Fixed-Background Limits and Backreaction

Fixed-background QFT predicts the state, spectrum, greybody-filtered flux, and renormalized stress tensor for a prescribed black-hole geometry. It does not determine how that geometry changes after the flux has removed an appreciable fraction of the mass. A different approximation is needed when local quantum stress is no longer a small correction, cumulative mass loss is no longer negligible, or stress-tensor fluctuations invalidate a mean-field description.

Helpful background. Flat-Spacetime Quantum Field Theory provides the test-field expansion. Partial Differential Equations and Green Functions supplies state evolution on a chosen metric. Causality and Global Structure develops event and apparent horizons. Quantum Fields in Curved Spacetime provides renormalized observables. Black-Hole Geometry and Mechanics supplies mass and surface-gravity scales. Black-Hole Perturbation Theory provides linear response around a fixed background.

Three small parameters, not one boundary

A local mean-stress criterion compares the quantum source with the background curvature scale $L^{-2}$. In an orthonormal frame, a schematic dimensionless measure is

$$\epsilon_{\mathrm{local}} \sim 8\pi G L^2 \left|\langle T_{\hat a\hat b}\rangle_{\mathrm{ren}}\right|.$$

Fixed geometry requires this to be small in the region and state being used. Near a horizon the orthonormal frame and state regularity matter; a small Schwarzschild-coordinate component is not sufficient.

A cumulative criterion compares the observation interval with the mass-loss time:

$$\epsilon_{\mathrm{cum}} =\frac{|\Delta M|}{M} \sim\frac{|\dot M|\Delta t}{M}.$$

Even a tiny instantaneous flux produces order-one change if integrated for the evaporation time. For a large four-dimensional Schwarzschild black hole, dimensional scaling gives $t_{\mathrm{dyn}}\sim GM$ and $t_{\mathrm{evap}}\sim G^2M^3$ up to species and greybody coefficients, so one may have a long adiabatic window with $t_{\mathrm{dyn}}\ll\Delta t\ll t_{\mathrm{evap}}$.

A fluctuation criterion tests the metric response to the stress-tensor noise on a specified sampling scale. A ratio of variance to squared mean can be a warning, but is not a universal criterion: the Minkowski vacuum has zero renormalized mean stress and nonzero smeared fluctuations, so that ratio is undefined even in a well-controlled background. The noise must be propagated through the gravitational response and compared with the accuracy required of the geometry; unstable modes and the sampling scale matter Hu and Verdaguer 2008, §§3.1–3.3, PDF.

Evidence through 2026-08-19

The evidence cutoff for this assessment is 2026-08-19. The rows describe nested approximations rather than rival calculations of one quantity.

Position or regime	Evidence and result	Assumption, limitation, or adverse case
Fixed collapse background	Exponential ray tracing gives the late Hawking temperature and state on a prescribed geometry.	Valid over intervals with negligible geometry change; it does not solve for mass loss or a moving horizon.
Adiabatically evolving exterior	Instantaneous flux and greybody data can drive a slowly varying mass parameter.	Requires separation between dynamical and evaporation times and controlled state transport; it already belongs to semiclassical backreaction.
Mean-field semiclassical gravity	The renormalized expectation value sources a self-consistent metric equation.	Requires a declared finite gravitational scheme and controlled induced metric fluctuations on the relevant scale; existence and stability are separate problems.
Stochastic or fluctuation-sensitive regime	Noise kernels can quantify departures from the mean geometry.	Smearing, state, gauge, and scale matter; it does not by itself supply microscopic quantum gravity or an endpoint.
Fine-grained information observables	Entropies and correlations ask whether radiation purifies and how information is recovered.	These are not determined by a fixed-background occupation spectrum or mean stress tensor alone.

The fixed-background result is established by the collapse calculation in Hawking 1975, §2, Eqs. (2.16)–(2.29), pp. 208–211. The local near-horizon scaling formulation in Fredenhagen and Haag 1990, §§2–4, pp. 273–284 strengthens the field-theoretic basis but likewise does not evolve the metric.

When the fixed-background approximation fails

The competing positions emphasize different discriminators. A local-curvature criterion asks when the instantaneous quantum source is large. An adiabatic criterion asks whether the background changes slowly per light-crossing time. A cumulative criterion asks whether a small correction has integrated to order one. A fluctuation criterion asks whether the mean stress tensor is representative.

These positions are complementary, but they need not trigger simultaneously. A macroscopic black hole has a tiny luminosity per dynamical time while still losing all its mass over a sufficiently long interval. This is the decisive adverse case against extrapolating one stationary Hawking spectrum through the entire lifetime. Conversely, a locally large stress near a singular or inner-horizon region can invalidate fixed geometry long before the asymptotic mass changes substantially.

Continue with The Adiabatic Evaporation Approximation. There the mass becomes a slowly varying function, the state and flux are transported consistently, and the expansion parameter is stated explicitly.

The mean-field equation and its inputs

Semiclassical gravity begins from an equation of the form

$$G_{ab}+\Lambda g_{ab} +\alpha H^{(1)}_{ab}+\beta H^{(2)}_{ab} =8\pi G\,\langle T_{ab}\rangle_{\mathrm{ren}}[g,\omega].$$

This is an integro-differential self-consistency problem: the metric determines the quantum state evolution and subtraction terms, while the renormalized stress tensor changes the metric. Conservation must hold on both sides, and the finite curvature couplings must match the stress-tensor scheme.

Here the local tensors can be defined by

$$H^{(1)}_{ab} =-\frac{2}{\sqrt{-g}}\frac{\delta}{\delta g^{ab}} \int d^4x\,\sqrt{-g}\,R^2, \qquad H^{(2)}_{ab} =-\frac{2}{\sqrt{-g}}\frac{\delta}{\delta g^{ab}} \int d^4x\,\sqrt{-g}\,R_{cd}R^{cd}.$$

Variations have compact support, or boundary terms are fixed separately. With this definition the displayed coefficients $\alpha,\beta$ absorb the factors from moving the corresponding action terms to the left-hand side and have dimensions of length squared. They are not implicitly the dimensionless coefficients in the gravitational action.

A fixed-background calculation supplies the right side evaluated on one $g$ and state. Inserting that result into a prescribed $M(u)$ without checking state transport, higher-derivative terms, and residuals is an ansatz, not a self-consistent solution. Stress-tensor fluctuations, noise, and initial-state dependence add further layers.

Fine-grained Page curves, island prescriptions, interior reconstruction, and microscopic endpoints are outside this equation’s ordinary mean-field content. They belong to black-hole information or quantum-gravity treatments even when Hawking flux is an input.

What remains unresolved

There is no universal numerical threshold that transfers every black-hole problem from fixed QFT to one next theory. The appropriate parameter depends on geometry, state class, observable, smearing scale, field content, and desired accuracy. Near an endpoint, all candidate expansions may fail together.

Open questions include the controlled range of semiclassical evolution in strongly fluctuating states, the relation between event and trapping horizons in long-time quantum evolution, and the appropriate description when higher-curvature terms become comparable. A stationary late-time spectrum alone does not establish an endpoint or a unitary completion.

What could change these conclusions

The conclusions could change if a new theorem proves long-time existence and error bounds for a self-consistent evaporating solution, or a no-go result rules out the assumed adiabatic state transport. An independent calculation that changes the relative size of mean stress and noise, a counterexample with order-one local response during negligible mass loss, or a regulator and boundary analysis that changes the finite coupling hierarchy would also change the interpretation. Experimental observation is not expected for astrophysical evaporation, but any measurement constraining horizon-scale quantum stress would change the empirical regime map.

Exercise

A model has $|\dot M|t_{\mathrm{dyn}}/M=10^{-12}$. Is a fixed Schwarzschild background valid for an interval $10^{13}t_{\mathrm{dyn}}$? State the relevant calculation rather than answering from the instantaneous ratio.

Solution

The instantaneous adiabatic ratio is small, but the cumulative estimate is $|\Delta M|/M\sim10^{-12}\times10^{13}=10$. The stationary fixed background is therefore inconsistent over that interval; it would predict more than an order-one mass change while holding the mass fixed. One must evolve the geometry, at least adiabatically, and recompute the changing temperature and greybody flux. The example shows why a local small parameter cannot replace the cumulative-timescale check.

References

• Fredenhagen, Klaus, and Rudolf Haag. “On the Derivation of Hawking Radiation Associated with the Formation of a Black Hole.” Communications in Mathematical Physics 127 (1990): 273–284. DOI.
• Hawking, Stephen W. “Particle Creation by Black Holes.” Communications in Mathematical Physics 43 (1975): 199–220. DOI.
• Hu, B. L., and Enric Verdaguer. “Stochastic Gravity: Theory and Applications.” Living Reviews in Relativity 11 (2008): 3. DOI. Open PDF.