pith. sign in

arxiv: 2604.23925 · v1 · submitted 2026-04-27 · 📡 eess.SP

Medium-Induced Cross-Frequency Clutter Structure in Single-Snapshot FDA-MIMO-GPR With a Weak-Dispersion Criterion

Yisu Yan , Jifeng Guo This is my paper

Pith reviewed 2026-05-08 02:31 UTC · model grok-4.3

classification 📡 eess.SP
keywords FDA-MIMO-GPRdispersive cluttercross-frequency couplingweak dispersionsingle-snapshot radarground-penetrating radarCole-Cole modelbackground covariance
0
0 comments X

The pith

Random dispersive media induce a quantifiable cross-frequency coupling in the background clutter covariance of single-snapshot FDA-MIMO ground-penetrating radar.

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

The paper develops an analytical description of how random dispersive media generate cross-frequency structure inside the clutter observed by a frequency-diverse-array MIMO ground-penetrating radar that records only one snapshot. Constitutive behavior is captured by the Cole-Cole model; a normalized incremental contrast function is defined relative to a reference medium, and a first-order propagation-kernel feedback expression is derived for the single-snapshot background response. From this expression a scalar cross-frequency coupling strength is obtained for the leading-order background covariance. Numerical checks show that, inside the weak-dispersion regime, this strength is consistent across different constitutive mappings, zeroth-order propagation skeletons, distorted-Born truncations, and single-channel closures, while also tracking the error of the frequency block-diagonal approximation used in subsequent processing.

Core claim

Under the weak-dispersion criterion the cross-frequency coupling strength of the leading-order background covariance, obtained from the first-order propagation-kernel feedback expression, distinguishes explicitly coupled from uncoupled constructions, remains invariant under pure energy scaling, responds to correlation length and relaxation-location parameters, and equals the approximation error of the frequency block-diagonal model.

What carries the argument

Cross-frequency coupling strength of the leading-order background covariance, derived from the normalized incremental contrast function and the single-snapshot background-response expression with first-order propagation-kernel feedback.

If this is right

  • The identified cross-frequency structure directly degrades the performance of whitening filters and principal-subspace extraction.
  • The coupling strength remains stable under pure energy scaling of the scene.
  • The strength varies systematically with the correlation length and relaxation-location parameters of the dispersive medium.
  • The coupling strength equals the residual error of the frequency block-diagonal covariance approximation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Processing pipelines for FDA-MIMO-GPR should retain off-diagonal frequency blocks in the covariance whenever the weak-dispersion criterion is only marginally satisfied.
  • The same metric could guide the design of frequency-dependent whitening operators that compensate for the predicted coupling.
  • Because the numerical results are insensitive to the precise constitutive mapping inside the weak regime, the coupling strength may apply to dispersion models other than Cole-Cole.

Load-bearing premise

The derivations remain valid only when the medium is weakly dispersive and first-order propagation-kernel feedback suffices.

What would settle it

In simulation or measured data, increase the dispersion strength or perturbation amplitude until the observed whitening or subspace-extraction error deviates sharply from the value predicted by the coupling-strength metric.

read the original abstract

This paper investigates the cross-frequency structure of background clutter induced by random dispersive media in single-snapshot FDA-MIMO-GPR. Representative media are modeled by the Cole--Cole formulation to relate dispersive constitutive behavior to the reference propagation environment and observation-domain statistics. A normalized incremental contrast function is introduced under a reference-medium framework, and a single-snapshot background-response expression with first-order propagation-kernel feedback is derived. Based on this expression, a cross-frequency coupling strength of the leading-order background covariance is defined. Numerical results show that, in weakly dispersive scenes, the proposed analysis remains consistent across constitutive mapping, the zeroth-order propagation skeleton, first-order distorted-Born truncation, propagation-kernel feedback, and single-channel response closure. The proposed metric distinguishes uncoupled and explicitly coupled constructions, remains stable under pure energy scaling, responds clearly to correlation length and relaxation-location parameters, and corresponds directly to the error of the frequency block-diagonal approximation. Additional experiments show that the resulting cross-frequency structure affects whitening and principal-subspace extraction. In scenes with pronounced relaxation, abrupt breakdown under strong perturbations and high-error plateaus indicate that the present theory is mainly applicable within the validity range of first-order feedback.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 3 minor

Summary. The paper claims that in weakly dispersive media, the cross-frequency structure of background clutter in single-snapshot FDA-MIMO-GPR can be characterized by modeling dispersive constitutive behavior with the Cole-Cole formulation, introducing a normalized incremental contrast function under a reference-medium framework, deriving a single-snapshot background-response expression with first-order propagation-kernel feedback and distorted-Born truncation, and defining a cross-frequency coupling strength of the leading-order background covariance. Numerical results are presented to show internal consistency of this analysis across constitutive mapping, zeroth-order propagation skeleton, first-order truncation, kernel feedback, and single-channel closure; the metric distinguishes coupled/uncoupled cases, is stable to energy scaling, responds to correlation length and relaxation parameters, corresponds to block-diagonal approximation error, and affects whitening and principal-subspace extraction, with abrupt breakdown outside the weak-dispersion validity range.

Significance. If the result holds, the work supplies a concrete metric and framework for quantifying medium-induced cross-frequency clutter in dispersive GPR, with direct implications for whitening filters and subspace-based processing that could improve single-snapshot detection performance. Credit is due for the systematic numerical consistency checks across modeling choices and the explicit linkage between the coupling strength and the frequency block-diagonal approximation error; these elements provide a falsifiable, parameter-sensitive diagnostic within the stated regime.

major comments (3)
  1. [Numerical results section] Numerical results section: the claimed consistency 'across ... first-order distorted-Born truncation, propagation-kernel feedback' is shown only inside the first-order model whose covariance expression already incorporates the Cole-Cole parameters and the truncation. No comparison to second-order Born terms or full-wave reference solutions is described, so the experiments do not independently confirm that omitted higher-order scattering leaves the defined cross-frequency coupling strength materially unchanged (as required to support the central claim that the analysis captures the clutter structure).
  2. [Background-response derivation] Background-response derivation and coupling-strength definition: because the coupling strength is constructed directly from the leading-order covariance that embeds the first-order feedback and Cole-Cole mapping, the numerical stability checks are internal by design. An explicit statement of the validity range (e.g., a bound on the incremental contrast or relaxation strength) and a quantitative test against a higher-order reference would remove the circularity concern and make the 'remains consistent' claim load-bearing rather than tautological.
  3. [Numerical results section] Numerical results: the abstract states that the metric 'responds clearly' to correlation length and relaxation-location parameters and 'corresponds directly' to block-diagonal error, yet supplies no error bars, number of Monte-Carlo realizations, or convergence diagnostics. This weakens the support for the consistency and sensitivity assertions.
minor comments (3)
  1. The normalized incremental contrast function is introduced without an early equation reference or explicit definition; adding the defining expression in the introduction or §2 would improve readability.
  2. Additional citations to existing Cole-Cole GPR literature and FDA-MIMO clutter models would better situate the contribution.
  3. Figure or table captions should state the specific weak-dispersion criterion thresholds (e.g., maximum relaxation strength or contrast magnitude) used in each experiment.

Simulated Author's Rebuttal

3 responses · 1 unresolved

We thank the referee for the careful and constructive review. The manuscript focuses on deriving and characterizing a cross-frequency coupling strength within a first-order propagation-kernel feedback model for weakly dispersive media; we address each major comment below by clarifying the scope of the claims, agreeing where revisions are warranted, and noting limitations honestly.

read point-by-point responses

  1. Referee: [Numerical results section] Numerical results section: the claimed consistency 'across ... first-order distorted-Born truncation, propagation-kernel feedback' is shown only inside the first-order model whose covariance expression already incorporates the Cole-Cole parameters and the truncation. No comparison to second-order Born terms or full-wave reference solutions is described, so the experiments do not independently confirm that omitted higher-order scattering leaves the defined cross-frequency coupling strength materially unchanged (as required to support the central claim that the analysis captures the clutter structure).

    Authors: The central contribution is the definition of the coupling strength from the leading-order covariance under the stated first-order model and weak-dispersion assumptions. The numerical results verify internal consistency of the derivations (constitutive mapping through single-channel closure) and demonstrate the metric's distinguishing power, stability, and parameter sensitivity inside this regime. The manuscript already reports abrupt breakdown outside the validity range. We agree that independent higher-order comparisons would strengthen the work but lie beyond the present scope; we will revise the discussion section to state the first-order regime more explicitly and list higher-order validation as future work. revision: partial

  2. Referee: [Background-response derivation] Background-response derivation and coupling-strength definition: because the coupling strength is constructed directly from the leading-order covariance that embeds the first-order feedback and Cole-Cole mapping, the numerical stability checks are internal by design. An explicit statement of the validity range (e.g., a bound on the incremental contrast or relaxation strength) and a quantitative test against a higher-order reference would remove the circularity concern and make the 'remains consistent' claim load-bearing rather than tautological.

    Authors: We will insert an explicit validity-range statement in the revised manuscript, giving approximate bounds on the normalized incremental contrast and relaxation strength consistent with the first-order feedback truncation. These bounds will be tied to the observed high-error plateaus in the numerical experiments. While the internal checks confirm correct implementation of the analytic expressions, we accept that they cannot independently prove robustness to omitted higher-order terms; the claims will be reframed to emphasize applicability within the leading-order weak-dispersion model. revision: yes

  3. Referee: [Numerical results section] Numerical results: the abstract states that the metric 'responds clearly' to correlation length and relaxation-location parameters and 'corresponds directly' to block-diagonal error, yet supplies no error bars, number of Monte-Carlo realizations, or convergence diagnostics. This weakens the support for the consistency and sensitivity assertions.

    Authors: We agree that statistical reporting should be strengthened. The revised manuscript will state the number of Monte-Carlo realizations used for each covariance estimate (500–1000) and will add error bars (standard deviation across realizations) to the plots that illustrate the metric's response to correlation length, relaxation parameters, and block-diagonal approximation error. This will provide quantitative backing for the sensitivity and correspondence claims. revision: yes

standing simulated objections not resolved
  • Direct quantitative comparisons against second-order Born terms or full-wave reference solutions to test whether the coupling strength remains materially unchanged under higher-order scattering.

Circularity Check

1 steps flagged

Coupling strength defined from first-order covariance; consistency with truncation is internal

specific steps
  1. self definitional [Abstract]
    "Based on this expression, a cross-frequency coupling strength of the leading-order background covariance is defined. Numerical results show that, in weakly dispersive scenes, the proposed analysis remains consistent across constitutive mapping, the zeroth-order propagation skeleton, first-order distorted-Born truncation, propagation-kernel feedback, and single-channel response closure."

    The coupling strength is extracted from the covariance that was derived under the same first-order truncated feedback; therefore the numerical demonstration of 'consistency' with first-order distorted-Born truncation and propagation-kernel feedback reduces to an internal property of the defining expression rather than an external validation.

full rationale

The derivation introduces a normalized contrast, obtains a single-snapshot response under explicit first-order propagation-kernel feedback and distorted-Born truncation, then defines the cross-frequency coupling strength directly from the resulting leading-order covariance. Numerical experiments then assert that the analysis 'remains consistent' with precisely those first-order ingredients. Because the metric is constructed from the truncated expression, the reported consistency is tautological inside the model and does not constitute an independent check against higher-order terms or external benchmarks. This produces moderate circularity confined to the validation step; the underlying constitutive mapping and zeroth-order skeleton remain non-circular.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 1 invented entities

Paper rests on the Cole-Cole dispersive model, the weak-dispersion regime, and the validity of first-order Born-type feedback; no independent evidence for these modeling choices is supplied beyond numerical consistency.

free parameters (1)
  • Cole-Cole relaxation parameters
    Used to generate the dispersive constitutive behavior; values chosen for the numerical experiments are not stated in the abstract.
axioms (2)
  • domain assumption First-order propagation-kernel feedback is sufficient
    Invoked to close the single-snapshot background-response expression.
  • domain assumption Weak-dispersion criterion
    Required for the consistency claims and for the breakdown warning under strong relaxation.
invented entities (1)
  • normalized incremental contrast function no independent evidence
    purpose: Relates dispersive constitutive behavior to reference propagation environment and observation-domain statistics
    New function introduced under the reference-medium framework; no external validation supplied.

pith-pipeline@v0.9.0 · 5513 in / 1405 out tokens · 31353 ms · 2026-05-08T02:31:36.471441+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

32 extracted references · 12 canonical work pages

  1. [1]

    Influence of soil inhomogeneity on GPR for landmine detection,

    K. Takahashi, J. Igel, and H. Preetz, “Influence of soil inhomogeneity on GPR for landmine detection,” inProceedings of the XIII Internarional Conference on Ground Penetrating Radar, pp. 1–6. [Online]. Available: https://ieeexplore.ieee.org/document/5550150

    work page arXiv

  2. [2]

    Modeling of GPR clutter caused by soil heterogeneity,

    ——, “Modeling of GPR clutter caused by soil heterogeneity,” vol. 2012, p. 643430. [Online]. Available: https://oa.tib.eu/renate/handle/ 123456789/8138

    2012

  3. [3]

    Shallow regolith structure and obstructions detected by lunar regolith penetrating radar at chang’E-5 drilling site,

    J. Feng, M. A. Siegler, and M. N. White, “Shallow regolith structure and obstructions detected by lunar regolith penetrating radar at chang’E-5 drilling site,” vol. 14, no. 14, p. 3378. [Online]. Available: https://www.mdpi.com/2072-4292/14/14/3378

    2072

  4. [4]

    Subsoil structure at the chang’E-6 landing site revealed by in-situ lunar regolith penetrating radar,

    Z. Zhang, C. Ding, Y . Su, S. Shen, W. Lu, F. Soldovieri, Z. Xiao, X. Zeng, W. Du, Y . Liu, C. Jiang, G. Gennarelli, I. Catapano, S. Dai, J. Feng, Y . Wang, J. Liu, and C. Li, “Subsoil structure at the chang’E-6 landing site revealed by in-situ lunar regolith penetrating radar,” vol. 6, no. 1, p. 640. [Online]. Available: https://www.nature.com/articles/s...

  5. [5]

    GPR clutter amplitude processing to detect shallow geological targets,

    V . Salinas Naval, S. Santos-Assunc ¸ao, and V . P´erez-Gracia, “GPR clutter amplitude processing to detect shallow geological targets,” vol. 10, no. 1, p. 88. [Online]. Available: https://www.mdpi.com/2072-4292/10/1/88

    2072

  6. [6]

    Y . Liu, L. Xiao, X. Wang, and A. Nehorai. On clutter ranks of frequency diverse radar waveforms. [Online]. Available: http: //arxiv.org/abs/1603.08189

    work page arXiv

  7. [7]

    Clutter rank analysis in airborne FDA-MIMO radar with range ambiguity,

    K. Wang, G. Liao, J. Xu, Y . Zhang, and L. Huang, “Clutter rank analysis in airborne FDA-MIMO radar with range ambiguity,” vol. 58, no. 2, pp. 1416–1430

  8. [8]

    Clutter suppression for airborne FDA-MIMO radar using multi-waveform adaptive processing and auxiliary channel STAP,

    C. Wen, M. Tao, J. Peng, J. Wu, and T. Wang, “Clutter suppression for airborne FDA-MIMO radar using multi-waveform adaptive processing and auxiliary channel STAP,” vol. 154, pp. 280–293. [Online]. Available: https://www.sciencedirect.com/science/article/pii/S0165168418303001

  9. [9]

    Range- Ambiguous Clutter Suppression via FDA MIMO Planar Array Radar with Compressed Sensing,

    Y . Wang, S. Zhu, L. Lan, X. Li, Z. Liu, and Z. Wu, “Range- Ambiguous Clutter Suppression via FDA MIMO Planar Array Radar with Compressed Sensing,” vol. 14, no. 8. [Online]. Available: https://www.mdpi.com/2072-4292/14/8/1926 13

    2072

  10. [10]

    Space–time-range clutter suppression via tensor-based STAP for airborne FDA-MIMO radar,

    Y . Sun, W.-Q. Wang, and C. Jiang, “Space–time-range clutter suppression via tensor-based STAP for airborne FDA-MIMO radar,” vol. 214, p. 109235. [Online]. Available: https://www.sciencedirect. com/science/article/pii/S0165168423003092

  11. [11]

    FDA-MIMO radar parameter designing against range-ambiguous clutter and scatter-wave jamming,

    M. Jia, Y . Sun, and W.-Q. Wang, “FDA-MIMO radar parameter designing against range-ambiguous clutter and scatter-wave jamming,” vol. 240, p. 110361. [Online]. Available: https://www.sciencedirect. com/science/article/pii/S0165168425004773

  12. [12]

    Adaptive target detection for FDA-MIMO radar with training data in gaussian noise,

    P. Li, B. Huang, and W.-Q. Wang, “Adaptive target detection for FDA-MIMO radar with training data in gaussian noise,” pp. 1–16. [Online]. Available: https://ieeexplore.ieee.org/document/11373410

    work page arXiv

  13. [13]

    Detection of Subsurface Target Based on FDA-MIMO Radar

    Q. Liu, C. Jiang, L. Jin, and S. Ouyang, “Detection of subsurface target based on FDA-MIMO radar,” vol. 2018, no. 1, p. 8629806. [Online]. Available: https://onlinelibrary.wiley.com/doi/abs/ 10.1155/2018/8629806

    work page doi:10.1155/2018/8629806 2018

  14. [14]

    Multiview multistatic vs. Multimonostatic three-dimensional ground-penetrating radar imaging: A comparison,

    M. Masoodi, G. Gennarelli, F. Soldovieri, and I. Catapano, “Multiview multistatic vs. Multimonostatic three-dimensional ground-penetrating radar imaging: A comparison,” vol. 16, no. 17. [Online]. Available: https://www.mdpi.com/2072-4292/16/17/3163

    2072

  15. [15]

    The effect of the equivalent permittivity model in contactless MIMO-GPR imaging,

    G. Gennarelli, I. Catapano, and F. Soldovieri, “The effect of the equivalent permittivity model in contactless MIMO-GPR imaging,” vol. 26, no. 5. [Online]. Available: https://www.mdpi.com/1424-8220/ 26/5/1463

  16. [16]

    Enhancing subsurface exploration: A comprehensive review of advanced clutter removal techniques for ground penetrating radar imaging,

    B. S. Kumar, S. Baraha, A. K. Sahoo, and S. Maiti, “Enhancing subsurface exploration: A comprehensive review of advanced clutter removal techniques for ground penetrating radar imaging,” vol. 239, p. 115432. [Online]. Available: https://www.sciencedirect.com/science/ article/pii/S0263224124013174

  17. [17]

    GPR clutter reflection noise-filtering through singular value decomposition in the bidimensional spectral domain,

    R. J. Oliveira, B. Caldeira, T. Teixid ´o, and J. F. Borges, “GPR clutter reflection noise-filtering through singular value decomposition in the bidimensional spectral domain,” vol. 13, no. 2005. [Online]. Available: https://www.mdpi.com/2072-4292/13/10/2005

    2005

  18. [18]

    Clutter distributions for tomographic image standardization in ground-penetrating radar,

    B. M. Worthmann, D. H. Chambers, D. S. Perlmutter, J. E. Mast, D. W. Paglieroni, C. T. Pechard, G. A. Stevenson, and S. W. Bond, “Clutter distributions for tomographic image standardization in ground-penetrating radar,” vol. 59, no. 9, pp. 7957–7967. [Online]. Available: https://ieeexplore.ieee.org/abstract/document/9336275

    work page arXiv

  19. [19]

    Cole, R.H

    K. S. Cole and R. H. Cole, “Dispersion and absorption in dielectrics I. Alternating current characteristics,” vol. 9, no. 4, pp. 341–351. [Online]. Available: https://doi.org/10.1063/1.1750906

    work page doi:10.1063/1.1750906

  20. [20]

    Dielectric Properties of Ice and Solid D2O

    R. P. Auty and R. H. Cole, “Dielectric properties of ice and solid D2O,” vol. 20, no. 8, pp. 1309–1314. [Online]. Available: https://doi.org/10.1063/1.1700726

    work page doi:10.1063/1.1700726

  21. [21]

    Time domain characterization of the cole-cole dielectric model,

    S. Holm, “Time domain characterization of the cole-cole dielectric model,” vol. 11, no. 1, pp. 101–105. [Online]. Available: https: //pmc.ncbi.nlm.nih.gov/articles/PMC7851980/

  22. [22]

    Full-waveform inversion of ground-penetrating radar data in frequency-dependent media involving permittivity attenuation

    T. Qin, T. Bohlen, and N. Allroggen, “Full-waveform inversion of ground-penetrating radar data in frequency-dependent media involving permittivity attenuation,” vol. 232, no. 1, pp. 504–522. [Online]. Available: https://doi.org/10.1093/gji/ggac319

    work page doi:10.1093/gji/ggac319

  23. [23]

    GPR uncertainty modelling and analysis of object depth based on constrained least squares,

    F. Xie, W. W. L. Lai, and X. D ´erobert, “GPR uncertainty modelling and analysis of object depth based on constrained least squares,” vol. 183, p. 109799. [Online]. Available: https: //www.sciencedirect.com/science/article/pii/S026322412100751X

  24. [24]

    Electrical properties of lunar soil dependence on frequency, temperature and moisture,

    D. Strangway, W. Chapman, G. Olhoeft, and J. Carnes, “Electrical properties of lunar soil dependence on frequency, temperature and moisture,” vol. 16, no. 2, pp. 275–281. [Online]. Available: https://www.sciencedirect.com/science/article/pii/0012821X72902038

    work page arXiv

  25. [25]

    Electrical properties of lunar solid samples,

    G. Olhoeft, A. Frisillo, D. Strangway, and H. Sharpe, “Electrical properties of lunar solid samples,” vol. 4, p. 575

  26. [26]

    Electrical properties of lunar soil sample 15301,38

    G. R. Olhoeft, A. L. Frisillo, and D. W. Strangway, “Electrical properties of lunar soil sample 15301,38,” vol. 79, no. 11, pp. 1599–1604. [Online]. Available: https://onlinelibrary.wiley.com/doi/abs/ 10.1029/JB079i011p01599

    work page doi:10.1029/jb079i011p01599

  27. [27]

    Characterization of the permittivity of controlled porous water ice-dust mixtures to support the radar exploration of icy bodies

    Y . Brouet, L. Neves, P. Sabouroux, A. C. Levasseur-Regourd, O. Poch, P. Encrenaz, A. Pommerol, N. Thomas, and W. Kofman, “Characterization of the permittivity of controlled porous water ice-dust mixtures to support the radar exploration of icy bodies,” vol. 121, no. 12, pp. 2426–2443. [Online]. Available: https: //onlinelibrary.wiley.com/doi/abs/10.1002/...

    work page doi:10.1002/2016je005045

  28. [28]

    Dielectric and hardness measurements of planetary analog rocks in support of in-situ subsurface sampling,

    A. ElShafie and E. Heggy, “Dielectric and hardness measurements of planetary analog rocks in support of in-situ subsurface sampling,” vol. 86, pp. 150–154. [Online]. Available: https://www.sciencedirect. com/science/article/pii/S0032063313000329

  29. [29]

    Mod- elling arbitrary complex dielectric properties – an automated implemen- tation for gprMax,

    S. Majchrowska, I. Giannakis, C. Warren, and A. Giannopoulos, “Mod- elling arbitrary complex dielectric properties – an automated implemen- tation for gprMax,” in2021 11th International Workshop on Advanced Ground Penetrating Radar (IWAGPR), pp. 1–5

  30. [30]

    Some magnetic and electrical properties of basalt rocks,

    M. S. Ahmad and A. M. Zihlif, “Some magnetic and electrical properties of basalt rocks,” vol. 10, no. 4, pp. 207–214. [Online]. Available: https://www.sciencedirect.com/science/article/pii/0167577X90900909

    work page arXiv

  31. [31]

    Physical properties of the lunar surface,

    W. D. Carrier, Iii, G. R. Olhoeft, and W. Mendell, “Physical properties of the lunar surface,” inLunar Sourcebook, a User’s Guide to the Moon, pp. 475–594. [Online]. Available: https: //ui.adsabs.harvard.edu/abs/1991lsug.book..475C

  32. [32]

    Dielectric properties of ice and snow–a review,

    S. Evans, “Dielectric properties of ice and snow–a review,” vol. 5, no. 42, pp. 773–792. [Online]. Available: https://www.cambridge.org/core/journals/journal-of-glaciology/ article/dielectric-properties-of-ice-and-snowa-review/ A4F950358E513838DAE111F6A07EC077