{"id":"e50a40f0-03c7-41a0-a7fe-7809c31b7607","arxiv_id":"1909.01307","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A modal decomposition of the round jet far field shows that the first Lumley mode dominates shear-stress production and that many modes draw energy directly from the mean flow, including in the -5/3 spectral range.","lead":"Using particle image velocimetry data from a turbulent round jet, this paper splits the energy equation into modes and shows that many modes receive energy directly from the mean flow, even in the wavenumber range where the spectrum shows the classic -5/3 slope. The result matters because it suggests the standard energy cascade picture is incomplete in high-shear jet regions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The SADFM basis is asserted, not validated: Eq. (3.15) contains an undefined frequency ω, and no reconstruction of the two-point correlation tensor is shown, so the modal-production claim is conditional on an unverified eigenbasis.","rationale":"The reader's CONDITIONAL verdict is appropriate. My stress-test converges on the same weakest assumption: the SADFM basis is the load-bearing element of the paper. The production analysis, spectral reconstructions, and the headline claims all depend on the eigenfunctions in Eq. (3.15). The paper asserts this basis from Part 1 but does not demonstrate that it is the eigenbasis of the measured correlation tensor. The missing temporal frequency ω compounds this: as written, Eq. (3.15) has no time variable, and PIV is typically not time-resolved at jet turbulence scales, so the 'Lumley Decomposition' may not be the space-time decomposition it claims to be. If the basis is wrong, the modal production values and the -7/3 single-mode reconstruction are not physically meaningful. I do not see an internal inconsistency in the Galerkin projection itself; the concern is one of validation and completeness. A direct reconstruction test would settle it. Thus the verdict remains CONDITIONAL, pending the basis validation and data availability.","tokens_in":20731,"tokens_out":9558,"duration_ms":93105,"concrete_test":"Reconstruct the measured two-point correlation tensor R^i_·j(ξ,θ,φ;ξ',θ',φ') from the SADFM eigenfunctions (3.15) with the reported eigenvalues and compare it directly to the correlation tensor computed from the PIV data before Parzen windowing. Compute the relative L2 error over the resolved domain and wavenumber range, including the -5/3 region. If the error exceeds a few percent, or if the reconstruction requires the time dimension (ω) that the data cannot supply, recompute the energy-normalized production (3.25) and the shear-stress reconstructions using a numerically computed POD basis from the same data and check whether the conclusions in the abstract survive.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim—that a wide range of LD modes receive significant energy directly from the mean flow even in the -5/3 range—is computed from the modal production term (3.21b), P = λ φ^i_α φ^{j*}_α ∇_j⟨V^i⟩. This expression is only the production of the true dynamical modes if the Φ_α in Eq. (3.15) are the actual eigenfunctions of the sampled two-point correlation tensor. The paper provides no validation of this. The SADFM form is asserted from Hodžić et al. (2019) with a stretched ξ-Fourier factor and a numerical θ-factor, but no residual or reconstruction-error of the two-point correlation is reported; Figure 13 validates only single-point statistics. Additionally, Eq. (3.15) contains a frequency ω with no time variable (e^{i(ω+κξ+mφ−2ξ)}), and the paper never states how ω is obtained from the PIV data, which are not described as time-resolved. If the basis is incomplete or the temporal dimension is mishandled, the ENP values in Figs. 8–10 and the single-mode reconstruction of the -7/3 cross-spectrum could be artifacts of the assumed basis rather than properties of the jet flow.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper, the second part of a two-part study, uses a semi-analytical Lumley Decomposition (LD) based on stretched amplitude decaying Fourier modes (SADFM) to analyze the dynamics of the turbulent round jet far field from PIV data. The authors derive a Galerkin projection of the turbulent kinetic energy transport equation in curvilinear coordinates, identify the modal production term, and use it to compute energy-normalized production (ENP) spectra. They report three central results: (i) a wide range of LD modes receive significant energy directly from the mean flow, even at wavenumbers where the averaged spectra exhibit the -5/3 slope; (ii) the -7/3 range of the cross-spectrum is reconstructed by the first mode alone in regions of high mean shear; and (iii) the first two modes almost fully reconstruct the shear-stress profile. The paper interprets these results as supporting the hypothesis of Wänström (2009) that multiple modes tap directly into mean-flow energy, in contrast to a purely Richardson-like cascade.","tokens_in":21003,"tokens_out":4426,"duration_ms":45157,"significance":"If the central claims are correct, the paper provides a substantive challenge to the usual interpretation of the -5/3 spectral range as a conservative cascade range in a inhomogeneous shear flow, and it supplies a tensor-calculus framework for modal TKE budgets that could be reused in other flows. The reconstruction of the -7/3 cross-spectrum by a single mode and the rapid convergence of the shear-stress reconstruction are striking and potentially useful results for reduced-order modeling. The analytical derivation of the modal transport equation in general coordinates is a clean formal contribution, and the authors are appropriately transparent about which terms can be reconstructed from the present data and about the effect of the Parzen window. However, the dynamical conclusions rest on the validity of the SADFM eigenfunction basis and on the interpretation of the ENP metric; neither of these is currently validated or quantified with respect to uncertainty, which limits the strength of the claims.","major_comments":[{"comment":"The eigenfunction basis is asserted rather than validated, and the frequency ω is undefined. The production term (3.21b) and all ENP results in Figs. 8–10 are computed by projecting onto the SADFM basis Φ_α defined in (3.15). These are production rates of the actual dynamical LD modes only if the SADFM functions are eigenfunctions of the sampled two-point correlation tensor. The paper does not report any residual or reconstruction error of the two-point correlation tensor; Fig. 13 validates only single-point statistics. Moreover, Eq. (3.15) contains e^{i(ω+κξ+mφ−2ξ)}, but no time variable appears in the expression, and the PIV data are not described as time-resolved. The authors should state how ω is obtained from the data, or remove it from the ansatz, and should provide a direct check of the basis, e.g., the reconstruction error of R^i_{·j}, a comparison with a numerical eigen-decomposition of the sampled correlation tensor, or an explicit reference to the validation in the companion paper (Hodžić et al. 2019). Until this is supplied, the central claim in Section 4 that 'a wide range of modes obtain a substantial part of their energy directly from the mean flow' is conditional on an unverified ansatz.","section":"§3.1, Eq. (3.15)"},{"comment":"The claim of near-constant ENP levels over the -5/3 range is made without uncertainty quantification. For example, the text reports a standard deviation of about 8.5% for α=1 over κ∈[26,300], but this is a scatter about the mean of a single realization, not an estimate of sampling uncertainty. Likewise, the fitted -5/3 and -7/3 slopes in Figs. 3–5 and 12 are presented without confidence bands or a stated fitting procedure. Since the central conclusion concerns the relative importance of direct production versus cascade transport in the -5/3 range, the authors should provide at least bootstrap confidence intervals on P_{ρλ,θ} and on the spectral slopes, or otherwise quantify the uncertainty from the finite PIV sample. Without this, the distinction between single-mode reconstruction of the -7/3 range and multi-mode reconstruction of the -5/3 range cannot be assessed quantitatively.","section":"§3.6, Figs. 8 and 10"},{"comment":"The ENP metric normalizes production by the eigenvalue λ, but 'significant' is not tied to any comparison point. In the -5/3 range the eigenvalues of high-wavenumber modes are small, so even a modest absolute production rate can yield a large normalized value. The abstract and conclusions state that modes obtain 'significant amounts of energy directly from the mean flow,' yet no benchmark is given against which this significance is judged. The authors should compare the production term to the dissipation rate and/or to the spectral energy flux across the -5/3 range, and discuss whether the inferred direct production changes the leading-order budget. Without such a comparison, the claim of 'significant' direct production remains a statement about the normalization convention rather than about the dynamical balance.","section":"§3.6, Eq. (3.25)"}],"minor_comments":[{"comment":"There are typos throughout, e.g., 'the the experimental setup' (p. 2), 'commmonly' (p. 7), and 'cumulutive' (Fig. 13 caption).","section":"§1, p. 2"},{"comment":"The sentence 'If α = β = γ expression (3.21 b) defines the non-linear energy transfer within a given mode' should refer to Eq. (3.21c), since (3.21b) is the production term; the current text appears to mislabel the equation reference.","section":"§3.3, p. 7"},{"comment":"The text refers to 'figure 14' before Figure 13 has been introduced; the appendix figures should be renumbered or cross-referenced in the order of appearance.","section":"§3.4, p. 11"},{"comment":"The symbol θ_{1/2} is used to normalize the transverse coordinate in Figs. 3–5 and throughout the text, but it is not defined in this paper; please define it explicitly.","section":"§3.2, p. 5"}],"recommendation":"major_revision","confidential_remarks":"The central issue is whether the SADFM basis from the companion paper has been validated as the eigenbasis of the sampled two-point correlation tensor. If the companion paper (Hodžić et al. 2019) already provides such a validation, the authors should cite the specific section and summarize the result; if not, the dynamical conclusions in the present manuscript remain conditional on an ansatz. I would also encourage the editor to check that the companion paper has been or is being published in a citable venue, as the present manuscript relies on it heavily."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this paper adds a real, useful diagnostic—modal production budgets for the far-field round jet—and reports two striking observations: the -7/3 cross-spectrum in high-shear regions is essentially one mode, and the first two modes rebuild the shear stress. If those observations survive scrutiny, they matter for spectral closures and for how we teach the -5/3 range in shear flows. The paper is also honest about the Wänström lineage: the central hypothesis is hers, and the authors frame this as a quantitative test, not a first suggestion.\n\nWhat is genuinely new is the ENP analysis itself: P/(ρλ) as a function of mode number, wavenumber, and θ. The near-constant ENP levels across a wide range of κ including the -5/3 range are an interesting, nontrivial datum. The modal self-similarity quantification via the θ-inner product is also a clean way to make that notion precise. The tensor Galerkin projection is formal but appears correct, with the caveat that the α-summation in Eq. (3.21b) is handled by fiat.\n\nThe soft spots are real but not fatal. The biggest is that Eq. (3.15) contains an unexplained temporal frequency ω in the SADFM basis, with no time variable and no statement of how ω is obtained from PIV data that are not described as time-resolved. That needs a fix before the analysis is fully comprehensible. More worrying, the paper validates the SADFM reconstruction only through single-point statistics and spectral reconstructions, not through the two-point correlation tensor from which the modes are supposed to come. Without a residual or a direct comparison to the sampled R_ij, the modal production levels are conditional on an asserted basis. A projection onto any basis will produce some P; the claim that these are the true Lumley modes needs evidence. I would also want error bars or at least a convergence estimate for the ENP and spectral slopes; the 8.5% standard deviation quoted for mode 1 is a start but not an uncertainty.\n\nFinally, the circularity concern is overstated in one respect: any POD analysis reconstructs the same data it was built from. That is the method. The issue is not circularity but completeness—does the basis actually span the measured correlation? On that, the paper is currently silent.\n\nRecommendation: yes, send it to review. The core observation is important enough and the methodology serious enough to deserve referee time, but the authors should be pushed to validate the basis, clarify ω, and provide data or code.","headline":"A genuine modal production budget for the jet far field, worth referee time, but the SADFM basis—especially the unexplained ω in Eq. (3.15)—needs validation before the central claim is fully consequential.","tokens_in":21511,"tokens_out":3145,"would_cite":false,"duration_ms":33970,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The far-field round jet's Lumley modes draw a substantial share of their energy directly from the mean flow, including in the wavenumber band where the spectrum shows a -5/3 slope.","keywords":["Lumley decomposition","proper orthogonal decomposition","turbulent round jet","far-field turbulence","turbulence kinetic energy production","energy cascade","shear stress spectrum","stretched amplitude decaying Fourier modes"],"falsifier":"Recompute the energy-normalized production on a time-resolved far-field jet dataset or a large-eddy simulation and check whether the production-to-energy ratio for modes at wavenumbers $\\kappa \\in [20,300]$ stays near its high-shear plateau; if the plateau disappears away from the half-width region, the claim that a wide range of modes are directly fed by the mean flow would fail.","tokens_in":20560,"feed_emoji":"🌊","tokens_out":10094,"duration_ms":89267,"temperature":0.7,"pith_summary":"This paper tries to establish that in the far field of a turbulent round jet, many Lumley-decomposition modes receive a significant fraction of their kinetic energy directly from the mean flow, even at wavenumbers where the averaged spectrum exhibits the $-5/3$ slope. Using the stretched amplitude decaying Fourier modes from the authors' earlier kinematics paper, it projects the turbulence kinetic energy equation onto the modes and reconstructs the production term mode by mode. The result is a per-mode energy budget in which shear-stress production is dominated by the first two modes, the first mode alone rebuilds the $-7/3$ cross-spectrum range in the high-shear half-width region, and the energy-normalized production stays nearly constant across a wide range of modes and wavenumbers. If correct, this means the inertial range of a jet is not just a passive cascade: individual scales are also being fed by the mean shear.","feed_headline":"Jet modes feed directly on mean flow even at -5/3 scales","feed_subtitle":"Per-mode energy budget shows shear-stress production sits in the first two modes, not an energy cascade.","key_machinery":"The argument is carried by the modal production term (3.21b), $\\lambda\\phi^i_\\alpha\\phi^{j*}_\\alpha\\nabla_j\\langle V^i\\rangle$, which measures how much turbulence kinetic energy a given Lumley mode draws per unit time from the mean-velocity gradients. Dividing by the modal energy $\\lambda$ gives the energy-normalized production (3.25), whose near-constancy across mode number and wavenumber is the quantitative evidence for direct mean-flow feeding. These objects are computed on the stretched amplitude decaying Fourier modes (SADFM) of Part 1, a basis in stretched spherical coordinates whose streamwise part is a decaying Fourier wave, combined with a Galerkin projection of the energy equation in curvilinear coordinates.","core_discovery":"The central discovery is a modal energy budget for the far-field round jet in which scale-by-scale feeding from the mean flow is the norm. When the production term of the turbulence kinetic energy equation is expanded in the Lumley eigenfunctions, the energy-normalized production $P_{\\rho\\lambda}$ remains at significant and nearly constant levels over a wide range of mode numbers and wavenumbers, including the $\\kappa$-range $20 < \\kappa < 300$ where the averaged one-dimensional spectra follow a $-5/3$ power law. The same modal budget shows that the first mode alone reconstructs the $-7/3$ range of the cross-spectrum near the half-width, where mean shear is high, and the first two modes almost completely reconstruct the Reynolds shear-stress profile. These observations are presented as support for the hypothesis that multiple modes extract energy directly from the mean flow, contradicting the idea that the $-5/3$ range is exclusively an inertial range fed only by a Richardson-style cascade.","pith_inferences":["The paper does not examine other inhomogeneous shear flows, but if the near-constant energy-normalized production is robust, the same direct mean-flow feeding should also show up in boundary layers and wakes, where spectral modes in the inertial range are often assumed to be cascade-dominated.","The paper does not specify how the temporal frequency $\\omega$ in the mode definition is obtained from PIV data; if it is inferred indirectly, the production plateau could be partly sensitive to that choice, so a time-resolved dataset would be the cleanest test.","A practical modeling direction left implicit in the paper is to assign each mode its measured production-to-energy ratio from mean-flow gradients rather than treating the inertial range as a passive drain, which could be formulated as a testable subgrid closure."],"forward_implications":["A two-mode truncation of the Lumley decomposition should be enough to reproduce the Reynolds shear-stress profile in the jet far field, since modes 1 and 2 nearly reconstruct it.","The $-7/3$ cross-spectrum range in high-shear regions is a property of the first mode by itself, so low-order models that keep the first mode can capture the shear-stress spectrum without higher modes.","Energy production in the $-5/3$ range implies that cascade-only closures miss a real source term; models must account for mode-specific mean-flow production.","Modal self-similarity across wavenumbers, manifested as a collapse of the modal spectral building blocks, offers a possible scaling law for transferring modes between wavenumber bands."],"supporting_citations":[{"why":"Supplies the SADFM eigenfunction basis, the PIV dataset E1, and the tensor formulation of the Lumley decomposition that the production analysis expands.","marker":"Hodžić et al. (2019)"},{"why":"Poses the hypothesis the paper tests: that multiple modes can extract a constant and significant share of their energy directly from the mean flow.","marker":"Wänström (2009)"},{"why":"Provides the cross-spectral similarity model for constant-shear flows that the paper extends to the modal building blocks of the spectra.","marker":"Lumley (1967)"},{"why":"Supplies the one-dimensional -5/3 spectral model used to interpret sign changes in the modal spectral flux.","marker":"Tennekes & Lumley (1972)"},{"why":"Gives the jet velocity-decay constant and momentum-flux scaling used to nondimensionalize the reconstructed stresses.","marker":"Hussein et al. (1994)"},{"why":"Establishes two-point similarity of the scaled velocity correlation, supporting the streamwise Fourier-type mode construction.","marker":"Ewing et al. (2007)"},{"why":"Documents earlier far-field mode evolution and the modal self-similarity that the present work quantifies.","marker":"Gamard et al. (2004)"}],"fun_headline_variants":["Jet far-field modes draw energy from mean flow at all scales","Shear-stress budget pinned to first two Lumley modes","Lumley decomposition: cascade not sole source for -5/3 scales","Mean flow directly powers jet modes across wavenumber range"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the assumption that the stretched amplitude decaying Fourier modes from the earlier paper really are the eigenfunctions of the measured correlation tensor, and that the temporal frequency written into those modes is available from PIV data, although the paper does not explain how.","fun_headline_variants_meta":{"raw":{"variants":["Jet far-field modes draw energy from mean flow at all scales","Shear-stress budget pinned to first two Lumley modes","Lumley decomposition: cascade not sole source for -5/3 scales","Mean flow directly powers jet modes across wavenumber range"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000793,"raw_usage":{"total_tokens":3487,"prompt_tokens":936,"completion_tokens":2551,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":552,"completion_tokens_details":{"reasoning_tokens":2478}},"tokens_in":552,"tokens_out":2551,"duration_ms":18223,"temperature":1.0,"reasoning_tokens":2478,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T05:21:32.462403+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the energy-normalized production on a time-resolved far-field jet dataset or a large-eddy simulation and check whether the production-to-energy ratio for modes at wavenumbers $\\kappa \\in [20,300]$ stays near its high-shear plateau; if the plateau disappears away from the half-width region, the claim that a wide range of modes are directly fed by the mean flow would fail.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the cross-spectral similarity model for constant-shear flows that the paper extends to the modal building blocks of the spectra."},{"cited_title":"& Lumley, J","cited_arxiv_id":null,"evidence_quote":"Supplies the one-dimensional -5/3 spectral model used to interpret sign changes in the modal spectral flux."},{"cited_title":"J., Capp, S","cited_arxiv_id":null,"evidence_quote":"Gives the jet velocity-decay constant and momentum-flux scaling used to nondimensionalize the reconstructed stresses."},{"cited_title":"K., Pedersen, J","cited_arxiv_id":null,"evidence_quote":"Establishes two-point similarity of the scaled velocity correlation, supporting the streamwise Fourier-type mode construction."},{"cited_title":"& George, W","cited_arxiv_id":null,"evidence_quote":"Documents earlier far-field mode evolution and the modal self-similarity that the present work quantifies."}],"review_version":1}