Lilian Berton
Didier Vega Oliveros
ABSTRACT
Wildfires happen in almost all biomes worldwide, and they can cause profound environmental and socio-economical impacts. Fire regimes are emergent properties of a complex system (e.g., climate-vegetation-fire interactions), which are influenced by many factors, such as climate warming, geography, vegetation, and human activity. In particular, anthropogenic land use can alter wildfires' incidence, size, season, and severity. Consequently, it is paramount to understand fire regimes to characterize similar spatiotemporal wildfire activity between natural or human ignition dynamics. This study proposes a framework combining complex networks and machine learning to classify similar wildfire regions. From global satellite data over the past two decades, we extract graph-based features to characterize spatiotemporal fire regimes from different Tropical/Temperate/Boreal forests around the globe. With the extracted high-level features, we can classify how dynamically similar the wildfire regimes are between six studied regions: Amazon basin (AMZ), Australia (AUS), northwest U.S. (NUS), northeast China (NCN), and central Africa (CAF) and Asia (CAS). Results from well-known classification algorithms indicate that the AMZ and CAF wildfire dynamics are more similar than the other regions. Although we found that wildfires in NUS are more similar to the AUS regime, NUS has around 33% of similar behavior to wildfires in the AMZ.
- 2020. MCD14DL - MODIS/Aqua+Terra Thermal Anomalies/Fire locations 1km FIRMS V006 NRT. Google Scholar
Cross Ref
- N. Andela, D. C. Morton, L. Giglio, Y. Chen, G. R. van der Werf, P. S. Kasibhatla, R. S. DeFries, G. J. Collatz, S. Hantson, S. Kloster, D. Bachelet, M. Forrest, G. Lasslop, F. Li, S. Mangeon, J. R. Melton, C. Yue, and J. T. Randerson. 2017. A human-driven decline in global burned area. Science 356, 6345 (2017), 1356--1362. Google ScholarCross Ref
- Sally Archibald, Caroline E. R. Lehmann, Jose L. Gómez-Dans, and Ross A. Bradstock. 2013. Defining pyromes and global syndromes of fire regimes. Proceedings of the National Academy of Sciences 110, 16 (2013), 6442--6447. Google ScholarCross Ref
- Rui Ba, Chen Chen, Jing Yuan, Weiguo Song, and Siuming Lo. 2019. SmokeNet: Satellite smoke scene detection using convolutional neural network with spatial and channel-wise attention. Remote Sensing 11, 14 (2019), 1702.Google ScholarCross Ref
- Lilian Berton, Alneu de Andrade Lopes, and Didier A Vega-Oliveros. 2018. A comparison of graph construction methods for semi-supervised learning. In 2018 International Joint Conference on Neural Networks (IJCNN). IEEE, 1--8.Google ScholarCross Ref
- Matthias M Boer, Victor Resco De Dios, Elisa Z Stefaniak, and Ross A Bradstock. 2021. A hydroclimatic model for the distribution of fire on Earth. Environmental Research Communications 3, 3 (mar 2021), 035001. Google ScholarCross Ref
- Geographic Area Coordination Center. 2020. National Large Incident Year to Date Report. https://gacc.nifc.gov/sacc/Google Scholar
- Olivier Chapelle, Bernhard Schlkopf, and Alexander Zien. 2010. Semi-Supervised Learning (1st ed.). The MIT Press.Google ScholarDigital Library
- Matheus A. De Castro Santos, Didier A. Vega-Oliveros, Liang Zhao, and Lilian Berton. 2020. Classifying El Niño-Southern Oscillation Combining Network Science and Machine Learning. IEEE Access 8 (2020), 55711--55723. Google ScholarCross Ref
- Leonardo N Ferreira, Didier A Vega-Oliveros, Moshé Cotacallapa, Manoel F Cardoso, Marcos G Quiles, Liang Zhao, and Elbert EN Macau. 2020. Spatiotemporal data analysis with chronological networks. Nature Communications 11, 1 (2020), 1--11.Google ScholarCross Ref
- Leonardo N. Ferreira, Didier A. Vega-Oliveros, Liang Zhao, Manoel F. Cardoso, and Elbert E.N. Macau. 2020. Global fire season severity analysis and forecasting. Computers & Geosciences 134 (2020), 104339.Google ScholarDigital Library
- Mike Flannigan. 2015. Fire evolution split by continent. Nature Geoscience 8 (Mar 2015), 167--168. Google ScholarCross Ref
- Omid Ghorbanzadeh, Khalil Valizadeh Kamran, Thomas Blaschke, Jagannath Aryal, Amin Naboureh, Jamshid Einali, and Jinhu Bian. 2019. Spatial prediction of wildfire susceptibility using field survey GPS data and machine learning approaches. Fire 2, 3 (2019), 43.Google ScholarCross Ref
- Louis Giglio, Wilfrid Schroeder, and Christopher O. Justice. 2016. The collection 6 MODIS active fire detection algorithm and fire products. Remote Sensing of Environment 178 (2016), 31 -- 41. Google ScholarCross Ref
- Nature Publishing Group. 2019. The complexities of wildfires. Nature Geoscience 12 (Feb 2019), 81. Google ScholarCross Ref
- Piyush Jain, Sean CP Coogan, Sriram Ganapathi Subramanian, Mark Crowley, Steve Taylor, and Mike D Flannigan. 2020. A review of machine learning applications in wildfire science and management. Environmental Reviews 28, 4 (2020), 478--505.Google ScholarCross Ref
- Janice Nelson. 2020. Geoscience Australia's Oliver Discusses Use of Landsat during Country's Historic Fires. https://www.usgs.gov/center-news/Google Scholar
- M. E. J. Newman. 2010. Networks: an introduction. Oxford University Press, Oxford; New York.Google Scholar
- Australian Government Bureau of Metereology. 2021. Bushfire weather. http://www.bom.gov.au/weather-services/fire-weather-centre/bushfire-weather/index.shtmlGoogle Scholar
- Thanh Cong Phan and Thanh Tam Nguyen. 2019. Remote Sensing Meets Deep Learning: Exploiting Spatio-Temporal-Spectral Satellite Images for Early Wildfire Detection. Technical Report.Google Scholar
- Zixing Song, Xiangli Yang, Zenglin Xu, and Irwin King. 2022. Graph-based semi-supervised learning: A comprehensive review. IEEE Transactions on Neural Networks and Learning Systems (2022).Google Scholar
- Mahyat Shafapour Tehrany, Simon Jones, Farzin Shabani, Francisco Martínez-Álvarez, and Dieu Tien Bui. 2019. A novel ensemble modeling approach for the spatial prediction of tropical forest fire susceptibility using logitboost machine learning classifier and multi-source geospatial data. Theoretical and Applied Climatology 137, 1 (2019), 637--653.Google ScholarCross Ref
- Marj Tonini, Mirko D'Andrea, Guido Biondi, Silvia Degli Esposti, Andrea Trucchia, and Paolo Fiorucci. 2020. A Machine Learning-Based Approach for Wildfire Susceptibility Mapping. The Case Study of the Liguria Region in Italy. Geosciences 10, 3 (2020), 105.Google ScholarCross Ref
- Didier A Vega-Oliveros, Moshé Cotacallapa, Leonardo N Ferreira, Marcos G Quiles, Liang Zhao, Elbert EN Macau, and Manoel F Cardoso. 2019. From spatio-temporal data to chronological networks: An application to wildfire analysis. In Proceedings of the 34th ACM/SIGAPP Symposium on Applied Computing. 675--682.Google ScholarDigital Library
- Wikipedia. 2019. 2019 Amazon rainforest wildfires. https://en.wikipedia.org/wiki/2019_Amazon_rainforest_wildfiresGoogle Scholar
- Wikipedia. 2020. 2020 California wildfires. https://en.wikipedia.org/wiki/2020_California_wildfiresGoogle Scholar
- Yuto Yamaguchi, Christos Faloutsos, and Hiroyuki Kitagawa. 2016. Camlp: Confidence-aware modulated label propagation. In SIAM International Conference on Data Mining. SIAM, 513--521.Google ScholarCross Ref
- Yi Zhao, Jiale Ma, Xiaohui Li, and Jie Zhang. 2018. Saliency detection and deep learning-based wildfire identification in UAV imagery. Sensors 18, 3 (2018), 712.Google ScholarCross Ref
Index Terms
- Graph-based semi-supervised classification for similar wildfire dynamics
Recommendations
A Bayesian belief network analysis of factors influencing wildfire occurrence in Swaziland
The impacts of wildfires on ecosystems and the factors contributing to their occurrence are increasingly receiving global attention. Advances in satellite remote sensing and information technology provide an opportunity to study these complex ...
Assessment of the Ecological Destruction of Wildfire Based on Remote Sensing Image
ICIGP '21: Proceedings of the 2021 4th International Conference on Image and Graphics ProcessingThe largest forest fire in history broke out in Australia in July 2019. The wildfire burned to Kangaroo Island in Australia on January 3, 2020, bringing a serious ecological disaster to Kangaroo Island. In order to analyze the severity of damage caused ...
The Dynamic Temperate and Boreal Fire and Forest-Ecosystem Simulator (DYNAFFOREST): Development and evaluation
AbstractFire is a dominant disturbance in temperate and boreal biomes, and increasing burned area with climate change may fundamentally alter forests. Improved information about how fire-induced changes to forests may feedback to affect ...
Highlights- Feedbacks between fires and forest are not well captured in broad-scale models.
Comments