{"id":"4c2b07ae-b569-433e-ba50-44edf23af94c","arxiv_id":"1908.05134","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"In the far field of a turbulent round jet, the Lumley decomposition in stretched spherical coordinates yields streamwise eigenfunctions that are Fourier modes with linearly growing wavelength and amplitude decaying as the -3/2 power of distance from the virtual origin.","lead":"This paper uses a tensor version of the Lumley decomposition in stretched spherical coordinates to derive streamwise basis functions for turbulence in the round jet far field. These modes, called stretched amplitude-decaying Fourier modes, have linearly growing wavelength and amplitude that decays as the minus three-halves power of distance, and they reproduce the classic -5/3 and -7/3 spectral slopes from two PIV datasets.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"SADFM deduction rests on unverified ξ-homogeneity of the scaled two-point correlation; the paper's own data could test it.","rationale":"The paper develops a coherent tensor formulation of a weighted Lumley decomposition and supports it with two independent PIV datasets, so the conditional verdict is appropriate. The single most load-bearing step is the homogeneity-after-scaling assumption stated just before Eq. (3.45): unless the scaled two-point correlation depends only on ζ = ξ' - ξ, the Fourier ansatz does not diagonalize the LD integral and the SADFM lose their status as deduced eigenfunctions. This is exactly the weakest assumption identified by the reader, and it is empirically checkable with the datasets already analyzed in the paper. I considered whether the hand-chosen inner-product weight is more load-bearing, since Appendix E shows it filters roughly 49% of the TKE and the basis is not optimal in the unweighted L2 norm. However, the paper explicitly frames the decomposition as an L2_w problem, so the weight is a disclosed modeling choice rather than a hidden flaw; the unverified ξ-homogeneity is the step that even the weighted claim depends on. The absence of a direct verification of Eq. (3.41) is what keeps the result conditional, and the proposed collapse test would either upgrade or refute the central claim.","tokens_in":27825,"tokens_out":12950,"duration_ms":146178,"concrete_test":"Using datasets E1 and E2, bin all point pairs by absolute streamwise coordinates ξ and ξ' in the self-similar domain, form the scaled two-point correlation <tilde v^ξ(ξ, θ0) tilde v^ξ(ξ', θ0)> for fixed θ0/θ_{1/2} values such as 0, 0.5, and 1, and plot it against ζ = ξ' - ξ for at least three disjoint absolute-ξ bins. A collapse of these curves within the PIV uncertainty estimated from the ~10^4 independent realizations would validate the homogeneity premise behind Eq. (3.41); a systematic spread with absolute ξ would falsify it and show that the SADFM are not eigenfunctions of the measured LD operator.","verdict_should_be":"UNCHANGED","load_bearing_attack":"At Eq. (3.41) the two-point correlation tensor of the scaled contravariant field is taken to be invariant under the streamwise separation ζ = ξ' - ξ, and this premise licenses the Fourier ansatz (3.45). If that invariance is not realized by the actual jet, the SADFM are not eigenfunctions of the weighted LD integral (3.40), and the analytical e^{-3ξ/2} amplitude is only an orthogonalizing factor rather than a deduced optimal mode. The paper does not test the premise: Section 6 projects data onto the SADFM and reports spectra, but projecting onto a basis only measures energy in that basis, not whether the basis diagonalizes the two-point correlation. The citation of Ewing et al. (2007) provides prior support, but the present PIV datasets, which contain pairs spanning the full ξ-range, are not used to produce the required collapse test. A secondary caveat is that the measured domain Ωξ = [0, 1.21] is finite: even with exact ξ-homogeneity, the finite-interval integral operator is Toeplitz rather than circulant, so strict Fourier eigenfunctions are an infinite/periodic-domain idealization whose finite-domain error is not quantified. The central claim is therefore best read as conditional on a directly verifiable property of the measured two-point statistics.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a tensor-calculus formulation of the Lumley decomposition (LD) in curvilinear coordinates and applies it to the self-similar far field of an axisymmetric round jet expressed in stretched spherical coordinates (SSC). The central analytical claim is that, after introducing a positive inner-product weight w = e^{-ξ}, the streamwise optimal eigenfunctions are stretched amplitude-decaying Fourier modes (SADFM) whose physical form is e^{iκξ-3ξ/2}, corresponding to an amplitude decay as (x-x0)^{-3/2} and linearly growing wavelength. The derivation is carried out in Section 3 and Appendices B-D, the transverse modes are solved numerically in the θ-direction, and the bases are used to produce spatial spectra from two independent PIV datasets, E1 and E2. The measured single-point profiles collapse reasonably, the spectra exhibit reported -5/3 and -7/3 regions, and the first LD modes contain a large fraction of the energy.","tokens_in":28104,"tokens_out":6545,"duration_ms":69181,"significance":"If the central premise is validated, the paper is a useful and original contribution: it extends Fourier-type POD bases to a flow direction that is not statistically homogeneous, for a class of flows admitting equilibrium similarity. The tensor formulation is clearly laid out, the algebraic derivations in Appendices B-C are transparent, and the two independent datasets provide a welcome cross-check of single-point statistics. The prediction of the -3/2 amplitude decay and the linear wavelength growth is explicit and falsifiable. However, the main analytical result is conditional on a homogeneity-after-scaling assumption that is not directly tested in the present manuscript, and the relation between weighted-space optimality and physical-space L2 orthogonality needs sharper statement. The significance is therefore real but prospective until the key premise is checked.","major_comments":[{"comment":"The deduction of the SADFM rests on the statement that the correlation tensor rR^j_·ĵ in Eq. (3.41) is invariant with respect to ζ = ξ'-ξ, which licenses the Fourier ansatz (3.45). This premise is imported from Ewing et al. (2007) and is not directly verified with the present two-point data. Projecting the measured fields onto the SADFM in Section 6 measures only the energy captured by the basis, not whether the basis diagonalizes the measured two-point correlation tensor. Because both datasets cover the full ξ-range, the authors can test the premise directly by computing the scaled two-point correlation as a function of ζ for several anchor positions ξ; if it does not collapse, the SADFM are not eigenfunctions of the actual LD operator. This is the load-bearing assumption of the paper and should be addressed before the central claim is accepted.","section":"Section 3.3, Eq. (3.41) and the sentence preceding Eq. (3.45)"},{"comment":"Even if exact ξ-homogeneity of Eq. (3.41) holds, the measured domain Ωξ=[0,1.21] is finite, so the integral operator with a homogeneous kernel is Toeplitz rather than circulant in ξ. Strict Fourier eigenfunctions are therefore an infinite-domain or periodic-domain idealization. Appendix C establishes orthogonality of the SADFM on the finite interval, but orthogonality alone does not make them eigenfunctions of the finite-interval LD integral. The paper should quantify the finite-domain error, for example by comparing the Fourier projection with numerical eigenfunctions of the truncated Toeplitz operator in ξ in the same way that the θ-direction is treated numerically, or by showing that boundary terms are negligible over Ωξ. Without this, the -3/2 amplitude prediction is not tied to the actual finite measurement domain.","section":"Section 4, Ωξ=[0,1.21], and Appendix C"},{"comment":"The paper should distinguish more sharply between optimality in the weighted space L2_w and in the physical space L2. The SADFM are eigenfunctions of the weighted LD integral (3.40) after choosing w=e^{-ξ}; the transformation χ=e^{-ξ/2}Φ in Eq. (3.52) is a unitary equivalence that makes the modes orthonormal in L2, but it does not by itself show that the χ-modes are eigenfunctions of the unweighted LD operator. The projection coefficients in Eqs. (6.6)-(6.7) are nevertheless presented as expansions of the physical field. The authors should state explicitly that the optimality claim applies to the weighted inner product, or supply the additional argument showing optimality in L2. This distinction matters because the abstract and conclusions present the SADFM as 'the optimal eigenfunctions' without specifying the space.","section":"Section 3.3, Eqs. (3.36), (3.52), and Section 6.3, Eqs. (6.6)-(6.7)"},{"comment":"The -5/3 and -7/3 spectral ranges are central to the claim that the SADFM reproduce homogeneous-turbulence scaling, but the figure displays no fitted reference lines or uncertainty estimates, and the stated ranges (κ∈[20:300] for -5/3 and κ∈[20:250] for -7/3) are not supported by a quantitative fitting procedure. The authors should provide the fit details and show compensated spectra, particularly for the cross-spectrum away from the centerline where the -7/3 range is reported to disappear. This is needed for the reader to judge whether the observed ranges are significant beyond visual inspection.","section":"Section 6.3, Fig. 9"}],"minor_comments":[{"comment":"The sentence reporting the conservative window sizes is incomplete: '≈ 1.25 and Lξ« 1.1 and' appears to contain a typo, and the stated values should be reconciled with Lξ=1.21 given in Section 4.","section":"Section 5.1"},{"comment":"The notation rUc for the contravariant centerline velocity should be defined more carefully, since it has different dimensions from the physical centerline velocity Uc; currently the reader has to infer the scaling from the reconstruction argument in Eq. (B.5).","section":"Section 3.3, Eq. (3.33) and Appendix B"},{"comment":"Several captions do not identify the subpanels, and the text refers to panels (a)-(h) without a consistent labeling in the figure files; for example, Fig. 9 has no subpanel labels in the caption.","section":"Figures 6-9"},{"comment":"The symbol θ1/2 is used in figures and text but is never formally defined; please add its definition, e.g., the location where the mean streamwise velocity falls to half its centerline value.","section":"Section 4 and Eq. (6.6)"}],"recommendation":"major_revision","confidential_remarks":"The paper is a reasonable fit for the journal's scope and the derivation is internally consistent under its stated assumptions. My recommendation of major revision is driven by the unverified two-point homogeneity premise, which is directly testable with the authors' own data, and by the finite-domain issue. I would be willing to review a revision that adds these tests and sharpens the optimality claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The thing to know about this paper: the analytical part is mostly honest and self-consistent, but the main result is not as deduced as it looks. The authors introduce a weighted inner product in stretched spherical coordinates, choose the weight to cancel the coordinate volume element, and then recover amplitude-decaying Fourier modes. That is a legitimate construction, but the -3/2 decay is put in through the weight, not discovered from the turbulence. The paper admits this, which I respect, but the abstract and conclusions lean harder on the word \"deduced\" than the derivation supports.\n\nWhat is genuinely new and valuable: the tensor formulation of the Lumley decomposition in curvilinear coordinates is clean and should be useful beyond this jet problem. The SADFM idea is sensible—if the scaled two-point correlation is streamwise homogeneous, then Fourier modes are the eigenfunctions, and the L2-transformed basis has a known analytic form. The authors do not invent this premise; they take it from Ewing et al. (2007). But they do not test it with their own PIV data, even though they have the data to do so. Projecting the field onto the SADFM and plotting spectra measures energy in that basis; it does not show that the basis diagonalizes the correlation. That is the load-bearing gap.\n\nThe experimental side is solid as far as it goes. Two independent datasets agree at the single-point level, the mean velocity and Reynolds stress profiles in stretched spherical coordinates collapse well, and the entropy/energy production decomposition is a nice touch. The -5/3 and -7/3 spectral slopes appear in the right places, but without uncertainty estimates or a stated fit range they are suggestive, not demonstrative. The finite ξ-domain is also a real caveat: on an interval, the integral operator is Toeplitz, not circulant, so strict Fourier eigenfunctions require a periodic or infinite-domain idealization. Windowing masks the leakage but does not repair the eigenfunction claim.\n\nBottom line: this is a worthwhile paper with a novel methodological kernel, and it deserves a serious referee. The logic is transparent, the writing is careful, and the limitations are mostly acknowledged. But the central claim is conditional on a directly checkable property of the two-point correlation, and the authors should be asked to check it before publication. I would send it to review, with a request for that test and for uncertainty estimates on the spectral slopes.","headline":"The SADFM result is a genuine, clearly-derived idea, but its central premise—homogeneity of the scaled two-point correlation—is not tested with data that could directly check it.","tokens_in":28655,"tokens_out":2732,"would_cite":true,"duration_ms":32977,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["47.27.wg","47.27.-i"],"model":"deepseek-v4-flash","headline":"This paper claims that the streamwise optimal Lumley-decomposition eigenfunctions of the self-similar round-jet far field are stretched amplitude-decaying Fourier modes, with wavelength growing linearly downstream and amplitude decaying…","keywords":["Lumley decomposition","proper orthogonal decomposition","stretched spherical coordinates","self-similar turbulent jet","stretched amplitude-decaying Fourier modes","turbulent energy spectra","equilibrium similarity","jet far field"],"falsifier":"Measure the two-point correlation tensor of the scaled contravariant velocity in stretched spherical coordinates over an extended streamwise range and check whether the correlation depends only on the separation $\\zeta$ and not on absolute $\\xi$; if it does depend on absolute $\\xi$, the derived SADFM are not eigenfunctions of the actual Lumley operator.","tokens_in":27596,"feed_emoji":"🌊","tokens_out":7580,"duration_ms":72271,"temperature":0.7,"pith_summary":"This paper claims that in the self-similar far field of a round turbulent jet, the energy-optimal basis functions in the streamwise direction are stretched amplitude-decaying Fourier modes: ordinary Fourier oscillations in a logarithmically stretched coordinate, multiplied by a $(C e^{\\xi})^{-3/2}$ decay. The argument recasts the Lumley decomposition in tensor form, works in stretched spherical coordinates, and introduces a streamwise-decaying weight $w=e^{-\\xi}$ in the inner product so the two-point correlation becomes translationally invariant along the stretched streamwise coordinate. If correct, this gives an analytical description of the streamwise part of the optimal basis for the jet far field, relating spatially growing wavelength and decaying energy to turbulent spectra. It also extends Fourier-based decomposition to inhomogeneous directions in flows that admit equilibrium similarity.","feed_headline":"Jet far-field energy modes are stretched, decaying waves","feed_subtitle":"In a self-similar round jet, optimal streamwise modes are stretched, decaying Fourier waves.","key_machinery":"The central object is the weighted Lumley-decomposition integral in tensor form, $\\mathcal{R}\\Phi = \\langle v(\\Phi,v)_w\\rangle = \\lambda\\Phi$, restricted to the weighted space $L^2_w$ with weight $w=e^{-\\xi}$ in stretched spherical coordinates. The weight is chosen to cancel exactly the geometric factors $e^{3\\xi}$ (volume element) and $e^{2\\xi}$ (metric) introduced by the stretched coordinates, making the scaled two-point correlation invariant under streamwise separation and allowing a Fourier ansatz along $\\xi$. The transformation $\\chi = e^{-\\xi/2}\\Phi$ then restores orthogonality in the unweighted $L^2$ space and fixes the $-3/2$ amplitude decay, producing the stretched amplitude-decaying Fourier modes that carry the argument.","core_discovery":"After transforming the round jet far field to stretched spherical coordinates $\\xi=\\ln(r/C)$, and using a weighted inner product with $w=e^{-\\xi}$, the Lumley decomposition integral becomes translationally invariant in $\\xi$. The paper deduces that the physical streamwise eigenfunctions take the stretched amplitude-decaying Fourier form $\\chi^\\xi_\\alpha = \\psi^\\xi_\\alpha(\\theta) e^{i(\\omega t + \\kappa\\xi + m\\phi)} / ((C e^{\\xi})^{3/2} \\sqrt{A} \\sin\\theta)$, so in terms of distance from the virtual origin, $\\tilde{x}=(x-x_0)/C$, the streamwise evolution is $\\tilde{x}^{i\\kappa - 3/2}$. Thus the wavelength increases linearly with downstream distance while the amplitude decays as the $-3/2$ power of distance, which the authors describe as reversed wave shoaling. Energy spectra computed by projecting experimental PIV fields onto these modes show a $-5/3$ range and a $-7/3$ cross-spectrum slope, the same scaling laws normally associated with homogeneous and constant-shear turbulence.","pith_inferences":["If the same weighted-LD construction works in planar wakes, mixing layers, or boundary layers with equilibrium similarity, the SADFM form would be a generic consequence of similarity scaling rather than of spherical geometry; this is directly testable by repeating the derivation in those coordinates.","The paper's Appendix E shows that the weight $e^{-\\xi}$ removes roughly half the resolved turbulent kinetic energy in the chosen window; a per-mode correction would allow eigenvalue spectra from weighted and unweighted decompositions to be compared quantitatively.","Because the $-5/3$ and $-7/3$ slopes survive in the SADFM basis, one could test whether these slopes are more robust under SADFM projection than under ordinary Fourier projection when the streamwise window is shortened, a question the current dataset could in principle settle."],"forward_implications":["The streamwise part of the optimal basis for the self-similar jet far field is known analytically, so modal analysis along the jet does not require numerically computed streamwise eigenfunctions.","Energy spectra projected onto SADFM exhibit $-5/3$ and $-7/3$ power-law regions, indicating that these classic scaling exponents are not tied to strict homogeneity or to ordinary trigonometric Fourier modes.","The eigenvalue problem is solved analytically in the streamwise and azimuthal directions, leaving only the transverse direction to numerical diagonalization; the first mode holds 38.3% of the energy and the first seven modes hold 80%.","The tensor formulation and the weight-function construction apply to any flow admitting equilibrium similarity, enabling Fourier-based decomposition along inhomogeneous flow directions in such flows."],"supporting_citations":[{"why":"Introduces the Lumley decomposition and the Fourier-based eigenfunction approach for homogeneous and stationary directions, which this paper extends.","marker":"Lumley (1967b)"},{"why":"Supplies the two-point similarity scaling of the round jet that justifies the homogeneity-after-scaling assumption in the stretched coordinates.","marker":"Ewing et al. (2007)"},{"why":"Provides the centerline velocity decay and virtual-origin scaling used to define the stretched spherical coordinates and the characteristic velocity.","marker":"Hussein et al. (1994)"},{"why":"Formulates the weighted Lumley decomposition integral and the self-adjoint operator framework that the tensor formulation builds on.","marker":"Holmes et al. (2012)"},{"why":"Provides the equilibrium similarity theory that underlies scaling the jet statistics and the extension to other self-similar flows.","marker":"George (1989)"},{"why":"Predicts the $-7/3$ cross-spectrum slope under uniform mean shear, which the SADFM spectra are compared against.","marker":"Lumley (1967a)"},{"why":"Previously measured full-field spatial spectra in the same jet facility and demonstrated the windowing and scaling effects that the current analysis refines.","marker":"Wänström (2009)"}],"fun_headline_variants":["Reversed wave shoaling: jet far-field modes stretch then decay","Stretched, decaying Fourier modes emerge from jet self-similarity","Round jet far field: wavelength grows, amplitude decays as -3/2 power","Jet far-field hidden order: stretched decaying Fourier waves","Anti-shoaling waves: jet far-field modes stretch and decay"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The derivation assumes the two-point correlation of the centerline-scaled contravariant velocity is independent of absolute position in the stretched streamwise coordinate $\\xi$, so the correlation depends only on the separation $\\zeta=\\xi'-\\xi$ and the Fourier ansatz applies.","fun_headline_variants_meta":{"raw":{"variants":["Reversed wave shoaling: jet far-field modes stretch then decay","Stretched, decaying Fourier modes emerge from jet self-similarity","Round jet far field: wavelength grows, amplitude decays as -3/2 power","Jet far-field hidden order: stretched decaying Fourier waves","Anti-shoaling waves: jet far-field modes stretch and decay"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000964,"raw_usage":{"total_tokens":4155,"prompt_tokens":1050,"completion_tokens":3105,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":666,"completion_tokens_details":{"reasoning_tokens":3014}},"tokens_in":666,"tokens_out":3105,"duration_ms":19800,"temperature":1.0,"reasoning_tokens":3014,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:23:35.601218+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the two-point correlation tensor of the scaled contravariant velocity in stretched spherical coordinates over an extended streamwise range and check whether the correlation depends only on the separation $\\zeta$ and not on absolute $\\xi$; if it does depend on absolute $\\xi$, the derived SADFM are not eigenfunctions of the actual Lumley operator.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Formulates the weighted Lumley decomposition integral and the self-adjoint operator framework that the tensor formulation builds on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the equilibrium similarity theory that underlies scaling the jet statistics and the extension to other self-similar flows."}],"review_version":1}