An enhanced method for predicting and analysing forest fires using an attention-based CNN model

Abstract

Prediction, prevention, and control of forest fires are crucial on at all scales. Developing effective fire detection systems can aid in their control. This study proposes a novel CNN (convolutional neural network) using an attention blocks module which combines an attention module with numerous input layers to enhance the performance of neural networks. The suggested model focuses on predicting the damage affected/burned areas due to possible wildfires and evaluating the multilateral interactions between the pertinent factors. The results show the impacts of CNN using attention blocks for feature extraction and to better understand how ecosystems are affected by meteorological factors. For selected meteorological data, RMSE 12.08 and MAE 7.45 values provide higher predictive power for selecting relevant and necessary features to provide optimal performance with less operational and computational costs. These findings show that the suggested strategy is reliable and effective for planning and managing fire-prone regions as well as for predicting forest fire damage.

Keywords

CNN / Attention module / Fire prediction / Ecosystem / Damage prediction

References

[1]	Al Janabi S, Al Shourbaji I, Salman MA (2018) Assessing the suitability of soft computing approaches for forest fire prediction. Appl Comput Inform 14:214–224. https://doi.org/10.1016/J.ACI.2017.09.006
[2]	Alonso-Betanzos A, Fontenla-Romero O, Guijarro-Berdi?as B, Hernández-Pereira E, Paz-Andrade MI, Jimenez E, Legido JL, Carballas T (2003) An intelligent system for forest fire risk prediction and firefighting management in Galicia. Expert Syst Appl 25:545–554
[3]	Arrue BC, Ollero A, Ramiro J (2000) An intelligent system for false alarm reduction in infrared forest-fire detection. IEEE Intell Syst Appl 15:64–73
[4]	Bolton T, Zanna L (2019) Applications of deep learning to ocean data inference and subgrid parameterization. J Adv Model Earth Syst 11:376–399. https://doi.org/10.1029/2018MS001472
[5]	Camp A, Oliver C, Hessburg P, Everett R (1997) Predicting late-successional fire refugia pre-dating European settlement in the Wenatchee mountains. For Ecol Manag 95:63–77. https://doi.org/10.1016/S0378-1127(97)00006-6
[6]	Coffield SR, Graff CA, Chen Y (2019) Machine learning to predict final fire size at the time of ignition. Int J Wildland Fire 28:861–873. https://doi.org/10.1071/WF19023
[7]	Cortez P, Morais A (2007) A data mining approach to predict forest fires using meteorological data. In: Portuguese Conference on Artificial Intelligence. Guimar?es, Portugal, pp 512–523
[8]	Dimitrakopoulos AP, Vlahou M, Anagnostopoulou CG (2011) Impact of drought on wildland fires in Greece: implications of climate change? Clim Chang 109:331–347
[9]	Guo Y, Liu Y, Oerlemans A (2016) Deep learning for visual understanding: a review. Neurocomputing 187:27–48. https://doi.org/10.1016/J.NEUCOM.2015.09.116
[10]	Hinton GE, Salakhutdinov RR (2006) Reducing the dimensionality of data with neural networks. Science 313:504–507
[11]	Hodges JL, Lattimer BY (2019) Wildland fire spread modeling using convolutional neural networks. Fire Technol 55:2115–2142. https://doi.org/10.1007/S10694-019-00846-4
[12]	Jafari Goldarag Y, Mohammadzadeh A, Ardakani AS (2016) Fire risk assessment using neural network and logistic regression. J Indian Soc Remote Sens 44:885–894. https://doi.org/10.1007/S12524-016-0557-6
[13]	Júnior JSS, Paulo JR, Mendes J (2022) Automatic forest fire danger rating calibration: exploring clustering techniques for regionally customizable fire danger classification. Expert Syst Appl 193:116380. https://doi.org/10.1016/J.ESWA.2021.116380
[14]	Kala CP (2023) Environmental and socioeconomic impacts of forest fires: a call for multilateral cooperation and management interventions. Nat Hazards Res 3:286–294. https://doi.org/10.1016/J.NHRES.2023.04.003
[15]	Koutsias N, Xanthopoulos G, Founda D (2013) On the relationships between forest fires and weather conditions in Greece from long-term national observations (1894–2010). Int J Wildland Fire 22:493–507
[16]	Krizhevsky A, Sutskever I, Hinton GE (2017) ImageNet classification with deep convolutional neural networks. Commun ACM 60:84–90. https://doi.org/10.1145/3065386
[17]	Li Z, Huang Y, Li X, Xu L (2021) Wildland fire burned areas prediction using long short-term memory neural network with attention mechanism. Fire Technol 57:1–23. https://doi.org/10.1007/S10694-020-01028-3/FIGURES/3
[18]	Liu J, Gong X (2019) Attention mechanism enhanced LSTM with residual architecture and its application for protein-protein interaction residue pairs prediction. BMC Bioinform 20:1–11. https://doi.org/10.1186/S12859-019-3199-1
[19]	Nebot à, Mugica F, Pellizzaro G (2021) Forest fire forecasting using fuzzy logic models. Forests 12:1005. https://doi.org/10.3390/F12081005
[20]	North MP, Stephens SL, Collins BM, Agee JK, Aplet G, Franklin JF, Fulé PZ (2015) Reform forest fire management. Science 349(6254):1280–1281. https://doi.org/10.1126/science.aab2356
[21]	Oliveira S, Oehler F, San-Miguel-Ayanz J (2012) Modeling spatial patterns of fire occurrence in Mediterranean Europe using multiple regression and random forest. For Ecol Manag 275:117–129. https://doi.org/10.1016/J.FORECO.2012.03.003
[22]	Pausas JG, Fernandez-Munoz S (2012) Fire regime changes in the Western Mediterranean Basin: from fuel-limited to draught-driven fire regime. Clim Chang 110:215–226
[23]	Peng Y, Wang Y (2019) Real-time forest smoke detection using hand-designed features and deep learning. Comput Electron Agric 167:105029. https://doi.org/10.1016/J.COMPAG.2019.105029
[24]	Pourghasemi H, reza, Beheshtirad M, Pradhan B, (2016) A comparative assessment of prediction capabilities of modified analytical hierarchy process (M-AHP) and Mamdani fuzzy logic models using Netcad-GIS for forest fire susceptibility mapping. Geomat Nat Haz Risk 7:861–885. https://doi.org/10.1080/19475705.2014.984247
[25]	Sakr GE, Elhajj IH, Mitri G, Wejinya UC 2010 Artificial intelligence for forest fire prediction. In: IEEE/ASME international conference on advanced intelligent mechatronics, AIM 1311–1316
[26]	Silva SJ, Heald CL, Ravela S, Mammarella I, Munger JW (2019) A deep learning parameterization for ozone dry deposition velocities. Geophys Res Lett 46:983–989. https://doi.org/10.1029/2018GL081049
[27]	Subramanian SG, Crowley M (2018) Using spatial reinforcement learning to build forest wildfire dynamics models from satellite images. Front ICT 5:6. https://doi.org/10.3389/FICT.2018.00006
[28]	Taylor SW, Alexander ME (2006) Science, technology, and human factors in fire danger rating: THE Canadian experience. Int J Wildland Fire 15:121–135. https://doi.org/10.1071/WF05021
[29]	Tien Bui D, Pradhan B, Nampak H, Quang-Thanh B, Quynh-An T, Quoc-Phi N (2016) Hybrid artificial intelligence approach based on neural fuzzy inference model and metaheuristic optimization for flood susceptibility modeling in a high-frequency tropical cyclone area using GIS. J Hydrol (amst) 540:317–330. https://doi.org/10.1016/J.JHYDROL.2016.06.027
[30]	Tien Bui D, Van LH, Hoang ND (2018) GIS-based spatial prediction of tropical forest fire danger using a new hybrid machine learning method. Ecol Inform 48:104–116. https://doi.org/10.1016/j.ecoinf.2018.08.008
[31]	Turco M, Bedia J, Di Liberto F, Fiorucci P, von Hardenberg J, Koutsias N, Llasat MC, Xystrakis F, Provenzale A (2016) Decreasing fires in Mediterranean Europe. PLoS ONE 11(3):e0150663. https://doi.org/10.1371/JOURNAL.PONE.0150663
[32]	Vaswani A, Shazeer N (2017) Attention is all you need. Adv Neural Inf Process Syst 30:6000–6010
[33]	Wang L, Zhao Q, Wen Z, Qu J (2018) RAFFIA: Short-term forest fire danger rating prediction via multiclass logistic regression. Sustainability 10(12):4620. https://doi.org/10.3390/su10124620
[34]	Wang Y, Dang L, Ren J (2019) Forest fire image recognition based on convolutional neural network. J Algorithm Comput Technol 13:1748302619887689. https://doi.org/10.1177/1748302619887689
[35]	Woo S, Park J, Lee JY, Kweon IS (2018) CBAM: Convolutional block attention module. In: Ferrari, V., Hebert, M., Sminchisescu, C., Weiss, Y. (eds) Computer Vision–ECCV 2018. Lecture notes in computer science, 11211. Springer, Cham, https://doi.org/10.1007/978-3-030-01234-2_1
[36]	Wu Z, He HS, Yang J (2014) Relative effects of climatic and local factors on fire occurrence in boreal forest landscapes of northeastern China. Sci Total Environ 493:472–480. https://doi.org/10.1016/J.SCITOTENV.2014.06.011
[37]	Yang X, Wang Y, Liu X, Liu Y (2022) High-precision real-time forest fire video detection using one-class model. Forests 13:18–26
[38]	Ying L, Han J, Du Y, Shen Z (2018) Forest fire characteristics in China: Spatial patterns and determinants with thresholds. For Ecol Manag 424:345–354. https://doi.org/10.1016/j.foreco.2018.05.020
[39]	You Y, Lu C, Wang W, Tang CK (2019) Relative CNN-RNN: Learning relative atmospheric visibility from images. IEEE Trans Image Process 28:45–55. https://doi.org/10.1109/TIP.2018.2857219
[40]	Zhang G, Wang M, Liu K (2019a) Forest fire susceptibility modeling using a convolutional neural network for Yunnan province of China. Int J Disaster Risk Sci 10:386–403. https://doi.org/10.1007/S13753-019-00233-1/FIGURES/10
[41]	Zhang H, Goodfellow I, Metaxas D (2019b) Self-attention generative adversarial networks. In: International conference on machine learning, Beach, CA, USA. pp 7354–7363 2019b
[42]	Zhang H, Goodfellow I, Metaxas D, Odena A (2019c) Self-attention generative adversarial networks. In: Proceedings of the 36th international conference on machine learning, Long Beach, California, PMLR 97