REVIEW 4 major objections 2 minor 16 references
Lab-scale contrails show ice scattering is more sensitive to water vapor than to soot load.
Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →
T0 review · grok-4.5
2026-07-13 20:57 UTC pith:27PARCAR
load-bearing objection Abstract promises a useful aircraft-relevant contrail lab and a water-vapor-over-soot ranking, but the supplied full text is a different paper, so the claims cannot be checked. the 4 major comments →
Effects of fuel and soot concentrations on the inception and development of contrails
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
Under matched laboratory cruise conditions, contrail scattering intensity increases with soot number concentration and exhaust water vapor, yet comparisons of ethylene and propane exhausts show that scattering propensity is more sensitive to water vapor content than to soot concentration; the same measurements supply the first experimental cross-section images of ice nucleation across turbulent shear layers and depolarization evidence of aspherical ice habits.
What carries the argument
A novel laboratory contrail tunnel fed by an inverted co-flow soot generator (ethylene or propane at set global equivalence ratios), operated at 20.8 kPa and 190 K, with SMPS/TEM soot characterization, instantaneous and averaged scattering intensity (and depolarization) diagnostics, and complementary FANS simulations with a two-equation soot–ice particulate model.
Load-bearing premise
The inverted co-flow soot generator and contrail tunnel at 20.8 kPa and 190 K must sufficiently reproduce long-haul aircraft exhaust–ambient mixing for the measured scattering and depolarization to stand in for real contrail ice inception and crystal habit.
What would settle it
If, under the same facility settings, ethylene and propane exhausts with deliberately matched water-vapor content but deliberately different soot number concentrations produced equal average scattering intensities—or if independent in-flight or remote-sensing campaigns found soot load dominating water vapor under cruise conditions—the claimed water-vapor sensitivity would be falsified.
If this is right
- Contrail optical properties used in climate models can be treated as more strongly controlled by exhaust humidity than by soot number for the fuel range studied.
- The laboratory cross-section images supply quantitative targets for microphysical–turbulence coupling in contrail simulations.
- Depolarization measurements give a direct lab route to ice-habit asphericity that can constrain remote-sensing retrievals.
- Fuel and combustion-parameter choices that change water-vapor yield can be ranked for contrail scattering impact without full flight tests.
Where Pith is reading between the lines
- If water vapor dominates soot load, hydrogen-rich or high-water-yield fuels may form optically stronger contrails even when soot is reduced, which would reverse simple soot-only mitigation strategies.
- The same tunnel geometry could be reused to test sustainable aviation fuels or ammonia blends by swapping only the soot-generator fuel stream.
- Shear-layer nucleation snapshots suggest that mixing-layer thickness, not only mean supersaturation, should become an explicit input in reduced-order contrail models.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The abstract of arXiv:2603.21226 claims a novel laboratory contrail facility (inverted co-flow soot generator feeding a tunnel at 20.8 kPa and 190 K), ethylene/propane exhausts at varied equivalence ratio, SMPS/TEM soot characterization, instantaneous and mean scattering plus depolarization, and complementary FANS two-equation soot–ice simulations. It asserts the first experimental contrail cross-section snapshots of ice nucleation across shear layers and that scattering propensity is more sensitive to exhaust water vapor than to soot number. The supplied full manuscript body, however, is an entirely different paper (LSTM-PINN benchmark for two-dimensional steady electrohydrodynamic shock-like flows; eight manufactured cases; Standard/ResAtt/LSTM-PINN comparison; Tables 2–14). No soot generator, contrail tunnel, optical diagnostics, fuel comparison, or FANS ice model appears in the body. The abstract claims therefore cannot be checked against methods, data, or analysis in the review package.
Significance. If the abstract claims were supported by controlled experiments and validated simulations, the work would matter for aviation climate research: a lab platform that isolates fuel type, soot loading, and water vapor under cruise-like T–p, plus first resolved cross-section views of turbulent mixing versus ice microphysics, would address questions that remote sensing and in-flight campaigns struggle with. The water-vapor-versus-soot ranking would be a useful, falsifiable design insight. Those strengths cannot be credited here because the load-bearing evidence is absent from the manuscript body. (The body text that is present is a separate, potentially useful PINN benchmark with open code, but it is not the paper under the stated title and abstract.)
major comments (4)
- Title, abstract, and paper_id (2603.21226, contrails / physics.flu-dyn) do not match the full manuscript body, which is a complete LSTM-PINN electrohydrodynamic benchmark (highlights, §§1–6, Cases 01–08, Tables 2–14, arXiv:2603.21227-style content). There is no coherent manuscript on contrail inception to evaluate. This is a document-integrity failure, not a local revision item.
- Central empirical claim—that scattering propensity is more sensitive to exhaust water vapor than to soot concentration for ethylene vs propane—has no supporting matrix, figures, error bars, or controls in the supplied text. Equivalence-ratio set, soot number/size (SMPS), morphology (TEM), water-vapor accounting, optical calibration, and any ethylene–propane comparison tables are missing. The claim cannot be assessed for internal consistency or effect size.
- The facility-emulation premise (inverted co-flow exhaust mixed into 20.8 kPa / 190 K ambient as a long-haul cruise analogue) and the “first experimental snapshots” of ice nucleation across shear layers are stated only in the abstract. No tunnel geometry, mixing diagnostics, imaging setup, or cross-section results appear in the body, so the aircraft-relevance and novelty claims are unsupported in the review package.
- Complementary FANS + two-equation particulate (soot/ice) simulations are cited in the abstract as part of the evidence chain but never appear (no equations, grid, nucleation model, or experiment–simulation comparison). Without them, the abstract’s modeling contribution cannot be refereed.
minor comments (2)
- Abstract closing sentence has a grammar error (“our study present”). Irrelevant until a matching full text is provided.
- If the intended submission is the LSTM-PINN EHD paper that occupies the body, it should be submitted under its own title/abstract and ID; the present package conflates two unrelated works.
Circularity Check
No load-bearing circular derivation: LSTM-PINN EHD benchmark is an empirical architecture comparison against manufactured analytical solutions; only mild self-citation of the inherited backbone.
specific steps
-
self citation load bearing
[Section 3.1 (LSTM-PINN Architecture); also Introduction citing Tao et al. [8, 11]]
"The network design adopted in this study is directly motivated by the LSTM-PINN architecture proposed by Tao et al. [8]. In this work, the algorithmic framework and the basic design idea of the recurrent physics-informed backbone are therefore inherited from the LSTM-PINN method reported by Tao et al. [8]."
The method under evaluation is taken from overlapping-author prior work rather than re-derived here. This is ordinary self-citation of a backbone, not a uniqueness or uniqueness-imported step that forces the eight-case accuracy ranking; the ranking itself is still measured against independent analytical references under a shared loss. Flagged only as mild, non-load-bearing self-citation.
full rationale
The supplied full manuscript is the LSTM-PINN electrohydrodynamic benchmark (not the contrail abstract header). Its central claims are (i) a unified four-variable residual formulation of steady EHD and (ii) an empirical ranking that LSTM-PINN beats Standard PINN and ResAtt-PINN on eight manufactured-solution cases under identical losses, sampling, and equations. Those rankings are measured RMSE/MSE/MAE against analytically constructed reference fields, not quantities forced by a fit or by definition of the inputs. The governing operators, residual construction (Eqs. 10–16), and loss are the same for all three models, so superiority is not self-definitional. The only mild circularity-adjacent element is inheritance of the LSTM backbone from the authors’ prior LSTM-PINN papers [8, 11]; that is ordinary method reuse and is not used as a uniqueness theorem that forbids alternatives or that forces the reported errors. No fitted parameter is renamed a prediction; no ansatz is smuggled in as a theorem; no known empirical law is merely relabeled. Score 1 reflects non-load-bearing self-citation of the architecture under test, not a circular derivation chain.
Axiom & Free-Parameter Ledger
axioms (4)
- domain assumption Laboratory ambient conditions of 20.8 kPa and 190 K plus mixing of inverted co-flow soot-generator exhaust adequately represent long-haul aircraft cruise contrail inception.
- domain assumption Soot particles from ethylene and propane flames act as the primary ice-nucleating particulate matter controlling contrail inception in the tunnel.
- domain assumption Instantaneous and averaged optical scattering intensity (and depolarization) are valid proxies for ice crystal amount and aspheric habit under the reported conditions.
- domain assumption Favre-averaged Navier–Stokes with a two-equation model for soot and ice is an adequate complementary description of the tunnel flow and particulate evolution.
read the original abstract
Fundamental questions related to the roles of fuel type, combustion parameters, and turbulence transport interactions in the inception and growth of contrails have remained intractable in remote sensing and in-flight measurements. Consequently, we developed a novel laboratory-scale facility for studying the inception, growth and persistence of contrails for aircraft-relevant conditions. The set of exhaust conditions, generated using an inverted co-flow soot generator at a set of global equivalence ratio for two fuels - ethylene and propane, is supplied to the contrail tunnel which then mixes with an ambient flow emulating long-haul aircraft cruise conditions (\SI{20.8}{kPa} and \SI{190}{K}). Detailed soot characterization using a scanning mobility particle sizer and transmission electron microscopy is coupled with measurements of instantaneous and averaged scattering intensities from the generated contrails. The experimental results are complemented by numerical simulations of the contrail tunnel using solutions of the Favre-averaged Navier-Stokes (FANS) equation and a two-equation model for handling particulate matter, including soot and ice. Results show the first experimental snapshots of a contrail cross section, highlighting the interaction of turbulent mixing and microphysical growth scales involved in ice nucleation across the shear layers. As expected, the average scattering intensities of contrails increase with soot number concentrations and water vapor content. Comparisons between ethylene and propane exhausts indicate that the scattering propensity of contrails is more sensitive to exhaust water vapor content than to soot concentrations. Finally, depolarization measurements are used to show asphericity in ice crystal habits. Thus, our study present a unique window into contrail formation, theoretical modeling and simulation.
Reference graph
Works this paper leans on
-
[1]
Woias, Micropumps—past, progress and future prospects, Sensors and Actuators B: Chemical 105 (2005) 28–38
P. Woias, Micropumps—past, progress and future prospects, Sensors and Actuators B: Chemical 105 (2005) 28–38
2005
-
[2]
X. Wang, C. Cheng, S. Wang, S. Liu, Electroosmotic pumps and their applications in microfluidic systems: a review, Microfluidics and Nanofluidics 6 (2009) 145–162
2009
-
[3]
S.Zeng,C.-H.Chen,J.C.Mikkelsen,J.G.Santiago, Fabricationandcharacterizationofelectroosmoticmicropumps, SensorsandActuators B: Chemical 79 (2001) 107–114
2001
-
[4]
Iranshahi, T
K. Iranshahi, T. Defraeye, R. M. Rossi, U. C. Müller, Electrohydrodynamics and its applications: Recent advances and future perspectives, International Journal of Heat and Mass Transfer 232 (2024) 125895
2024
-
[5]
Raissi, P
M. Raissi, P. Perdikaris, G. E. Karniadakis, Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations, Journal of Computational Physics 378 (2019) 686–707
2019
-
[6]
C.Zhao,F.Zhang,W.Lou,X.Wang,J.Yang, Acomprehensivereviewofadvancesinphysics-informedneuralnetworksandtheirapplications in complex fluid dynamics, Physics of Fluids 36 (2024) 101301
2024
-
[7]
J.Abbasi,A.D.Jagtap,B.Moseley,A.Hiorth,P.Ø.Andersen, Challengesandadvancementsinmodelingshockfrontswithphysics-informed neural networks: A review and benchmarking study, Neurocomputing 657 (2025) 131440
2025
-
[8]
Z. Tao, K. Xu, F. Liu, Lstm-pinn: An hybrid method for prediction of steady-state electrohydrodynamic flow, Journal of Computational Physics 548 (2026) 114586
2026
-
[9]
F. Xing, J. Li, Z. Tao, F. Liu, Y. Tan, Modeling dynamic gas–liquid interfaces in underwater explosions using interval-constrained physics- informed neural networks, Physics of Fluids 37 (2025) 116126
2025
-
[10]
A.Kashefi,T.Mukerji, Physics-informedKANpointnet:Deeplearningforsimultaneoussolutionstoinverseproblemsinincompressibleflow on numerous irregular geometries, arXiv preprint arXiv:2504.06327 (2025)
Pith/arXiv arXiv 2025
-
[11]
Z. Tao, An lstm-pinn hybrid method to the specific problem of population forecasting, arXiv preprint arXiv:2505.01819 (2025)
Pith/arXiv arXiv 2025
-
[12]
F. Xing, J. Li, Z. Tao, F. Liu, Y. Tan, Modeling dynamic gas–liquid interfaces in underwater explosions using interval-constrained physics- informed neural networks, Physics of Fluids 37 (2025)
2025
-
[13]
Z. Tao, F. Liu, J. Li, G. Chen, Analytical and neural network approaches for solving two-dimensional nonlinear transient heat conduction, arXiv preprint arXiv:2504.02845 (2025)
Pith/arXiv arXiv 2025
-
[14]
Zhang, X
Z. Zhang, X. Yan, P. Liu, K. Zhang, R. Han, S. Wang, A physics-informed convolutional neural network for the simulation and prediction of two-phase darcy flows in heterogeneous porous media, Journal of Computational Physics 477 (2023) 111919
2023
-
[15]
76721–76774
Z.Hao,J.Yao,C.Su,H.Su,Z.Wang,F.Lu,Z.Xia,Y.Zhang,S.Liu,L.Lu,J.Zhu,Pinnacle:Acomprehensivebenchmarkofphysics-informed neural networks for solving pdes, in: Advances in Neural Information Processing Systems, volume 37, 2024, pp. 76721–76774
2024
-
[16]
Hochreiter, J
S. Hochreiter, J. Schmidhuber, Long short-term memory, Neural Computation 9 (1997) 1735–1780. ����� �� ������������ ��������� �� ������������ �� �� �� ������ ��������� ����� ����� �� ��������� Table 13 ��������� ������� ����� ��� ��� ��������� ������ Case Category Parameters Case 01 Density� � � ����,� � � ����,� � � ����,� � � ����,� � � ���� Case 01 Vel...
1997
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.