International Journal of Information Technology and Computer Science (IJITCS)
IJITCS Vol. 17, No. 1, 8 Feb. 2025
Cover page and Table of Contents: PDF (size: 1804KB)
Machine Learning based Wildfire Area Estimation Leveraging Weather Forecast Data
PDF (1804KB), PP.1-15
Views: 0 Downloads: 0
Author(s)
Index Terms
Wildfires, Area Estimation, Weather Forecast Data, Machine Learning, Ensemble Learning, Burned Area Prediction
Abstract
Wildfires are increasingly destructive natural disasters, annually consuming millions of acres of forests and vegetation globally. The complex interactions among fuels, topography, and meteorological factors, including temperature, precipitation, humidity, and wind, govern wildfire ignition and spread. This research presents a framework that integrates satellite remote sensing and numerical weather prediction model data to refine estimations of final wildfire sizes. A key strength of our approach is the use of comprehensive geospatial datasets from the IBM PAIRS platform, which provides a robust foundation for our predictions. We implement machine learning techniques through the AutoGluon automated machine learning toolkit to determine the optimal model for burned area prediction. AutoGluon automates the process of feature engineering, model selection, and hyperparameter tuning, evaluating a diverse range of algorithms, including neural networks, gradient boosting, and ensemble methods, to identify the most effective predictor for wildfire area estimation. The system features an intuitive interface developed in Gradio, which allows the incorporation of key input parameters, such as vegetation indices and weather variables, to customize wildfire projections. Interactive Plotly visualizations categorize the predicted fire severity levels across regions. This study demonstrates the value of synergizing Earth observations from spaceborne instruments and forecast data from numerical models to strengthen real-time wildfire monitoring and postfire impact assessment capabilities for improved disaster management. We optimize an ensemble model by comparing various algorithms to minimize the root mean squared error between the predicted and actual burned areas, achieving improved predictive performance over any individual model. The final metric reveals that our optimized WeightedEnsemble model achieved a root mean squared error (RMSE) of 1.564 km2 on the test data, indicating an average deviation of approximately 1.2 km2 in the predictions.
Cite This Paper
Saket Sultania, Rohit Sonawane, Prashasti Kanikar, "Machine Learning based Wildfire Area Estimation Leveraging Weather Forecast Data", International Journal of Information Technology and Computer Science(IJITCS), Vol.17, No.1, pp.1-15, 2025. DOI:10.5815/ijitcs.2025.01.01
Reference
[1]“Wildfire Statistics.” Accessed: Aug. 03, 2024. [Online]. Available: https://crsreports.congress.gov/product/pdf/IF/IF10244/59
[2]‘2023: A year of intense global wildfire activity | Copernicus’. Accessed: Aug. 03, 2024. [Online]. Available: https://atmosphere.copernicus.eu/2023-year-intense-global-wildfire-activity
[3]‘Global fire monitoring | Copernicus’. Accessed: Aug. 03, 2024. [Online]. Available: https://atmosphere.copernicus.eu/global-fire-monitoring
[4]H. Setchell, ‘Global Fire Assimilation System’, ECMWF. Accessed: Aug. 03, 2024. [Online]. Available: https://www.ecmwf.int/en/forecasts/dataset/global-fire-assimilation-system
[5]H. Liang, M. Zhang, and H. Wang, ‘A Neural Network Model for Wildfire Scale Prediction Using Meteorological Factors’, IEEE Access, vol. 7, pp. 176746–176755, 2019, doi: 10.1109/ACCESS.2019.2957837.
[6]M. R. Nosouhi, K. Sood, N. Kumar, T. Wevill, and C. Thapa, ‘Bushfire Risk Detection Using Internet of Things: An Application Scenario’, IEEE Internet Things J., vol. 9, no. 7, pp. 5266–5274, Apr. 2022, doi: 10.1109/JIOT.2021.3110256.
[7]N. Sazib, J. D. Bolten, and I. E. Mladenova, ‘Leveraging NASA Soil Moisture Active Passive for Assessing Fire Susceptibility and Potential Impacts Over Australia and California’, IEEE J. Sel. Top. Appl. Earth Obs. Remote Sens., vol. 15, pp. 779–787, 2022, doi: 10.1109/JSTARS.2021.3136756.
[8]F. N. Ismail and S. Amarasoma, ‘One-class Classification-Based Machine Learning Model for Estimating the Probability of Wildfire Risk’, Procedia Comput. Sci., vol. 222, pp. 341–352, Jan. 2023, doi: 10.1016/j.procs.2023.08.173.
[9]A. Zaidi, ‘Predicting wildfires in Algerian forests using machine learning models’, Heliyon, vol. 9, no. 7, p. e18064, Jul. 2023, doi: 10.1016/j.heliyon.2023.e18064.
[10]Y. O. Sayad, H. Mousannif, and H. Al Moatassime, ‘Predictive modeling of wildfires: A new dataset and machine learning approach’, Fire Saf. J., vol. 104, pp. 130–146, Mar. 2019, doi: 10.1016/j.firesaf.2019.01.006.
[11]T. Preeti, S. Kanakaraddi, A. Beelagi, S. Malagi, and A. Sudi, ‘Forest Fire Prediction Using Machine Learning Techniques’, in 2021 International Conference on Intelligent Technologies (CONIT), Jun. 2021, pp. 1–6. doi: 10.1109/CONIT51480.2021.9498448.
[12]M. Naderpour, H. M. Rizeei, and F. Ramezani, ‘Wildfire Prediction: Handling Uncertainties Using Integrated Bayesian Networks and Fuzzy Logic’, in 2020 IEEE International Conference on Fuzzy Systems (FUZZ-IEEE), Jul. 2020, pp. 1–7. doi: 10.1109/FUZZ48607.2020.9177700.
[13]A. Malik, N. Jalin, S. Rani, P. Singhal, S. Jain, and J. Gao, ‘Wildfire Risk Prediction and Detection using Machine Learning in San Diego, California’, in 2021 IEEE SmartWorld, Ubiquitous Intelligence & Computing, Advanced & Trusted Computing, Scalable Computing & Communications, Internet of People and Smart City Innovation (SmartWorld/SCALCOM/UIC/ATC/IOP/SCI), Oct. 2021, pp. 622–629. doi: 10.1109/SWC50871.2021.00092.
[14]A. S. Oliver, U. Ashwanthika, and R. Aswitha, ‘Detection of Forest Fire using Convolutional Neural Networks’, in 2020 7th International Conference on Smart Structures and Systems (ICSSS), Jul. 2020, pp. 1–6. doi: 10.1109/ICSSS49621.2020.9202200.
[15]K. Revathi, T. Tamilselvi, R. Arunkumar, and T. Divya, ‘Spot Fire: An Intelligent Forest Fire Detection System Design With Machine Learning’, in 2022 International Conference on Automation, Computing and Renewable Systems (ICACRS), Dec. 2022, pp. 532–537. doi: 10.1109/ICACRS55517.2022.10029044.
[16]A. Singh, R. Yadav, G. Sudhamshu, A. Basnet, and R. Ali, ‘Wildfire Spread Prediction using Machine Learning Algorithms’, in 2023 14th International Conference on Computing Communication and Networking Technologies (ICCCNT), Jul. 2023, pp. 1–5. doi: 10.1109/ICCCNT56998.2023.10308041.
[17]V. Bharathi and C. PeddaReddy, ‘Comparison of Forest Fire Prediction System using Machine Learning Alogrithms’, in 2023 3rd International Conference on Advance Computing and Innovative Technologies in Engineering (ICACITE), May 2023, pp. 1539–1542. doi: 10.1109/ICACITE57410.2023.10182818.
[18]C. Ran and L. Fang, ‘Prediction model of forest fire area based on the improved Extreme Gradient Boosting’, in 2021 2nd International Conference on Computer Communication and Network Security (CCNS), Jul. 2021, pp. 16–19. doi: 10.1109/CCNS53852.2021.00011.
[19]N. Omar, A. Al-zebari, and A. Sengur, ‘Deep Learning Approach to Predict Forest Fires Using Meteorological Measurements’, in 2021 2nd International Informatics and Software Engineering Conference (IISEC), Dec. 2021, pp. 1–4. doi: 10.1109/IISEC54230.2021.9672446.
[20]H. Hamann and J. S. May 2021, ‘Predicting Wildfires with Weather Forecast Data’, IBM Developer. Accessed: Sep. 07, 2024. [Online]. Available: https://developer.ibm.com/exchanges/data/all/spot-challenge-wildfires/
[21]V. Vinodhini, M. R. S. Kumar, S. Sankar, D. Pandey, B. K. Pandey, and V. K. Nassa, ‘IoT-based early forest fire detection using MLP and AROC method’, Int. J. Glob. Warm., vol. 27, no. 1, pp. 55–70, Jan. 2022, doi: 10.1504/IJGW.2022.122794.
[22]D. Gao, J. Xin, and F. Zhang, ‘A decision tree algorithm for forest fire prediction based on wireless sensor networks’, Int. J. Embed. Syst., vol. 13, no. 4, pp. 422–430, Jan. 2020, doi: 10.1504/IJES.2020.110656.