Wildfire in a Narrow Gully: A Geometric Reduction Approach

Enrico Valdinoci; Lorenzo De Gaspari; Serena Dipierro

arxiv: 2604.12192 · v1 · submitted 2026-04-14 · 🧮 math.AP

Wildfire in a Narrow Gully: A Geometric Reduction Approach

Lorenzo De Gaspari , Serena Dipierro , Enrico Valdinoci This is my paper

Pith reviewed 2026-05-10 16:17 UTC · model grok-4.3

classification 🧮 math.AP

keywords wildfire modeldimensional reductionnonlocal parabolic equationFermi coordinatesnarrow gullygeometric evolutionasymptotic analysisreflection technique

0 comments

The pith

In a narrow gully a wildfire model reduces to a geometric equation along its curved axis.

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

The paper studies a nonlocal parabolic equation modeling bushfire spread where fuel sits inside a gully and rocky hillsides impose insulating boundaries. As the gully width shrinks to zero the three-dimensional problem collapses to a one-dimensional evolution posed only along the gully's possibly curved centerline. The reduction keeps Dirichlet data at the gully ends while the side insulation disappears. This matters because it replaces a degenerating domain problem with a simpler geometric flow that still captures ignition thresholds and nonlocal interactions.

Core claim

In the narrow-gully limit the original nonlocal parabolic equation with Neumann conditions on the lateral boundaries and Dirichlet conditions at the terminals converges to a geometric equation along the gully axis equipped only with inner and outer Dirichlet data. The convergence is obtained by passing to Fermi coordinates that straighten the possibly curved geometry and by deriving uniform parabolic estimates via a reflection technique that remains valid while the domain collapses and the boundary conditions change.

What carries the argument

Fermi coordinates along the gully axis together with a reflection technique that produces uniform bounds for the nonlocal parabolic operator, allowing passage to the limit despite domain degeneration.

If this is right

Fire propagation in the reduced model is governed solely by the geometry of the centerline and the prescribed temperatures at the two ends.
No transverse Neumann condition survives in the limit, so the reduced problem is purely Dirichlet at the terminals.
The same limit procedure applies to non-straight gullies because Fermi coordinates adapt to curvature.
Ignition interactions encoded in the nonlocal kernel are preserved in the reduced equation after the limit is taken.

Where Pith is reading between the lines

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

The reflection method developed here could be tested on other parabolic equations with degenerating lateral boundaries, such as diffusion in thin tubes or cracks.
The reduced one-dimensional model offers a computationally cheaper way to simulate long-term fire spread along valleys once the centerline geometry is known.
Extensions to time-dependent or stochastic gully shapes would require checking whether the uniform bounds survive when the axis itself moves.

Load-bearing premise

The reflection technique and tailored parabolic estimates remain uniform even as the domain narrows and the lateral boundary conditions switch from Neumann to absent.

What would settle it

Numerical solutions of the full three-dimensional model computed for a sequence of successively narrower gullies that fail to approach the solution of the reduced one-dimensional geometric equation would falsify the reduction.

Figures

Figures reproduced from arXiv: 2604.12192 by Enrico Valdinoci, Lorenzo De Gaspari, Serena Dipierro.

**Figure 1.** Figure 1: North slope of Mann Gulch, August 6, 1949. Retrieval of victim’s bodies [USFS49]. 1.1. Documented history. Wildfires have a documented history of occurring in gullies, ravines, and canyons. These fires are particularly dangerous because of how fire and wind interact with the terrain. The specific configuration of a location can enhance natural phenomena, intensifying the fire’s behavior. For example, gu… view at source ↗

**Figure 2.** Figure 2: The Fourmile Canyon Fire had burned more than 6,000 acres in 2010. This image showcasing burn scars was taken by the Advanced Land Imager on NASA’s Earth Observing-1 [NAS10]. 1.2. Topographical scenario. We stress that the topographical scenario considered in this paper differs from the one that has already been extensively studied in the literature. Indeed, most existing studies focus on bushfire spread… view at source ↗

**Figure 3.** Figure 3: Smoke filled canyons, Arizona. The image represents the northern rim of the Grand Canyon in 2019. A wildfire burnt more than 19,000 acres. The image was taken almost a month after the initial incident, as the land was still burning, by an astronaut onboard the International Space Station [NAS19]. Similar patterns are observed in desert mountain environments such as the Sonoran Desert and the Flinders … view at source ↗

**Figure 4.** Figure 4: Sketch of the geometry of a gully in dimension n = 2. We denote with {κj (x)}j the principal curvatures of S at each point x ∈ S and we define (1.2) L0 := inf x∈S 1⩽j⩽n−1 1 |κj (x)| , that is, L0 is the smallest curvature radius of S, which may also be equal to +∞, in which case S is contained in a hyperplane. 1.4. Mathematical description of bushfire propagation in a gully. We now formulate an adaptation … view at source ↗

read the original abstract

We consider a bushfire model in a gully. The biological scenario under consideration involves flammable fuel (trees, leaves, etc.) concentrated within the gully, surrounded by rocky hillslopes containing little or no burnable material. The mathematical formulation of the problem is a nonlocal evolution equation of parabolic type. The nonlocality arises from an ignition mechanism that becomes active when the temperature reaches the ignition threshold and is modeled via a kernel interaction with limitrophe areas. The rocky hillsides of the gully impose insulating boundary conditions of Neumann type, while the entrance and exit of the gully are modeled by (not necessarily homogeneous) Dirichlet boundary data, corresponding to prescribed environmental temperatures on the gully's terminals. Given the geometry of the domain, in the asymptotic regime of a narrow gully the model undergoes a dimensional reduction and can be analyzed through a geometric equation posed along the (not necessarily straight) axis of the gully. The reduced equation is supplemented with inner and outer Dirichlet boundary conditions (with no Neumann condition remaining in the limit). The analysis relies on the use of Fermi coordinates to capture the potentially curvilinear geometry of the gully, as well as on parabolic estimates tailored to the specific equation in order to properly account for the ignition interactions. These estimates are delicate, as the domain degenerates and the boundary conditions vary in the limit. To overcome these difficulties, we develop a bespoke reflection technique that provides uniform bounds and enables the passage to the limit.

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

They reduce the nonlocal wildfire equation in a narrow gully to a 1D geometric model along the axis via Fermi coordinates and a reflection trick, but the uniform bounds on the ignition term as width vanishes need checking in the details.

read the letter

The paper's core move is to start from a nonlocal parabolic equation modeling ignition in a gully with Neumann sides and Dirichlet ends, then pass to the limit as the width shrinks to zero. The result is a reduced 1D equation posed on the (possibly curved) centerline, now with only inner and outer Dirichlet conditions and no remaining lateral boundary condition. Fermi coordinates handle the geometry, and a custom reflection is introduced to control the nonlocal kernel uniformly despite the degeneration. That reduction itself is the new piece; prior work on thin domains or wildfire models does not appear to have produced exactly this 1D limit with the ignition term preserved this way. The setup is clean and the motivation from the biological scenario is straightforward. The estimates are flagged as delicate, which matches the situation: the domain collapses, the boundary conditions change, and the kernel sees exterior regions that must be reflected consistently. The stress-test point about possible ε-dependent mismatch in how the reflected kernel interacts with limitrophe areas is worth watching. If the reflection is constructed so that the exterior contributions converge without extra error terms blowing up, the bounds go through; the abstract claims they do, but the passage to the limit will stand or fall on whether the paper supplies explicit control on those errors rather than just existence of the technique. Minor gaps in the sketch would be fixable in revision. This is aimed at people working on asymptotic analysis of nonlocal equations in degenerating domains or on geometric models for wildfire spread. A reader already comfortable with Fermi coordinates and parabolic estimates in thin geometries will see the value in the adaptation. The work is coherent on its own terms and shows honest engagement with the technical obstacles, so it deserves a serious referee who can check the reflection construction and the limit passage in detail. I would send it out for review rather than desk reject.

Referee Report

2 major / 2 minor

Summary. The paper considers a nonlocal parabolic PDE for wildfire spread in a narrow gully geometry, where fuel is concentrated inside the gully with Neumann conditions on the sides and Dirichlet data at the ends. It claims that as the gully width ε tends to zero, the model reduces via Fermi coordinates to a 1D geometric equation along the (possibly curved) axis, with the nonlocal ignition kernel preserved in the limit; the proof relies on tailored parabolic estimates and a bespoke reflection technique to obtain ε-uniform bounds despite domain degeneration and the change in lateral boundary conditions.

Significance. If the uniform bounds and passage to the limit are rigorously justified, the result supplies a mathematically controlled dimensional reduction for nonlocal ignition models in degenerate domains, which could inform simplified 1D simulations of fire propagation in confined geometries. The combination of Fermi coordinates with a reflection method adapted to the nonlocal term is a technical contribution that addresses a genuine analytic difficulty.

major comments (2)

[§4] §4 (Reflection technique and uniform estimates): the argument that the bespoke reflection produces ε-independent control on the nonlocal ignition integral (presumably the term involving the kernel acting on limitrophe regions) is not fully convincing, because the transverse collapse alters the measure of the reflected exterior; the error between the reflected kernel interaction and the true exterior interaction appears to retain an O(ε) or worse remainder that is not shown to vanish uniformly when the ignition threshold is crossed.
[Theorem 1.1] Theorem 1.1 (main reduction statement): the claimed strong convergence of the solution u_ε to the 1D limit u_0 in the appropriate topology is stated, but the proof sketch in §5 does not explicitly quantify how the changing boundary conditions (Neumann to no lateral condition) interact with the nonlocal term to preserve the ignition threshold; an additional compactness or monotonicity argument may be needed to close the limit.

minor comments (2)

[§2] The notation for the Fermi coordinate system (r,s) is introduced without an explicit diagram or reference to the standard definition of the signed distance function; adding a short figure would clarify the curvilinear axis.
[§3] Several estimates in §3 cite 'standard parabolic theory' without specifying the precise reference or the precise form of the maximum principle used for the nonlocal equation.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the detailed and constructive report. The comments on the reflection estimates and the limit passage are well-taken and point to places where additional explicit arguments will improve clarity. We address each major comment below and will revise the manuscript accordingly.

read point-by-point responses

Referee: [§4] §4 (Reflection technique and uniform estimates): the argument that the bespoke reflection produces ε-independent control on the nonlocal ignition integral (presumably the term involving the kernel acting on limitrophe regions) is not fully convincing, because the transverse collapse alters the measure of the reflected exterior; the error between the reflected kernel interaction and the true exterior interaction appears to retain an O(ε) or worse remainder that is not shown to vanish uniformly when the ignition threshold is crossed.

Authors: We appreciate the referee highlighting the need for a sharper error control in the reflected nonlocal term. The reflection is performed after rescaling the transverse coordinate by ε in Fermi coordinates, so that the measure discrepancy between the reflected exterior and the true exterior is O(ε) in the transverse direction. Because the ignition kernel is assumed integrable and Lipschitz continuous (as stated in the model hypotheses), this discrepancy produces an error bounded by Cε times the uniform L^∞ bound on the solution, which vanishes uniformly as ε→0 independently of whether the ignition threshold has been crossed. The current proof sketch in §4 uses this fact implicitly via the maximum principle, but we agree the estimate was not written out in full detail. In the revision we will insert a short lemma (new Lemma 4.3) that quantifies the O(ε) remainder and shows it tends to zero uniformly on compact time intervals, thereby confirming ε-independent control on the ignition integral. revision: yes
Referee: [Theorem 1.1] Theorem 1.1 (main reduction statement): the claimed strong convergence of the solution u_ε to the 1D limit u_0 in the appropriate topology is stated, but the proof sketch in §5 does not explicitly quantify how the changing boundary conditions (Neumann to no lateral condition) interact with the nonlocal term to preserve the ignition threshold; an additional compactness or monotonicity argument may be needed to close the limit.

Authors: The referee is correct that the interaction of the vanishing lateral Neumann conditions with the nonlocal ignition operator requires a more explicit justification to pass to the limit while preserving the threshold. In §5 we obtain uniform bounds and weak convergence from the estimates of §4, then identify the limit equation by testing against test functions supported away from the lateral boundaries. Because the kernel is integrable, the nonlocal term converges strongly in L^1_loc once the domain converges in the Hausdorff sense under the Fermi coordinate change; the ignition threshold is therefore inherited by the limit. Nevertheless, to make the strong convergence and threshold preservation fully rigorous, we will add a compactness argument in the revised §5 (using the uniform parabolic estimates to apply an Aubin-Lions-type lemma in the scaled variables) together with a monotonicity observation on the ignition function. This will close the limit without changing the statement of Theorem 1.1. revision: yes

Circularity Check

0 steps flagged

No circularity: reduction obtained via independent asymptotic analysis and estimates

full rationale

The derivation proceeds from the original nonlocal parabolic PDE by introducing Fermi coordinates to handle the curvilinear gully geometry, then establishing uniform bounds via tailored parabolic estimates and a bespoke reflection technique to pass to the limit as the transverse width ε→0. This yields the reduced 1D geometric equation along the axis with adjusted boundary conditions. None of the load-bearing steps reduce by definition, fitting, or self-citation to the target result; the estimates address the degeneration and nonlocality directly without presupposing the limit equation. The procedure is self-contained against external mathematical tools and does not rename or smuggle prior results.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard tools from parabolic PDE theory and differential geometry adapted to the narrow limit; no free parameters or new entities are introduced.

axioms (2)

standard math Standard parabolic regularity estimates apply to the nonlocal ignition equation
Invoked to obtain uniform bounds before taking the narrow-gully limit.
domain assumption Fermi coordinates can be introduced along the (possibly curved) gully axis
Used to capture the geometry in the asymptotic reduction.

pith-pipeline@v0.9.0 · 5566 in / 1220 out tokens · 61085 ms · 2026-05-10T16:17:32.449790+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[1]

Image ID: Austock000221021

Location: Dimboola, Wimmera, Victoria. Image ID: Austock000221021. Accessed: 27 Feb 2026. ©Original copyright holder. Used only as an external reference.https://www.austockphoto.com.au/ image/aerial-view-of-a-green-creek-bed-winding-through-a-JtvYz. [Cha23c] ,Aerial view of vegetation in dry creek beds through in an arid outback landscape, 2023. Location:...

work page 2026
[2]

Reprinted in: E

MR4968074 [Fer22] Enrico Fermi,Sopra i fenomeni che avvengono in vicinanza di una linea oraria, Rendiconti dell’Accademia Nazionale dei Lincei, Classe di Scienze Fisiche, Matematiche e Naturali31(1922), 21–23. Reprinted in: E. Fermi,Collected Papers, vol. 1, University of Chicago Press (1962). [Gau65] Carl Friedrich Gauss,General investigations of curved ...

work page 1922

[1] [1]

Image ID: Austock000221021

Location: Dimboola, Wimmera, Victoria. Image ID: Austock000221021. Accessed: 27 Feb 2026. ©Original copyright holder. Used only as an external reference.https://www.austockphoto.com.au/ image/aerial-view-of-a-green-creek-bed-winding-through-a-JtvYz. [Cha23c] ,Aerial view of vegetation in dry creek beds through in an arid outback landscape, 2023. Location:...

work page 2026

[2] [2]

Reprinted in: E

MR4968074 [Fer22] Enrico Fermi,Sopra i fenomeni che avvengono in vicinanza di una linea oraria, Rendiconti dell’Accademia Nazionale dei Lincei, Classe di Scienze Fisiche, Matematiche e Naturali31(1922), 21–23. Reprinted in: E. Fermi,Collected Papers, vol. 1, University of Chicago Press (1962). [Gau65] Carl Friedrich Gauss,General investigations of curved ...

work page 1922