Pith. sign in

REVIEW 4 major objections 4 minor 14 references

SPP location with spherical ray tracing by refractive index

T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper tries to establish that single-point satellite positioning can be reformulated as a spherical ray-tracing time-of-flight inversion, replacing straight-line pseudorange corrections with analytic travel-time derivatives in a…

desk verdict Abstract promises an LSQR inversion the Application admits was never run; the forward-model results are modest, but the code/data and derivative formulas give a salvageable kernel. read the letter →

arxiv 2506.19859 v1 pith:T2TK4AJI submitted 2025-06-10 physics.ao-ph astro-ph.EPphysics.geo-ph

Hui Qian , Xiaosan Zhu , Dongliang Liu This is my paper
classification physics.ao-phastro-ph.EPphysics.geo-ph
keywords singlepointpositioningsphericalraytracingrefractiveindexLSQRinversionGNSStroposphericdelayionospherictraveltime
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Single-point satellite positioning normally treats radio signals as straight lines and patches the result with empirical ionospheric and tropospheric corrections. This paper tries to replace that two-step approach with a single refraction-based inversion: the signal path is traced as a curve through a spherically layered atmosphere, and the travel-time equations are differentiated analytically with respect to station coordinates, ray parameter, and layer velocities. Those derivatives feed a linearized LSQR inversion that estimates station coordinates, receiver clock bias, and the electromagnetic wave velocity at the bottom of the ionosphere and troposphere. On a GPS P1 observation from 20 July 2006, the ray-bending positioning with a tropospheric refractive-index model reduces the uncorrected position error from 8 m to 5 m, compared with 4.6 m for conventional corrections. This matters because atmospheric correction becomes part of the geometric model itself rather than a separate empirical patch, which could improve robustness in dynamic or extreme atmospheric conditions.

What carries the argument

The load-bearing object is the power-law layer model $v = a r^{1-b}$ with $a$ and $b$ fixed by the endpoint radii and velocities, together with the ray parameter $p$, which is constant across layers in a spherically symmetric medium. For a layer the paper derives the travel time $\Delta t = (-u_1+u_2)/b$ and horizontal arc $\Delta w$ with $u_i = \sqrt{-p^2+r_i^2/v_i^2}$, plus analytic first derivatives $\partial w/\partial p$, $\partial w/\partial r_1$, $\partial w/\partial r_2$, $\partial w/\partial v_1$, $\partial w/\partial v_2$ and the corresponding $\partial t$ derivatives, including chain-rule compensation terms for interpolated layers. These closed-form partial derivatives are what make a linearized LSQR inversion of station coordinates, clock biases, and refractive-index parameters possible, and the summation of layer arcs at constant $p$ is the geometric constraint that ties the ray tracing to the observed travel times.

What would settle it

Numerically integrate the travel time and horizontal arc for a known multi-layer power-law profile and compare the analytic partial derivatives with finite differences; any mismatch beyond floating-point error would falsify the derivation on which the LSQR inversion depends.

Watch

Extended reading notes

Core claim

The paper claims that GNSS single-point positioning can be reformulated as a time-of-flight problem in spherical coordinates, in which the straight-line pseudorange is replaced by the integral of the signal's travel time along a curved ray through a layered atmosphere. Within each layer the velocity follows $v = a r^{1-b}$, determined by the radii and velocities at the layer's two boundaries; this yields closed-form expressions for the travel time $\Delta t = (-u_1+u_2)/b$, the horizontal arc $\Delta w$, and the derivatives of both with respect to the ray parameter $p$, the radii $r_1, r_2$, and the velocities $v_1, v_2$. Because $p$ is constant along a ray, the total horizontal distance is the sum of layer arcs, and damped least-squares (LSQR) inversion solves for the unknowns. The authors demonstrate the claim on a GPS P1 dataset: with a tropospheric refractive-index profile the position error drops from 8 m to 5 m, and they argue that directly inverting the near-surface refractive indices would remove the remaining profile inaccuracies.

Load-bearing premise

The method assumes the atmosphere is horizontally stratified into spherical layers whose velocity follows the power-law form $v = a r^{1-b}$, with the ray parameter $p$ constant across layers, which real horizontal refractivity gradients (especially in the troposphere) violate.

Editorial extensions

If this is right

  • If the method is correct, single-frequency GNSS receivers could compute position, clock bias, and atmospheric velocity parameters in one damped-least-squares solve, without separate Klobuchar or empirical tropospheric corrections.
  • The analytic derivatives allow the inversion to be computed fast enough for real-time use, since no numerical ray shooting is needed.
  • The formulation folds ionospheric and tropospheric delays into the same geometric model, so their effects are estimated consistently rather than added as independent corrections.
  • With six or more simultaneous observations, the bottom-of-ionosphere and bottom-of-troposphere velocities become observable unknowns, turning positioning into a by-product of atmospheric profiling.
  • The demonstrated improvement (8 m to 5 m on the tested P1 data) indicates that even without full inversion, spherical ray tracing reduces straight-line positioning bias.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural test is to run the LSQR inversion on a multi-constellation dataset with surveyed ground coordinates and collocated radiosonde or water-vapor radiometer measurements; agreement between the inverted bottom-of-troposphere velocities and measured refractivity would validate the physical layer model.
  • The same layer-integral algebra could be extended to an ellipsoidal or horizontally graded refractivity model, treating the 1D spherical solution as the zeroth-order term of a gradient expansion, which would address the main real-world limitation of constant ray parameter.
  • Because the paper's own ionospheric layer test shows negligible contribution to position error, the practical value of full inversion likely concentrates in the troposphere; a focused inversion of tropospheric velocity fields may give the largest accuracy gain per added unknown.
  • The closed-form derivatives could be reused as the forward model in a tomographic sounder that assimilates multi-satellite travel times into a 3D refractivity grid, joining positioning with atmospheric sensing.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes a spherical-coordinate ray-tracing method for single-point positioning, in which the atmosphere is modeled as one-dimensional layers with power-law velocity profiles v = a r^(1-b), and the authors derive travel-time and arc-length derivative formulas for inverting station coordinates, clock biases, and near-surface refractive indices. The application section compares straight-line positioning with forward ray tracing using a prescribed tropospheric refractive-index profile and the IRI-2020 ionospheric model, reporting position errors of 8 m, 5 m, and 4.6 m for the uncorrected, troposphere-only, and combined cases. The central claimed contribution, stated in the abstract, is that a linearized LSQR inversion of station coordinates, receiver clock biases, and velocities was performed; however, the Application section explicitly states that such inversions were not pursued, and no inversion results are presented.

Significance. If the proposed inversion were actually implemented and validated, the idea of replacing empirical ionospheric and tropospheric corrections with a single refractive-index ray-tracing inversion would be a conceptually interesting contribution to single-point positioning. The forward-modeling apparatus and the derivative formulas, if correct and clearly derived, could provide a foundation for such a method, and the availability of code and data via figshare is a positive feature. However, as it stands, the manuscript does not demonstrate the claimed inversion, and the reported accuracy improvements are direct consequences of the assumed refractive-index profiles rather than independent tests of the method. The present work therefore does not substantiate its central claim.

major comments (4)
  1. [Abstract and Application] The abstract states that "a linearized LSQR method is employed to invert station coordinates, receiver clock biases, and electromagnetic wave velocities at the bottom of the ionosphere and troposphere," but the Application section explicitly says that "such inversions are feasible, they demand higher satellite numbers and data quality, which were not pursued in this study." No inversion results, residual statistics, or estimated parameters are reported anywhere. This is a direct internal contradiction that invalidates the abstract's flagship claim.
  2. [Application, Figs. 1-4] The reported improvements from 8 m to 5 m to 4.6 m are obtained by inserting a hand-assigned tropospheric refractive-index profile and the IRI-2020 ionospheric profile into the forward ray tracer. No comparison is made against standard empirical corrections (e.g., Klobuchar plus a mapping function), no sensitivity analysis is given for the assumed profile parameters, and no uncertainty quantification is provided. Therefore these numbers cannot support the conclusion that the proposed ray-tracing method is accurate or that it "effectively realizes combined tropospheric and ionospheric corrections."
  3. [Application, data description] The numerical experiment is described only as "GPS observation P1 data acquired on UTC July 20, 2006," with no statement of the number of satellites, the number of epochs, elevation cutoff, or satellite geometry. The abstract's reference to "over six sets of observations" has no counterpart in the data description, and the discussion of inversion feasibility in the Application section cannot be assessed without this information.
  4. [Methodology, derivative formulas] The central derivative derivations are presented in a form that cannot be checked as printed: several equations contain garbled symbols and undefined quantities (for example, the expressions for ∂v0/∂r1 and the "additional compensation terms" use Δv, Δr, and layer variables without consistent definitions), and the chain-rule terms are not fully specified. Since the proposed LSQR inversion depends entirely on these partial derivatives, the derivation must be rewritten in a clear, self-contained form before the method can be evaluated.
minor comments (4)
  1. [Introduction and Discussion] The text contains repeated duplicated passages and truncated sentences, such as the cutoff "can exceed 2.5 meters under certain atmospheric conditions (Boehm" in the Introduction, which interrupts the reference to Boehm et al. (2006).
  2. [Figures] The figure captions are inconsistent (e.g., Fig. 2 is captioned both "Error of GPS observation positioning with troposphere refraction index" and appears with several identical repeats), and the figures lack axis labels and units, making it impossible to interpret the plotted errors quantitatively.
  3. [References] Several references are incomplete or malformed, including "Kedar, S., & P. L. (2003)" and "Hoque, M. R., & N. B. (2018)"; these entries do not provide full author lists and should be corrected.
  4. [Methodology, notation] The layer exponent b is defined in the text as b = 1 - Δv/Δr but later appears in denominators and in conjunction with Δr and Δv without a clear restatement; the notation should be unified and all symbols defined at first use.

Circularity Check

2 steps flagged · score 6.0 of 10

Abstract's LSQR-inversion claim is contradicted by the Application's own admission that such inversions were not pursued; the reported 8 to 5 to 4.6 m gains are deterministic outputs of prescribed refractivity profiles, so the empirical predictions reduce to the chosen inputs.

  1. other [Abstract; Application final paragraph; Application data description.]
    "Subsequently, a linearized LSQR method is employed to invert station coordinates, receiver clock biases, and electromagnetic wave velocities at the bottom of the ionosphere and troposphere using over six sets of observations. ... While such inversions are feasible, they demand higher satellite numbers and data quality, which were not pursued in this study."

    The abstract's flagship claim, that the LSQR inversion defines the method and its results, is contradicted by the paper's own Application section, which states the inversions 'were not pursued.' No inverted station coordinates, clock biases, or bottom-layer velocities are reported; the numerical experiment only performs forward ray tracing with prescribed profiles. The 'over six sets of observations' also has no counterpart in the Application, which describes a single P1 dataset from July 20, 2006. The claimed derivation chain therefore terminates in an unexecuted step, so the central 'result' is asserted rather than derived, and the reviewer instruction to flag admitted limitations applies directly here.

  2. fitted input called prediction [Application, second paragraph and closing paragraph; Discussion of Fig. 2 through Fig. 4.]
    "When adopting a tropospheric model with a refractive index profile defined by 293 at 10 km below ground level, 269.1694 at ground level, and 0 at 12 km altitude, the positioning accuracy was significantly improved. The position error was reduced to 5 meters ... Simultaneous application of conventional tropospheric and ionospheric corrections yielded further reduced errors: a position error of 4.6 meters and pseudorange error of 8 meters."

    The reported accuracy gains are deterministic outputs of the prescribed refractivity inputs inserted into the ray tracer, not independent predictions. The profile values (293, 269.1694, 0, and the IRI-2020 electron densities) are assumed, not measured or inverted, yet the paper presents the resulting 8 m to 5 m to 4.6 m improvements as confirmation that 'the proposed refractive index profiling algorithm effectively realizes combined tropospheric and ionospheric corrections.' The validation reduces to evaluating the chosen inputs: the same conventional empirical models that the abstract claims to replace are the very source of the reported improvement, so the empirical 'prediction' is forced by construction from the inputs.

full rationale

The partial-derivative chain in the Methodology, including the travel-time and arc formulas Δt = −(u1 + u2)/b, Δw, ∂w/∂p, and the ∂/∂r, ∂/∂v, ∂/∂z derivatives, is self-contained: the formulas are mathematical consequences of the assumed power-law layer model v = a r^(1−b) and the constancy of the ray parameter p, and they are checkable against the released code and test data, so no circularity attaches to that derivation itself. The paper contains no self-citations and invokes no uniqueness theorem, so the self-citation patterns do not arise. However, two load-bearing steps fail the input/result separation. First, the abstract claims a linearized LSQR inversion of station coordinates, receiver clock biases, and bottom-layer velocities using over six sets of observations, but the Application states such inversions 'were not pursued' and shows no inversion outputs; the central claimed contribution is therefore asserted rather than executed, and the 'over six sets' has no counterpart in the single P1 dataset described. Second, the empirical validation is a forward computation in which the chosen tropospheric profile and IRI-2020 are inserted into the ray tracer; the 'improvements' are deterministic consequences of those inputs, so the claim that the algorithm effectively realizes combined tropospheric and ionospheric corrections reduces to testing the very conventional models it claims to replace. The accuracy numbers are thus input-forced rather than independent predictions, giving partial circularity and an unsupported central claim.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on a small number of imposed modeling choices: a spherical 1D atmosphere, a power-law intra-layer velocity profile, a hand-assigned tropospheric profile, and the fitted ray parameter. There are no invented physical entities, but the tropospheric profile is a free parameter and the inversion result is not demonstrated.

free parameters (3)
  • Tropospheric refractive index profile anchor values = 293 at 10 km below ground, 269.1694 at ground, 0 at 12 km altitude
    These three numbers are hand-assigned in the Application section and define the tropospheric model whose ray-tracing correction produces the reported accuracy improvement.
  • Ray parameter p = not reported
    p is estimated per ray by damped least squares from the horizontal-distance constraint (Methodology); the reported positioning depends on these fitted values.
  • Damped least squares damping parameter = not reported
    The inversion in Methodology requires a damping choice, which is never stated; the solution depends on it.
assumptions (4)
  • domain assumption Refractive index from satellite to top of atmosphere is 1.
    Stated in the first line of Methodology; it ignores any refractive effect above the modeled layers, so the top-of-atmosphere travel time is set to geometric free-space time.
  • ad hoc to paper Velocity within each layer follows v = a r^(1-b) with a and b fixed by the layer's endpoint velocity and radius values.
    This power-law form is imposed in Methodology; it is not derived from atmospheric physics or validated against measurements.
  • domain assumption The atmosphere is spherically symmetric and horizontally stratified; horizontal refractivity gradients are neglected.
    The entire layer model and the constancy of ray parameter p rely on 1D spherical symmetry, which real tropospheric and ionospheric structures violate.
  • standard math Ray parameter p is constant along a ray across layers.
    This is Snell's law for spherical layering and is standard, but it holds only under the spherical-symmetry assumption above.

how reviews work

0 comments
Cite this review

Pith. "Pith review of SPP location with spherical ray tracing by refractive index." pith.science (2026). https://pith.science/paper/T2TK4AJI

@misc{pith2026250619859,
  author       = {Pith},
  title        = {Pith review of: SPP location with spherical ray tracing by refractive index},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/T2TK4AJI}},
  note         = {Machine review of arXiv:2506.19859}
}
read the original abstract

Atmospheric layer structure is a primary factor affecting the precision of single-point satellite positioning. The assumption of electromagnetic wave rectilinear propagation hinders the accurate implementation of ionospheric and tropospheric corrections, whereas curvilinear positioning methods fully account for ray deflection. This study aims to derive partial derivative formulas for theoretical travel time with respect to latitude, longitude, elevation, and velocity models by formulating electromagnetic wave travel time equations under a coordinate-based one-dimensional layered velocity model. Subsequently, a linearized LSQR method is employed to invert station coordinates, receiver clock biases, and electromagnetic wave velocities at the bottom of the ionosphere and troposphere using over six sets of observations. This replaces conventional ionospheric/tropospheric pseudorange corrections in single-point positioning, establishing a novel spherical coordinate refraction travel time positioning method. The classical straight-line pseudorange positioning is reformulated into a time-of-flight positioning approach, and the positioning accuracy differences between straight-line and spherical coordinate refraction travel time methods are compared. By integrating classical ionospheric and tropospheric models to construct corresponding refractive index models and combining them with curvilinear ray tracing methods, the inherent theoretical limitations of positioning can be effectively mitigated.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

  1. [1]

    Boehm, J., Niell, A., Tregoning, P., & Schuh, H. (2006). Global Mapping Function (GMF): A new empirical mapping function based on numerical weather model data. Geophysical Research Letters, 33(7)

    work page 2006

  2. [2]

    Ishimaru, A. (1997). Wave propagation and scattering in random media. Academic Press

    work page 1997

  3. [3]

    Klobuchar, J. A. (1987). Ionospheric time - delay algorithm for single - frequency GPS users. IEEE Transactions on Aerospace and Electronic Systems, AES - 23(3), 325 - 331

    work page 1987

  4. [4]

    Misra, P., & Enge, P. (2006). Global Positioning System: Signals, Measurements, and Performance. Ganga - Jamuna Press

    work page 2006

  5. [5]

    Nava, B., Coisson, P., & Radicella, S. M. (2008). A new version of the NeQuick ionosphere electron density model. Journal of Atmospheric and Solar - Terrestrial Physics, 70(15), 1856 - 1862

    work page 2008

  6. [6]

    Ibrahim, G., & El-Rabbany, A. (2011). Performance Analysis of NOAA Tropospheric Signal Delay Model. Measurement Science and Technology, 22(11), 115107. doi:10.1088/0957-0233/22/11/115107

    work page doi:10.1088/0957-0233/22/11/115107 2011

  7. [7]

    Kedar, S., & P. L. (2003). The Effect of the Second Order GPS Ionospheric Correction on Receiver Positions. Geophysical Research Letters, 30(15). doi:10.1029/2003gl017639

    work page doi:10.1029/2003gl017639 2003

  8. [8]

    Wang, K., Liang, S., & Zhu, L. (2008). Impact of Surface Meteorological Measurements on GPS Height Determination. Geophysical Research Letters, 35(23). doi:10.1029/2008gl035929

    work page doi:10.1029/2008gl035929 2008

Show all 14 references
  1. [9]

    Hoque, M. R., & N. B. (2018). Positioning Performance of the NTCM Model Driven by GPS Klobuchar Model Parameters. Journal of Space Weather and Space Climate, 8, A20. doi:10.1051/swsc/2018009

    2018

  2. [10]

    Singh, S., & Reilly, T. (2006). Improved Positioning by Addition of Atmospheric Corrections to Local Area Differential GPS. Radio Science, 41(5). doi:10.1029/2005rs003339

    2006 doi

  3. [11]

    Wadge, G., & C. T. (2002). Atmospheric Models, GPS and InSAR Measurements of the Tropospheric Water Vapour Field over Mount Etna. Geophysical Research Letters, 29(19). doi:10.1029/2002gl015159

    2002 doi

  4. [12]

    Zhou, K., & Wu, S. (2014). Real-Time Positioning Accuracy Analysis of Single-Frequency GPS Based on Ionosphere Grid Products. Applied Mechanics and Materials, 644–650, 4591. doi:10.4028/www.scientific.net/amm.644-650.4591

    2014 doi

  5. [13]

    Zhang, Y., & C. W. (2021). On the Assessment GPS-Based WRFDA for InSAR Atmospheric Correction: A Case Study in Pearl River Delta Region of China. Remote Sensing, 13(16), 3280. doi:10.3390/rs13163280

    2021 doi

  6. [14]

    Cross P., Fielding,E.J

    Li, Z., Muller,J. Cross P., Fielding,E.J.. (2005). Interferometric Synthetic Aperture Radar (InSAR) Atmospheric Correction: GPS, Moderate Resolution Imaging Spectroradiometer (MODIS), and InSAR Integration. Journal of Geophysical Research: Solid Earth, 110(B3). doi:10.1029/200...

    2005 doi

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.