a binding-energy verdict. Its integrated energy quantities are also different from power, which requires an interval duration.

A feedback model must connect the appropriate channels to the observed state and losses. The fraction of a stellar luminosity that interacts with gas, the cooling conditions and the direction of a flow can matter as much as the nominal input. A bright source alone does not establish complete coupling.

A young star beside a rim does not settle the cause

A bright rim, shell or age sequence can motivate a feedback hypothesis. Yet young stars and dense gas can be spatially associated even without the proposed trigger. An image showing proximity is not a direct observation of what the same material would have done without that feedback.

One possible claim is that feedback caused some stars to form earlier. A stronger claim is that it increased the total amount of star formation that would otherwise have occurred. Those are distinct causal statements and can differ between a local region and a larger system.

Krumholz and collaborators discuss these distinctions and the observational difficulty of establishing triggering. Kinematics, ages, gas structure, comparisons with suitable regions and models of alternative histories can add evidence. Even a high stellar-to-gas ratio might reflect gas removal rather than additional stars.

The next lesson considers magnetic fields and nonthermal motions. They contribute further physical constraints and stresses, but a field measurement or a broad line must also be connected to a conditional model before it determines a collapse or support claim.

Worked examples

Try the reasoning

1. Is cumulative birth mass 40M0 the same quantity as current retained stellar and remnant mass 29M0?

   Show the answer and reasoning

   No.

   The supplied 11M0 of returns accounts for the difference.

2. Can the integrated formation mass be called a formation rate without an interval duration?

   Show the answer and reasoning

   No.

   A rate requires mass per time; no durations are specified here.

3. Does crossing the chosen boundary guarantee that gas is unbound from the galaxy?

   Show the answer and reasoning

   No.

   Its gravitational environment, velocity, trajectory and later flows have not been supplied.

4. Does a young star beside a bright feedback-driven rim alone prove extra star formation was triggered?

   Show the answer and reasoning

   No.

   A causal claim needs evidence against alternative histories and must distinguish earlier formation from additional total formation.

What to carry forward

• Feedback transfers radiation, momentum, energy and material through several conditional channels.
• Birth mass, retained stellar mass, remaining gas and boundary export need separate entries.
• Balanced budgets and spatial associations do not by themselves establish escape or causal triggering.