Averaged Null Energy and QNEI Geometric Hypotheses

ANEC, a finite quantum null energy inequality, and a compactly weighted null average are not the same claim. ANEC concerns a complete affine null geodesic and an infinite-averaging limit. A QNEI is a lower bound for a declared finite sampler, often with the null tensor contracted along a timelike curve. In four dimensions the second can help prove the first under limiting and geometric hypotheses even though no state-independent lower bound exists for every compact null-geodesic sampler.

Required background. Worldline, spacetime-volume, and null-energy bounds separates the averaging domains. Hadamard states and their wavefront characterization supplies the admissible ultraviolet state class.

Helpful background. Bisognano–Wichmann geometric modular action gives a distinct positivity route in special geometries. Positivity, spectrum, covariance, and locality hypotheses helps identify the exact implication. The averaged null energy condition gives information-theoretic uses, while ANEC in curved spacetime develops the geometric application.

The complete-geodesic statement

Let $\gamma:\mathbb R\to M$ be an affinely parametrized complete null geodesic with tangent $k^a$. Because an unweighted integral may not converge absolutely, fix a real cutoff $f\in C_0^\infty(\mathbb R)$ with $f=1$ near zero and define

$$\mathcal E_L(\omega)=\int_{\mathbb R} f(\lambda/L)^2 \langle T_{ab}k^ak^b\rangle_\omega(\gamma(\lambda)) \,\mathrm d\lambda.$$

One precise ANEC convention is

$$\liminf_{L\to\infty}\mathcal E_L(\omega)\geq0.$$

The affine normalization of $k^a$ must be held fixed; rescaling the affine parameter rescales both the tangent and the integral. The state must make the pulled-back expectation a distribution on the line, and the large-$L$ limit must be controlled. “Hadamard” controls short distances but does not by itself guarantee infrared convergence of the complete integral.

For the minimally coupled free scalar in four-dimensional Minkowski space, ANEC holds on standard finite-particle domains and extensions satisfying the required decay and limiting conditions. The result follows by writing the normal-ordered null stress in momentum space: the long affine integral enforces equality of the momentum component conjugate to the null coordinate, and the surviving quadratic form is nonnegative. Klinkhammer states and proves the flat-space result with the operator-domain restrictions in Klinkhammer 1991, §§ II–III, pp. 2544–2548. It is not an assertion for arbitrary distributions called states or for nonminimally coupled fields.

From timelike QNEIs to an achronal limit

A complementary proof approaches $\gamma$ by long timelike curves inside a tubular neighborhood. Let $u_\epsilon^a$ be future timelike and tend, after a declared rescaling, to $k^a$. Apply a QNEI to

$$\int g_{L,\epsilon}(\tau)^2 \langle T_{ab}\ell^a\ell^b\rangle (\gamma_\epsilon(\tau))\,\mathrm d\tau,$$

where $\ell^a$ is the parallel null direction. Choose $L\to\infty$ and $\epsilon\to0$ together so that the timelike curves remain in the controlled tube, their samplers converge to $f(\lambda/L)$ with the correct Jacobian, and the QNEI lower bound tends to zero. Continuity of the renormalized expectation across the tube then gives the ANEC liminf.

This construction makes the geometry visible. In the theorem of Fewster, Olum, and Pfenning, the complete null geodesic lies in a flat tubular neighborhood whose intrinsic causal relations agree with those inherited from the full globally hyperbolic spacetime; the scalar is minimally coupled and the state satisfies the stated Hadamard and convergence conditions Fewster, Olum, and Pfenning 2007, Theorem II.1 and § III, pp. 025007-3–025007-10. The causal-isolation condition prevents shortcuts outside the tube from turning the null line into a chronal curve. Boundaries may be present only when the tube remains a fixed positive distance from them and the induced causal structure is unchanged.

The worked Minkowski scaling family is continued on averaged null energy conditions. An independent check is dimensional: after the sampler broadens, every derivative contribution in the timelike QNEI bound must vanish. If a chosen normalization leaves a nonzero negative constant, it cannot prove ANEC.

Completeness, achronality, and state hypotheses

Completeness supplies both tails needed for the broad-sampling limit. Achronality prevents two points of the geodesic from being timelike related; in curved spacetime conjugate points can destroy it. The field equation and coupling determine whether the null contraction has the positive structure used in the QNEI. The Hadamard condition licenses renormalization, while additional infrared or falloff assumptions license the limit. These roles are logically distinct.

The conclusion is correspondingly narrow. A proved ANEC on one complete achronal geodesic does not imply the pointwise null energy condition, a finite-segment lower bound, QNEC, or a global focusing theorem. Geometry and Einstein’s equation enter only when the result is handed to a gravitational argument.

Failure boundary: isolate a negative pulse

Take a state whose null energy contains a negative pulse followed later by compensating positive energy. A compact segment enclosing only the negative pulse has a negative integral. Extending the sampler to the complete geodesic can restore nonnegativity, but truncation removes the compensation. Thus applying ANEC to a finite segment is an invalid converse.

Dropping achronality creates another failure: a complete null geodesic with conjugate points may admit timelike shortcuts, so the tubular causal comparison used above breaks. The QNEI along each timelike test curve may remain correct, yet the limiting proof no longer licenses ANEC on that geodesic. The strongest surviving statement is the original finite timelike QNEI.

Exercises

Show that the sign of the complete ANEC integral is invariant under a positive affine rescaling.

Solution

Let $\lambda'=a\lambda$ with $a>0$. Then $k'^a=\mathrm dx^a/\mathrm d\lambda'=a^{-1}k^a$ and $\mathrm d\lambda'=a\,\mathrm d\lambda$. Hence $\int T_{ab}k'^ak'^b\mathrm d\lambda'=a^{-1}\int T_{ab}k^ak^b\mathrm d\lambda$. The integral changes by a positive factor, so its sign is invariant. A negative $a$ reverses time orientation and is not the declared affine normalization.

References

• Fewster, Christopher J., Ken D. Olum, and Michael J. Pfenning. “Averaged Null Energy Condition in Spacetimes with Boundaries.” Physical Review D 75 (2007): 025007. DOI; Open PDF.
• Fewster, Christopher J., and Thomas A. Roman. “Null Energy Conditions in Quantum Field Theory.” Physical Review D 67 (2003): 044003; erratum 80 (2009): 069903. DOI; Open PDF.
• Klinkhammer, Gunnar. “Averaged Energy Conditions for Free Scalar Fields in Flat Spacetime.” Physical Review D 43 (1991): 2542–2548. DOI.