A high-fidelity ensemble simulation framework for interrogating wildland-fire behaviour and benchmarking machine learning models

Qing Wang; Matthias Ihme; Cenk Gazen; Yi-Fan Chen; John Anderson

doi:10.1071/WF24097

RESEARCH ARTICLE (Open Access)

Previous Next Contents Vol 33(12)

A high-fidelity ensemble simulation framework for interrogating wildland-fire behaviour and benchmarking machine learning models

Qing Wang

^A ^* , Matthias Ihme ^A ^B ^C , Cenk Gazen ^A , Yi-Fan Chen ^A and John Anderson ^A

+ Author Affiliations

- Author Affiliations

^A Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, USA.

^B Department of Mechanical Engineering, Stanford University, Stanford, CA 94305, USA.

^C Department of Photon Science, SLAC National Accelerator Laboratory, Menlo Park, CA 94025, USA.

^* Correspondence to: wqing@google.com

International Journal of Wildland Fire 33, WF24097 https://doi.org/10.1071/WF24097

Submitted: 12 June 2024 Accepted: 24 October 2024 Published: 22 November 2024

© 2024 The Author(s) (or their employer(s)). Published by CSIRO Publishing on behalf of IAWF. This is an open access article distributed under the Creative Commons Attribution 4.0 International License (CC BY).

Abstract

Background

Wildfire research uses ensemble methods to analyse fire behaviours and assess uncertainties. Nonetheless, current research methods are either confined to simple models or complex simulations with limitations. Modern computing tools could allow for efficient, high-fidelity ensemble simulations.

Aims

This study proposes a high-fidelity ensemble wildfire simulation framework for studying wildfire behaviour, assessing fire risks, analysing uncertainties, and training machine learning (ML) models.

Methods

We present a simulation framework that integrates the Swirl-Fire large-eddy simulation tool for wildfire predictions with the Vizier optimisation platform for automated run-time management of ensemble simulations and large-scale batch processing. All simulations are executed on tensor-processing units to enhance computational efficiency.

Key results

A dataset of 117 simulations is created, each with 1.35 billion mesh points. The simulations are compared to existing experimental data and show good agreement in terms of fire rate of spread. Analysis is performed for fire acceleration, mean rate of spread, and fireline intensity.

Conclusions

Strong coupling between wind speed and slope is observed for fire-spread rate and intermittency. A critical Froude number that delineates fires from plume-dominated to wind-dominated is identified and confirmed with literature observations.

Implications

The ensemble simulation framework is efficient in facilitating large-scale parametric wildfire studies.

Keywords: ensemble simulations, fire propagation, fire/atmospheric coupling, large-eddy simulation, tensor processing units, TensorFlow, wildfire modelling, wildland fire prediction.

References

Abadi M, Agarwal A, Barham P, Brevdo E, Chen Z, Citro C, Corrado GS, Davis A, Dean J, Devin M, Ghemawat S, Goodfellow I, Harp A, Irving G, Isard M, Jia Y, Jozefowicz R, Kaiser L, Kudlur M, Levenberg J, Man´e D, Monga R, Moore S, Murray D, Olah C, Schuster M, Shlens J, Steiner B, Sutskever I, Talwar K, Tucker P, Vanhoucke V, Vasudevan V, Vi´egas F, Vinyals O, Warden P, Wattenberg M, Wicke M, Yu Y, Zheng X (2015) TensorFlow: Large-scale machine learning on heterogeneous systems. Software available from tensorflow.org

Abatzoglou JT, Williams AP (2016) Impact of anthropogenic climate change on wildfire across western US forests. The Proceedings of the National Academy of Sciences USA 113(42), 11770-11775.
| Crossref | Google Scholar | PubMed |

Albini FA (1976) ‘Estimating wildfire behavior and effects. Vol. 30.’ (Department of Agriculture, Forest Service, Intermountain Forest and Range Experiment Station, USA)

Alexander ME (1982) Calculating and interpreting forest fire intensities. Canadian Journal of Botany 60, 349-357.
| Crossref | Google Scholar |

Alexander ME, Cruz MG (2013) Limitations on the accuracy of model predictions of wildland fire behaviour: a state-of-the-knowledge overview. Forestry Chronicle 89(03), 372-383.
| Crossref | Google Scholar |

Allaire F, Filippi J-B, Mallet V (2020) Generation and evaluation of an ensemble of wildland fire simulations. International Journal of Wildland Fire 29(2), 160-173.
| Crossref | Google Scholar |

Anderson K, Flannigan M, Reuter G (2005) Using ensemble techniques in fire growth modelling. In ‘Sixth Symposium on Fire and Forest Meteorology’, 24–27 October 2005. (Canmore, AB, Canada).

Atchley AL, Linn RR, Jonko A, Hoffman C, Hyman JD, Pimont F, Sieg C, Middleton RS (2021) Effects of fuel spatial distribution on wildland fire behaviour. International Journal of Wildland Fire 30(3), 179-189.
| Crossref | Google Scholar |

Barbero R, Abatzoglou JT, Larkin NK, Kolden CA, Stocks B (2015) Climate change presents increased potential for very large fires in the contiguous United States. International Journal of Wildland Fire 24(7), 892-899.
| Crossref | Google Scholar |

Benali A, Ervilha AR, Sá ACL, Fernandes PM, Pinto RMS, Trigo RM, Pereira JMC (2016) Deciphering the impact of uncertainty on the accuracy of large wildfire spread simulations. Science of The Total Environment 569–570, 73-85.
| Crossref | Google Scholar | PubMed |

Benali A, Sá ACL, Ervilha AR, Trigo RM, Fernandes PM, Pereira JMC (2017) Fire spread predictions: Sweeping uncertainty under the rug. Science of The Total Environment 592, 187-196.
| Crossref | Google Scholar | PubMed |

Butler BW, Anderson WR, Catchpole EA (2007) Influence of slope on fire spread rate. In: Butler, Bret BW, Cook W, comps. 2007. ‘The fire environment —innovations, management, and policy’. (Eds BW Butler, W Cook, FL Destin) pp. 75–83. (US Forest Service)

Cheney NP, Gould JS (1997) Letter to the editor: fire growth and acceleration. International Journal of Wildland Fire 7(1), 1-5.
| Crossref | Google Scholar |

Cheney NP, Gould JS, Catchpole WR (1993) The influence of fuel, weather and fire shape variables on fire-spread in grasslands. International Journal of Wildland Fire 3(1), 31-44.
| Crossref | Google Scholar |

Chung WT, Akoush B, Sharma P, Tamkin A, Jung KS, Chen JH, Guo J, Brouzet D, Talei M, Savard B, Poludnenko AY, Ihme M (2023) Turbulence in Focus: Benchmarking Scaling Behavior of 3D Volumetric Super-Resolution with BLASTNet 2.0 Data. Advances in Neural Information Processing Systems 36.

Clements CB, Seto D (2015) Observations of fire–atmosphere interactions and near-surface heat transport on a slope. Boundary-Layer Meteorology 154(3), 409-426.
| Crossref | Google Scholar |

Clements CB, Zhong S, Goodrick SL, Li J, Potter B, Bian X, Heilman W, Charney J, Perna R, Jang M, Lee D, Patel M, Street S, Aumann G (2007) Observing the dynamics of wildland grass fires: FireFlux–A field validation experiment. Bulletin of the American Meteorological Society 88(9), 1369-1382.
| Crossref | Google Scholar |

Clements CB, Davis B, Seto D, Contezac J, Kochanski A, Fillipi J-B, Lareau N, Barboni B, Butler B, Krueger S, Ottmar R, Vihnanek R, Heilman WE, Flynn J, Jenkins MA, Mandel J, Teske C, Jimenez D, O’Brien J, Lefer B (2014) Overview of the 2013 FireFlux-II grass fire field experiment. In: ‘Advances in Forest Fire Research’. (Ed. Viegas, DX) pp. 392–400. (Imprensa da Universidade de Coimbra)

Clements CB, Kochanski AK, Seto D, Davis B, Camacho C, Lareau NP, Contezac J, Restaino J, Heilman WE, Krueger SK, Butler B, Ottmar RD, Vihnanek R, Flynn J, Filippi J-B, Barboni T, Hall DE, Mandel J, Jenkins MA, O’Brien J, Hornsby B, Teske C (2019) The FireFlux II experiment: a model-guided field experiment to improve understanding of fire–atmosphere interactions and fire spread. International Journal of Wildland Fire 278(4), 308-326.
| Crossref | Google Scholar |

Cruz MG (2010) Monte Carlo-based ensemble method for prediction of grassland fire spread. International Journal of Wildland Fire 19(4), 521-530.
| Crossref | Google Scholar |

Cruz MG, Alexander ME (2013) Uncertainty associated with model predictions of surface and crown fire rates of spread. Environmental Modelling & Software 47, 16-28.
| Crossref | Google Scholar |

Cruz MG, Hurley RJ, Bessell R, Sullivan AL (2020) Fire behaviour in wheat crops – effect of fuel structure on rate of fire spread. International Journal of Wildland Fire 29(3), 258-271.
| Crossref | Google Scholar |

Eftekharian E, Ghodrat M, He Y, Ong RH, Kwok KCS, Zhao M, Samali B (2019) Investigation of terrain slope effects on wind enhancement by a line source fire. Case Studies in Thermal Engineering 14, 100467.
| Crossref | Google Scholar |

Finney MA (1998) ‘FARSITE, Fire Area Simulator–model development and evaluation. No. 4.’ (US Department of Agriculture, Forest Service, Rocky Mountain Research Station)

Finney MA (2006) An overview of FlamMap fire modeling capabilities. In: Andrews, Patricia L.; Butler, Bret W., comps. 2006. ‘Fuels Management-How to Measure Success: Conference Proceedings. Vol. 41’. pp. 213–220. (US Forest Service)

Finney MA, Grenfell IC, McHugh CW, Seli RC, Trethewey D, Stratton RD, Brittain S (2011) A method for ensemble wildland fire simulation. Environmental Modeling and Assessment 16(2), 153-167.
| Crossref | Google Scholar |

Finney MA, Cohen JD, Forthofer JM, McAllister SS, Gollner MJ, Gorham DJ, Saito K, Akafuah NK, Adam BA, English JD, Dickinson RE (2015) Role of buoyant flame dynamics in wildfire spread. Proceedings of the National Academy of Sciences of the United States of America 112(32), 9833-9838.
| Crossref | Google Scholar | PubMed |

Finney MA, McAllister SS, Forthofer JM, Grumstrup TP (2021) ‘Wildland Fire Behaviour: Dynamics, Principles and Processes.’ (CSIRO Publishing: Melbourne, Australia)

Ganteaume A, Camia A, Jappiot M, San-Miguel-Ayanz J, Long-Fournel M, Lampin C (2013) A review of the main driving factors of forest fire ignition over Europe. Environmental Management 51, 651-662.
| Crossref | Google Scholar | PubMed |

Giorgi F, Francisco R (2000) Evaluating uncertainties in the prediction of regional climate change. Geophysical Research Letters 27(9), 1295-1298.
| Crossref | Google Scholar |

Golovin D, Solnik B, Moitra S, Kochanski G, Karro J, Sculley D (2017) Google Vizier: A service for Black-Box optimization, In ‘Proceedings of the 23rd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining’. KDD ’17. pp. 1487–1495. (Association for Computing Machinery: New York, NY, USA)

Ihme M, Chung WT, Mishra AA (2022) Combustion machine learning: principles, progress and prospects. Progress in Energy and Combustion Science 91, 101010.
| Crossref | Google Scholar |

Innocent J, Sutherland D, Khan N, Moinuddin K (2023) Physics-based simulations of grassfire propagation on sloped terrain at field scale: motivations, model reliability, rate of spread and fire intensity. International Journal of Wildland Fire 32(4), 496-512.
| Crossref | Google Scholar |

Jouppi NP, Yoon DH, Kurian G, Li S, Patil N, Laudon J, Young C, Patterson D (2020) A domain-specific supercomputer for training deep neural networks. Communications of the ACM 63(7), 67-78.
| Crossref | Google Scholar |

Jouppi NP, Yoon DH, Ashcraft M, Gottscho M, Jablin TB, Kurian G, Laudon J, Li S, Ma P, Ma X, Norrie T, Patil N, Prasad S, Young C, Zhou Z, Patterson D (2021) Ten lessons from three generations shaped google’s TPUv4i: Industrial product. In ‘2021 ACM/IEEE 48th Annual International Symposium on Computer Architecture (ISCA)’. pp. 1–14. (Institute of Electrical and Electronics Engineers)

Ju X, Gollner MJ, Wang Y, Tang W, Zhao K, Ren X, Yang L (2019) Downstream radiative and convective heating from methane and propane fires with cross wind. Combustion and Flame 204, 1-12.
| Crossref | Google Scholar |

Klemp JB, Lilly DK (1978) Numerical simulation of hydrostatic mountain waves. Journal of the Atmospheric Sciences 35(1), 78-107.
| Crossref | Google Scholar |

Leutbecher M, Palmer TN (2008) Ensemble forecasting. Journal of Computational Physics 227(7), 3515-3539.
| Crossref | Google Scholar |

Lewis JM (2005) Roots of ensemble forecasting. Monthly Weather Review 133(7), 1865-1885.
| Crossref | Google Scholar |

Linn RR (1997) A Transport Model for Prediction of Wildfire Behavior (No. LA-13334-T). PhD Thesis, Los Alamos National Lab, NM, USA.

Linn RR, Reisner J, Colman JJ, Winterkamp JL (2002) Studying wildfire behavior using FIRETEC. International Journal of Wildland Fire 11(4), 233-246.
| Crossref | Google Scholar |

Linn RR, Winterkamp JL, Edminster C, Colman JJ, Smith WS (2007) Coupled influences of topography and wind on wildland fire behaviour. International Journal of Wildland Fire 16(2), 183-195.
| Crossref | Google Scholar |

Linn RR, Winterkamp JL, Weise DR, Edminster C (2010) A numerical study of slope and fuel structure effects on coupled wildfire behaviour. International Journal of Wildland Fire 19(2), 179-201.
| Crossref | Google Scholar |

Liu N (2023) Wildland surface fire spread: mechanism transformation and behavior transition. Fire Safety Journal 141, 103974.
| Crossref | Google Scholar |

Liu N, Wu J, Chen H, Zhang L, Deng Z, Satoh K, Viegas DX, Raposo JR (2015) Upslope spread of a linear flame front over a pine needle fuel bed: the role of convection cooling. Proceedings of the Combustion Institute 35(3), 2691-2698.
| Crossref | Google Scholar |

Mandel J, Bennethum LS, Beezley JD, Coen JL, Douglas CC, Kim M, Vodacek A (2008) A wildland fire model with data assimilation. Mathematics and Computers in Simulation 79(3), 584-606.
| Crossref | Google Scholar |

Moinuddin KAM, Sutherland D, Mell WE (2018) Simulation study of grass fire using a physics-based model: striving towards numerical rigour and the effect of grass height on the rate of spread. International Journal of Wildland Fire 27(12), 800-814.
| Crossref | Google Scholar |

Morandini F, Silvani X (2010) Experimental investigation of the physical mechanisms governing the spread of wildfires. International Journal of Wildland Fire 19, 570-582.
| Crossref | Google Scholar |

Morvan D, Frangieh N (2018) Wildland fires behaviour: wind effect versus Byram’s convective number and consequences upon the regime of propagation. International Journal of Wildland Fire 27(9), 636-641.
| Crossref | Google Scholar |

Murphy JM, Sexton DMH, Barnett DN, Jones GS, Webb MJ, Collins M, Stainforth DA (2004) Quantification of modelling uncertainties in a large ensemble of climate change simulations. Nature 430(7001), 768-772.
| Crossref | Google Scholar | PubMed |

Nelson Jr, RM (2002) An effective wind speed for models of fire spread. International Journal of Wildland Fire 11(2), 153-161.
| Crossref | Google Scholar |

Noble IR, Bary GAV, Gill AM (1980) McArthur’s fire-danger meters expressed as equations. Australian Journal of Ecology 5, 201-220.
| Crossref | Google Scholar |

Parisien MA, Kafka VG, Hirsch KG, Todd JB, Lavoie SG, Maczek PD (2005) Mapping wildfire susceptibility with the BURN-P3 simulation model (information report). Technical Report. (Natural Resources Canada, Canadian Forest Service: Edmonton)

Paugam R, Wooster MJ, Mell WE, Rochoux MC, Filippi J-B, Rücker G, Frauenberger O, Lorenz E, Schroeder W, Main B, Govender N (2021) Orthorectification of helicopter-borne high resolution experimental burn observation from infra red handheld imagers. Remote Sensing 13(23), 4913.
| Crossref | Google Scholar |

Pimont F, Dupuy J-L, Linn RR (2012) Coupled slope and wind effects on fire spread with influences of fire size: a numerical study using FIRETEC. International Journal of Wildland Fire 21(7), 828-842.
| Crossref | Google Scholar |

Pinto RMS, Benali A, Sá ACL, Fernandes PM, Soares PMM, Cardoso RM, Trigo RM, Pereira JMC (2016) Probabilistic fire spread forecast as a management tool in an operational setting. SpringerPlus 5(1), 1205.
| Crossref | Google Scholar | PubMed |

Riley K, Thompson M (2017) An uncertainty analysis of wildfire modeling. In ‘Natural Hazard Uncertainty Assessment: Modeling and Decision Support’. Chapter 13. Geophysical Monograph 223. (Eds K Riley, P Webley, M Thompson) pp. 193–213. (John Wiley & Sons)

Rochoux M, Emery C, Ricci S, Cuenot B, Trouve A (2014) Towards predictive simulation of wildfire spread at regional scale using ensemble-based data assimilation to correct the fire front position. Fire Safety Journal 11, 1443-1456.
| Crossref | Google Scholar |

Rothermel RC (1972) A mathematical model for predicting fire spread in wildland fuels. Technical Report. Research Paper INT-115. (USDA Forest Service, Intermountain Forest and Range Experiment Station: Ogden, Utah)

Sánchez-Monroy X, Mell WE, Torres-Arenas J, Butler BW (2019) Fire spread upslope: Numerical simulation of laboratory experiments. Fire Safety Journal 108, 102844.
| Crossref | Google Scholar |

Silvani X, Morandini F, Dupuy J-L (2012) Effects of slope on fire spread observed through video images and multiple-point thermal measurements. Experimental Thermal and Fluid Science 41, 99-111.
| Crossref | Google Scholar |

Sullivan AL (2007) Convective Froude number and Byram’s energy criterion of Australian experimental grassland fires. Proceedings of the Combustion Institute 31(2), 2557-2564.
| Crossref | Google Scholar |

Sullivan AL (2009) Wildland surface fire spread modelling, 1990–2007. 2: Empirical and quasi-empirical models. International Journal of Wildland Fire 18, 369-386.
| Crossref | Google Scholar |

Tay BH, Ananda AL (1990) A survey of remote procedure calls. Operating Systems Review 24(3), 68-79.
| Crossref | Google Scholar |

Thomas D, Butry D, Gilbert S, Webb D, Fung J (2017) ‘The costs and losses of wildfires: A literature review’. NIST Special Publication 1215. (National Institute of Standards and Technology)

Valero MM, Jofre L, Torres R (2021) Multifidelity prediction in wildfire spread simulation: modeling, uncertainty quantification and sensitivity analysis. Environmental Modelling & Software 141, 105050.
| Crossref | Google Scholar |

Viegas DX (2004a) A mathematical model for forest fires blowup. Combustion Science and Technology 177(1), 27-51.
| Crossref | Google Scholar |

Viegas DX (2004b) Slope and wind effects on fire propagation. International Journal of Wildland Fire 13(2), 143-156.
| Crossref | Google Scholar |

Wang Q, Ihme M, Chen Y-F, Anderson J (2022) A TensorFlow simulation framework for scientific computing of fluid flows on tensor processing units. Computer Physics Communications 274, 108292.
| Crossref | Google Scholar |

Wang Q, Ihme M, Linn RR, Chen Y-F, Yang V, Sha F, Clements C, McDanold JS, Anderson J (2023) A high-resolution large-eddy simulation framework for wildland fire predictions using TensorFlow. International Journal of Wildland Fire 32(12), 1711-1725.
| Crossref | Google Scholar |

Weise DR, Biging GS (1994) Effects of wind velocity and slope on fire behavior. Fire Safety Science 4, 1041-1051.
| Crossref | Google Scholar |

Weise DR, Biging GS (1996) Effects of wind velocity and slope on flame properties. Canadian Journal of Forest Research 26(10), 1849-1858.
| Crossref | Google Scholar |

Wu H, Levinson D (2021) The ensemble approach to forecasting: a review and synthesis. Transportation Research Part C: Emerging Technologies 132, 103357.
| Crossref | Google Scholar |

Xie X, Liu N, Lei J, Shan Y, Zhang L, Chen H, Yuan X, Li H (2017) Upslope fire spread over a pine needle fuel bed in a trench associated with eruptive fire. Proceedings of the Combustion Institute 36(2), 3037-3044.
| Crossref | Google Scholar |

Yoon J, Kravitz B, Rasch P (2015) Extreme fire season in California: a glimpse into the future?’. Bulletin of the American Meteorological Society 96, S5-S9.
| Crossref | Google Scholar |

Zhang N, Zheng ZC (2007) An improved direct-forcing immersed-boundary method for finite difference applications. Journal of Computational Physics 221(1), 250-268.
| Crossref | Google Scholar |