Lagrangian Particle Classification and Lagrangian Flux Identities for a Moving Hypersurface

Baiyun Huang; Lingyun Ding; Qinghai Zhang; Shuang Hu

REVIEW 2 major objections 4 minor 36 references

Lagrangian Particle Classification and Lagrangian Flux Identities for a Moving Hypersurface

T0 review · 2 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 The paper proves that the flux of a conserved scalar through a moving hypersurface equals two time-independent Lagrangian expressions: a weighted sum of integrals over donating regions and a surface integral over a generating cycle, and…

desk verdict Real advance in Lagrangian flux calculation, but the headline flux identity is proved only for smooth scalars while the introduction advertises the discontinuous-VOF case. read the letter →

arxiv 2506.04125 v1 pith:3FFBD7IK submitted 2025-06-04 math.NA cs.NA

Lingyun Ding , Shuang Hu , Baiyun Huang , Qinghai Zhang This is my paper

classification math.NAcs.NA MSC 37K2570H3376M12

keywords Lagrangianfluxcalculationdonatingregionstopologicaldegreefluxingindexscalarconservationlawsgeneratingcyclemovinghypersurfacefinitevolumemethods

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

The paper aims to solve the Lagrangian flux calculation problem in any dimension $m\ge 2$: given a moving hypersurface $S(t)$, a velocity $u$, and a scalar $f$ obeying the conservation law $\partial_t f + \nabla\cdot(fu)=0$, it identifies the Eulerian flux through $S(t)$ between $t_0$ and $t_0+k$ with purely spatial integrals evaluated at $t_0$. The first identity expresses that flux as a weighted sum of integrals over donating regions, the sets of initial particle positions whose fluxing index equals $n$. The second identity expresses it as a surface integral over a closed generating cycle formed by $S(t_0)$, the backward image of $S(t_0+k)$, and the streak surface traced by the boundary. Along the way the fluxing index of a particle is shown to be a topological degree, which generalizes the two-dimensional winding-number construction and is what makes the identities valid for self-intersecting cycles. The paper then turns the cycle form into a second-, fourth-, and sixth-order 3D algorithm with proved convergence.

What carries the argument

The generating cycle is the central object: $G_D(t_0,k) = S(t_0) \cup \phi^{-k}_{t_0+k}(S(t_0+k)) \cup \Psi_{\partial S}(t_0,k)$, where the streak surface $\Psi_{\partial S}$ collects the backward trajectories of the boundary of the moving surface. This cycle is the image under $\chi$ of the boundary of the $m$-cube, $\chi(\partial B^m)$, which makes it a possibly self-intersecting topological cycle with a well-defined intrinsic orientation. The proof machinery is the topological degree, the area formula converting integrals over $B^m$ into weighted sums over degree-level sets, and custom versions of the divergence theorem and Reynolds transport theorem for self-intersecting cycles. The degree reduces to the winding number in two dimensions; the customized theorems let one differentiate the flux quantity in time, use the conservation law to remove the divergence term, and close the identity by integrating over the time interval.

What would settle it

Use the LFC3D implementation on step-function initial data $f=1$ inside a ball, with the unsteady incompressible flow of equation (5.3) and a fixed planar surface $S: z=1/4$, and compare the generating-cycle integral against an independently computed Eulerian flux. If the difference does not vanish as $h$ and $\Delta t$ are refined, then Theorem 3.18 does not hold for the discontinuous scalars that motivated the work.

Watch

Extended reading notes

Core claim

The central claim is Theorem 3.18: for a $C^1$ velocity $u$, a scalar $f$ satisfying $\partial_t f + \nabla\cdot(fu)=0$, and a moving hypersurface $S(t)$, the Eulerian flux equals both the summation over nonzero indices $n$ of $n$ times the integral of $f(x,t_0)$ over the donating region $D^n_S(t_0,k)$, and the surface integral of $F\cdot n$ over the generating cycle $G_D(t_0,k)$, where $F$ is any vector field with divergence $f(x,t_0)$. The fluxing index of a particle $p$, the signed number of times the particle crosses $S(t)$, is shown in Theorem 3.10 to equal the topological degree $\deg(\chi, B^m, p)$ of the map $\chi(z,\tau)=\phi^{-\tau k}_{t_0+\tau k}(S(z,t_0+\tau k))$. These results, together with the index-by-index equality of flux sets and donating regions, are the analytical solution the paper proposes for Lagrangian particle classification and Lagrangian flux calculation.

Load-bearing premise

The proof of the flux identity differentiates the scalar pointwise, so its central assumption is that the scalar is smooth; if the scalar is only discontinuous, the case that matters for volume-of-fluid methods, the paper does not supply the approximation argument needed to extend the theorem.

Editorial extensions

If this is right

The Eulerian flux of a conserved scalar through a moving surface can be computed without evolving $f$ in time; only $f$ at the initial instant and the flow map are needed.
Flux sets are computable from a topological degree, so particle crossing counts are available without tracing individual pathlines.
The cycle form of the identity avoids decomposing self-intersecting donating-region boundaries, which removes the ill-conditioned step in earlier two-dimensional approaches and generalizes to any space dimension $m\ge 2$.
The proposed LFC3D algorithm achieves second-, fourth-, and sixth-order accuracy in the reported tests, with CPU time growing slower than a full three-dimensional finite-volume solve would require.
Because the identity covers moving and self-intersecting hypersurfaces, it also settles the orientation issue raised by eversions of closed surfaces in $\mathbb{R}^3$, where extrinsic orientations fail.

Reading between the lines

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

If the identity extends to discontinuous $f$ by approximation, it would provide a rigorous three-dimensional truncation-error analysis for unsplit volume-of-fluid advection, a connection the paper motivates but does not prove.
The generating-cycle representation suggests a fully Lagrangian finite-volume method in which fluxes over arbitrarily long time intervals are assembled by solving ODEs for the surface skeleton rather than by evolving the scalar field; the paper notes this CFL-free possibility as future work.
For flows on curved manifolds, where no global extrinsic orientation exists, the intrinsic parametrization used here is the natural starting point, so the same proof structure may carry over with the Euclidean flow map replaced by geodesic flow.

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.

Desk Editor's Note

read the letter

Bottom line: this is a genuine advance in Lagrangian flux calculation, and it deserves a serious referee. The topological-degree characterization of fluxing indices (Theorem 3.10), the two flux identities in (3.28), and the first 3D high-order LFC algorithm are all new and mostly well supported. The orientation discussion via sphere eversion is not just decoration; it identifies a real obstruction that the 2D winding-number framework hides. The customized divergence theorem for self-intersecting cycles is a neat piece of work, and the algorithm's design follows naturally from the surface-integral flux identity. The numerical tests are extensive and show the claimed second-, fourth-, and sixth-order rates for smooth scalars, including on self-intersecting generating cycles. The convergence proof in Theorem 4.2 is more sketched than fully detailed, but the structure is credible and the numerical evidence is consistent.

The main soft spot is the regularity gap in Theorem 3.18. The theorem states no regularity for f, and the introduction explicitly says the identity applies even when f is discontinuous in space, citing the VOF context. But the proof of the first equality in (3.28), through (3.31), substitutes the pointwise conservation law and invokes the Reynolds transport theorem, which is stated only for C^1 scalar fields. For a weak, discontinuous solution there is no pointwise equation, and no mollification or BV approximation argument is supplied. The numerical experiments use only smooth or constant f, so they do not exercise the advertised discontinuous case. This is a real gap, not a fatal one: the smooth case is proven, and the identity likely extends by approximation, but the paper currently overclaims. The second equality in (3.28) has the same issue because F is an antiderivative of f and needs to be C^1 for Theorem 3.14.

Other concerns are minor. The non-monotone second-order rates in Tables 5.3 and 5.4 are unexplained; they look pre-asymptotic, but the authors should say so. The sign in (5.2) is actually fine as written, since the quotient is coarse-over-fine, but that is easy to misread. The citation pattern for prior LFC work is appropriate, including the alternative approach of Karrasch and colleagues.

Who this is for: people working on computational transport, finite volume methods, and VOF, and anyone who wants a higher-dimensional generalization of winding-number-based flux identities. Recommendation: send to peer review. A major revision should add a precise regularity hypothesis to Theorem 3.18 and an approximation argument for discontinuous f, or explicitly restrict the claim to smooth f and reposition the VOF motivation as future work. With that fixed, the paper would be a solid contribution.

Referee Report

2 major / 4 minor

Summary. The paper develops a Lagrangian theory of flux calculation for moving hypersurfaces in R^m, m≥2. It defines the fluxing index of a Lagrangian particle, characterizes flux sets as topological degrees of a composite map (Theorem 3.10), constructs donating regions via a generating cycle (Definitions 3.12 and 3.17), and proves two flux identities (Theorem 3.18) that identify the Eulerian flux through a moving hypersurface with a weighted sum of initial-time integrals over donating regions and with a surface integral over the generating cycle. It then proposes a three-dimensional LFC algorithm based on the surface-integral identity, states a convergence result (Theorem 4.2), and supports it with numerical experiments showing second-, fourth-, and sixth-order accuracy.

Significance. If established in full generality, the proposed flux identities would provide exact, time-independent Lagrangian expressions for Eulerian fluxes in any dimension m≥2, generalizing the two-dimensional LFC framework to three and higher dimensions while avoiding the ill-conditioned intersection computations of earlier donating-region methods. The paper's use of topological degree, its customized divergence and Reynolds transport theorems for self-intersecting cycles, and its reproducible MATLAB code are concrete strengths. The surface-integral identity (3.28) and the accompanying algorithm are potentially valuable for finite-volume and VOF flux calculations.

major comments (2)

[§3.5, Theorem 3.18, Eq. (3.31)] The proof of the first equality in (3.28) uses the pointwise conservation law ∂τ f = −∇·(f u) in the third line of (3.31), and the preceding steps rely on Theorem 3.16 and Theorem 3.14, both stated for C1 integrands. The theorem states no regularity assumption on f, and the Introduction (after Eq. (1.6)) explicitly claims the identity applies "even when f is discontinuous in space," citing [32] and VOF methods. For a weak solution of (1.4), the pointwise substitution is unjustified, the surface flux (1.5) may lack a classical trace, and the area-formula terms need a limiting argument. No mollification or BV approximation is supplied, and the Section 5 tests use only smooth or constant f. The central identity is therefore proved only for classical (at least C1) scalars, and the advertised discontinuous VOF setting is unsupported. Please either add a rigorous approximation argument or restate the theorem with explicit regularity and adjust the Introduction and Conclusion accordingly.
[§4, Theorem 4.2, proof around (4.8)–(4.10)] The convergence proof for Algorithm 3 contains two gaps. First, (4.10) asserts Iκ(\tilde P, pij) = Iκ(\tilde P, f) from interpolation at the spline knots (ui, vj), but the quadrature formula (4.4) evaluates the integrand at Gauss–Legendre nodes, which are not the interpolation sites; equality at the knots does not imply equality at the quadrature abscissae. A correct estimate of |Iκ(\tilde P, f) − Iκ(\tilde P, pij)| must be added. Second, in (4.8) the term ∫ |F(xuv)(JP − J_{\tilde P})| dudv is bounded by K2 O(h^{2κ−2}), but a pointwise bound on the Jacobian difference is only O(h^{κ−1}), so the displayed estimate needs a justification (e.g., an integration-by-parts argument exploiting the structure of the interpolation error). As written, the claimed κth-order accuracy of Algorithm 3 is not fully proved.

minor comments (4)

[§3.2, Theorem 3.10] Theorem 3.10 and Definitions 3.5–3.6: the fluxing index in (3.5) counts only transverse crossings, but the equality with the topological degree is stated for all p. Add the hypothesis that p ∉ χ(∂B^m) and that all preimages are regular, or discuss the measure-zero exceptional set.
[§3.4, Theorem 3.14] Theorem 3.14 is stated for φ∈C1, but its proof differentiates the cofactors Ki,j of dφ, which requires φ∈C2; either strengthen the hypothesis or add a standard density argument.
[§3.3, Lemma 3.13 and Definition 3.12] The map χ in (3.7) is defined on the open cube B^m, while (3.18) evaluates it on ∂B^m; specify the continuous extension used to define the generating cycle.
[§2.3, sphere eversion example] The sphere-eversion example computes the flux as 8π/3; it would help to state explicitly that the flux is independent of the particular eversion and that the example is meant to illustrate the failure of extrinsic orientation.

Circularity Check

0 steps flagged · score 1.0 of 10

No meaningful circularity: the flux identities follow from external area-formula and divergence-theorem inputs, not from a fitted definition; the only flagged concern is a regularity gap in the discontinuous-f claim supported by a self-citation.

full rationale

The central derivation is self-contained. Theorem 3.18's first equality is obtained by differentiating the area-formula identity (3.27), applying the paper's own Reynolds transport theorem (Theorem 3.16), substituting the given conservation law ∂_τ f = -∇·(fu), and applying the paper's cycle divergence theorem (Theorem 3.14); the second equality follows from Theorem 3.14 and Lemma 3.13. The donating regions D^n_S are defined independently as level sets of the topological degree of χ in Definition 3.17, and the generating cycle is χ(∂B^m) by Lemma 3.13, so the weighted sum on the right of (3.28) is not chosen to match the left-hand Eulerian flux. Theorem 3.10 also derives the fluxing index from the sign identity (3.6), Jacobi's formula, and degree theory rather than assuming it. The prior work of the same group [33–35] is used as motivation and context, not as the proof of the m-dimensional result; the m-dimensional theorems are proved in the text from stated assumptions and external classical results (area formula [7,10], degree theory [4]). The only noteworthy caveat is a regularity/evidence gap rather than circularity: the introduction claims the flux identity 'applies even when f is discontinuous in space; see [32] for such an analysis,' but the proof in (3.31) substitutes the pointwise conservation law and uses C^1 integrands in Theorems 3.14 and 3.16, and Theorem 3.18 states no regularity for f. This makes the advertised VOF/discontinuous-f extension depend on self-citation [32] and on an approximation argument not supplied here; however, no fitted parameter or definitional reduction is involved, so the circularity score remains minimal. Numerical tests in Section 5 use smooth or constant f and verify the proved identity, so there is no fitted-input-as-prediction pattern.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The paper introduces no free parameters and no new physical entities. It relies on standard results (topological degree, area formula, Jacobi's formula, Cauchy-Lipschitz, Sard's theorem) and on two regularity assumptions: hypersurfaces are regular, and f is smooth enough for the pointwise conservation law in the proof of Theorem 3.18. The latter is an unflagged assumption.

assumptions (7)

standard math Topological degree theory (existence, uniqueness, additivity, product formula) as in Deimling.
Used in Section 2.2 and Theorem 3.10 to define and compute fluxing indices.
standard math Area formula for Lipschitz maps (Giaquinta-Modica-Soucek; Krantz-Parks).
Used in (3.27) to convert integrals over B^m into sums over donating regions.
standard math Jacobi's formula for the Jacobian determinant of a flow (Lemma 3.8).
Used to prove the flow map preserves orientation (Lemma 3.9).
standard math Cauchy-Lipschitz theorem for ODEs.
Used in Lemma 3.9 to exclude vanishing Jacobian via uniqueness.
standard math Sard's theorem that singular values have measure zero.
Used in Definition 2.6 and Definition 3.4 to ignore singular points.
domain assumption All hypersurfaces and cycles are regular (C^1 with rank m-1).
Stated in Section 3.1; the paper argues approximation by regular hypersurfaces but does not carry that through the proofs.
ad hoc to paper The scalar field f is differentiable enough for the pointwise identity partial_tau f = -div(f u) in the proof of Theorem 3.18.
Needed in the derivation of (3.31); the theorem statement omits this assumption and the introduction claims applicability to discontinuous f.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Lagrangian Particle Classification and Lagrangian Flux Identities for a Moving Hypersurface." pith.science (2026). https://pith.science/paper/3FFBD7IK

@misc{pith2026250604125,
  author       = {Pith},
  title        = {Pith review of: Lagrangian Particle Classification and Lagrangian Flux Identities for a Moving Hypersurface},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3FFBD7IK}},
  note         = {Machine review of arXiv:2506.04125}
}

abstract

For a moving hypersurface in the flow of a nonautonomous ordinary differential equation in $n$-dimensional Euclidean spaces, the fluxing index of a passively-advected Lagrangian particle is the total number of times it crosses the moving hypersurface within a time interval. The problem of Lagrangian particle classification is to decompose the phase space into flux sets, equivalence classes of Lagrangian particles at the initial time. In the context of scalar conservation laws, the problem of Lagrangian flux calculation (LFC) is to find flux identities that relate the Eulerian flux of a scalar through the moving hypersurface, a spatiotemporal integral over the moving surface in a given time interval, to spatial integrals over donating regions at the initial time of the interval. In this work, we implicitly characterize flux sets via topological degrees, explicitly construct donating regions, prove the equivalence of flux sets and donating regions, and establish two flux identities; these analytical results constitute our solutions to the aforementioned problems. Based on a flux identity suitable for numerical calculation, we further proposed a new LFC algorithm, proved its convergence, and demonstrated its efficiency, good conditioning, and high-order accuracy by results of various numerical tests.

Figures

Figures reproduced from arXiv: 2506.04125 by the authors.

**Figure 3.1.** Fluxing indices np of a Lagrangian particle p through a static surface S. A shaded region represents S, a dotted line the pathline Φ+k t0 (p), and a marker “×” a crossing point Φ+k t0 (p) ∩ S. By Definition 3.5, the above fluxing index can be expressed as np(u, t0, k, S) = X τ∈T× sign[v×(p, τ ) · n(zp, t0 + τ k)], (3.5) where T× := τ ∈ (0, 1) | ϕ +τk t0 (p) ∈ S(t0 + τ k) [PITH_FULL_IMAGE:figures/full_fig_p011_3_1.png] view at source ↗

**Figure 3.2.** The generating cycle GD of a moving square S(z1, z2, t) = (z1 − t, z2, 0) in the flow of u(x, y, z, t) ≡ (0, 0, −1) over the time interval (0, 1). The squares ABCD, A˜B˜C˜D˜, and A1B1C1D1 represent S(0), S(1), and ϕ −1 1 (S(1)), respectively. The boundary of the parallelepiped ABCD − A1B1C1D1 constitutes GD. The four dotted lines are the pathlines of A1, B1, C1 and D1. Theorem 3.14. Define a cycle S := φ(∂B m) with … view at source ↗

**Figure 5.1.** The generating cycle constructed with κ = 6 and h = 1 256 for the fixed surface (5.5) over the time interval [t0, te] = [0, 3 2 ] in the flow of (5.3) where T = 3. The surface color indicates the altitude. To satisfy the range of u, v ∈ (0, 1)2 assumed in Algorithm 3, we convert (5.6) to the following equivalent form S(u, v, t) = (2u − 1) sin πt 2 , (2v − 1) sin πt 2 , t 2 72 [9(2u − 1)2 + 4(2v − 1)2 ] . The evo… view at source ↗

Figures from the paper (3 more)

**Figure 5.2.** Figure 5.2: The moving surface (5.6) (the first row) and the corresponding generating cycles [PITH_FULL_IMAGE:figures/full_fig_p024_5_2.png]

**Figure 5.3.** Figure 5.3: The generating cycle GS (the left subplot) and part of its self-intersection in (5.10) (the right subplot) constructed by Algorithm 2 with κ = 6 and h = 1 256 for the fixed surface (5.9) in the compressible flow (5.7) during the time interval [0, 1]. The color indica…

**Figure 5.4.** Figure 5.4: The generating cycles constructed by Algorithm 2 with [PITH_FULL_IMAGE:figures/full_fig_p026_5_4.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

[32]

Q. Zhang. On a family of unsplit advection algorithms for volume-of-fluid methods.SIAM Journal on Numerical Analysis, 51(5):2822–2850, 2013

work page 2013
[1]

J. B. Bell, C. N. Dawson, and G. R. Shubin. An unsplit, higher order Godunov method for scalar conservation laws in multiple dimensions.Journal of Computational Physics, 74(1):1–24, 1988

work page 1988
[2]

A. J. Chorin and J. E. Marsden.A Mathematical Introduction to Fluid Mechanics. Springer, New York, 3rd edition, 1993. ISBN:978-0387979182

work page 1993
[3]

P. Colella. Multidimensional upwind methods for hyperbolic conservation laws.Journal of Computational Physics, 87(1):171–200, 1990

work page 1990
[4]

Deimling.Nonlinear Functional Analysis

K. Deimling.Nonlinear Functional Analysis. Dover Publications, New York, 2010. ISBN: 978-0486474410

work page 2010
[5]

L. Ding, R. Hunt, R. M. McLaughlin, and H. Woodie. Enhanced diffusivity and skewness of a diffusing tracer in the presence of an oscillating wall.Research in the Mathematical Sciences, 8(3):1–29, 2021

work page 2021
[6]

G. K. Francis and B. Morin. Arnold Shapiro’s eversion of the sphere.The Mathematical Intelligencer, 2(4):200–203, 1980

work page 1980
[7]

Giaquinta, G

M. Giaquinta, G. Modica, and J. Soucek.Cartesian Currents in the Calculus of Variations I: Cartesian Currents. Ergebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge / A Series of Modern Surveys in Mathematics. Springer Berlin Heidelberg, London, 1998

work page 1998

Show all 36 references

[8]

Hofherr and D

F. Hofherr and D. Karrasch. Lagrangian transport through surfaces in compressible flows. SIAM Journal on Applied Dynamical Systems, 17(1):526–546, 2018

2018
[9]

Karrasch

D. Karrasch. Lagrangian transport through surfaces in volume-preserving flows.SIAM Journal on Applied Mathematics, 76(3):1178–1190, 2016. 27

2016
[10]

S. G. Krantz and H. R. Parks.Geometric integration theory. Springer Science & Business Media, New York, 2008

2008
[11]

J. D. Lambert et al.Numerical Methods for Ordinary Differential Systems, volume 146. Wiley, New York, 1991

1991
[12]

R. J. Leveque. High-resolution conservative algorithms for advection in incompressible flow. SIAM Journal on Numerical Analysis, 33(2):627–665, 1996

1996
[13]

R. J. LeVeque et al.Finite Volume Methods for Hyperbolic Problems. Cambridge university press, Cambridge, 2002

2002
[14]

Levy and W

S. Levy and W. P. Thurston.Making Waves: A Guide to the Ideas behind Outside In. A K Pe- ters, Wellesley, Massachusetts, 1995. https://www.youtube.com/watch?v=wO61D9x6lNY

1995
[15]

Lin and Q

Z. Lin and Q. Zhang. High-order finite-volume solutions of the steady-state advection–diffusion equation with nonlinear Robin boundary conditions.Journal of Computational Physics, 345:358–372, 2017

2017
[16]

Lipinski and K

D. Lipinski and K. Mohseni. A ridge tracking algorithm and error estimate for efficient computa- tion of Lagrangian coherent structures.Chaos: An Interdisciplinary Journal of Nonlinear Science, 20(1):017504, 2010

2010
[17]

Litman.Sphere Eversion: Analysis of a Veridical Paradox

M. Litman.Sphere Eversion: Analysis of a Veridical Paradox. PhD thesis, Pennsylvania State University, 2017

2017
[18]

Outerelo and J

E. Outerelo and J. M. Ruiz.Mapping Degree Theory. American Mathematical Society, Provi- dence RI, 2009

2009
[19]

Rudin.Real and Complex Analysis

W. Rudin.Real and Complex Analysis. McGraw Hill, Singapore, 3rd edition, 1986. ISBN:978- 0070542341

1986
[20]

E. N. Sarmin and L. Chudov. On the stability of the numerical integration of systems of ordinary differential equations arising in the use of the straight line method.USSR Computational Mathematics and Mathematical Physics, 3(6):1537–1543, 1963

1963
[21]

W. E. Schiesser and G. W. Griffiths.A Compendium of Partial Differential Equation Models: Method of Lines Analysis with Matlab. Cambridge University Press, New York, 2009

2009
[22]

S. C. Shadden, F. Lekien, and J. E. Marsden. Definition and properties of Lagrangian coherent structures from finite-time Lyapunov exponents in two-dimensional aperiodic flows.Physica D: Nonlinear Phenomena, 212(3-4):271–304, 2005

2005
[23]

S. Smale. A classification of immersions of the two-sphere.Transactions of the American Mathematical Society, 90(2):281–290, 1959

1959
[24]

Sommariva and M

A. Sommariva and M. Vianello. Compression of multivariate discrete measures and applications. Numerical Functional Analysis and Optimization, 36(9):1198–1223, 2015

2015
[25]

Sudhakar, J

Y. Sudhakar, J. M. De Almeida, and W. A. Wall. An accurate, robust, and easy-to-implement method for integration over arbitrary polyhedra: application to embedded interface meth- ods.Journal of Computational Physics, 273:393–415, 2014

2014
[26]

Sudhakar, A

Y. Sudhakar, A. Sommariva, M. Vianello, and W. A. Wall. On the use of compressed poly- hedral quadrature formulas in embedded interface methods.SIAM Journal on Scientific Computing, 39(3):B571–B587, 2017

2017
[27]

Sudhakar and W

Y. Sudhakar and W. A. Wall. Quadrature schemes for arbitrary convex/concave volumes and integration of weak form in enriched partition of unity methods.Computer Methods in Applied Mechanics and Engineering, 258:39–54, 2013

2013
[28]

L. W. Tu.An Introduction to Manifolds. Springer, New York, 2nd edition, 2011. ISBN: 978-1-4419-7399-3

2011
[29]

J. H. Verner. Explicit Runge–Kutta methods with estimates of the local truncation error.SIAM Journal on Numerical Analysis, 15(4):772–790, 1978

1978
[30]

H. Whitney. On regular closed curves in the plane.Composito Math., 4:276–284, 1937

1937
[31]

Q. Zhang. Highly accurate Lagrangian flux calculation via algebraic quadratures on spline- approximated donating regions.Computer Methods in Applied Mechanics and Engineering, 264:191–204, 2013

2013
[33]

Q. Zhang. On donating regions: Lagrangian flux through a fixed curve.SIAM Review, 55(3):443–461, 2013

2013
[34]

Q. Zhang. On generalized donating regions: Classifying Lagrangian fluxing particles through a fixed curve in the plane.Journal of Mathematical Analysis and Applications, 424(2):861– 877, 2015

2015
[35]

Zhang and L

Q. Zhang and L. Ding. Lagrangian flux calculation through a fixed planar curve for scalar conservation laws.SIAM Journal on Scientific Computing, 41(6):A3596–A3623, 2019

2019
[36]

Zhang, W

Z. Zhang, W. Wang, and B. Qiu. Oceanic mass transport by mesoscale eddies.Science, 345(6194):322–324, 2014. 28

2014

Pith tools

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

[1] [32]

Q. Zhang. On a family of unsplit advection algorithms for volume-of-fluid methods.SIAM Journal on Numerical Analysis, 51(5):2822–2850, 2013

work page 2013

[2] [1]

J. B. Bell, C. N. Dawson, and G. R. Shubin. An unsplit, higher order Godunov method for scalar conservation laws in multiple dimensions.Journal of Computational Physics, 74(1):1–24, 1988

work page 1988

[3] [2]

A. J. Chorin and J. E. Marsden.A Mathematical Introduction to Fluid Mechanics. Springer, New York, 3rd edition, 1993. ISBN:978-0387979182

work page 1993

[4] [3]

P. Colella. Multidimensional upwind methods for hyperbolic conservation laws.Journal of Computational Physics, 87(1):171–200, 1990

work page 1990

[5] [4]

Deimling.Nonlinear Functional Analysis

K. Deimling.Nonlinear Functional Analysis. Dover Publications, New York, 2010. ISBN: 978-0486474410

work page 2010

[6] [5]

L. Ding, R. Hunt, R. M. McLaughlin, and H. Woodie. Enhanced diffusivity and skewness of a diffusing tracer in the presence of an oscillating wall.Research in the Mathematical Sciences, 8(3):1–29, 2021

work page 2021

[7] [6]

G. K. Francis and B. Morin. Arnold Shapiro’s eversion of the sphere.The Mathematical Intelligencer, 2(4):200–203, 1980

work page 1980

[8] [7]

Giaquinta, G

M. Giaquinta, G. Modica, and J. Soucek.Cartesian Currents in the Calculus of Variations I: Cartesian Currents. Ergebnisse der Mathematik und ihrer Grenzgebiete. 3. Folge / A Series of Modern Surveys in Mathematics. Springer Berlin Heidelberg, London, 1998

work page 1998