pith. sign in

arxiv: 2509.25983 · v3 · pith:VH3EAZR3new · submitted 2025-09-30 · ⚛️ physics.flu-dyn

Rare-event detection in a backward-facing-step flow using live optical-flow velocimetry: observation of an upstream jet burst

Juan Pimienta , Jean-Luc Aider This is my paper

Pith reviewed 2026-05-22 12:44 UTC · model grok-4.3

classification ⚛️ physics.flu-dyn
keywords rare eventsbackward-facing stepoptical flow velocimetryupstream jet burstturbulent separated flowvortex dynamicsKelvin-Helmholtz instability
0
0 comments X

The pith

Live optical flow velocimetry enables the first direct detection of an upstream jet burst in backward-facing step flow.

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

The authors seek to demonstrate that rare extreme events in turbulent flows can be captured by using live real-time analysis of velocity fields to automatically record data when unusual conditions occur. In experiments with a backward-facing step at moderate Reynolds number, they ran continuous monitoring for 1.5 hours and set thresholds on velocity components to flag potential rare events. If the method works, it reveals a jet-like flow intrusion into the recirculation zone that starts from the breakdown of a vortex and is maintained by pairs of rotating vortices, along with increased energy and rotation measures. This matters because such events likely control important aspects of mixing and flow transition that standard short measurements miss.

Core claim

The paper reports the experimental observation of an upstream-directed jet burst in a backward-facing step flow at Reynolds number based on step height of 2100. This event was captured using long-duration live optical flow velocimetry, where velocity probes at a fixed location triggered recording upon exceeding large negative deviation thresholds. The burst appears as a jet intrusion into the recirculation region, initiated by the collapse of a merged Kelvin-Helmholtz vortex and sustained by counter-rotating vortices, and it coincides with heavy-tailed velocity statistics and amplified fluctuating kinetic energy and enstrophy.

What carries the argument

Live Optical Flow Velocimetry (L-OFV) for continuous real-time velocity field analysis that triggers high-speed or high-resolution capture on data-driven extreme events defined by Z-scores of velocity components.

If this is right

  • The approach establishes a platform for detecting rare events in separated shear layers.
  • The observed jet burst is linked to specific vortex dynamics including Kelvin-Helmholtz vortex collapse.
  • Heavy-tailed statistics appear in the velocity probe data during the event.
  • Amplification of fluctuating kinetic energy and enstrophy accompanies the jet burst.

Where Pith is reading between the lines

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

  • Extending the monitoring duration or using multiple probe locations could increase the chance of capturing additional events of this type.
  • Similar live triggering techniques might apply to other turbulent flow setups where rare events influence overall behavior.
  • If the mechanism is general, it could inform models of transition and mixing in separated flows.

Load-bearing premise

The chosen probe position and the specific thresholds for negative velocity deviations correctly flag the upstream jet bursts rather than other phenomena or missing them entirely.

What would settle it

Repeating the experiment with the same setup but failing to detect any jet bursts despite long monitoring times, or finding that the velocity signatures at other locations show different extreme behaviors, would indicate the reported event is not representative or the trigger is insufficient.

Figures

Figures reproduced from arXiv: 2509.25983 by Jean-Luc Aider, Juan Pimienta.

Figure 1
Figure 1. Figure 1: FIG. 1: 3D sketch of the BFS model used in the experiments, showing the measurement [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: a) Sketch showing the main instabilities and main structures generated [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Instantaneous velocity field magnitude in the vertical symmetry plane ( [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: Time series and PDF of the streamwise component [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5: Time series and PDF of the wall-normal component [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: Time series and PDF of the kinetic energy associated to the streamwise [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7: Time series and PDF of the kinetic energy associated to the wall-normal [PITH_FULL_IMAGE:figures/full_fig_p012_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8: Comparison of the time series of the velocity components from the probes located [PITH_FULL_IMAGE:figures/full_fig_p014_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9: Comparison of the time-series of the energy associated to each of the velocity [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10: Phase space trajectories [PITH_FULL_IMAGE:figures/full_fig_p016_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: FIG. 11: Fluctuating kinetic energy snapshots and streamline plots. The snapshots are [PITH_FULL_IMAGE:figures/full_fig_p017_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIG. 12: PDF comparison of the time series of the probes from the initial long observation, [PITH_FULL_IMAGE:figures/full_fig_p019_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIG. 13: PDF comparison of the probes time series from the detected event, the baseline [PITH_FULL_IMAGE:figures/full_fig_p020_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIG. 14: Joint PDF of space averaged fluctuation of kinetic energy and enstrophy for the [PITH_FULL_IMAGE:figures/full_fig_p021_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: FIG. 15: Time series of global and local quantities, expressed in terms of their [PITH_FULL_IMAGE:figures/full_fig_p022_15.png] view at source ↗
read the original abstract

Rare and extreme events in turbulent flows play a critical role in transport, mixing and transition, yet are notoriously difficult to capture experimentally. Here we report, to our knowledge, the first direct experimental detection of an upstream-directed jet burst in a backward-facing step (BFS) flow at $Re_h=2100$, using long-duration Live Optical Flow Velocimetry (L-OFV). Continuous monitoring over 1.5 h enabled a data-driven definition of extremes as rare velocity probes excursions deep into the observed distribution's tails; in practice, large negative events ($u: Z < -6$, $v: Z < -5$ at $(x,y) = (2h,h / 2)$, where $|Z| > > 0$ stands for large deviations from the mean value) triggered the live capture of surrounding velocity fields. The recording is triggered when the probes surpass the defined threshold, using live analysis of the velocity fields. The detected event features a jet-like intrusion into the recirculation region initiated by the collapse of a merged Kelvin-Helmholtz vortex and sustained by counter-rotating vortices, and is accompanied with heavy-tailed probe statistics and simultaneous amplification of fluctuating kinetic energy and enstrophy. While a single event was recorded, underscoring its rarity, the results establish L-OFV as a viable platform for rare-event detection in separated shear layers and document a previously unreported mechanism of upstream jet bursting in BFS flow.

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 manuscript reports the first direct experimental detection of an upstream-directed jet burst in a backward-facing step flow at Re_h=2100 using long-duration Live Optical Flow Velocimetry (L-OFV). Continuous 1.5-hour monitoring with live analysis enabled data-driven thresholds on velocity probes (u: Z < -6, v: Z < -5 at (x,y)=(2h, h/2)) to trigger capture of surrounding fields. The single recorded event is interpreted as a jet-like intrusion into the recirculation region initiated by collapse of a merged Kelvin-Helmholtz vortex and sustained by counter-rotating vortices, accompanied by heavy-tailed probe statistics and simultaneous amplification of fluctuating kinetic energy and enstrophy. The work positions L-OFV as a viable platform for rare-event detection in separated shear layers.

Significance. If the L-OFV measurements prove accurate for extreme velocities and the mechanism can be placed on firmer statistical footing, the result would be significant for documenting a previously unreported upstream jet-burst process in BFS flows and for establishing live optical-flow triggering as a practical tool for capturing rare events in turbulent separated shear layers. The long-duration monitoring approach directly addresses the experimental challenge of observing low-probability phenomena.

major comments (3)
  1. [§4] §4 (Results, event description): The central claim of a previously unreported jet-burst mechanism rests on qualitative interpretation of velocity fields from a single captured sequence; no quantitative vortex identification (e.g., swirling strength or circulation integrals) or comparison to prior BFS literature is provided to substantiate the merged KH-vortex collapse and counter-rotating vortex sustenance.
  2. [§3] §3 (L-OFV implementation and validation): No cross-validation or error quantification of the optical-flow algorithm is reported against DNS or conventional PIV for |Z| > 5 excursions inside the recirculation zone, which is load-bearing because the headline observation depends on faithful capture of these extreme velocities rather than optical-flow bias or seeding artifacts.
  3. [§3.2] §3.2 (Probe location and thresholds): The choice of probe position (2h, h/2) and data-driven thresholds (u: Z < -6, v: Z < -5) is described as arising from the observed distribution tails, yet no a-priori justification, sensitivity analysis to nearby locations, or occurrence-rate statistics are supplied; this post-hoc selection risks systematic bias in event detection.
minor comments (2)
  1. [Abstract] Abstract: The symbol Z is introduced as 'large deviations from the mean value' without an explicit formula (e.g., whether it is a standardized score using local or global standard deviation); this should be stated once for clarity.
  2. Figure captions (assumed for event visualization): Ensure that instantaneous velocity vectors or vorticity contours are scaled consistently with the probe time series so readers can directly relate the reported |Z| excursions to the visualized structures.

Simulated Author's Rebuttal

3 responses · 1 unresolved

We thank the referee for the careful reading and constructive comments. We respond to each major point below, indicating revisions made to the manuscript.

read point-by-point responses

  1. Referee: [§4] §4 (Results, event description): The central claim of a previously unreported jet-burst mechanism rests on qualitative interpretation of velocity fields from a single captured sequence; no quantitative vortex identification (e.g., swirling strength or circulation integrals) or comparison to prior BFS literature is provided to substantiate the merged KH-vortex collapse and counter-rotating vortex sustenance.

    Authors: The description in the original manuscript was based on direct inspection of the instantaneous velocity and vorticity fields during the event. To strengthen the quantitative support, we have added swirling-strength isosurfaces to identify the merged Kelvin-Helmholtz vortices and their subsequent collapse, together with circulation integrals around the counter-rotating structures that sustain the jet intrusion. We have also inserted a brief comparison to existing BFS literature on vortex pairing and reattachment dynamics (e.g., references to studies reporting similar vortex interactions in the recirculation zone). These additions appear in the revised §4. revision: yes

  2. Referee: [§3] §3 (L-OFV implementation and validation): No cross-validation or error quantification of the optical-flow algorithm is reported against DNS or conventional PIV for |Z| > 5 excursions inside the recirculation zone, which is load-bearing because the headline observation depends on faithful capture of these extreme velocities rather than optical-flow bias or seeding artifacts.

    Authors: We acknowledge that explicit cross-validation for extreme excursions was not included. The L-OFV implementation follows the same algorithm validated in our earlier moderate-velocity studies; for the present data we have added an uncertainty estimate based on local seeding density and a consistency check against the expected physical velocity scales inside the BFS recirculation region. A full DNS or PIV benchmark for |Z| > 5 in this geometry is not available from our existing datasets. revision: partial

  3. Referee: [§3.2] §3.2 (Probe location and thresholds): The choice of probe position (2h, h/2) and data-driven thresholds (u: Z < -6, v: Z < -5) is described as arising from the observed distribution tails, yet no a-priori justification, sensitivity analysis to nearby locations, or occurrence-rate statistics are supplied; this post-hoc selection risks systematic bias in event detection.

    Authors: The probe location was chosen because it lies within the primary shear-layer region where vortex dynamics are most active, consistent with prior BFS flow visualizations. We have now performed a sensitivity study by repeating the threshold analysis at four neighboring locations (±0.25h); the same extreme event is recovered with only minor changes in threshold values. Occurrence-rate statistics remain limited by the single recorded event in 1.5 h of monitoring; the heavy-tailed probe distributions are nevertheless reported as supporting evidence of rarity. revision: partial

standing simulated objections not resolved
  • Direct cross-validation of L-OFV against DNS or conventional PIV specifically for |Z| > 5 velocity excursions in the recirculation zone.

Circularity Check

0 steps flagged

No significant circularity in this direct experimental observation paper

full rationale

This paper reports a direct experimental detection of a rare upstream jet burst using live optical-flow velocimetry over 1.5 hours of monitoring. Thresholds (u: Z < -6, v: Z < -5 at a fixed probe location) are defined empirically from the tails of the observed velocity distribution to trigger capture; the recorded fields are then interpreted as showing a jet-like intrusion initiated by Kelvin-Helmholtz vortex collapse. No mathematical derivation, first-principles prediction, or fitted parameter is presented whose output reduces by construction to the input data or to a self-citation chain. The central claim is an empirical observation rather than a derived result, making the analysis self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The work is observational and relies on standard fluid-dynamics assumptions plus data-driven thresholds chosen after inspecting the velocity distribution.

free parameters (1)
  • Extreme-event thresholds = u: Z < -6, v: Z < -5 at (x,y)=(2h, h/2)
    Z-score cutoffs (u: Z < -6, v: Z < -5) and probe location chosen from observed tails to trigger recording.
axioms (1)
  • domain assumption Live optical-flow velocimetry produces sufficiently accurate instantaneous velocity fields to identify rare events
    Invoked when the live analysis is used to trigger capture of surrounding fields.

pith-pipeline@v0.9.0 · 5797 in / 1357 out tokens · 39844 ms · 2026-05-22T12:44:10.822352+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

38 extracted references · 38 canonical work pages

  1. [1]

    (2) The analysis was limited to the streamwise region 0 ≤x/h≤10, where the relevant dynamics occur. Figure 11 shows ten FKE instantaneous contour plots superposed to the streamlines of the flow, over the time interval ranging fromt = 3.65 tot = 5.15 s, centered around the detected 16 (a) (b) (c) (d) (e) (f) (g) (h) (i) (j) FIG. 11: Fluctuating kinetic ene...

    work page 1955

  2. [2]

    F. M. Longin, From value at risk to stress testing: The extreme value approach, Journal of Banking & Finance 24, 1097 (2000)

    work page 2000

  3. [3]

    N. N. Taleb, The Black Swan:: The Impact of the Highly Improbable: With a New Section: On Robustness and Fragility: 2 (Incerto) (Random House, New York, 2007)

    work page 2007

  4. [4]

    S. I. Seneviratne, X. Zhang, M. Adnan, W. Badi, C. Dereczynski, A. D. Luca, S. Ghosh, I. Iskandar, J. Kossin, S. Lewis, et al. , Weather and climate extreme events in a changing climate (Cambridge University Press, 2021)

    work page 2021

  5. [5]

    P. R. Grant, B. R. Grant, R. B. Huey, M. T. Johnson, A. H. Knoll, and J. Schmitt, Evolu- tion caused by extreme events, Philosophical Transactions of the Royal Society B: Biological Sciences 372, 20160146 (2017)

    work page 2017

  6. [6]

    C. E. Walters, M. M. Mesl´ e, and I. M. Hall, Modelling the global spread of diseases: A review of current practice and capability, Epidemics 25, 1 (2018)

    work page 2018

  7. [7]

    Marani, G

    M. Marani, G. G. Katul, W. K. Pan, and A. J. Parolari, Intensity and frequency of extreme novel epidemics, Proceedings of the National Academy of Sciences 118, e2105482118 (2021)

    work page 2021

  8. [8]

    Lemoult, J.-L

    G. Lemoult, J.-L. Aider, and J. E. Wesfreid, Experimental scaling law for the subcritical transition to turbulence in plane poiseuille flow, Physical Review E—Statistical, Nonlinear, and Soft Matter Physics 85, 025303 (2012)

    work page 2012

  9. [9]

    Daviaud, J

    F. Daviaud, J. Hegseth, and P. Berg´ e, Subcritical transition to turbulence in plane couette flow, Physical review letters 69, 2511 (1992)

    work page 1992

  10. [10]

    Duguet, P

    Y. Duguet, P. Schlatter, and D. S. Henningson, Formation of turbulent patterns near the onset of transition in plane couette flow, Journal of Fluid Mechanics 650, 119 (2010)

    work page 2010

  11. [11]

    Klotz, G

    L. Klotz, G. Lemoult, I. Frontczak, L. S. Tuckerman, and J. E. Wesfreid, Couette-poiseuille flow experiment with zero mean advection velocity: Subcritical transition to turbulence (2017)

    work page 2017

  12. [12]

    Lemoult, J

    G. Lemoult, J. L. Aider, and J. E. Wesfreid, Turbulent spots in a channel: large-scale flow and self-sustainability, Journal of Fluid Mechanics 731 (2013). 24

    work page 2013

  13. [13]

    Lemoult, K

    G. Lemoult, K. Gumowski, J.-L. Aider, and J. E. Wesfreid, Turbulent spots in channel flow: an experimental study: large-scale flow, inner structure and low-order model, The European Physical Journal E 37, 1 (2014)

    work page 2014

  14. [14]

    Gom´ e, L

    S. Gom´ e, L. Tuckerman, and D. Barkley, Extreme events in transitional turbulence, Philo- sophical Transactions of the Royal Society A 10.1098/rsta.2021.0036 (2021)

    work page doi:10.1098/rsta.2021.0036 2021

  15. [15]

    M. P. Hack and O. T. Schmidt, Extreme events in wall turbulence, Journal of Fluid Mechanics 907, A9 (2021)

    work page 2021

  16. [16]

    H. K. Moffatt, H. Moffatt, and H. K. Moffatt, Extreme events in turbulent flow, Journal of Fluid Mechanics 10.1017/jfm.2020.1079 (2021)

    work page doi:10.1017/jfm.2020.1079 2020

  17. [17]

    Bouchet, J

    F. Bouchet, J. Rolland, and E. Simonnet, Rare event algorithm links transitions in turbulent flows with activated nucleations, Physical review letters 122, 074502 (2019)

    work page 2019

  18. [18]

    Gayout, M

    A. Gayout, M. Bourgoin, and N. Plihon, Rare event-triggered transitions in aerodynamic bifurcation, Physical Review Letters 126, 104501 (2021)

    work page 2021

  19. [19]

    I. K. Atallah, L. Pastur, R. Monchaux, and L. Zimmer, From low-frequency oscillations to markovian bistable stall dynamics, Physical Review Fluids (2024)

    work page 2024

  20. [20]

    Varon, Y

    E. Varon, Y. Eulalie, S. Edwige, P. Gilotte, and J.-L. Aider, Chaotic dynamics of large-scale structures in a turbulent wake, Physical Review Fluids 2, 034604 (2017)

    work page 2017

  21. [21]

    Lestang, F

    T. Lestang, F. Bouchet, and E. L´ evˆ eque, Numerical study of extreme mechanical force exerted by a turbulent flow on a bluff body by direct and rare-event sampling techniques, Journal of Fluid Mechanics 895, A19 (2020)

    work page 2020

  22. [22]

    Giardina, J

    C. Giardina, J. Kurchan, V. Lecomte, and J. Tailleur, Simulating rare events in dynamical processes, Journal of statistical physics 145, 787 (2011)

    work page 2011

  23. [23]

    P. K. Yeung, P.-K. Yeung, P.-K. Yeung, P.-K. Yeung, X. Zhai, X. M. Zhai, X. M. Zhai, K. R. Sreenivasan, and K. R. Sreenivasan, Extreme events in computational turbulence, Proceedings of the National Academy of Sciences of the United States of America 10.1073/pnas.1517368112 (2015)

    work page doi:10.1073/pnas.1517368112 2015

  24. [24]

    Lohse and S

    D. Lohse and S. Grossmann, Intermittency in turbulence, Physica A: Statistical Mechanics and its Applications 194, 519 (1993)

    work page 1993

  25. [25]

    K. N. Rao, R. Narasimha, and M. A. B. Narayanan, The ‘bursting’ phenomenon in a turbulent boundary layer, Journal of Fluid Mechanics 48, 339 (1971)

    work page 1971

  26. [26]

    J. F. Morrison, H. M. Tsai, and P. M. D. Bradshaw, Conditional-sampling schemes for tur- 25 bulent flow, based on the variable-interval time averaging (vita) algorithm, Experiments in Fluids 7, 173 (1988)

    work page 1988

  27. [27]

    ¨Orl¨ u and R

    R. ¨Orl¨ u and R. Vinuesa, Instantaneous wall-shear-stress measurements: advances and appli- cation to near-wall extreme events, Measurement Science and Technology 31, 112001 (2020)

    work page 2020

  28. [28]

    R. D. Whalley, D. Dennis, Graham, and R. J. Poole, An experimental investigation into spatiotemporal intermittencies in turbulent channel flow close to transition, Experiments in Fluids 60, 1 (2019)

    work page 2019

  29. [29]

    B. D. Lucas and T. Kanade, An iterative image registration technique with an application to stereo vision, in International Joint Conference on Artificial Intelligence (1981)

    work page 1981

  30. [30]

    Gautier and J

    N. Gautier and J. L. Aider, Real-time planar flow velocity measurements using an optical flow algorithm implemented on GPUs, Journal of Visualization 18, 277 (2015)

    work page 2015

  31. [31]

    Gautier and J

    N. Gautier and J. L. Aider, Control of the separated flow downstream of a backward-facing step using visual feedback, Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences 469 (2013)

    work page 2013

  32. [32]

    Armaly, F

    B. Armaly, F. Durst, J. Pereira, and B. Schonung, Experimental and theoretical investigation of backward-facing step flow, Journal of Fluid Mechanics 127, 473 (1983)

    work page 1983

  33. [33]

    H. Le, P. Moin, and J. Kim, Direct numerical simulation of turbulent flow over a backward- facing step, Journal of fluid mechanics 330, 349 (1997)

    work page 1997

  34. [34]

    Kostas, J

    J. Kostas, J. Soria, and M. Chong, Particle image velocimetry measurements of a backward- facing step flow, Experiments in fluids 33, 838 (2002)

    work page 2002

  35. [35]

    J. L. Aider, A. Danet, and M. R. Lesieur, Large-eddy simulation applied to study the influence of upstream conditions on the time-dependant and averaged characteristics of a backward- facing step flow, Journal of Turbulence 8, 51 (2007)

    work page 2007

  36. [36]

    Tailleur and J

    J. Tailleur and J. Kurchan, Probing rare physical trajectories with lyapunov weighted dynam- ics, Nature Physics 3, 203 (2006)

    work page 2006

  37. [37]

    She, Intermittency and non-gaussian statistics in turbulence, Fluid Dynamics Research 8, 143 (1991)

    Z. She, Intermittency and non-gaussian statistics in turbulence, Fluid Dynamics Research 8, 143 (1991)

    work page 1991

  38. [38]

    Li and C

    Y. Li and C. Meneveau, Origin of non-gaussian statistics in hydrodynamic turbulence., Phys- ical review letters 95 16, 164502 (2005). 26

    work page 2005