A splitting mixed finite element method for a stochastic Keller-Segel system with multiplicative noise

Liet Vo; Thoa Thieu

REVIEW 2 major objections 5 minor 26 references

A splitting mixed finite element method for a stochastic Keller-Segel system with multiplicative noise

T0 review · 2 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read This paper proves a splitting mixed finite element method for a stochastic Keller-Segel system with multiplicative noise, establishing strong error estimates and convergence in probability at rate (k+h^2)^alpha.

desk verdict The main convergence theorem has an invalid absorption step that breaks the Gronwall argument, but the splitting method and stability analysis are worth refereeing if the proof can be fixed. read the letter →

arxiv 2608.07798 v1 pith:XR2SGV6K submitted 2026-08-07 math.NA cs.NAmath.PR

Thoa Thieu , Liet Vo This is my paper · ORCID

classification math.NAcs.NAmath.PR MSC 65N1265N1565N30

keywords stochasticKeller-SegelmultiplicativenoisesplittingmixedfiniteelementerrorestimatesconvergenceinprobabilitylogisticgrowthlocalizationtechniqueimplicitEuler

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 develops and analyzes a splitting mixed finite element method for a stochastic Keller-Segel chemotaxis system with logistic growth driven by multiplicative Wiener noise, and claims the first rigorous finite element error analysis for this class of equations. The method introduces the chemical gradient $\sigma$ = grad v as an auxiliary variable and uses a time-lagged splitting, so each time step solves three successive subproblems with continuous piecewise linear elements for all unknowns, avoiding the inf-sup stability condition and fully coupled nonlinear solves. The main theorem bounds the localized expected squared errors of u, $\sigma$, and v by C ln(1/k)(k+$h^{2}$), and the corollary upgrades this to convergence in probability with rate (k+$h^{2}$)^$\alpha$ for any $\alpha$ < 1/2. A sympathetic reader would care because the result turns a coupled nonlinear stochastic system with strong noise-flux interaction into a cheap, decoupled scheme with quantitative accuracy guarantees.

What carries the argument

The load-bearing object is the auxiliary variable $\sigma$ = grad v, which rewrites the nonlinear chemotaxis flux grad·(u grad v) as first-order products, together with a time-lagged splitting that computes $sigma^{{m+1}}$_h from the previous density u^m_h, then updates u, then recovers v by a linear relation. This removes the LBB constraint and permits equal-order P1 spaces. The proof machinery also includes the projection operators P_u, P_sigma, P_v, the localization sets Omega_rho,m where the exact and discrete densities are bounded by rho = ln ln(1/k)/C, Holder continuity estimates for the exact solution, and a discrete Gronwall argument after absorbing interaction terms for small k.

What would settle it

One concrete check: exhibit a data set satisfying the paper's assumptions on B and u0 for which the mild solution is not a variational weak solution with the assumed $H^{2}$ spatial regularity, or for which the Holder estimate E[||grad(u(t)-u(s))||^2] <= C|t-s| fails; then Theorem 3.1 would have no exact solution with the required regularity to compare against. A numerical search for such a case could start with low-regularity initial data and measure the Holder exponent of the reference solution.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central claim is Theorem 3.1: under the stated regularity assumptions, for every small time step k, the fully discrete approximations satisfy max_m E[1_{Omega_rho,m} ||u(t_m)-u^m_h||^2_L2] + k sum_m E[1_{Omega_rho,m} ||grad(u(t_m)-u^m_h)||^2_L2] <= C ln(1/k)(k+$h^{2}$), with analogous bounds for the auxiliary gradient $\sigma$ = grad v in $H^{1}$ and for v in $L^{2}$. Corollary 3.1 then concludes convergence in probability at rate (k+$h^{2}$)^$\alpha$ for every $\alpha$ in (0,1/2). The argument splits the error into projection and discrete parts, derives an error equation from the mixed variational formulation, and controls the nonlinear chemotaxis term and the multiplicative stochastic forcing on the good-event sets Omega_rho,m, where both the exact and discrete densities stay bounded.

Load-bearing premise

The load-bearing premise is that a variational weak solution of the system exists with the spatial regularity u in $L^{2}$($\Omega$; L^infty(0,T;$H^{2}$)) cap $L^{6}$($\Omega$;L^infty(0,T;$H^{1}$)) and the Holder-in-time estimates of Lemma 2.2; the paper's Remark 2.1 concedes it does not prove this existence, asserting it follows from the mild-solution theory it cites.

Editorial extensions

If this is right

The scheme achieves optimal strong error bounds using only continuous piecewise linear elements for u, sigma, and v, so low-order codes suffice for a stochastic chemotaxis model.
Because each timestep decouples into three successively solved subproblems, the cost per step drops relative to fully coupled mixed methods, directly benefiting Monte Carlo estimation over many sample paths.
The error bounds yield explicit rates for convergence in probability, so users can choose k and h to hit a target accuracy with confidence.
The gradient variable sigma converges in H^1 and v in L^2, meaning the method delivers a directly usable approximation of the chemotactic gradient, not just the density.
The numerical experiments indicate the method reproduces the global boundedness of solutions caused by logistic growth, matching the analytic well-posedness picture.

Reading between the lines

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

A sharpened Aubin-Nitsche duality argument could plausibly raise the proven L^2 spatial rate from O(h) to O(h^2), as the paper's own Test 2 data suggest; the paper does not prove this.
The localization to Omega_rho,m means the error bound is probabilistic: the current proof leaves open whether full L^p(Omega) strong convergence holds without the logarithmic factor; removing the localization is a natural next step.
The same auxiliary-variable-plus-time-lag template likely transfers to related stochastic reaction-diffusion and chemotaxis-fluid systems, since the structure that causes the difficulty (multiplicative noise hitting a nonlinear flux) appears there too.
Because the splitting makes each subproblem smaller and independent of the others, it may combine well with adaptive mesh refinement or reduced-basis methods for stochastic sampling, though the paper does not explore this.

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

The headline is unwelcome but simple: the paper's advertised error estimate is not proved as written. The gap is the absorption step between (3.37) and (3.38) in Theorem 3.1. The right-hand side contains C k (1+sqrt(rho)+rho) E[1_{Omega_m}||eps^{m+1}||^2], and the proof tries to absorb it into the left-hand side term E[1_{Omega_{m+1}}||eps^{m+1}||^2]. Since Omega_{m+1} is a subset of Omega_m, that left-hand term is the smaller one, so the absorption goes the wrong way. The discrete Gronwall step that produces the O(k+h^2) rates therefore does not follow, and Corollary 3.1's convergence-in-probability is unsupported. This is an internal gap in the main claim, not a minor technicality.

The paper is not without merit. The splitting mixed method itself—time-lagged, no inf-sup condition, P1 elements for all unknowns—is a sensible extension of the earlier transport-noise work, and the decoupling genuinely helps Monte Carlo efficiency. The high-moment stability estimates in Lemma 2.1 and the Holder-type regularity in Lemma 2.2 look plausible and may be correct. The numerical experiments follow the expected rates and the qualitative behavior of the logistic damping is consistent with the analytical theory. Credit where due: the scheme is well motivated and the stability part is largely coherent.

Two further soft spots, in proportion. First, the theorem assumes the variational solution has H^2/H^1 regularity and Holder continuity; Remark 2.1 concedes existence of such a solution is not proved. The error analysis could be vacuous if that regularity cannot be obtained from the mild solution theory. Second, no code or data are provided, and errors are measured against a reference solution, which is standard but not a manufactured test.

Who is this for? SPDE numerics researchers working on splitting schemes; the method is interesting and the failure mode is instructive. If the absorption step can be repaired—for instance by keeping the same indicator on both sides or restructuring the localization—the paper could be a solid contribution.

My recommendation: send it to peer review. The flaw is central but identifiable, and the method and stability analysis deserve a serious referee. The reviewer should be asked to verify the absorption step and to press the authors on the regularity assumptions. If the main theorem cannot be fixed, the paper would need substantial revision before publication.

Referee Report

2 major / 5 minor

Summary. The paper proposes a splitting mixed finite element method for a stochastic Keller–Segel system with logistic growth and multiplicative noise. The method introduces an auxiliary variable representing the chemical gradient, uses a time-lagged splitting to decouple the original coupled system, and employs continuous piecewise linear finite elements for all unknowns, thereby avoiding the LBB condition. The authors claim optimal strong error estimates of order O(ln(1/k)(k+h^2)) and convergence in probability, supported by stability lemmas and numerical experiments.

Significance. If the main error estimate were valid, the paper would be a useful contribution as the first rigorous numerical analysis for stochastic Keller–Segel equations with multiplicative noise, and the splitting approach is computationally attractive and well motivated. The stability analysis in Lemma 3.1 and the qualitative numerical experiments are coherent and informative. However, the central convergence theorem contains a critical gap in the absorption step, so the advertised error estimates are not established.

major comments (2)

[Section 3.2, between (3.37) and (3.38)] The absorption step that removes the term C k (1+sqrt(rho)+rho) E[1_{Omega_{rho,m}} ||eps_u^{m+1}||^2_{L2}] from the right-hand side of (3.37) is invalid. By (3.15), Omega_{rho,m+1} is a subset of Omega_{rho,m}, so 1_{Omega_{rho,m+1}} <= 1_{Omega_{rho,m}} and therefore E[1_{Omega_{rho,m+1}} X] <= E[1_{Omega_{rho,m}} X] for every nonnegative random variable X. The quantity being absorbed is thus larger than the only L2 term on the left-hand side at level m+1, and the k-weighted gradient term on the left uses the indicator 1_{Omega_{rho,m}} and cannot control the L2 error at time m+1. Consequently, (3.38) does not follow from (3.37), and the discrete Gronwall step leading to (3.40)–(3.41) is unjustified. Theorem 3.1 and Corollary 3.1 are unsupported.
[Section 2.2, Remark 2.1, and Theorem 3.1] The error analysis is conditional on existence and regularity of a variational solution that the paper does not establish. Remark 2.1 explicitly states that the authors “do not establish the existence and uniqueness of variational (weak) solutions,” and the claim that the mild solution of [4, Theorem 2.1] becomes a variational solution is only sketched with a reference to [5, Theorem 6.5]. The higher regularity assumed in Theorem 3.1 (u in L^2(Omega; L^infty(0,T;H^2(D))) intersect L^6(Omega; L^infty(0,T;H^1(D))), together with the Holder estimates in Lemma 2.2) is not derived from the cited well-posedness results. Since the error in Theorem 3.1 is measured against this hypothetical solution, the announced “first rigorous error analysis” is not self-contained; a proof of the regularity assumptions, or an explicit statement that the theorem is conditional on an unproved assumption, is needed.

minor comments (5)

[Section 3.2, Eq. (3.20)] The factor nu in the first right-hand side term of (3.20) is undefined; the term should presumably be the integral of (grad u(t_{m+1}) - grad u(s), grad phi_h) without the multiplier nu.
[Lemma 2.2 and Section 3.2] The inequality “1 >= p < infinity” appears where “1 <= p < infinity” is meant; the direction of the inequality is incorrect.
[Title and Section 4] The title contains “MUL TIPLICA TIVE NOISE” and Section 4 uses “T est 1,” “T est 2,” “T est 3”; these spacing errors should be corrected.
[Remark 3.1 and Corollary 3.1] The localization parameter rho is fixed as rho = ln ln(1/k)/bC in Remark 3.1, but Corollary 3.1 says “provided that the localization parameter rho is chosen sufficiently large” and the proof later “lets rho tend to infinity”; since rho is already a function of k, the intended meaning is that rho tends to infinity as k tends to 0, and this should be stated more clearly.
[Section 3.2, Eq. (3.41)] The passage from E^{C(1+sqrt(rho)+rho)T} with rho = ln ln(1/k) to the final factor C ln(1/k) is not immediate; the exponential yields a power of ln(1/k) whose exponent depends on the constant bC. The paper should specify bC and justify the claimed logarithmic factor.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main error estimates are derived from the stated variational formulation and discrete scheme without reducing to the inputs; the noted proof and regularity gaps are correctness issues, not circularity.

full rationale

The derivation of Theorem 3.1 is not circular. The error equations (3.20)-(3.21) are obtained by subtracting the fully discrete splitting scheme (3.6)-(3.8) from the continuous variational formulation (2.25), and the subsequent estimates bound each term using the stated regularity assumptions, the projection estimates (3.5), Lemma 2.1, Lemma 2.2, and the Lipschitz and linear-growth assumptions on B. No parameter is fitted to the data whose error is later reported; the localization sequence Omega_{rho,m} in (3.15) is used as a proof device, and its probability tending to one is justified by Markov's inequality from Lemmas 2.1 and 3.1 rather than from Theorem 3.1. References to the same author's earlier work [24] are for the time-lagged splitting idea and for computational practice, not for any theorem on which the main error bound depends, so they are not load-bearing circularity. Two non-circular weaknesses should be weighed separately: Remark 2.1 concedes that the existence of a variational solution satisfying the required regularity is not proved in this paper and is only asserted through [4, Theorem 2.1] and [5, Theorem 6.5], and the absorption step between (3.37) and (3.38) is suspect because Omega_{rho,m+1} subset Omega_{rho,m} makes the right-hand indicator the larger one; however, these concern completeness or correctness of the proof, not equivalence of the claimed result to its inputs.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The proof rests on standard SPDE regularity, finite element interpolation, and elliptic estimates, plus an unproven weak-solution regularity assumption. The localization parameter rho is a proof device, not a physical constant. No new entities are introduced.

free parameters (1)

localization radius rho(k) = rho(k) = ln(ln(1/k))/bC, with bC implicit in the proof
Chosen in Remark 3.1 to define the nested good sets Omega_{rho,m}; bC is said to be determined in the proof of Theorem 3.1 and the final error constant depends on it. Not fitted to data, but a proof device that affects the stated rate.

assumptions (5)

domain assumption Existence and uniqueness of a variational (weak) solution (u,v) of (2.8) with sufficient regularity.
Remark 2.1 does not prove this; it asserts the mild solution of [4, Theorem 2.1] can be converted to a weak solution using [5, Theorem 6.5], and the error analysis then assumes u in L^2(Omega;L^infty(0,T;H^2)) intersect L^6(Omega;L^infty(0,T;H^1)).
domain assumption Noise coefficient B maps H^m to H^m and satisfies Lipschitz and linear growth bounds (2.4)-(2.5) for m=0,1.
Used throughout Lemma 2.1, Lemma 2.2, Lemma 3.1 and Theorem 3.1; standard for SPDE numerics but not verified for any specific B beyond the prototype B(u)=u/2 in the experiments.
standard math Elliptic regularity for the Neumann problem -Delta v + v = u: ||v||_{W^{2,r}} <= C||u||_{L^r} and ||v||_infinity + ||grad v||_infinity <= C||u||_{L^r} in 2D.
Invoked via (2.2)-(2.3) in Lemma 2.1 and Lemma 2.2; valid on smooth bounded domains.
standard math Ladyzhenskaya inequality in 2D and Sobolev embedding H^1 into L^q for all q less than infinity.
Used throughout the stability and error analysis to control products of functions in L^4 and H^1.
standard math The projection estimates (3.5) for the operators P_u, P_sigma and P_v on quasi-uniform P1 meshes.
Requires H^2 regularity of the projected functions; standard finite element approximation theory, but the H^2 regularity of the stochastic solution is itself an assumption.

how reviews work

0 comments

Cite this review

Pith. "Pith review of A splitting mixed finite element method for a stochastic Keller-Segel system with multiplicative noise." pith.science (2026). https://pith.science/paper/XR2SGV6K

@misc{pith2026260807798,
  author       = {Pith},
  title        = {Pith review of: A splitting mixed finite element method for a stochastic Keller-Segel system with multiplicative noise},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XR2SGV6K}},
  note         = {Machine review of arXiv:2608.07798}
}

read the original abstract

In this paper, we propose and analyze a splitting mixed finite element method for a stochastic Keller--Segel system with logistic growth driven by multiplicative noise. By introducing an auxiliary variable representing the chemical gradient together with a time-lagged splitting strategy, the proposed method decouples the original coupled system into a sequence of simpler subproblems. Consequently, it eliminates the Ladyzhenskaya--Babu\v{s}ka--Brezzi stability constraint, permits the use of continuous piecewise linear finite element spaces for all unknowns, and avoids solving a fully coupled nonlinear system at each time step, thereby significantly reducing the computational cost. Combined with an implicit Euler time discretization, the proposed approach yields a fully discrete numerical scheme for the stochastic Keller--Segel system. Using a localization technique together with suitable stochastic stability arguments, we establish optimal strong error estimates for the fully discrete approximations and prove convergence in probability with explicit convergence rates. Numerical experiments verify the theoretical convergence rates and demonstrate that the proposed method successfully captures the global boundedness induced by the logistic growth term as well as the influence of multiplicative noise on chemotactic aggregation.

Figures

Figures reproduced from arXiv: 2608.07798 by the authors.

**Figure 1.** Evolution of the expected numerical solution E[uM h ] for the stochastic Keller–Segel system with B(u) = 2u at the final times T = 0.005, 0.01, 0.02, and 0.05. The expected solution remains bounded and its peak value decreases monotonically over time. 5. Conclusion In this paper, we developed and analyzed a splitting mixed finite element method for a stochastic Keller–Segel system with logistic growth driven by mul… view at source ↗

**Figure 2.** Evolution of the deterministic numerical solution obtained by setting B(u) = 0 at the final times T = 0.005, 0.01, 0.02, and 0.05. Similar to the expected stochastic solution, the peak value decreases monotonically as time increases [PITH_FULL_IMAGE:figures/full_fig_p029_2.png] view at source ↗

**Figure 3.** Evolution of the pathwise numerical solution corresponding to the first realization (Sample 1) at the final times T = 0.005, 0.01, 0.02, and 0.05. 29 [PITH_FULL_IMAGE:figures/full_fig_p029_3.png] view at source ↗

Figures from the paper (2 more)

**Figure 4.** Figure 4: Evolution of the pathwise numerical solution corresponding to the second realization (Sample 2) at the final times T = 0.005, 0.01, 0.02, and 0.05 [PITH_FULL_IMAGE:figures/full_fig_p030_4.png]

**Figure 5.** Figure 5: Evolution of the pathwise numerical solution corresponding to the third realization (Sample 3) at the final times T = 0.005, 0.01, 0.02, and 0.05. 30 [PITH_FULL_IMAGE:figures/full_fig_p030_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 25 canonical work pages

[1]

Amrouche and N

C. Amrouche and N. E. H. Seloula. Lp-theory for vector potentials and Sobolev’s inequalities for vector fields: application to the Stokes equations with pressure boundary conditions.Mathematical Models and Methods in Applied Sciences, 23(01):37–92, 2013

work page 2013
[2]

X. Cao. Boundedness in a quasilinear parabolic–parabolic Keller–Segel system with logistic source.Journal of Mathematical Analysis and Applications, 412(1):181–188, 2014

work page 2014
[3]

Carelli and A

E. Carelli and A. Prohl. Rates of convergence for discretizations of the stochastic incompressible Navier–Stokes equations.SIAM Journal on Numerical Analysis, 50(5):2467–2496, 2012

work page 2012
[4]

Y. Chen, J. Zhai, and T. Zhang. Well-posedness of stochastic chemotaxis system.Journal of Differential Equa- tions, 442:113531, 2025

work page 2025
[5]

Da Prato and J

G. Da Prato and J. Zabczyk.Stochastic equations in infinite dimensions. Cambridge university press, 2014

work page 2014
[6]

Duarte-Rodr ´ ıguez, M.´A

A. Duarte-Rodr ´ ıguez, M.´A. Rodr ´ ıguez-Bellido, D. A. Rueda-G´ omez, and´E. J. Villamizar-Roa. Numerical anal- ysis for a chemotaxis-Navier–Stokes system.ESAIM: Mathematical Modelling and Numerical Analysis, 55:S417– S445, 2021

work page 2021
[7]

L. C. Evans.Partial differential equations, volume 19. American mathematical society, 2022

2022
[8]

Feng and L

X. Feng and L. Vo. High moment and pathwise error estimates for fully discrete mixed finite element approxi- mattions of stochastic Navier-Stokes equations with additive noise.Communications in Computational Physics, 36(3):821–849, 2024

work page 2024

Show all 26 references

[9]

Hillen and K

T. Hillen and K. J. Painter. A user’s guide to PDE models for chemotaxis.Journal of mathematical biology, 58(1):183–217, 2009

2009
[10]

Huang and J

H. Huang and J. Qiu. The microscopic derivation and well-posedness of the stochastic Keller–Segel equation. Journal of nonlinear science, 31(1):6, 2021

2021
[11]

E. F. Keller and L. A. Segel. Initiation of slime mold aggregation viewed as an instability.Journal of Theoretical Biology, 26(3):399–415, 1970

1970
[12]

E. F. Keller and L. A. Segel. Model for chemotaxis.Journal of theoretical biology, 30(2):225–234, 1971

1971
[13]

Lankeit and M

J. Lankeit and M. Winkler. A generalized solution concept for the Keller–Segel system with logarithmic sen- sitivity: global solvability for large nonradial data.Nonlinear Differential Equations and Applications NoDEA, 24(4):49, 2017

2017
[14]

G. J. Lord, C. E. Powell, and T. Shardlow. An introduction to computational stochastic PDEs.Cambridge Texts Appl. Math, 50, 2014

2014
[15]

Mayorcas and M

A. Mayorcas and M. Tomaˇ sevi´ c. Blow-up for a stochastic model of chemotaxis driven by conservative noise on R2.Journal of Evolution Equations, 23(3):57, 2023

2023
[16]

Mimura and T

M. Mimura and T. Tsujikawa. Aggregating pattern dynamics in a chemotaxis model including growth.Physica A: Statistical Mechanics and its Applications, 230(3-4):499–543, 1996

1996
[17]

Misiats, O

O. Misiats, O. Stanzhytskyi, and I. Topaloglu. On global existence and blowup of solutions of stochastic Keller– Segel type equation.Nonlinear Differential Equations and Applications NoDEA, 29(1):3, 2022

2022
[18]

Mukherjee, E

D. Mukherjee, E. Hausenblas, and T. Tran. Uniqueness of the stochastic Keller–Segel model in one dimension. Stochastic Processes and their Applications, 200:104999, 2026

2026
[19]

H. D. Nguyen, T. Thieu, and L. Vo. Finite element approximations of the stochastic Benjamin-Bona-Mahony equation with multiplicative noise.arXiv preprint arXiv:2603.08496, 2026

2026
[20]

H. D. Nguyen and L. Vo. Fully discrete finite element methods for the stochastic Kuramoto-Sivashinsky equation with multiplicative noise.arXiv preprint arXiv:2510.03670, 2025

2025 arXiv
[21]

Osaki and A

K. Osaki and A. Yagi. Finite dimensional attractor for one-dimensional Keller-Segel equations.Funkcialaj ekvacioj serio internacia, 44(3):441–470, 2001

2001
[22]

J. I. Tello and M. Winkler. A chemotaxis system with logistic source.Communications in Partial Differential Equations, 32(6):849–877, 2007

2007
[23]

M. J. Tindall, P. K. Maini, S. L. Porter, and J. P. Armitage. Overview of mathematical approaches used to model bacterial chemotaxis II: bacterial populations.Bulletin of mathematical biology, 70(6):1570–1607, 2008

2008
[24]

L. Vo. Analysis of fully discrete Crank-Nicolson finite element methods for a stochastic Keller-Segel chemotaxis system with gradient-type multiplicative noise.arXiv preprint arXiv:2507.15103, 2025

2025 arXiv
[25]

L. Vo. A mixed finite element method for the stochastic Boussinesq equations with multiplicative noise.Com- munications in Computational Physics, 2026

2026
[26]

Zhang, J

J. Zhang, J. Zhu, and R. Zhang. Characteristic splitting mixed finite element analysis of Keller–Segel chemotaxis models.Applied Mathematics and Computation, 278:33–44, 2016. 31

2016

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Amrouche and N

C. Amrouche and N. E. H. Seloula. Lp-theory for vector potentials and Sobolev’s inequalities for vector fields: application to the Stokes equations with pressure boundary conditions.Mathematical Models and Methods in Applied Sciences, 23(01):37–92, 2013

work page 2013

[2] [2]

X. Cao. Boundedness in a quasilinear parabolic–parabolic Keller–Segel system with logistic source.Journal of Mathematical Analysis and Applications, 412(1):181–188, 2014

work page 2014

[3] [3]

Carelli and A

E. Carelli and A. Prohl. Rates of convergence for discretizations of the stochastic incompressible Navier–Stokes equations.SIAM Journal on Numerical Analysis, 50(5):2467–2496, 2012

work page 2012

[4] [4]

Y. Chen, J. Zhai, and T. Zhang. Well-posedness of stochastic chemotaxis system.Journal of Differential Equa- tions, 442:113531, 2025

work page 2025

[5] [5]

Da Prato and J

G. Da Prato and J. Zabczyk.Stochastic equations in infinite dimensions. Cambridge university press, 2014

work page 2014

[6] [6]

Duarte-Rodr ´ ıguez, M.´A

A. Duarte-Rodr ´ ıguez, M.´A. Rodr ´ ıguez-Bellido, D. A. Rueda-G´ omez, and´E. J. Villamizar-Roa. Numerical anal- ysis for a chemotaxis-Navier–Stokes system.ESAIM: Mathematical Modelling and Numerical Analysis, 55:S417– S445, 2021

work page 2021

[7] [7]

L. C. Evans.Partial differential equations, volume 19. American mathematical society, 2022

2022

[8] [8]

Feng and L

X. Feng and L. Vo. High moment and pathwise error estimates for fully discrete mixed finite element approxi- mattions of stochastic Navier-Stokes equations with additive noise.Communications in Computational Physics, 36(3):821–849, 2024

work page 2024