arxiv: 2604.26599 · v1 · submitted 2026-04-29 · ❄️ cond-mat.dis-nn

Recognition: unknown

Effective length scales, dispersion relations, and discrete densities of states for Laplacian eigenvectors on complex networks

Per Arne Rikvold

Authors on Pith no claims yet

Pith reviewed 2026-05-07 10:43 UTC · model grok-4.3

classification ❄️ cond-mat.dis-nn

keywords Laplacian eigenvectorseffective length scalesdispersion relationscomplex networkssign-changing edgesdiscrete densities of statesgraph diffusion

0 comments

The pith

A ratio of total edges to sign-changing edges defines effective length scales for Laplacian eigenvectors on networks.

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

The paper adapts a condensed-matter technique for estimating correlation lengths to graph Laplacians by treating each eigenvector as a pattern of positive and negative values on vertices. The effective length is computed as twice the total number of edges divided by the number of edges that link vertices with opposite signs in that eigenvector. This length, paired with the eigenvalue that sets the inverse time scale, produces dispersion relations and discrete densities of states. The method is demonstrated on nine networks ranging from trees with shortcuts to the roundworm nervous system and an electrical power grid, separating modes that are spread across the network from those that are localized.

Core claim

Effective length scales for the eigenvectors of a graph Laplacian can be estimated as the ratio of twice the total number of edges in the network to the number of edges that connect vertices carrying eigenvector values of opposite sign. These length scales, when combined with the eigenvalues, permit the construction of dispersion relations that characterize diffusion or oscillation processes and allow the identification of both distributed and localized eigenvectors through their associated discrete densities of states.

What carries the argument

The effective length scale obtained as twice the total edge count divided by the count of sign-changing edges for a given Laplacian eigenvector, which converts the inverse-time eigenvalue into a wave-number-like quantity for dispersion relations.

If this is right

Dispersion relations become constructible for any network once its Laplacian eigenvectors are known, without requiring a continuous embedding.
Eigenvectors can be classified as extended or localized by comparing their computed length scales to the overall network diameter.
Discrete densities of states can be compiled for diffusion or vibration processes on empirical networks such as food webs or power grids.
The approach extends the volume-to-interface concept from disordered materials directly to discrete graph structures.

Where Pith is reading between the lines

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

The same sign-change counting idea could be tested on other graph operators such as adjacency or normalized Laplacians to see whether analogous dispersion relations emerge.
Comparing these lengths to participation ratios or inverse participation ratios on the same eigenvectors might reveal whether the two localization diagnostics agree on real-world networks.
If the length scales prove robust, they could serve as a quick diagnostic for how far perturbations or signals propagate before decaying in empirical graphs.

Load-bearing premise

The ratio of twice the total edges to the number of sign-changing edges directly supplies a physically usable length scale for building dispersion relations without separate calibration against known analytic cases.

What would settle it

Apply the method to a regular lattice graph or cycle graph whose exact dispersion relation is known from Fourier analysis; mismatch between the derived length scales and the analytic wave numbers would disprove the claim.

Figures

Figures reproduced from arXiv: 2604.26599 by Per Arne Rikvold.

**Figure 1.** Figure 1: FIG. 1. Unweighted, undirected line graph with view at source ↗

**Figure 2.** Figure 2: FIG. 2. A small-world network generated from the view at source ↗

**Figure 3.** Figure 3: FIG. 3. Cayley tree graph with a branching ratio of 3 and 4 layers, view at source ↗

**Figure 4.** Figure 4: FIG. 4. A small-world network [27] generated from the Cayley tree graph with a branching ratio view at source ↗

**Figure 5.** Figure 5: FIG. 5. Neuronal network of the roundworm view at source ↗

**Figure 6.** Figure 6: FIG. 6. Undirected, unweighted graph generated from the directed, weighted graph representing view at source ↗

**Figure 7.** Figure 7: FIG. 7. Undirected, unweighted graph representing long-lasting social relationships in a community view at source ↗

**Figure 8.** Figure 8: FIG. 8. Undirected, unweighted graph generated from the undirected, weighted graph representing view at source ↗

**Figure 9.** Figure 9: FIG. 9. Pore network extracted from a random packing of 3-mm diameter glass beads. Pores view at source ↗

**Figure 10.** Figure 10: FIG. 10. This plot shows together on a logarithmic scale the magnitudes of all the view at source ↗

read the original abstract

To construct dispersion relations for diffusion or oscillation processes on random networks, it is necessary to obtain effective length scales for the eigenvectors of a graph Laplacian matrix, whose eigenvalues represent inverse time scales. For this purpose, we adapt a method originally introduced in condensed-matter physics to estimate correlation lengths for disordered materials as the ratio of volume to interface area [P. Debye, H.R. Anderson and H. Brumberger, J. Appl. Phys. 28, 679 (1957)]. In a graph setting of vertices connected by edges, we interpret this as the ratio of twice the total number of edges to the number of edges connecting vertices bearing values of different sign on the particular eigenvector. After describing the method and the necessary concepts in pedagogical detail, we apply it to nine different graphs representing natural and artificial networks, including two tree graphs without and with random shortcuts, the nervous system of a roundworm, a food web, a social network of dolphins, an electrical power grid, and a model porous material. The results identify both distributed and localized eigenvectors. They are given in graphical format showing example eigenvectors, dispersion relations, and discrete densities of states, as well as tables summarizing the main numerical results.

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.

Desk Editor's Note private letter to a colleague

Adapts Debye volume-interface ratio to define length scales from eigenvector sign changes on graphs, applies it to nine networks, but skips calibration on cases with known exact modes.

read the letter

The main point is a direct translation of the Debye correlation-length estimator to graph Laplacians. They take twice the total edge count divided by the number of edges linking opposite-sign components of a given eigenvector and treat that ratio as an effective length. From there they build dispersion relations and discrete densities of states for nine networks ranging from trees to the C. elegans connectome, a dolphin social graph, a power grid, and a porous-material model. They also flag which modes look distributed and which look localized. The write-up walks through the definitions clearly and the calculations are parameter-free, which keeps the procedure transparent and reproducible from the adjacency list and the eigenvector signs alone. That is the concrete contribution. The examples are specific enough that a reader can see how the length scales behave across different topologies. The soft spot is the missing anchor on solvable graphs. On a path graph the eigenvectors are exact discrete sines whose wavelengths are known, and the sign-change count scales as N over mode number, yet the paper does not check whether the resulting length produces the correct proportionality when plotted against the eigenvalues. Without that step or a comparison to another length measure, the dispersion curves stay illustrative rather than quantitatively tied to the spectrum. The work is aimed at people already using Laplacian spectra for diffusion or oscillation models on networks who want a quick way to assign wavelengths to the modes. It is not a foundational theoretical result, but the concrete numbers on empirical graphs give it practical value. The logic is coherent and the data are real, so it belongs in peer review rather than a desk reject. Referees can ask for the calibration check if they want tighter validation.

Referee Report

2 major / 2 minor

Summary. The manuscript adapts the Debye volume-to-interface ratio from condensed-matter physics to define effective length scales for eigenvectors of the graph Laplacian. For each eigenvector, the length is taken as twice the total edge count divided by the number of edges connecting vertices of opposite sign. These lengths are inverted to assign wave numbers k, from which dispersion relations λ(k) and discrete densities of states are constructed. The procedure is applied to nine networks (trees with and without shortcuts, C. elegans connectome, food web, dolphin social network, power grid, and a porous-material model), with results presented as example eigenvectors, dispersion plots, and summary tables that distinguish distributed versus localized modes.

Significance. If the length-scale construction is valid, the work supplies a direct, topology-based route to dispersion relations for diffusive or oscillatory dynamics on arbitrary networks. This could be useful for interpreting time scales in biological, social, and infrastructural systems without requiring geometric embedding or additional fitting parameters. The explicit application to diverse empirical graphs and the identification of localized modes constitute concrete demonstrations of the method's reach.

major comments (2)

[Method section (Debye adaptation)] Method section (description of the Debye adaptation): the effective length l = 2E / N_sign is inserted directly into a dispersion relation via k ~ 1/l without calibration on any graph possessing known analytic eigenvectors and eigenvalues. On the path graph the k-th eigenvector has exactly (k-1) sign changes, so l scales as N/k; any mismatch in the prefactor that converts l to wave number would uniformly rescale the entire λ(k) curve. The manuscript contains no such benchmark, leaving the mapping from eigenvalue to wave vector unanchored.
[Results on empirical networks] Results on empirical networks (dispersion relations and tables): the reported dispersion curves and densities of states are not compared against independent length-scale diagnostics (participation ratio, inverse participation ratio, or known analytic limits). No sensitivity analysis to the precise definition of sign-changing edges or to amplitude weighting is supplied, so the quantitative support for the central claim that these lengths yield physically usable dispersion relations remains moderate.

minor comments (2)

[Figure captions] Figure captions and axis labels for the dispersion plots should explicitly state the assumed functional form relating k to the computed length (e.g., k = π/l or k = 2π/l) so that readers can reproduce the curves from the tabulated lengths.
[Abstract and main text] The abstract states that the method is applied to “nine different graphs”; the main text should include a short table listing the nine networks together with their vertex and edge counts for quick reference.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the positive evaluation of the work's potential utility and for the constructive comments. We respond point by point below, agreeing that additional validation will strengthen the manuscript, and we outline the specific revisions planned.

read point-by-point responses

Referee: Method section (Debye adaptation): the effective length l = 2E / N_sign is inserted directly into a dispersion relation via k ~ 1/l without calibration on any graph possessing known analytic eigenvectors and eigenvalues. On the path graph the k-th eigenvector has exactly (k-1) sign changes, so l scales as N/k; any mismatch in the prefactor that converts l to wave number would uniformly rescale the entire λ(k) curve. The manuscript contains no such benchmark, leaving the mapping from eigenvalue to wave vector unanchored.

Authors: We agree that an explicit benchmark on analytically tractable graphs is needed to anchor the wave-number mapping. The prefactor in l = 2E / N_sign follows directly from the Debye volume-to-interface construction (with the factor of two arising because each undirected edge contributes to the total 'volume' count), but we recognize that demonstrating consistency with known scalings is essential. In the revised manuscript we will add a new subsection (in Methods or early Results) that applies the procedure to the path graph and the cycle graph. This will show that the number of sign changes for the k-th eigenvector is exactly k-1 on the path, yielding l ~ N/k, and will compare the resulting λ(k) against the exact analytic dispersion (e.g., λ_m = 2 - 2 cos(π m / N) for the path). Any uniform rescaling factor can then be identified and discussed, thereby anchoring the mapping. revision: yes
Referee: Results on empirical networks (dispersion relations and tables): the reported dispersion curves and densities of states are not compared against independent length-scale diagnostics (participation ratio, inverse participation ratio, or known analytic limits). No sensitivity analysis to the precise definition of sign-changing edges or to amplitude weighting is supplied, so the quantitative support for the central claim that these lengths yield physically usable dispersion relations remains moderate.

Authors: We concur that cross-validation with standard localization diagnostics and a sensitivity check would strengthen the quantitative claims. We will revise the Results section to include direct comparisons of our effective lengths against both the participation ratio and inverse participation ratio for all nine networks; these will appear in an expanded table and/or supplementary figures, with explicit discussion of how the two approaches classify distributed versus localized modes. We will also add a brief sensitivity analysis (in the main text or supplementary material) that varies the sign-change threshold (strict zero versus a small amplitude cutoff) and confirms that the main dispersion features and density-of-states trends remain robust. Note that amplitude weighting is not part of the present definition, which counts binary sign changes; we will clarify this choice and its rationale. revision: yes

Circularity Check

0 steps flagged

No circularity: length scale defined independently of eigenvalues

full rationale

The paper defines the effective length scale directly as twice the total edge count divided by the number of sign-changing edges on each eigenvector. This construction uses only the graph topology and the sign pattern of the eigenvector components, without reference to the eigenvalue itself or any fitting procedure. Dispersion relations are then assembled by mapping wave numbers inversely to these lengths and plotting against the separately computed Laplacian eigenvalues. The core step cites an external 1957 reference (Debye et al.) rather than self-citation, and no uniqueness theorem, ansatz smuggling, or renaming of known results occurs. Because the sign-change count is an independent observable property of the eigenvector and the resulting λ(k) plots are not forced to match any input by definition, the derivation chain remains self-contained.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the direct transfer of the continuous-media correlation-length definition to discrete graphs via sign changes, plus standard properties of the graph Laplacian.

axioms (1)

domain assumption The ratio of twice the total number of edges to the number of sign-changing edges on an eigenvector yields an effective length scale analogous to volume over interface area in continuous disordered media.
This is the core interpretive step adapted from Debye et al. (1957) and invoked to enable dispersion relations.

pith-pipeline@v0.9.0 · 5513 in / 1323 out tokens · 55910 ms · 2026-05-07T10:43:36.382771+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

55 extracted references · 1 canonical work pages

[1]

broken bonds

These are connected to 3 neighbors in layer 3, etc. Thus, each layer contains three times the number of vertices as the previous one. Figures 3 (a)–(c) show this structure. Each subtree rooted at layerrhas three-fold permutational symmetry around its root [35]. Its global parameters are⟨d⟩= 1.95,ρ= 0.05,C= 0,D= 6, andL≈4.3615. We find the ratio,L/L rand ≈...

1918
[2]

Further basic properties of the normalized spectrum for a graphGwith Nvertices are listed in Lemma 1.7 of Ref

Eigenvalue spectrum of normalized Laplacians The walk-normalized Laplacian,W=D −1Lis related to the symmetrically normalized Laplacian,L=D −1/2LD−1/2, by the similarity transformation,L=D 1/2WD−1/2, and the two normalized forms therefore have the same set of eigenvalues,{λ i}, which is confined to the interval, [0,2]. Further basic properties of the norma...
[3]

As these will have randomly varying signs, this will increase the number of broken bonds in the affected eigenvectors

Verification of numerical criterion for exact zero vertex values Vertex values in some eigenvectors that are known to be exactly zero due to high symme- tries in the graph or duplication or addition of motifs [23, 25] are reported in the numerical calculations used here as very small, real numbers. As these will have randomly varying signs, this will incr...

2078
[4]

small-world

R. Monasson, Diffusion, localization and dispersion relations on “small-world” lattices, Eur. Phys. J. B12, 555 (1999)

1999
[5]

D. I. Shuman, S. K. Narang, P. Frossard, A. Ortega, and P. Vandergheynst, The emerging field of signal processing on graphs, IEEE Signal Processing MagazineMay 2013, 83 (2013)

2013
[6]

D. I. Shuman, B. Ricaud, and P. Vandergheynst, Vertex-frequency analysis on graphs, Appl. Comput. Harmon. Anal.40, 260 (2016)

2016
[7]

Ricaud, P

B. Ricaud, P. Borgnat, N. Tremblay, P. Goncalves, and P. Vandergheynst, Fourier could be a data scientist: From graph Fourier transform to signal processing on graphs, Comptes Rendus Physique20, 474 (2019)

2019
[8]

MacMillan and N

T. MacMillan and N. T. Ouellette, Lagrangian scale decomposition via the graph Fourier transform, Phys. Rev. Fluids7, 124401 (2022)

2022
[9]

Debye, H

P. Debye, H. R. Anderson, and H. Brumberger, Scattering by an inhomogeneous solid. II. The correlation function and its application, J. Appl. Phys.28, 679 (1957)

1957
[10]

Euler, Solutio problematis ad geometriam situs pertinentis, Commentarii Acadmiae Scien- 34 tiarum Imperialis Petropolitanae8, 128 (1736)

L. Euler, Solutio problematis ad geometriam situs pertinentis, Commentarii Acadmiae Scien- 34 tiarum Imperialis Petropolitanae8, 128 (1736)
[11]

Kirchoff, Ueber den Durchgang eines elektrischen Stromes durch eine Ebene, insbesondere durch eine kreisf¨ ormige, Ann

G. Kirchoff, Ueber den Durchgang eines elektrischen Stromes durch eine Ebene, insbesondere durch eine kreisf¨ ormige, Ann. Phys. Chem.64, 487 (1845)
[12]

Kinney, P

R. Kinney, P. Crucitti, R. Albert, and V. Latora, Modeling cascading failures in the North American power grid, Eur. Phys. J. B46, 101 (2005)

2005
[13]

J. Zhu, X. Ma, H. Yang, Y. Li, C. Tao, and H. Li, Comprehensive geographic networks analysis: Statistical, geometric and algebraic perspectives, Symmetry14, 797 (2022)

2022
[14]

Allheeib, K

N. Allheeib, K. Adhinugraha, D. Taniar, and M. Saiful Islam, Computing reverse nearest neighbourhood on road maps, World Wide Web25, 99 (2022)

2022
[15]

Reis and K

P. Reis and K. J. M˚ aløy, Drainage front width in a three-dimensional random porous medium under gravitational and capillary effects, Phys. Rev. Research7, 033244 (2025)

2025
[16]

L. S. G. Leite, S. Banerjee, Y. Wei, J. Elowitt, and A. E. Clark, Modern chemical graph theory, WIREs Comput. Mol. Sci.14, 1729 (2024)

2024
[17]

Arnatkeviciute, B

A. Arnatkeviciute, B. D. Fulcher, R. Pocock, and A. Fornito, Hub connectivity, neuronal diversity, and gene expression in the Caenorhabditis elegans connectome, PLoS Comput. Biol. 14, e1005989 (2018)

2018
[18]

A. G. Rossberg,Food Webs and Biodiversity(Wiley Blackwell, Oxford, 2013)

2013
[19]

Kuhlmann and S

M. Kuhlmann and S. Oepen, Squibs: Towards a catalogue of linguistic graph banks, Comp. Linguistics42, 819 (2016)

2016
[20]

M. E. J. Newman,Networks. An Introduction(Oxford University Press, Oxford, 2010)

2010
[21]

Masuda, M

N. Masuda, M. A. Porter, and R. Lambiotte, Random walks and diffusion on networks, Phys. Rep.716-717, 1 (2017)

2017
[22]

Chung,Spectral graph theory(American Mathematical Society, Providence, RI, 1996) Ch

F. Chung,Spectral graph theory(American Mathematical Society, Providence, RI, 1996) Ch. 1

1996
[23]

Moler and C

C. Moler and C. Van Loan, Nineteen dubious ways to compute the exponential of a matrix, twenty-five years later, SIAM Rev.45, 3 (2003)

2003
[24]

Arnaudon, R

A. Arnaudon, R. L. Peach, and M. Barahona, Scale-dependent measure of network centrality from diffusion dynamics, Phys. Rev. Res.2, 033104 (2020)

2020
[25]

Bolla, B

M. Bolla, B. Bullins, S. Chaturapruek, S. Chen, and K. Friedl, Spectral properties of modu- larity matrices, Lin. Alg. Appl.473, 359 (2015)

2015
[26]

Banerjee and J

A. Banerjee and J. Jost, On the spectrum of the normalized graph Laplacian, Lin. Alg. Appl. 428, 3015 (2008). 35

2008
[27]

Banerjee and J

A. Banerjee and J. Jost, Spectral plot properties: Towards a qualitative classification of networks, Netw. Heterog. Media3, 395 (2008)

2008
[28]

Banerjee and J

A. Banerjee and J. Jost, Spectral characterization of network structure and dynamics, in Dynamics on and of complex networks, edited by N. Ganguly (Birkhauser, Boston, 2009) pp. 118–132

2009
[29]

Sciriha, A characterization of singular graphs, El

I. Sciriha, A characterization of singular graphs, El. J. Lin. Algebra16, 451 (2007)

2007
[30]

D. J. Watts and S. H. Strogatz, Collective dynamics of small-world networks, Nature393, 440 (1998)

1998
[31]

M. I. N. Rosa and M. Ruzzene, Small-world disordered lattices: spectral gaps and diffusive transport, New. J. Phys.24, 073020 (2022)

2022
[32]

J. L. Grady and J. R. Polimeni,Discrete Calculus(Springer-Verlag, London, U.K., 2010) Ch. 8, Measuring Networks

2010
[33]

P. A. Rikvold and G. Stell, D-dimensional interpenetrable-sphere models of random two- phase media: Microstructure and an application to chromatography, J. Coll. Int. Sci.108, 158 (1985)

1985
[34]

R. T. Armstrong, J. E. McClure, V. Robins, Z. Liu, C. H. Arns, S. Schl¨ uter, and S. Berg, Porous media characterization using Minkowski functionals: Theories, applications and future directions, Transp. Por. Media130, 305 (2019)

2019
[35]

M. E. J. Newman, Models of the Small World, J. Stat. Phys.101, 819 (2000)

2000
[36]

M. E. J. Newman and D. J. Watts, Renormalization group analysis of the small-world network model, Phys. Lett. A263, 341 (1999)

1999
[37]

Tuncer and A

A. Tuncer and A. Erzan, Spectral renormalization group for the Gaussian model andψ 4 theory on nonspatial networks, Phys. Rev. E92, 022106 (2015)

2015
[38]

Tuncer and A

A. Tuncer and A. Erzan, Explicit construction of the eigenvectors and eigenvalues of the graph Laplacian on the Cayley tree, Linear Algebra Appl.586, 111 (2020)

2020
[39]

https://sites.google.com/view/dynamicconnectomelab/resources
[40]

Kaiser and C

M. Kaiser and C. C. Hilgetag, Non-optimal component placement, but short processing paths, due to long-distance projections in neural systems, PLoS Comput. Biol.2, e95 (2006)

2006
[41]

J. G. White, E. Southgate, J. N. Thompson, and S. Brenner, The structure of the nervous system of the nematode Caenorhabditis Elegans, Phil. Trans. R. Soc. London314, 1 (1986)

1986
[42]

R. R. Christian and J. J. Luczkovich, Organizing and understanding a winter’s seagrass food- 36 web network through effective trophic levels, Ecol. Model.117, 99 (1999)

1999
[43]

R. J. Williams, E. L. Berlow, J. A. Dunne, A.-L. Barab´ asi, and N. D. Martinez, Two degrees of separation in complex food webs, Proc. Natl. Acad. Sci. U.S.A.99, 12913 (2002)

2002
[44]

P. A. Rikvold and V. Sevim, Individual-based predator-prey model for biological coevolution: Fluctuations, stability, and community structure, Phys. Rev. E75, 051920 (2007)

2007
[45]

Lusseau, The emergent properties of a dolphin social network, Proc

D. Lusseau, The emergent properties of a dolphin social network, Proc. R. Soc. London B (suppl.)270, S186 (2003)

2003
[46]

Lusseau, K

D. Lusseau, K. Schneider, O. J. Boisseau, P. Haase, E. Slooten, and S. M. Dawson, The bottlenose dolphin community of Doubtful Sound features a large proportion of long-lasting associations, Behavioral Ecol. Sociobiol.54, 396 (2003)

2003
[47]

Newman, were obtained from https://web.archive.org/web/20170822170651/http://networkdata.ics.uci.edu/data/dolphins/dolphins.gml

Data with 62 individuals, as entered by M. Newman, were obtained from https://web.archive.org/web/20170822170651/http://networkdata.ics.uci.edu/data/dolphins/dolphins.gml

work page arXiv
[48]

S. Dale, T. Alquthami, T. Baldwin, O. Faruque, J. Langston, P. McLaren, R. Meeker, M. Steurer, and K. Schoder,Progress Report for the Institute for Energy Systems, Economics and Sustainability and the Florida Energy Systems Consortium, Tech. Rep. (Center for Ad- vanced Power Systems, Florida State University, Tallahassee, FL, 2009)

2009
[49]

Abou Hamad, B

I. Abou Hamad, B. Israels, P. A. Rikvold, and S. V. Poroseva, Spectral matrix methods for partitioning power grids: Applications to the Italian and Floridian high-voltage networks, Phys. Proc.4, 125 (2010)

2010
[50]

Abou Hamad, P

I. Abou Hamad, P. A. Rikvold, and S. V. Poroseva, Floridian high-voltage power-grid network partitioning and cluster optimization using simulated annealing., Phys. Proc.15, 2 (2011)

2011
[51]

P. A. Rikvold, I. Abou Hamad, B. Israels, and S. V. Poroseva, Modeling power grids, Phys. Proc.34, 119 (2012)

2012
[52]

Y. Xu, A. J. Gurfinkel, and P. A. Rikvold, Architecture of the Florida power grid as a complex network, Physica A401, 130 (2014)

2014
[53]

J. F. Brodin, K. Pierce, P. Reis, P. A. Rikvold, M. Moura, M. Jankov, and K. J. M˚ aløy, Interface instability of two-phase flow in a three-dimensional porous medium, Phys. Rev. Fluids10, 064003 (2025)

2025
[54]

J. T. Gostick, Versatile and efficient pore network extraction method using marker-based watershed segmentation, Phys. Rev. E96, 023307 (2017)

2017
[55]

J. T. Gostick, Z. A. Khan, T. G. Tranter, M. D. R. Kok, M. Agnaou, M. Sadeghi, and R. Jervis, 37 PoreSpy: A Python toolkit for quantitative analysis of porous media images, J. Open Source Software4, 1296 (2019). 38

2019