Pith. sign in

REVIEW 2 major objections 35 references

Optimal mean-time path planning for unmanned underwater vehicles: a Hamilton-Jacobi approach

T0 review · 2 major / 0 minor · reviewed 2026-07-12 · grok-4.5

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

Pith's one-line read A coupled Hamilton-Jacobi system yields the underwater path that minimizes mean travel time across an ensemble of disagreeing ocean forecasts.

desk verdict Clean, usable extension of deterministic HJ path planning to probability-weighted ensembles; the coupled system and alternating fast-sweeping scheme are the real additions, and the numerics show the mean path can leave every single-member optimum. read the letter →

arxiv 2607.03407 v1 pith:5P3SH3OC submitted 2026-07-03 math.OC

Jonathan Valyou , Jeremy Brandman , Sanghyun Lee This is my paper · ORCID
classification math.OC MSC 49L2035F2165N06
keywords Hamilton-Jacobiequationsoptimalcontrolpathplanningunmannedunderwatervehiclesensembleforecastsfastsweepingmethodmeanreachabilitytime
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

Unmanned underwater vehicles usually plan paths from a single ocean-current forecast, but real forecasts disagree. This paper shows how to replace that single forecast with a weighted ensemble and still obtain a globally optimal path: the path that minimizes average travel time. The average travel time and the travel times under each individual forecast satisfy a closed system of static Hamilton-Jacobi equations; the Hamiltonian minimizer supplies the heading the vehicle should follow. An extension of the fast-sweeping method solves the whole system efficiently. Numerical examples confirm that the resulting mean-time path can look quite different from every path that would be optimal for any single ensemble member, giving planners a practical way to hedge forecast uncertainty.

What carries the argument

The coupled static Hamilton-Jacobi system (Proposition 1) that links the mean travel-time function to the individual travel times through a probability-weighted Hamiltonian; the Single Sweep Set Alternating Lax-Friedrichs Fast Sweeping scheme then updates every unknown inside each sweep until the whole system converges.

What would settle it

On a two-member ensemble whose members are identical linear-in-time currents, the computed mean path and mean time must recover the known semi-analytic deterministic solution (and the classical single-equation fast-sweeping solution) to within discretization error; any systematic mismatch falsifies the claim.

Watch

Extended reading notes

Core claim

The minimum mean reachability time u and the individual reachability times T_i under an ensemble of ocean models jointly satisfy a system of time-independent Hamilton-Jacobi equations. The Hamiltonian minimizer of that system is the optimal control; integrating the control produces a path whose expected travel time is minimal, and that path can deviate substantially from every deterministic optimum associated with a single forecast.

Load-bearing premise

Every ocean forecast must keep current speed strictly below the vehicle’s top speed everywhere, so that every point remains reachable under every model; if any forecast violates the bound the equations become degenerate and a mean-time path is no longer guaranteed.

Editorial extensions

If this is right

  • Mission planners obtain one path that is optimal in the average sense without running separate deterministic optimizations and then combining them by hand.
  • When ensemble members differ strongly, the mean-optimal route need not resemble any individual deterministic optimum, so ignoring uncertainty can produce systematically longer expected transit times.
  • Placing all probability mass on a single forecast recovers the classical deterministic Hamilton-Jacobi path-planning equations exactly.
  • The alternating fast-sweeping extension makes the coupled system computationally practical for realistic two-dimensional domains.

Reading between the lines

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

  • The same construction extends immediately to energy-optimal planning by changing the running cost inside the Hamiltonian while leaving the sweeping scheme unchanged.
  • Hard obstacles can be inserted by simple domain and boundary-condition modifications; the PDE system itself needs no redesign.
  • A quantitative link between the spread among the individual T_i and the geometric deviation of the mean path would give a practical diagnostic for when ensemble planning is essential.
  • Existence and uniqueness of viscosity solutions for the coupled system remains open and would underwrite convergence of the sweeping algorithm.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 0 minor

Summary. The paper extends deterministic Hamilton–Jacobi path planning for UUVs to an ensemble of ocean-current forecasts. Using the dynamic programming principle, it derives a coupled system of static HJ equations (Proposition 1, eqs. 14–16) for the probability-weighted mean reachability time u and the individual ensemble travel times T_i; the Hamiltonian minimizer θ* supplies the feedback direction used to reconstruct the mean-optimal path by backward ODE integration. The authors generalize Lax–Friedrichs fast sweeping to this system via a Single Sweep Set Alternating scheme (SSSA-LFFS), verify near-linear convergence against a semi-analytic linear-in-time current, benchmark efficiency against a sweep-until-convergence alternative, and present 2D examples (including vortices and a double-gyre) showing that the mean-optimal path can deviate substantially from every single-member deterministic optimum.

Significance. If the derivation and numerics hold, the work supplies a practical, risk-neutral ensemble path planner that needs only forecast members and likelihoods, recovers the deterministic theory of Brandman & Olson as a special case, and demonstrates that mean-optimal routes need not interpolate individual optima when ensemble members disagree strongly. Strengths include a carefully written DP derivation, an explicit reduction check (p(1)=1), a reproducible semi-analytic verification with reported L^∞ rates, a clear efficiency comparison of two sweeping strategies (Tables 1–2), and illustrative examples that falsifiably show path deviation under uncertainty. The open viscosity theory for the system and the risk-neutral (not robust) objective are real limitations, but they are acknowledged and do not erase the algorithmic contribution for applied optimal control and ocean robotics.

major comments (2)
  1. Remark 2 (after Proposition 1) asserts that replacing the distinguished index “1” in (14) by any other ensemble index j yields PDE systems that are only “approximately equivalent” in numerical practice. Because the continuous formulation then appears to depend on an arbitrary labeling of ensemble members, this is load-bearing for well-posedness of the claimed system. Please either (i) prove that the continuous system is independent of the choice of distinguished index, or (ii) report quantitative comparisons (e.g., ||u^(j) - u^(k)|| and path Hausdorff distances) across all labelings for the examples in §4, and state clearly which labeling is used in each figure.
  2. Section 2.5 / eqs. (7), (14)–(15): for time-dependent currents, s_i,max depends on the unknown arrival time T_i(x), so the coefficients of the HJ system are themselves solution-dependent. The DP derivation treats this formally, but the paper never states the precise function space or fixed-point structure in which (14)–(16) is to be understood, nor any comparison/monotonicity property that would support uniqueness of the viscosity solution of the coupled system. A short well-posedness discussion (even partial: e.g., continuous dependence for frozen T_i, or a contraction argument under small time-dependence) is needed to underwrite the claim that the numerical solution approximates “the” mean reachability time.

Circularity Check

1 steps flagged · score 1.0 of 10

No significant circularity: mean-time HJ system is derived from an independent definition of mean reachability time via the dynamic-programming principle; the only mild self-reference is recovery of the deterministic special case from prior work by an overlapping author.

  1. self citation load bearing [Remark 1 (after Proposition 1) and Introduction]
    "Notice that when p(1)=1 and p(j)=0 for j=2,3,...,n, this system of Hamilton-Jacobi PDEs simplifies to the deterministic reachability time PDE as presented in [1]."

    The deterministic baseline is recovered as a special case of the new system and is cited from prior work by an overlapping author. The citation is not used to justify the multi-member derivation itself (which proceeds from the DPP applied to the independently defined mean), so the circularity is only mild and non-load-bearing.

full rationale

The mean reachability time is introduced by definition as the probability-weighted sum of individual transit times (eq. 5) and the value function u is the pointwise minimum of that quantity over paths (eq. 6). Proposition 1 then obtains the coupled static HJ system (14)–(16) by a standard dynamic-programming argument (Taylor expansion of the DPP, optimality of the maximal admissible speed, and the same argument applied to each T_i). The derivation does not presuppose the PDE system, nor does it fit any free parameter that is later re-labeled a prediction. When all probability mass is placed on a single ensemble member the system reduces exactly to the deterministic reachability equation of the authors’ earlier work [1]; that reduction is a consistency check, not a load-bearing premise. The numerical method is an extension of the publicly available Lax–Friedrichs Fast Sweeping scheme and is validated against a semi-analytic linear-current solution that is independent of the present paper. Consequently the central claim is self-contained; the single self-citation is non-circular and does not force the result.

Assumptions & free parameters 3 free parameters · 4 assumptions · 2 invented entities

The central claim rests on the classical viscosity-solution framework for Hamilton-Jacobi equations, the dynamic-programming principle, a uniform reachability bound that keeps the Hamiltonian non-degenerate, and the modeling choice that the objective is the probability-weighted mean (risk-neutral) rather than a robust or risk-sensitive criterion. Artificial viscosity and mesh parameters are numerical free parameters required for the discrete scheme but do not enter the continuous claim.

free parameters (3)
  • artificial viscosity η = (η_x1, η_x2)
    Chosen by hand (values 1.5 or 1.75) to stabilize the Lax-Friedrichs scheme; magnitude is O(1) while the solution is O(10^3). Affects numerical accuracy but not the continuous PDE claim.
  • ensemble probabilities p(i)
    Treated as known inputs supplied by the user; different choices produce different mean-optimal paths (Example 3a). Not fitted inside the paper.
  • vehicle max speed s_max
    Fixed to 1 m/s in all experiments; enters the definition of s_i,max and the reachability assumption.
assumptions (4)
  • standard math Dynamic programming principle for the mean reachability time
    Used in the derivation of Proposition 1 (Section 2.5) to obtain the coupled HJ system; standard in optimal-control theory.
  • standard math Viscosity-solution framework for Hamilton-Jacobi equations (vanishing viscosity)
    Invoked to justify uniqueness of the continuous solution and the design of the Lax-Friedrichs scheme (Section 3.1).
  • domain assumption max |v_c,i| < s_max for every ensemble member (reachability)
    Stated in (2),(4),(12)–(13); guarantees that every point is reachable and that s_i,max > 0, preventing degeneracy of the Hamiltonian.
  • ad hoc to paper Objective is the probability-weighted mean travel time (risk-neutral)
    Explicitly chosen in Section 1 and eq. (5); the authors note that the method does not address worst-case (robust) behavior.
invented entities (2)
  • coupled system of static Hamilton-Jacobi PDEs for mean reachability time
    purpose: Encodes the probability-weighted mean travel time and the individual ensemble travel times in a single time-independent PDE system whose solution yields both the value function and the optimal control.
    Derived in Proposition 1; reduces to the known deterministic eikonal-type equation when one probability equals 1. No independent experimental handle outside the numerical examples of the paper.
  • Single-Sweep-Set Alternating Lax-Friedrichs Fast Sweeping (SSSA-LFFS)
    purpose: Efficient numerical solver for the coupled HJ system that updates every unknown once per sweep set.
    Introduced in Algorithm 3 and shown empirically faster than the naïve sweep-until-convergence alternative; purely algorithmic construct.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Optimal mean-time path planning for unmanned underwater vehicles: a Hamilton-Jacobi approach." pith.science (2026). https://pith.science/paper/5P3SH3OC

@misc{pith2026260703407,
  author       = {Pith},
  title        = {Pith review of: Optimal mean-time path planning for unmanned underwater vehicles: a Hamilton-Jacobi approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5P3SH3OC}},
  note         = {Machine review of arXiv:2607.03407}
}
read the original abstract

Unmanned underwater vehicles (UUV) integrate ocean forecasts with path planning algorithms in order to identify energy- or time-minimizing paths that enable mission completion. Typically, a well-defined deterministic ocean forecast is assumed to be available for path planning; however, in practice, different ocean forecasts can disagree. In this paper, we extend previous work on deterministic optimal path planning to identify optimal mean-time paths when presented with an ensemble of possible ocean forecasts. In particular, we formulate a system of time-independent Hamilton-Jacobi partial differential equations that incorporates forecast uncertainty and yields the optimal mean reachability travel time and the necessary controls to find the associated optimal path. An efficient numerical solution of this system of PDEs is obtained through an extension of the Fast Sweeping Method; verification and benchmarking results are provided. Additional numerical examples illustrate the impact uncertainty can have on the optimal path; in particular, these results demonstrate that the vehicle's optimal path can deviate significantly from the deterministic optimal paths associated with the individual ensemble members.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 11 canonical work pages

  1. [1]

    In: OCEANS 2023-MTS/IEEE US Gulf Coast, pp

    Brandman, J., Olson, C.: Globally time-optimal path planning for unmanned underwater vehicles in three-dimensional current fields using Hamilton-Jacobi partial differential equations. In: OCEANS 2023-MTS/IEEE US Gulf Coast, pp. 1–10 (2023). https://doi.org/10.23919/OCEANS52994.2023.10337134 . IEEE

    doi:10.23919/oceans52994.2023.10337134 2023

  2. [2]

    Journal of Computational Physics196(1), 367–391 (2004) https://doi.org/10.1016/j.jcp.2003.11.007

    Kao, C.Y., Osher, S., Qian, J.: Lax–Friedrichs sweeping scheme for static Hamil- ton–Jacobi equations. Journal of Computational Physics196(1), 367–391 (2004) https://doi.org/10.1016/j.jcp.2003.11.007

    work page doi:10.1016/j.jcp.2003.11.007 2004

  3. [3]

    Journal of Computational Physics217(1), 176–199 (2006) https: //doi.org/10.1016/j.jcp.2006.02.010

    Lermusiaux, P.F.J.: Uncertainty estimation and prediction for interdisciplinary ocean dynamics. Journal of Computational Physics217(1), 176–199 (2006) https: //doi.org/10.1016/j.jcp.2006.02.010 . Uncertainty Quantification in Simulation Science

    work page doi:10.1016/j.jcp.2006.02.010 2006

  4. [4]

    Princeton University Press, Princeton, NJ, USA (1957)

    Bellman, R.: Dynamic Programming. Princeton University Press, Princeton, NJ, USA (1957)

    1957

  5. [5]

    Stochastic Modelling and Applied Probability, vol

    Fleming, W.H., Soner, H.M.: Controlled Markov Processes and Viscosity Solu- tions. Stochastic Modelling and Applied Probability, vol. 25. Springer, New York (2006)

    2006

  6. [6]

    Cambridge University Press, Cambridge, UK (1999)

    Sethian, J.A.: Level Set Methods and Fast Marching Methods: Evolving Interfaces in Computational Geometry, Fluid Mechanics, Computer Vision, and Materials Science, 2nd edn. Cambridge University Press, Cambridge, UK (1999)

    1999

  7. [7]

    IEEE Trans- actions on Automatic Control40(9), 1528–1538 (1995) https://doi.org/10.1109/ 9.412626

    Tsitsiklis, J.N.: Efficient algorithms for globally optimal trajectories. IEEE Trans- actions on Automatic Control40(9), 1528–1538 (1995) https://doi.org/10.1109/ 9.412626

    1995

  8. [8]

    John Wiley & Sons, New York (1965)

    Isaacs, R.: Differential Games. John Wiley & Sons, New York (1965)

    1965

Show all 35 references
  1. [9]

    Springer, Boston,MA (1995)

    Ba¸ sar, T., Bernhard, P.: H-infinity Optimal Control and Related Minimax Design Problems: a Dynamic Game Approach. Springer, Boston,MA (1995)

    1995

  2. [10]

    Ocean Dynamics53(4), 343–367 (2003)

    Evensen, G.: The Ensemble Kalman Filter: Theoretical formulation and practical implementation. Ocean Dynamics53(4), 343–367 (2003)

    2003

  3. [11]

    IEEE Transactions on Automatic Control51(5), 742–753 (2006)

    Calafiore, G.C., Campi, M.C.: The scenario approach to robust control design. IEEE Transactions on Automatic Control51(5), 742–753 (2006)

    2006

  4. [12]

    IEEE Control Systems Magazine36(6), 30–44 (2016) https: //doi.org/10.1109/MCS.2016.2602087

    Mesbah, A.: Stochastic model predictive control: An overview and perspectives for future research. IEEE Control Systems Magazine36(6), 30–44 (2016) https: //doi.org/10.1109/MCS.2016.2602087

    2016 doi

  5. [13]

    34 2005, pp

    Garau, B., Alvarez, A., Oliver, G.: Path planning of autonomous underwater vehicles in current fields with complex spatial variability: an A* approach, vol. 34 2005, pp. 194–198 (2005). https://doi.org/10.1109/ROBOT.2005.1570118

    2005 doi

  6. [14]

    Ocean Dynamics64, 1373–1397 (2014) https://doi.org/10.1007/s10236-014-0757-y

    Lolla, T., Lermusiaux, P., Ueckermann, M., Haley, P.: Time-optimal path plan- ning in dynamic flows using level set equations: theory and schemes. Ocean Dynamics64, 1373–1397 (2014) https://doi.org/10.1007/s10236-014-0757-y

    2014 doi

  7. [15]

    IEEE Transactions on Robotics23(2), 331– 341 (2007)

    Petres, C., Pailhas, Y., Patron, P., Petillot, Y., Evans, J., Lane, D.: Path planning for autonomous underwater vehicles. IEEE Transactions on Robotics23(2), 331– 341 (2007)

    2007

  8. [16]

    In: 2020 59th IEEE Conference on Decision and Control (CDC), pp

    Parkinson, C., Bertozzi, A.L., Osher, S.J.: A Hamilton-Jacobi formulation for time-optimal paths of rectangular nonholonomic vehicles. In: 2020 59th IEEE Conference on Decision and Control (CDC), pp. 4073–4078 (2020). https://doi. org/10.1109/CDC42340.2020.9303861

    2020 doi

  9. [17]

    https://arxiv.org/abs/1302.4987

    Wellman, M.P., Ford, M., Larson, K.: Path Planning under Time-Dependent Uncertainty (2013). https://arxiv.org/abs/1302.4987

    2013 arXiv

  10. [18]

    In: Proceedings

    Rathbun, D., Kragelund, S., Pongpunwattana, A., Capozzi, B.: An evolution based path planning algorithm for autonomous motion of a uav through uncertain environments. In: Proceedings. The 21st Digital Avionics Systems Conference, vol. 2, pp. 8–282 (2002). https://doi.org/10.11...

    2002 doi arXiv

  11. [19]

    IEEE Transactions on Automatic Control50(7), 947–957 (2005)

    Mitchell, I.M., Bayen, A.M., Tomlin, C.J.: A time-dependent Hamilton–Jacobi formulation of reachable sets for continuous dynamic games. IEEE Transactions on Automatic Control50(7), 947–957 (2005)

    2005

  12. [20]

    Computer Methods in Applied Mechanics and Engineering333, 218–237 (2018)

    Subramani, D.N., Wei, Q.J., Lermusiaux, P.F.: Stochastic time-optimal path- planning in uncertain, strong, and dynamic flows. Computer Methods in Applied Mechanics and Engineering333, 218–237 (2018)

    2018

  13. [21]

    Computer Methods in Applied Mechanics and Engineer- ing353, 391–415 (2019) https://doi.org/10.1016/j.cma.2019.04.033

    Subramani, D.N., Lermusiaux, P.F.J.: Risk-optimal path planning in stochastic dynamic environments. Computer Methods in Applied Mechanics and Engineer- ing353, 391–415 (2019) https://doi.org/10.1016/j.cma.2019.04.033

    2019 doi

  14. [22]

    SIAM Journal on Numerical Analysis41(1), 325–363 (2003) https://doi.org/10.1137/S0036142901392742 https://doi.org/10.1137/S0036142901392742

    Sethian, J.A., Vladimirsky, A.: Ordered upwind methods for static Hamilton– Jacobi equations: Theory and algorithms. SIAM Journal on Numerical Analysis41(1), 325–363 (2003) https://doi.org/10.1137/S0036142901392742 https://doi.org/10.1137/S0036142901392742

    2003 doi

  15. [23]

    Zhao, H.: A fast sweeping method for eikonal equations. Math. Comput.74, 603–627 (2004)

    2004

  16. [24]

    Transactions of the American Mathematical Society277(1), 1–42 (1983)

    Crandall, M.G., Lions, P.-L.: Viscosity solutions of Hamilton-Jacobi equations. Transactions of the American Mathematical Society277(1), 1–42 (1983). Accessed 2025-04-21 35

    1983

  17. [25]

    Nonlinear Analysis: Theory, Methods & Applications 10(4), 353–370 (1986) https://doi.org/10.1016/0362-546X(86)90133-1

    Crandall, M.G., Lions, P.-L.: On existence and uniqueness of solutions of Hamilton-Jacobi equations. Nonlinear Analysis: Theory, Methods & Applications 10(4), 353–370 (1986) https://doi.org/10.1016/0362-546X(86)90133-1

    1986 doi

  18. [26]

    American Mathematical Society, Providence, R.I

    Evans, L.C.: Partial Differential Equations. American Mathematical Society, Providence, R.I. (2010)

    2010

  19. [27]

    Journal of Computational Physics234, 452–471 (2013) https://doi.org/10.1016/j.jcp.2012.10.008

    Chen, W., Chou, C.-S., Kao, C.-Y.: Lax–Friedrichs fast sweeping methods for steady state problems for hyperbolic conservation laws. Journal of Computational Physics234, 452–471 (2013) https://doi.org/10.1016/j.jcp.2012.10.008

    2013 doi

  20. [28]

    SIAM Journal on Numerical Analysis29(3), 867–884 (1992) https://doi.org/10.1137/ 0729053 https://doi.org/10.1137/0729053

    Rouy, E., Tourin, A.: A viscosity solutions approach to shape-from-shading. SIAM Journal on Numerical Analysis29(3), 867–884 (1992) https://doi.org/10.1137/ 0729053 https://doi.org/10.1137/0729053

    1992 doi

  21. [29]

    SIAM Journal on Numerical Analysis36(3), 667–695 (1999)

    Bou´ e, M., Dupuis, P.: Markov chain approximations for deterministic control problems with affine dynamics and quadratic cost in the control. SIAM Journal on Numerical Analysis36(3), 667–695 (1999). Accessed 2026-03-29

    1999

  22. [30]

    SIAM Journal on Numeri- cal Analysis41(2), 673–694 (2003) https://doi.org/10.1137/S0036142901396533 https://doi.org/10.1137/S0036142901396533

    Tsai, Y.-H.R., Cheng, L.-T., Osher, S., Zhao, H.-K.: Fast sweeping algo- rithms for a class of Hamilton–Jacobi equations. SIAM Journal on Numeri- cal Analysis41(2), 673–694 (2003) https://doi.org/10.1137/S0036142901396533 https://doi.org/10.1137/S0036142901396533

    2003 doi

  23. [31]

    SIAM Journal on Numerical Analysis 42(6), 2612–2632 (2005) https://doi.org/10.1137/S0036142902419600 https://doi.org/10.1137/S0036142902419600

    Kao, C.-Y., Osher, S., Tsai, Y.-H.: Fast sweeping methods for static hamilton–jacobi equations. SIAM Journal on Numerical Analysis 42(6), 2612–2632 (2005) https://doi.org/10.1137/S0036142902419600 https://doi.org/10.1137/S0036142902419600

    2005 doi

  24. [32]

    Fomel, S., Luo, S., Zhao, H.: Fast sweeping method for the factored eikonal equation. J. Comput. Physics228, 6440–6455 (2009) https://doi.org/10.1016/j. jcp.2009.05.029

    2009 doi

  25. [33]

    Communications on Applied Mathematics and Computation6(1), 3–29 (2022) https://doi.org/10.1007/ s42967-022-00209-x

    Miksis, Z.M., Zhang, Y.-T.: Sparse-grid implementation of fixed-point fast sweeping WENO schemes for eikonal equations. Communications on Applied Mathematics and Computation6(1), 3–29 (2022) https://doi.org/10.1007/ s42967-022-00209-x

    2022

  26. [34]

    Physica D: Nonlinear Phenomena212(3–4), 271–304 (2005)

    Shadden, S.C., Lekien, F., Marsden, J.E.: Definition and properties of Lagrangian coherent structures from finite-time Lyapunov exponents in two-dimensional aperiodic flows. Physica D: Nonlinear Phenomena212(3–4), 271–304 (2005)

    2005

  27. [35]

    Ocean Dynamics66(10), 1231–1251 (2016) https: //doi.org/10.1007/s10236-016-0982-8 36

    Wang, T., Le Maˆ ıtre, O.P., Hoteit, I.,et al.: Path planning in uncertain flow fields using ensemble method. Ocean Dynamics66(10), 1231–1251 (2016) https: //doi.org/10.1007/s10236-016-0982-8 36

    2016 doi

Pith tools

Reviewed July 12, 2026 · model on record in the stance chip above.