pith. sign in

arxiv: 2504.16945 · v2 · submitted 2025-04-17 · ⚛️ physics.soc-ph · cs.SI

Graph Percolation as Decision Threshold for Risk Management in Cross-Country Thermal Soaring

John J. Bird This is my paper

Pith reviewed 2026-05-22 20:05 UTC · model grok-4.3

classification ⚛️ physics.soc-ph cs.SI
keywords graph percolationthermal soaringrisk managementupdraftsglide reachabilityflight logsdecision threshold
0
0 comments X

The pith

Graph percolation sets the threshold pilots use to decide when they must exploit a thermal to stay aloft.

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

The paper claims that the risk of failing to locate the next updraft during unpowered cross-country flight can be modeled by treating observed thermals as nodes in a graph and possible glides between them as edges. Percolation on this graph marks the critical connectivity point beyond which a pilot can expect to reach another updraft and continue flight indefinitely. Examination of actual pilot flight logs shows that experienced soarers almost never operate below this percolation threshold and will accept slower climbs to keep the graph connected. The result supplies a concrete, observation-based rule for balancing speed against the danger of being stranded.

Core claim

When thermal locations and their glide-reachability relations are cast as a graph, the percolation threshold of that graph corresponds to the minimum configuration allowing continued safe flight; flight-log analysis confirms that pilots maintain conditions above this threshold, trading climb rate for connectivity when needed.

What carries the argument

Graph percolation on a network whose nodes are in-flight thermal observations and whose edges connect thermals reachable by glide from one another; the threshold determines whether an infinite connected path remains available.

Load-bearing premise

A pilot's ability to continue safe flight is directly given by whether the observed thermal graph lies above its percolation threshold.

What would settle it

A verified flight log in which a pilot completes a long cross-country leg while the real-time thermal graph remains below the percolation threshold would refute the correspondence.

Figures

Figures reproduced from arXiv: 2504.16945 by John J. Bird.

Figure 1
Figure 1. Figure 1: Cross-country soaring requires exploiting a sequen [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: A collection of thermals which forms a random geometr [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 5
Figure 5. Figure 5: Normalized thermal strength distribution vs percol [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Percolation histogram and time history of altitude fo [PITH_FULL_IMAGE:figures/full_fig_p004_6.png] view at source ↗
Figure 8
Figure 8. Figure 8: Normalized thermal strength distribution vs percol [PITH_FULL_IMAGE:figures/full_fig_p005_8.png] view at source ↗
read the original abstract

Long range flight by fixed-wing aircraft without propulsion systems can be accomplished by "soaring" -- exploiting randomly located updrafts to gain altitude which is expended in gliding flight. As the location of updrafts is uncertain and cannot be determined except through in situ observation, aircraft exploiting this energy source are at risk of failing to find a subsequent updraft. Determining when an updraft must be exploited to continue flight is essential to managing risk and optimizing speed. Graph percolation offers a theoretical explanation for this risk, and a framework for evaluating it using information available to the operator of a soaring aircraft in flight. The utility of graph percolation as a risk measure is examined by analyzing flight logs from human soaring pilots. This analysis indicates that in sport soaring pilots rarely operate in a condition which does not satisfy graph percolation, identifies an apparent desired minimum node degree, and shows that pilots accept reduced climb rates in order to maintain percolation.

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.

Referee Report

3 major / 2 minor

Summary. The paper claims that graph percolation on a network of thermal updrafts (nodes) connected by glide-reachability edges provides a theoretical decision threshold for managing risk of failing to find subsequent lift in cross-country thermal soaring. Analysis of human pilot flight logs indicates that pilots rarely operate below the percolation threshold, identifies an apparent desired minimum node degree, and shows that pilots trade climb rate to maintain percolation.

Significance. If the mapping from percolation threshold to operational risk holds and is shown to be non-redundant with simpler heuristics, the work would supply a principled graph-theoretic framework for in-flight risk assessment using only locally observable information. This could inform both human training and autonomous soaring algorithms, with the empirical flight-log component providing a concrete test of the model against real decision data.

major comments (3)
  1. [Flight-log analysis] Flight-log analysis section: the 'apparent desired minimum node degree' is extracted from the same pilot data used to demonstrate that pilots rarely fly below the percolation threshold. This creates a data-driven threshold that reduces the claimed theoretical confirmation to an observed fit, raising circularity concerns for the central claim.
  2. [Results / flight log analysis] Validation of the decision threshold: no ablation or comparative test is presented against simpler baseline rules (e.g., maintaining a fixed minimum number of reachable thermals or a minimum local density). Without such a comparison, it remains unclear whether the percolation condition is the load-bearing predictor of pilot behavior or merely correlated with it.
  3. [Methods] Graph construction details: the precise criteria for placing thermal nodes and defining glide-reachability edges from in-situ observations are not specified with sufficient rigor (e.g., altitude margins, wind effects, or thermal strength thresholds), making it difficult to assess how the percolation threshold is computed or reproduced from available pilot information.
minor comments (2)
  1. [Abstract / Results] The abstract states that pilots 'accept reduced climb rates in order to maintain percolation,' but the corresponding quantitative trade-off (e.g., climb-rate penalty versus percolation margin) is not shown in a figure or table; adding such a plot would improve clarity.
  2. [Theory section] Notation for the minimum node degree threshold is introduced informally; defining it explicitly with an equation (e.g., k_min) and stating its relation to the percolation critical point would reduce ambiguity.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their constructive comments, which have helped us identify areas for improvement. We address each major comment in detail below, and plan to incorporate revisions accordingly.

read point-by-point responses

  1. Referee: Flight-log analysis section: the 'apparent desired minimum node degree' is extracted from the same pilot data used to demonstrate that pilots rarely fly below the percolation threshold. This creates a data-driven threshold that reduces the claimed theoretical confirmation to an observed fit, raising circularity concerns for the central claim.

    Authors: We thank the referee for highlighting this potential circularity. The percolation threshold itself is a purely theoretical quantity derived from graph theory and is independent of any flight data. The logs serve solely to test whether pilots operate above this threshold. The minimum node degree is presented as a separate empirical observation from the same dataset. We will revise the text to explicitly separate these elements and clarify that the primary validation concerns the percolation condition, not the derived degree value. revision: partial

  2. Referee: Validation of the decision threshold: no ablation or comparative test is presented against simpler baseline rules (e.g., maintaining a fixed minimum number of reachable thermals or a minimum local density). Without such a comparison, it remains unclear whether the percolation condition is the load-bearing predictor of pilot behavior or merely correlated with it.

    Authors: The referee correctly notes the absence of any ablation or baseline comparison. The current analysis demonstrates correlation between pilot behavior and the percolation threshold but does not test whether simpler heuristics (minimum reachable thermals or local density) suffice. We will add a comparative analysis in the revised manuscript to evaluate whether the percolation condition supplies non-redundant predictive value. revision: yes

  3. Referee: Graph construction details: the precise criteria for placing thermal nodes and defining glide-reachability edges from in-situ observations are not specified with sufficient rigor (e.g., altitude margins, wind effects, or thermal strength thresholds), making it difficult to assess how the percolation threshold is computed or reproduced from available pilot information.

    Authors: We agree that the Methods section lacks sufficient rigor on graph construction. The revised manuscript will expand this section with explicit criteria for node placement, edge definition (including altitude margins and wind corrections), and the thermal strength thresholds applied to the flight-log data. revision: yes

Circularity Check

1 steps flagged

Minimum node degree threshold fitted from same pilot data used to validate percolation maintenance

specific steps
  1. fitted input called prediction [Abstract / flight logs analysis]
    "This analysis indicates that in sport soaring pilots rarely operate in a condition which does not satisfy graph percolation, identifies an apparent desired minimum node degree, and shows that pilots accept reduced climb rates in order to maintain percolation."

    The minimum node degree is identified from the identical pilot flight logs that are then used to claim pilots maintain percolation. The threshold is therefore statistically forced by the same observations it is said to explain, reducing the 'theoretical explanation' to a post-hoc fit of the data.

full rationale

The paper presents graph percolation as a theoretical risk framework whose threshold marks safe continuation. However, the load-bearing 'apparent desired minimum node degree' is extracted directly from the flight-log dataset that is also used to demonstrate pilots rarely violate percolation and trade climb rate to maintain it. This makes the reported confirmation a data-driven fit rather than an independent test of the model. No equations or external benchmarks are shown to separate the threshold identification from the validation observations. The central claim therefore reduces to a fitted parameter renamed as theoretical prediction.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on modeling uncertain thermal locations as a graph whose connectivity threshold governs flight continuation risk; the minimum node degree is extracted from the same flight data used for validation, and the mapping from pilot observations to graph edges is treated as given by domain knowledge.

free parameters (1)
  • minimum node degree threshold
    Identified as 'apparent desired' from analysis of pilot flight logs; functions as the effective percolation criterion.
axioms (1)
  • domain assumption Thermal locations and glide reachability can be represented as nodes and edges in a graph using information available in flight.
    Invoked to justify applying percolation theory to the risk of failing to find the next updraft.

pith-pipeline@v0.9.0 · 5683 in / 1287 out tokens · 78483 ms · 2026-05-22T20:05:37.150196+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

18 extracted references · 18 canonical work pages

  1. [1]

    Reichmann, Cross-Country Soaring

    H. Reichmann, Cross-Country Soaring . Soaring Society of America, 1981

    work page 1981

  2. [2]

    Brigliadori and R

    L. Brigliadori and R. Brigliadori, Competing in Gliders: Winning With Your Mind. Bellavite, 2006

    work page 2006

  3. [3]

    N ew insights into autonomous soaring,

    C. Walton, I. Kaminer, V . Dobrokhodov, and K. D. Jones, “N ew insights into autonomous soaring,” 2017 56th IEEE Conference on Decision and Control (CDC), pp. 2059–2064, 2017

    work page 2017

  4. [4]

    Autonomous soaring: The mon- tague cross-country challenge,

    D. J. Edwards and L. M. Silverberg, “Autonomous soaring: The mon- tague cross-country challenge,” Journal of Aircraft , vol. 47, pp. 1763– 1769, 2010

    work page 2010

  5. [5]

    Guidance and control for an autonomous soar ing uav,

    M. J. Allen, “Guidance and control for an autonomous soar ing uav,” tech. rep., NASA Dryden Flight Research Center, 2007

    work page 2007

  6. [6]

    The aut osoar autonomous soaring aircraft part 2: Hardware implementati on and flight results,

    N. T. Depenbusch, J. J. Bird, and J. W. Langelaan, “The aut osoar autonomous soaring aircraft part 2: Hardware implementati on and flight results,” Journal of Field Robotics , pp. 1–24, 2017

    work page 2017

  7. [7]

    Optimum airspeed selector,

    P . B. J. MacCready, “Optimum airspeed selector,” Soaring, pp. 10–11, 1 1958

    work page 1958

  8. [8]

    Bounded ratio nality and risk strategy in thermal soaring,

    J. J. Bird, D. Sazhin, and J. W. Langelaan, “Bounded ratio nality and risk strategy in thermal soaring,” Technical Soaring, 2020

    work page 2020

  9. [9]

    Speed to fly with management of the risk of land ing out,

    Y . Fukada, “Speed to fly with management of the risk of land ing out,” Technical Soaring, vol. 24, 2000

    work page 2000

  10. [10]

    Maccready theory with uncertain lift and limited altitude,

    J. Cochrane, “Maccready theory with uncertain lift and limited altitude,” Technical Soaring, vol. 23, pp. 88–96, 1999

    work page 1999

  11. [11]

    Random plane networks,

    E. N. Gilbert, “Random plane networks,” Journal of the Society for Industrial and Applied Mathematics , vol. 9, pp. 533–543, 1961

    work page 1961

  12. [12]

    Random geometric graphs as a model for bounding the endurance of soaring aircr aft,

    N. T. Depenbusch, C. Lagoa, and J. W. Langelaan, “Random geometric graphs as a model for bounding the endurance of soaring aircr aft,” in 53rd IEEE Conference on Decision and Control , pp. 248–253, IEEE, 12 2014

    work page 2014

  13. [13]

    R. B. Stull, An Introduction to Boundary Layer Meteorology . Springer Netherlands, 1 ed., 1988

    work page 1988

  14. [14]

    Continuum p ercolation with steps in the square or the disc,

    P . Balister, B. Bollob´ as, and M. Walters, “Continuum p ercolation with steps in the square or the disc,” Random Structures and Algorithms , vol. 26, pp. 392–403, 2005

    work page 2005

  15. [15]

    Soaring behaviour and performance o f some east african birds, observed from a motor-glider,

    C. J. Pennycuick, “Soaring behaviour and performance o f some east african birds, observed from a motor-glider,” Ibis, vol. 114, pp. 178– 218, 1972

    work page 1972

  16. [16]

    Becker and P

    S. Becker and P . Bruce, Experimental Study of Paraglider Aerodynamics. PhD thesis, Imperial College London, 2017

    work page 2017

  17. [17]

    Weglide,

    WeGlide UG, “Weglide,” 2025

    work page 2025

  18. [18]

    K. Rohde-Brandenburger, “Determination of new handic aps for eval- uation of gliders in club class translation of: Berechnung v on neuen handicap-faktoren f¨ ur die wertung von segelflugzeugen in d er club- klasse,” tech. rep., F´ ed´ eration A´ eronautique Internationale, 2017. APPENDIX Continuum percolation typically considers spatially dis- tributed ran...

    work page 2017