{"id":"ab4e4a71-3238-4098-a4fc-1c18a288f438","arxiv_id":"2411.09671","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Nonlinear wave scattering data on a globally hyperbolic Lorentzian spacetime determine the metric up to conformal diffeomorphism.","lead":"The paper proves that the scattering data of semilinear waves on a curved spacetime determine the spacetime metric up to an overall scaling and coordinate changes. This is a step toward recovering the geometry of a medium or universe from waves that pass through it and interact with each other.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bridge from equality of scattering operators to the boundary conformal diffeomorphism required by Theorem 6.1 is only cited, not proved; the cited [60] concerns a timelike DN map, so the characteristic null-boundary case is not covered.","rationale":"The reader's CONDITIONAL verdict is well-founded. I concentrated on the step from equality of the nonlinear scattering operators to the hypotheses of Theorem 6.1, because without it no local reconstruction can be compared across the two spacetimes. Remark 6.2 is the only place this is justified, and it defers to [60] in a non-obvious setting. The other candidate gap, the termination of layer stripping in §5.3, is real but appears to be a technical compactness argument that could be repaired; the boundary identification step is more foundational. I therefore recommend leaving the verdict at CONDITIONAL rather than ACCEPT. If the boundary-determination lemma can be proved along the lines of [60] for characteristic S±, or if the explicit two-metric check confirms that linear scattering data determine the boundary conformal jets, then the proof gap would close and ACCEPT could be considered.","tokens_in":64040,"tokens_out":8741,"duration_ms":102693,"concrete_test":"Write out the first-order scattering operator L_j : H^1(S−)→H^1(S+) for g_j and prove or refute the following lemma: if L1=L2 as operators between the same coordinate null hypersurfaces, then there is a conformal diffeomorphism Φ of a neighborhood of S+ mapping g(1)|S+ to a positive multiple of g(2)|S+. The cleanest check is to specialize to g_j = e^{2ω_j}g0 with ω_1=0 and ω_2 a nonconstant smooth function supported in a small neighborhood of S+ and equal to 1 near i±, so the interiors are isometric. Compute the canonical relation and principal symbol of L_j; if the lens relation or the leading symbol differs with ω_2, then linear data do see the boundary conformal factor and the cited step is plausible. If L1=L2 while a nonconstant boundary conformal factor is invisible at S+, Remark 6.2 is false.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central chain is N(1)=N(2) ⇒ equality of the first-order linear scattering operator ⇒ a conformal diffeomorphism Φ:U(1)→U(2) matching earliest scattering light observation sets ⇒ Theorem 6.1 ⇒ local reconstruction ⇒ layer-stripping ⇒ conformal equivalence of the metrics. The hinge is the second implication, stated only in Remark 6.2: equality of N is said to imply the existence of Φ by first-order linearization and 'a similar idea as in [60]', but no proof is supplied. This is not a cosmetic omission. Theorem 6.1 is a comparison theorem that assumes Φ; Section 7 reconstructs the regular scattering light observation sets only relative to a fixed boundary patch and never shows that L(1)=L(2) determines the conformal class of g|S+ or produces a map Φ between boundary neighborhoods. Moreover, [60] treats the Dirichlet-to-Neumann map on a timelike boundary, whereas here S± are characteristic null hypersurfaces. The null boundary metric is degenerate, its normal is lightlike, and the linear scattering operator is a wave-front relation between characteristic surfaces, not an elliptic boundary map. Transferring the boundary-determination statement therefore needs a genuinely new argument. If this step fails, the proof cannot begin: there is no Φ to feed into Theorem 6.1, and the nonlinear detection schemes of §7.2–7.3 compare scattering relations on coordinate patches without establishing the conformal boundary identification needed for the two-metric comparison. A secondary gap, the strategy-only termination of layer stripping in §5.3, is real but appears repairable by compactness; the boundary identification step is more upstream and more foundational.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the inverse problem of determining a time-dependent globally hyperbolic Lorentzian metric on a region M diffeomorphic to a diamond in (-1,1)×R^3 from the nonlinear scattering operator of a semilinear wave equation with analytic nonlinearity. The main result, Theorem 1.2, asserts that equality of the scattering operators N(1)=N(2) on a fixed domain of scattering data implies that the two metrics are conformally diffeomorphic. The proof combines higher-order linearization of the scattering operator, construction of receding conormal waves, propagation and interaction of their singularities, reconstruction of earliest and regular scattering light observation sets, and a layer-stripping procedure that builds the conformal structure from the future null boundary inward. A substantial part of the microlocal analysis, including the forward problem, the interaction calculus, and the reconstruction from a three-to-one scattering relation, is developed in detail.","tokens_in":64397,"tokens_out":5142,"duration_ms":51414,"significance":"If the proof can be completed, the paper would be a significant advance: it replaces boundary measurements for Lorentzian inverse problems by scattering data posed on characteristic null infinity, and it extends the source-to-solution-map results of Kurylev–Lassas–Uhlmann to a global scattering setting without interior observations. The paper contains many carefully proven lemmas, especially in Sections 4 and 7, and the symbolic computation for three-wave interaction is a genuine strength. It also makes falsifiable predictions about which singularities of nonlinear interactions are detectable from the scattering operator. The central obstruction is not the nonlinear interaction calculus but an unproved boundary-determination step asserted in Remark 6.2 and the incomplete termination argument of the layer-stripping procedure in Section 5.3.4.","major_comments":[{"comment":"The step asserting that equality N(1)=N(2) yields a conformal diffeomorphism Φ:U(1)→U(2) relating the earliest scattering light observation sets is not proved; it is only justified by the phrase 'one can use a similar idea as in [60]'. This step is load-bearing: Theorem 6.1 assumes the existence of such a Φ, and Sections 7.2–7.3 only compare scattering relations and light observation sets locally on coordinate patches without constructing the boundary identification needed for a two-metric comparison. Moreover, [60] treats a Dirichlet-to-Neumann map on a timelike boundary, whereas S± here are characteristic null hypersurfaces with a degenerate induced metric and lightlike normal; the linear scattering operator is a wave-front relation between characteristic surfaces, not an elliptic boundary map. Transferring the boundary-determination statement therefore requires a genuinely new argument or a precise reference covering null boundaries. Without this step, the central chain N(1)=N(2) ⇒ Φ ⇒ Theorem 6.1 ⇒ local conformal reconstruction does not close.","section":"Remark 6.2"},{"comment":"The termination of the layer-stripping procedure is described as a strategy rather than proved. The final paragraphs of Step 4 assert that the step sizes T1,T2,... 'do not get too small' by a compactness argument and then propose an ε0/N search with the claim 'Such an N0 does exist', but no proof is supplied. The load-bearing points are: (i) the reconstructed diamond sets cover a fixed region I(T0) independent of the choices made in the local reconstructions, and (ii) the stabilization criterion D_N = D_{N+1} actually detects whether the reconstructed conformal structure equals the true one. Both points are nontrivial because the local reconstructions are only up to conformal diffeomorphisms and the proposed failure test relies on unanalyzed singularities of scattering observation sets. This gap affects the global conclusion of Theorem 1.2, not merely a technical convenience.","section":"Section 5.3.4"},{"comment":"The extension to general analytic nonlinearities omits cross-terms from lower-order coefficients. After choosing m0 as the smallest index with β_{m0}(q) ≠ 0, the text asserts U_{m0} = Q_s(β_{m0} v_1^{m0−2} v_2 v_3), but the formulas in (8.2) show that for m0 ≥ 4 the corresponding mixed derivatives also contain contributions such as β_2 v_1 A_3^{jkl} and β_2 A_2^{ij} A_2^{kl} in A_4, and more generally products of β_2 with β_{m0−2}. Unless these contributions are shown to be absent or microlocally negligible, for example by an inductive argument using the vanishing condition v_1 ∈ ˚I^μ(Λ_1), the principal symbol of U_{m0} need not be the pure β_{m0} term on which the detection argument relies. Since Theorem 1.2 is stated for general analytic nonlinearities with possibly nonzero β_2, this gap is load-bearing.","section":"Section 8.3"}],"minor_comments":[{"comment":"In the paragraph defining the scattering operator, 'we consider consider the set' contains a duplicated word.","section":"Introduction"},{"comment":"In the introductory paragraph of Section 5, 'By concatenating such local reconstructions, we cam eventually get' should read 'we can eventually get'.","section":"Section 5"},{"comment":"Definition 7.1 appears twice: once for sufficiency of V− and again for the three-to-one scattering relation. The second definition should be renumbered, and the reference in the proof of Lemma 7.6 to 'Condition (R7.1)' should be corrected.","section":"Section 7.1"},{"comment":"The proof of Lemma 6.17 contains two consecutive 'Proof.' blocks, the first ending with the sentence 'We can prove it using Lemma 6.4 and the proof in Section 6.3.2.' This appears to be a leftover and should be removed.","section":"Section 6.1.3"}],"recommendation":"major_revision","confidential_remarks":"This is a serious paper with a strong microlocal core, and the overall strategy is plausible. The main obstacle is the unproved boundary-determination step in Remark 6.2: without it, Theorem 6.1 cannot be applied to the scattering problem. The layer-stripping termination and the general-nonlinearity symbol argument in Section 8.3 also need repair. These are gaps in the current proof rather than errors in the already-established parts of the calculus, so a major revision with added proofs or precise references seems appropriate."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: Theorem 1.2 is the right kind of result and it is new. The claim that the nonlinear scattering operator on null infinity determines the conformal class of a time-dependent Lorentzian metric is not in [46], [27], or [34], and the way the authors combine multi-fold linearization, receding waves, and light-observation sets is a real step beyond those papers. The forward theory in Section 2 and the nonlinear interaction calculus in Sections 4, 7, and 8 are worked out carefully; the layer-stripping construction is intricate and mostly persuasive.\n\nWhere I part company with the reader's report: I do not see circularity. The cited comparison theorems are published, and self-citation is not the issue. The reader's conditional verdict is substantively right.\n\nThe soft spots are real, and the first is load-bearing. Remark 6.2 says equality of the nonlinear scattering operators implies a conformal diffeomorphism between boundary neighborhoods by first-order linearization and \"a similar idea as in [60]\". But [60] is a timelike-boundary Dirichlet-to-Neumann result. Here the boundaries are characteristic null hypersurfaces: the induced metric is degenerate, the normal is null, and the linear scattering operator is a relation between lightlike covectors, not an elliptic boundary map. That is not a cosmetic mismatch. Theorem 6.1 takes the boundary conformal diffeomorphism as an assumption; Section 7 reconstructs light observation sets on fixed coordinate patches but does not establish a common conformal boundary identification across the two metrics. If that step fails, there is no boundary map to feed into Theorem 6.1, and the nonlinear detection schemes in Section 7.2-7.3 never get off the ground.\n\nThe second gap, the termination argument in Section 5.3.4, is more minor. The strategy with epsilon_0/N and stabilizing N_0 is plausible and probably repairable by compactness, but it is described as a strategy rather than proved. I would flag it as a required expansion, not a deal-breaker.\n\nWho is this paper for? People working in inverse problems for nonlinear hyperbolic equations and Lorentzian geometry. The machinery sections are worth reading even if Theorem 1.2 has to be revised. My advice: send it to a serious referee. The claim is important enough and the infrastructure is substantial enough that referee time is justified. The referee report should insist on a full proof of Remark 6.2 and a clean termination argument for the layer stripping.","headline":"A genuinely new inverse result for nonlinear scattering on null infinity, but the proof hinges on an unproved boundary-determination step that has to be fixed before the theorem is established.","tokens_in":64906,"tokens_out":2021,"would_cite":true,"duration_ms":23231,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35P25","58J50"],"pacs":[],"model":"deepseek-v4-flash","headline":"Full nonlinear scattering data determine a Lorentzian metric up to conformal equivalence.","keywords":["inverse scattering","semilinear wave equation","Lorentzian metric","scattering operator","conformal diffeomorphism","receding waves","scattering light observation sets","layer stripping"],"falsifier":"Exhibit two metrics satisfying Assumption 1.1 whose linear scattering operators agree but whose earliest scattering light observation sets cannot be identified by any conformal diffeomorphism; such a pair would break the reduction in Remark 6.2 and Theorem 1.2 would not follow from the given argument.","tokens_in":63873,"feed_emoji":"📡","tokens_out":7690,"duration_ms":65761,"temperature":0.7,"pith_summary":"This paper tries to prove that the nonlinear scattering operator of a semilinear wave equation encodes the full geometry of the spacetime the waves travel through. The central claim is that if two metrics produce identical scattering operators for all small scattering data on past null infinity, then the two metrics must be conformally equivalent: they differ only by a smooth diffeomorphism and a positive pointwise rescaling. This matters because inverse scattering for time-dependent metrics is usually obstructed by gauge freedom, and the nonlinearity removes that obstruction by making waves interact. The paper shows the principle works for cubic nonlinearities and for any analytic nonlinearity that genuinely depends on the solution.","feed_headline":"Scattering data determine a Lorentzian metric up to conformal change","feed_subtitle":"How semilinear waves scatter on two spacetimes reveals their geometry, modulo conformal rescalings.","key_machinery":"The engine of the argument is multi-fold linearization combined with receding waves: specially built solutions of the linear wave equation whose singularities are conormal to a null hypersurface. When three such waves intersect transversally, the nonlinearity produces a new singularity whose principal symbol carries the metric and the nonlinearity coefficient; the resulting three-to-one scattering relations between lightlike vectors on $S_-$ and $S_+$ determine the regular scattering light observation sets of interior points. A layer-stripping procedure recovers the metric locally in regions free of cut points, and Theorem 6.1 converts equality of the observation sets into a conformal diffeomorphism.","core_discovery":"On a globally hyperbolic Lorentzian spacetime $M$ with past and future null infinity $S_-$ and $S_+$, the paper considers the semilinear wave equation $\\square_g u + F(T,X,u)=0$ with scattering data $u_-$ on $S_-$, and defines the nonlinear scattering operator $\\mathcal{N}$ by sending $u_-$ to the restriction of the solution on $S_+$. Theorem 1.2 asserts: if $\\mathcal{N}^{(1)}(u_-)=\\mathcal{N}^{(2)}(u_-)$ for every sufficiently small $u_-$, then there exists a smooth diffeomorphism $\\Psi\\colon M\\to M$ and a smooth function $\\gamma$ such that $\\Psi^* g^{(1)} = e^{2\\gamma} g^{(2)}$. The proof takes higher-order derivatives of $\\mathcal{N}$ at zero data; the nonlinear interaction of three receding waves creates new singularities that travel along null geodesics, and reading these singularities on $S_+$ reconstructs the scattering light observation sets, from which the conformal structure of the metric is recovered in layers.","pith_inferences":["A natural extension is that the same scattering data should also determine the nonlinearity coefficients, up to the natural conformal gauge, since the principal symbols of the produced singularities depend on the nonlinearity at the interaction point.","The proof suggests that only a finite-order jet of the scattering operator matters in each layer, so one could test on model spacetimes whether truncated scattering data still fix the conformal class.","The layer-stripping construction may extend to asymptotically Minkowski or asymptotically de Sitter settings with nontrapping null geodesics, where the same scattering light observation sets can be defined at conformal infinity."],"forward_implications":["Equality of the full nonlinear scattering operator forces the two spacetimes to be conformally diffeomorphic, leaving only the conformal factor and a diffeomorphism as ambiguity.","The cubic case already carries the reconstruction, and the general analytic nonlinearity reduces to it by using the first nonvanishing Taylor coefficient of the nonlinearity.","The reconstruction uses data only on past and future null infinity, so no receivers inside the spacetime are needed.","Because the reconstruction is layered, it handles caustics provided each small diamond region is free of cut points."],"supporting_citations":[{"why":"introduces the multi-fold linearization and nonlinear wave interaction method that this paper adapts from source-to-solution maps to scattering.","marker":"[46]"},{"why":"supplies the radiation-field scattering framework from which receding waves are drawn.","marker":"[29, 30]"},{"why":"provides the boundary light observation set reconstruction that Theorem 6.1 transfers to null infinity.","marker":"[34]"},{"why":"is cited as the result that the linear scattering operator determines the lens relation and boundary jets up to conformal diffeomorphisms, the load-bearing step in Remark 6.2.","marker":"[60]"},{"why":"supplies the three-to-one scattering relation framework for nonlinear hyperbolic equations with disjoint sources and receivers.","marker":"[27]"},{"why":"provides the paired Lagrangian distribution calculus used to compute singularities of nonlinear interactions on Lorentzian manifolds.","marker":"[50]"},{"why":"gives the null cone light observation set reconstruction of the topological, smooth, and conformal structure.","marker":"[66]"},{"why":"supplies the Lorentzian geometry facts about cut points and uniqueness of null geodesic segments used throughout the layer stripping.","marker":"[15]"}],"fun_headline_variants":["Scattering operator recovers conformal class of Lorentzian metric","Semilinear wave scattering pins down metric up to conformal factor","Inverse scattering: Lorentzian metrics identified modulo conformal rescaling","How waves unmask spacetime geometry: conformal structure from scattering","Scattering data reveal spacetime up to a conformal change"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the assumption, stated in Remark 6.2, that equality of the linearized scattering operators already determines the null-geodesic lens relation and the boundary jets of the metric up to a conformal diffeomorphism; the paper cites a similar result rather than proving it.","fun_headline_variants_meta":{"raw":{"variants":["Scattering operator recovers conformal class of Lorentzian metric","Semilinear wave scattering pins down metric up to conformal factor","Inverse scattering: Lorentzian metrics identified modulo conformal rescaling","How waves unmask spacetime geometry: conformal structure from scattering","Scattering data reveal spacetime up to a conformal change"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000203,"raw_usage":{"total_tokens":1292,"prompt_tokens":760,"completion_tokens":532,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":376,"completion_tokens_details":{"reasoning_tokens":446}},"tokens_in":376,"tokens_out":532,"duration_ms":5887,"temperature":1.0,"reasoning_tokens":446,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T20:24:02.086468+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exhibit two metrics satisfying Assumption 1.1 whose linear scattering operators agree but whose earliest scattering light observation sets cannot be identified by any conformal diffeomorphism; such a pair would break the reduction in Remark 6.2 and Theorem 1.2 would not follow from the given argument.","supporting_citations":[{"cited_title":"Inverse problems for Lorentzian manifolds and non- linear hyperbolic equations","cited_arxiv_id":null,"evidence_quote":"introduces the multi-fold linearization and nonlinear wave interaction method that this paper adapts from source-to-solution maps to scattering."},{"cited_title":"Reconstruction of Lorentzian manifolds from boundary light observation sets","cited_arxiv_id":null,"evidence_quote":"provides the boundary light observation set reconstruction that Theorem 6.1 transfers to null infinity."},{"cited_title":"The inverse problem for the Dirichlet-to-Neumann map on Lorentzian mani- folds","cited_arxiv_id":null,"evidence_quote":"is cited as the result that the linear scattering operator determines the lens relation and boundary jets up to conformal diffeomorphisms, the load-bearing step in Remark 6.2."},{"cited_title":"Inverse problems for nonlinear hyperbolic equations with disjoint sources and receivers","cited_arxiv_id":null,"evidence_quote":"supplies the three-to-one scattering relation framework for nonlinear hyperbolic equations with disjoint sources and receivers."},{"cited_title":"Inverse problems for semilinear wave equations on Lorentzian manifolds","cited_arxiv_id":null,"evidence_quote":"provides the paired Lagrangian distribution calculus used to compute singularities of nonlinear interactions on Lorentzian manifolds."},{"cited_title":"Reconstruction of lorentzian manifolds from null cone light observation sets","cited_arxiv_id":null,"evidence_quote":"gives the null cone light observation set reconstruction of the topological, smooth, and conformal structure."},{"cited_title":"Global Lorentzian Geometry, Second Edition","cited_arxiv_id":null,"evidence_quote":"supplies the Lorentzian geometry facts about cut points and uniqueness of null geodesic segments used throughout the layer stripping."}],"review_version":1}