Characterizing fire history on military land using machine learning and landsat imagery

Maura C. O’Grady; Adam G. Wells; Michael G. Just; Wade A. Wall

doi:10.1071/WF24214

RESEARCH ARTICLE (Open Access)

Previous Next Contents Vol 34(8)

Characterizing fire history on military land using machine learning and landsat imagery

Maura C. O’Grady

^A ^B , Adam G. Wells ^A , Michael G. Just ^A and Wade A. Wall ^A ^*

+ Author Affiliations

- Author Affiliations

^A US Army Corps of Engineers, Engineer Research and Development Center, Construction Engineering Research Laboratory, P.O. Box 9005, Champaign, IL 61826, USA.

^B University of Illinois in Champaign Urbana, Department of Plant Biology, 505 S. Goodwin Avenue, Urbana, IL 61801, USA.

^* Correspondence to: wade.a.wall@usace.army.mil

International Journal of Wildland Fire 34, WF24214 https://doi.org/10.1071/WF24214

Submitted: 12 December 2024 Accepted: 12 July 2025 Published: 6 August 2025

© 2025 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-NonCommercial-NoDerivatives 4.0 International License (CC BY-NC-ND)

Abstract

Background

In the past several decades, United States wildland fire occurrences have increased due to anthropogenic activities, shifts in precipitation and temperature patterns, and long-term fire suppression policies. Detailed records of local fire histories are needed to further understand ignition sources and the interaction between human activity, weather patterns and fire occurrences.

Aim

To estimate local fire histories, we delineate burned area on military installations over decadal time series (1984–2023) of Landsat imagery using random forest and boosted regression tree algorithms.

Methods

We trained and tested each model with 10 images from a manually delineated burned area dataset and applied them to Landsat images acquired from 1984 to 2023. We validated the model’s yearly summaries with the remaining manually delineated burned area dataset and compared success rates through confusion matrices, omission/commission error, sensitivity and specificity.

Key results

The mean accuracy for the random forest models across all four installations was 0.941, while the mean accuracy of boosted regression models was 0.935. There was no significant difference between random forest and boosted regression model performance.

Conclusions

We present a methodology which can be utilized by other Army personnel and local land managers to develop fire histories for local-scale management units, particularly those geared towards national defense institutions.

Keywords: burned area, classification, fire frequency, landsat, local scale, military land, random forest, XGBoost.

References

Ager AA, Day MA, Alcasena FJ, Evers CR, Short KC, Grenfell I (2021) Predicting paradise: modeling future wildfire disasters in the western US. The Science of The Total Environment 784, 147057.
| Crossref | Google Scholar | PubMed |

Bailey R (1980) ‘Description of the Ecoregions of the United States.’ (US Department of Agriculture, Forest Service)

Belgiu M, Drăguţ L (2016) Random forest in remote sensing: a review of applications and future directions. ISPRS Journal of Photogrammetry and Remote Sensing 114, 24-31.
| Crossref | Google Scholar |

Bot K, Borges JG (2022) A systematic review of applications of machine learning techniques for wildfire management decision support. Inventions 7(1), 15.
| Crossref | Google Scholar |

Breiman L (2001) Random forests. Machine Learning 45, 5-32.
| Crossref | Google Scholar |

Calkin DE, Thompson MP, Finney MA (2015) Negative consequences of positive feedbacks in US wildfire management. Forest Ecosystems 2, 1-10.
| Crossref | Google Scholar |

Cattau ME, Mahood AL, Balch JK, Wessman CA (2022) Modern pyromes: biogeographical patterns of fire characteristics across the contiguous United States. Fire 5(4), 95.
| Crossref | Google Scholar |

Chen T, Guestrin C (2016) Xgboost: A scalable tree boosting system. In ‘Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining’. pp. 785–794. (Association for Computing Machinery: New York, NY, USA). 10.1145/2939672.2939785

Chuvieco E, Martín MP, Palacios A (2002) Assessment of different spectral indices in the red-near-infrared spectral domain for burned land discrimination. International Journal of Remote Sensing 23(23), 5103-5110.
| Crossref | Google Scholar |

Chuvieco E, Aguado I, Salas J, García M, Yebra M, Oliva P (2020) Satellite remote sensing contributions to wildland fire science and management. Current Forestry Reports 6, 81-96.
| Crossref | Google Scholar |

Conedera M, Tinner W, Neff C, Meurer M, Dickens AF, Krebs P (2009) Reconstructing past fire regimes: methods, applications, and relevance to fire management and conservation. Quaternary Science Reviews 28(5-6), 555-576.
| Crossref | Google Scholar |

Dewitz J, U.S. Geological Survey (2021) National Land Cover Database (NLCD) 2019 Products (ver. 2.0, June 2021): U.S. Geological Survey data release. 10.5066/P9KZCM54

Duane MV, Cohen WB, Campbell JL, Hudiburg T, Turner DP, Weyermann DL (2010) Implications of alternative field-sampling designs on Landsat-based mapping of stand age and carbon stocks in Oregon forests. Forest Science 56(4), 405-416.
| Crossref | Google Scholar |

ESRI, Inc. (2022) ‘ArcGIS Pro [GIS Software], Version 3.0.0.’ (ESRI: Redlands, CA, USA)

Fagerland MW, Lydersen S, Laake P (2013) The McNemar test for binary matched-pairs data: mid-p and asymptotic are better than exact conditional. BMC Medical Research Methodology 13, 91.
| Crossref | Google Scholar | PubMed |

Forsyth GG, Van WBW (2008) The recent fire history of the Table Mountain National Park and implications for fire management: original research. Koedoe : African Protected Area Conservation and Science 50, 3-9.
| Crossref | Google Scholar |

Freeman J, Kobziar L, Rose EW, Cropper W (2017) A critique of the historical-fire-regime concept in conservation. Conservation Biology 31, 976-985.
| Crossref | Google Scholar | PubMed |

Gavin DG, Hu FS, Lertzman K, Corbett P (2006) Weak climatic control of stand‐scale fire history during the late Holocene. Ecology 87(7), 1722-1732.
| Crossref | Google Scholar | PubMed |

Giglio L, Boschetti L, Roy DP, Humber ML, Justice CO (2018) The Collection 6 MODIS burned area mapping algorithm and product. Remote Sensing of Environment 217, 72-85.
| Crossref | Google Scholar | PubMed |

Goodwin NR, Collett LJ (2014) Development of an automated method for mapping fire history captured in Landsat TM and ETM+time series across Queensland, Australia. Remote Sensing of Environment 148, 206-221.
| Crossref | Google Scholar |

Gorelick N, Hancher M, Dixon M, Ilyushchenko S, Thau D, Moore R (2017) Google Earth Engine: planetary-scale geospatial analysis for everyone. Remote Sensing of Environment 202, 18-27.
| Crossref | Google Scholar |

Hawbaker TJ, Vanderhoof MK, Beal YJ, Takacs JD, Schmidt GL, Falgout JT, Williams B, Fairaux NM, Caldwell MK, Picotte JJ, Howard SM, Stitt S, Dwyer JL (2017) Mapping burned areas using dense time-series of Landsat data. Remote Sensing of Environment 198, 504-522.
| Crossref | Google Scholar |

Hawbaker TJ, Vanderhoof MK, Schmidt GL, Beal YJ, Picotte JJ, Takacs JD, Falgout JT, Dwyer JL (2020) The Landsat Burned Area algorithm and products for the conterminous United States. Remote Sensing of Environment 244, 111801.
| Crossref | Google Scholar |

Hearst MA, Dumais ST, Osuna E, Platt J, Scholkopf B (1998) Support vector machines. IEEE Intelligent Systems and their Applications 13, 18-28.
| Crossref | Google Scholar |

Hoffmann WA, Orthen B, Nascimento PKVd (2003) Comparative fire ecology of tropical savanna and forest trees. Functional Ecology 17(6), 720-726.
| Crossref | Google Scholar |

Hohmann MG, Wall WA, Just MG, Huskins SD (2023) Multiple intrinsic and extrinsic drivers influence the quantity and quality components of seed dispersal effectiveness in the rare shrub Lindera subcoriacea. PLoS One 18, e0283810.
| Crossref | Google Scholar | PubMed |

Just MG, Hohmann MG, Hoffmann WA (2016) Where fire stops: vegetation structure and microclimate influence fire spread along an ecotonal gradient. Plant Ecology 217, 631-644.
| Crossref | Google Scholar |

Just MG, Wall WA, Huskins SD, Hohmann MG (2024) Effects of landscape heterogeneity and disperser movement on seed dispersal. Ecologies 5(2), 198-217.
| Crossref | Google Scholar |

Kauth RJ, Thomas GS (1976) The Tasselled Cap—A Graphic Description of the Spectral-Temporal Development of Agricultural Crops as Seen by LANDSAT. (LARS Symposia) http://docs.lib.purdue.edu/lars_symp/159

Kelly LT, Bennett AF, Clarke MF, McCarthy MA (2015) Optimal fire histories for biodiversity conservation. Conservation Biology 29, 473-481.
| Crossref | Google Scholar | PubMed |

Krebs P, Pezzatti GB, Mazzoleni S, et al. (2010) Fire regime: history and definition of a key concept in disturbance ecology. Theory in Biosciences 129, 53-69.
| Crossref | Google Scholar | PubMed |

Kuhn M (2008) Building predictive models in R using the caret package. Journal of Statistical Software 28(5), 1-26.
| Crossref | Google Scholar |

Kuhn M, Johnson K (2013) ‘Applied Predictive Modeling.’ (Springer: New York, NY, USA). 10.1007/978-1-4614-6849-3

Kyser GB, Wilson RG, Zhang J, DiTomaso JM (2013) Herbicide-assisted restoration of Great Basin sagebrush steppe infested with medusahead and downy brome. Rangeland Ecology & Management 66(5), 588-596.
| Google Scholar |

Leach MK, Givnish TJ (1999) Gradients in the composition, structure, and diversity of remnant oak savannas in southern Wisconsin. Ecological Monographs 69(3), 353-374.
| Crossref | Google Scholar |

Lee Y, Han D, Ahn MH, Im J, Lee SJ (2019) Retrieval of total precipitable water from Himawari-8 AHI data: a comparison of random forest, extreme gradient boosting, and deep neural network. Remote Sensing 11(15), 1741.
| Crossref | Google Scholar |

Lertzman K, Fall J (1998) From forest stands to landscapes: spatial scales and the roles of disturbances. In ‘Ecological Scale: Theory and Applications’. (Eds DL Peterson and V Thomas Parker) pp. 339–367. (Columbia University Press: New York, NY, USA)

Likas A, Vlassis N, Verbeek JJ (2003) The global k-means clustering algorithm. Pattern Recognition 36, 451-461.
| Crossref | Google Scholar |

Liu J, Heiskanen J, Maeda EE, Pellikka PK (2018) Burned area detection based on Landsat time series in savannas of southern Burkina Faso. International Journal of Applied Earth Observation and Geoinformation 64, 210-220.
| Crossref | Google Scholar |

López García MJ, Caselles V (1991) Mapping burns and natural reforestation using thematic Mapper data. Geocarto International 6(1), 31-37.
| Crossref | Google Scholar |

Marlon JR, Bartlein PJ, Gavin DG, Long CJ, Anderson RS, Briles CE, Brown KJ, Colombaroli D, Hallett DJ, Power MJ, Scharf EA, Walsh MK (2012) Long-term perspective on wildfires in the western USA. Proceedings of the National Academy of Sciences 109(9), 535-43.
| Crossref | Google Scholar | PubMed |

Masek JG, Vermote EF, Saleous NE, Wolfe R, Hall FG, Huemmrich KF, Lim TK (2006) A Landsat surface reflectance dataset for North America, 1990-2000. IEEE Geoscience and Remote Sensing Letters 3(1), 68-72.
| Crossref | Google Scholar |

Nagy RC, Fusco E, Bradley B, Abatzoglou JT, Balch J (2018) Human-related ignitions increase the number of large wildfires across US ecoregions. Fire 1(1), 4.
| Crossref | Google Scholar |

Pedregosa F, Varoquaux G, Gramfort A, Michel V, Thirion B, Grisel O (2011) Scikit-learn: machine learning in Python. Journal of Machine Learning Research 12, 2825-2830.
| Google Scholar |

Powell SL, Cohen WB, Healey SP, Kennedy RE, Moisen GG, Pierce KB, Ohmann JL (2010) Quantification of live aboveground forest biomass dynamics with Landsat time-series and field inventory data: a comparison of empirical modeling approaches. Remote Sensing of Environment 114(5), 1053-1068.
| Crossref | Google Scholar |

Price RA, Bourne M (2011) Effects of Wildfire and Prescribed Burning on Distributed Particles of Composition-B Explosive on Training Ranges. In ‘Environmental Chemistry of Explosives and Propellant Compounds in Soils and Marine Systems: Distributed Source Characterization and Remedial Technologies, Vol. 1069’. (Eds M Chappell, C Price, R George) pp. 363–377. (American Chemical Society) 10.1021/bk-2011-1069.ch019

Ramo R, Chuvieco E (2017) Developing a random forest algorithm for MODIS global burned area classification. Remote Sensing 9(11), 1193.
| Crossref | Google Scholar |

R Core Team (2023) R: A Language and Environment for Statistical Computing, Version 4.3.0. R Foundation for Statistical Computing, Vienna, Austria. Available at https://www.R-project.org/

Rollins MG (2009) LANDFIRE: a nationally consistent vegetation, wildland fire, and fuel assessment. International Journal of Wildland Fire 18(3), 235.
| Crossref | Google Scholar |

Roy DP, Huang H, Boschetti L, Giglio L, Yan L, Zhang HH, Li Z (2019) Landsat-8 and Sentinel-2 burned area mapping - A combined sensor multi-temporal change detection approach. Remote Sensing of Environment 231, 111254.
| Crossref | Google Scholar |

Royal EJ, Kross CS, Willson JD (2022) Legacy land use predicts occupancy patterns of prairie-associated herpetofauna in Western Arkansas. Landscape Ecology 38(2), 423-438.
| Crossref | Google Scholar |

Sahin EK (2020) Assessing the predictive capability of ensemble tree methods for landslide susceptibility mapping using XGBoost, gradient boosting machine, and random forest. SN Applied Sciences 2(7), 1308.
| Crossref | Google Scholar |

Santos JL, Hradsky BA, Keith DA, Rowe K, Senior KL, Sitters H, Kelly LT (2022) Beyond inappropriate fire regimes: a synthesis of fire-driven declines of threatened mammals in Australia. Conservation Letters 15, e12905.
| Crossref | Google Scholar |

Sayler K, Zanter K (2021) ‘Landsat 8 Collection 2 (C2) Level 2 Science Product (L2SP) Guide LSDS-1619 Version 2.0.’ (EROS Sioux Falls: South Dakota, USA)

Senande-Rivera M, Insua-Costa D, Miguez-Macho G (2022) Spatial and temporal expansion of global wildland fire activity in response to climate change. Nature Communications 13, 1208.
| Crossref | Google Scholar | PubMed |

Shao Z, Ahmad MN, Javed A (2024) Comparison of Random Forest and XGBoost Classifiers using Integrated Optical and SAR Features for mapping urban impervious surface. Remote Sensing 16(4), 665.
| Crossref | Google Scholar |

Stein BA, Scott C, Benton N (2008) Federal lands and endangered species: the role of military and other federal lands in sustaining biodiversity. BioScience 58(4), 339-347.
| Crossref | Google Scholar |

Syphard AD, Keeley JE (2015) Location, timing and extent of wildfire vary by cause of ignition. International Journal of Wildland Fire 24, 37-47.
| Crossref | Google Scholar |

Trigg S, Flasse S (2001) An evaluation of different bi-spectral spaces for discriminating burned shrub-savannah. International Journal of Remote Sensing 22(13), 2641-2647.
| Crossref | Google Scholar |

Tucker CJ (1979) Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment 8(2), 127-150.
| Crossref | Google Scholar |

U.S. Geological Survey (2015) Landsat surface reflectance data (ver. 1.1, 27 March, 2019). U.S. Geological Survey Fact Sheet 2015-3034, 1 p. 10.3133/fs20153034

Van Rossum G, Drake Jr FL (1995) ‘Python Reference Manual.’ (Centrum voor Wiskunde en Informatica: Amsterdam, Netherlands)

Veraverbeke S, Lhermitte S, Verstraeten WW, Goossens R (2011) Evaluation of pre/post-fire differenced spectral indices for assessing burn severity in a Mediterranean environment with Landsat Thematic Mapper. International Journal of Remote Sensing 32(12), 3521-3537.
| Crossref | Google Scholar |

Wall WA, Hohmann MG, Just MG, Hoffmann WA (2021a) Characterizing past fire occurrence in longleaf pine ecosystems with the Mid-Infrared Burn Index and a Random Forest classifier. Forest Ecology and Management 500, 119635.
| Crossref | Google Scholar |

Wall WA, Walker AS, Gray JB, Hohmann MG (2021b) Fire effects on the vital rates and stochastic population growth rate of the rare shrub Lindera subcoriacea Wofford. Plant Ecology 222, 119-131.
| Crossref | Google Scholar |

Westerling AL, Hidalgo HG, Cayan DR, Swetnam TW (2006) Warming and earlier spring increase western US forest wildfire activity. Science 313(5789), 940-943.
| Crossref | Google Scholar | PubMed |

Wulder MA, Masek JG, Cohen WB, Loveland TR, Woodcock CE (2012) Opening the archive: how free data has enabled the science and monitoring promise of Landsat. Remote Sensing of Environment 122, 2-10.
| Crossref | Google Scholar |