Assessment of China’s forest fire occurrence with deep learning, geographic information and multisource data

doi:10.1007/s11676-022-01559-1

Abstract

Abstract:

Considerable economic losses and ecological damage can be caused by forest fires, and compared to suppression, prevention is a much smarter strategy. Accordingly, this study focuses on developing a novel framework to assess forest fire risks and policy decisions on forest fire management in China. This framework integrated deep learning algorithms, geographic information, and multisource data. Compared to conventional approaches, our framework featured timesaving, easy implementation, and importantly, the use of deep learning that vividly integrates various factors from the environment and human activities. Information on 96,594 forest fire points from 2001 to 2019 was collected on Moderate Resolution Imaging Spectroradiometer (MODIS) fire hotspots from 2001 to 2019 from NASA's Fire Information Resource Management System. The information was classified into factors such as topography, climate, vegetation, and society. The prediction of forest fire risk was generated using a fully connected network model, and spatial autocorrelation used to analyze the spatial aggregation correlation of active fire hotspots in the whole area of China. The results show that high accuracy prediction of fire risks was achieved (accuracy 87.4%, positive predictive value 87.1%, sensitivity 88.9%, area under curve (AUC) 94.1%). Based on this, it was found that Chinese forest fire risk shows significant autocorrelation and agglomeration both in seasons and regions. For example, forest fire risk usually raises dramatically in spring and winter, and decreases in autumn and summer. Compared to the national average, Yunnan Province, Guangdong Province, and the Greater Hinggan Mountains region of Heilongjiang Province have higher fire risks. In contrast, a large region in central China has been recognized as having a long-term, low risk of forest fires. All forest risks in each region were recorded into the database and could contribute to the forest fire prevention. The successful assessment of forest fire risks in this study provides a comprehensive knowledge of fire risks in China over the last 20 years. Deep learning showed its advantage in integrating multiple factors in predicting forest fire risks. This technical framework is expected to be a feasible evaluation tool for the occurrence of forest fires in China.

Key words: Forest fires, Deep learning, Spatial autocorrelation, Risk zoning, Management strategies

Yakui Shao, Zhichao Wang, Zhongke Feng, Linhao Sun, Xuanhan Yang, Jun Zheng, Tiantian Ma. Assessment of China’s forest fire occurrence with deep learning, geographic information and multisource data[J]. JOURNAL OF FORESTRY RESEARCH, 2023, 34(4): 963-976.

Yakui Shao, Zhichao Wang, Zhongke Feng, Linhao Sun, Xuanhan Yang, Jun Zheng, Tiantian Ma. [J]. 林业研究(英文版), 2023, 34(4): 963-976.

References 78

1	Adab H, Kanniah KD, Solaimani K. Modeling forest fire risk in the northeast of Iran using remote sensing and GIS techniques. Nat Hazards, 2013, 65(3): 1723-1743, DOI
2	Alom MZ, Taha TM, Yakopcic C, Westberg S, Sidike P, Nasrin MS, Hasan M, Essen BCV, Awwal AAS, Asari VK. A state-of-the-art survey on deep learning theory and architectures. Electronics, 2019, 8(3): 292, DOI
3	Anselin L. Local indicators of spatial association—LISA. Geogr Anal, 1995, 27(2): 93-115, DOI
4	Astiani D, Curran LM, Burhanuddin TM, Gusmayanti E. Fire-driven biomass and peat carbon losses and post-fire soil co2 emission in a west kalimantan peatland forest. J Trop For Sci, 2018, 30(4): 570-575
5	Banerjee P. Maximum entropy-based forest fire likelihood mapping: analysing the trends, distribution, and drivers of forest fires in Sikkim Himalaya. Scand J Forest Res, 2021, 36: 275-288, DOI
6	Bisong E (2019) Building machine learning and deep learning models on Google Cloud Platform: a comprehensive guide for beginners. Apress, Berkeley, pp 581–598. [Google Scholar] [CrossRef]
7	Bo M, Mercalli L, Pognant F, Cat BD, Clerico M. Urban air pollution, climate change and wildfires: the case study of an extended forest fire episode in northern Italy favoured by drought and warm weather conditions. Energy Rep, 2020, 6: 781-786, DOI
8	Brown KJ, Hebda NJ, Conder N, Golinksi KG, Hawkes B, Schoups G, Hebda RJ. Changing climate, vegetation, and fire disturbance in a sub-boreal pine-dominated forest, British Columbia. Canada Can J For Res, 2017, 47(5): 615-627, DOI
9	Deng O, Li Y, Feng Z, Zhang D. Model and zoning of forest fire risk in Heilongjiang province based on spatial Logistic. Trans Chin Soc Agr Eng, 2012, 28(8): 200-205
10	Eugenio FC, dos Santos AR, Fiedler NC, Ribeiro GA, da Silva AG, dos Santos ÁB, Paneto GG, Schettino VR. Applying GIS to develop a model for forest fire risk: a case study in Espírito Santo, Brazil. J Environ Manage, 2016, 173: 65-71, DOI
11	Fang K, Yao Q, Guo Z, Zheng B, Du J, Qi F, Yan P, Li J, Ou T, Liu J, Ou T, Liu J, He M, Trouet V. ENSO modulates wildfire activity in China. Nat Commun, 2021, 12(1): 1764, DOI
12	Ganteaume A, Camia A, Jappiot M, San-Miguel-Ayanz J, Long-Fournel M, Lampin C. A review of the main driving factors of forest fire ignition over Europe. Environ Manage, 2013, 51(3): 651-662, DOI
13	Gao JF. Middle and long term plan discussion of key problems to forest fire prevention in China. For Invent Plan, 2015, 40(1): 4 (in Chinese)
14	Gao C, Lin HL, Hu HQ, Song H (2020) A review of models of forest fire occurrence prediction in China. Chin J Appl Ecol 31 (09):3227–3240. (in Chinese). https://doi.org/10.13287/j.1001-9332.202009.014.
15	Garcia C, Woodard P, Titus S, Adamowicz W, Lee B. A logit model for predicting the daily occurrence of human caused forest fires. Int J Wildland Fire, 1995, 5(2): 101-111, DOI
16	Ghobadi GJ, Gholizadeh B, Dashliburun OM. Forest fire risk zone mapping from geographic information system in northern forests of Iran (Case study, Golestan province). Int J Agr Crop Sci, 2012, 4(12): 818-824
17	Gholamnia K, Gudiyangada T, Ghorbanzadeh O, Blaschke T. Comparisons of diverse machine learning approaches for wildfire susceptibility mapping. Symmetry, 2020, 12(4): 604, DOI
18	Gigliarano C, Figini S, Muliere P. Making classifier performance comparisons when ROC curves intersect. Comput Stat Data Anal, 2014, 77: 300-312, DOI
19	Giglio L, Descloitres J, Justice CO, Kaufman YJ. An enhanced contextual fire detection algorithm for MODIS. Remote Sens Environ, 2003, 87(2–3): 273-282, DOI
20	Guo FT, Su ZW, Wang GY, Sun L, Lin FF, Liu AQ. Wildfire ignition in the forests of southeast China: identifying drivers and spatial distribution to predict wildfire likelihood. Appl Geogr, 2016, 66: 12-21, DOI
21	Heo JP, Im CG, Ryu KH, Sung SW, Yoo C, Yang DR. Shallow fully connected neural network training by forcing linearization into valid region and balancing training rates. Processes, 2022, 10(6): 1157, DOI
22	Holden ZA, Jolly WM. Modeling topographic influences on fuel moisture and fire danger in complex terrain to improve wildland fire management decision support. Forest Ecol Manag, 2011, 262(12): 2133-2141, DOI
23	Hong HY, Tsangaratos P, Ilia I, Liu JZ, Zhu AX, Chong X. Applying genetic algorithms to set the optimal combination of forest fire related variables and model forest fire susceptibility based on data mining models. The case of Dayu County. China Sci Total Environ, 2018, 630: 1044-1056, DOI
24	Jaafari A, Davood MG, Eric KZ. A Bayesian modeling of wildfire probability in the Zagros Mountains. Iran Ecol Inform, 2017, 39: 32-44, DOI
25	Jahdi R, Salis M, Darvishsefat AA, Urdiroz FA, Etemad V, Mostafavi MA, Lozano OM, Spano D. Calibration of FARSITE fire area simulator in Iranian northern forests. Nat Hazards Earth Sys, 2015, 15: 443-459, DOI
26	Justice CO, Giglio L, Korontzi S, Owens J, Morisette JT, Roy D, Descloitres J, All Ea Ume S, Petitcolin F, Kaufman Y. The MODIS fire products. Remote Sens Environ, 2002, 83(1–2): 244-262, DOI
27	KöHl M, Lasco R, Cifuentes M, Jonsson Ö, Korhonen KT, Mundhenk P, Djn J, Stinson G. Changes in forest production, biomass and carbon: results from the 2015 UN FAO Global Forest Resource Assessment. Forest Ecol Manag, 2015, 352(352): 21-34, DOI
28	Lecun Y, Bengio Y, Hinton G. Deep learning. Nature, 2015, 521(7553): 436, DOI
29	Li XH, LV D. Elaborate forecast about fire risk grade in forest and grassland of inner mongolia based on intelligent grid. Meteorol Environ Res, 2021, 12(5): 39-42
30	Li P, Li WJ, Feng ZM, Xiao CW, Liu YY. Spatiotemporal dynamics of active fire frequency in Southeast Asia with the FIRMS Moderate Resolution Imaging Spectroradiometer (MODIS) and Visible Infrared Imaging Radiometer (VIIRS) data. Resources Sci, 2019, 41(8): 1526-1540 (in Chinese)
31	Li YD, Feng ZK, Chen SL, Zhao ZY, Wang FG. Application of the artificial neural network and support vector machines in forest fire prediction in the Guangxi Autonomous Region. China Discrete Dyn Nat Soc, 2020, 2020: 5612650
32	Liang HL, Wang WH, Guo FT, Lin FF, Lin YR. Comparing the application of logistic and geographically weighted logistic regression models for Fujian forest fire forecasting. Acta Ecol Sin, 2017, 37(12): 4128-4144 (in Chinese)
33	Linn R, Reisner J, Colman JJ, Winterkamp J. Studying wildfire behavior using FIRETEC. Int J Wildland Fire, 2002, 11(4): 233-246, DOI
60	Liyan S, Zhou LX, Liu ML, Yu Y. Research on forest fire prediction method based on deep learning. J For Eng, 2019, 4(03): 132-136, DOI
34	Long TT, Yin JY, Ou CR, Yang Q, Li Y, Wang QH. Comprehensive assessment and spatial pattern study on forest fire risk in Yunnan Province. Chin Safety Sci J, 2021, 31(9): 167-173 (in Chinese)
35	Lopes AMG, Cruz MG, Viegas DX. FireStation—an integrated software system for the numerical simulation of fire spread on complex topography. Environ Modell Softw, 2002, 17: 269-285, DOI
36	Ma W, Feng Z, Cheng Z, Chen S, Wang F. Identifying forest fire driving factors and related impacts in china using random forest algorithm. Forests, 2020, 11(5): 507, DOI
37	Ma WY, Feng ZK, Cheng ZX, Wang FG. Study on forest fire drivers and distribution pattern in Shanxi Province. J Central South Univ For Sci Tech, 2020, 40(9): 57-69 (in Chinese)
38	Mohajane M, Costache R, Karimi F, Bao PQ, Essahlaoui A, Nguyen H, Laneve G, Oudija F. Application of remote sensing and machine learning algorithms for forest fire mapping in a Mediterranean area. Ecol Indic, 2021, 129, DOI
39	Morales-Hidalgo D, Oswalt SN, Somanathan E. Status and trends in global primary forest, protected areas, and areas designated for conservation of biodiversity from the Global Forest Resources Assessment 2015. Forest Ecol Manag, 2015, 352: 68-77, DOI
40	Naderpour M, Rizeei HM, Ramezani F. Forest fire risk prediction: a spatial deep neural network-based framework. Remote Sens, 2021, 13(13): 2513, DOI
41	Oliveira S, Oehler F, San-Miguel-Ayanz J, Camia A, Pereira JMC. Modeling spatial patterns of fire occurrence in Mediterranean Europe using multiple regression and random forest. Forest Ecol Manag, 2012, 275: 117-129, DOI
42	Overmars KP, de Koning GHJ, Veldkamp A. Spatial autocorrelation in multi-scale land use models. Ecol Model, 2003, 164(2–3): 257-270, DOI
43	Pan P, Sun YJ, Ouyang XZ, Rao JF, Feng RQ, Yang ZY. Spatial variation of carbon density in Pinus massoniana forest in Jiangxi Province. China Chin J Appl Ecol, 2019, 30(6): 1885-1892 (in Chinese)
44	Parente J, Amraoui M, Menezes I, Pereira MG (2019) Drought in Portugal: Current regime, comparison of indices and impacts on extreme wildfires. Sci Total Environ 685 (OCT.1):150–173.
45	Pastor E, Zárate L, Planas E, Arnaldos J. Mathematical models and calculation systems for the study of wildland fire behaviour. Prog Energ Combust, 2003, 29(2): 139-153, DOI
46	Pencina MJ, D'Agostino RB Sr, D'Agostino RB Jr, Vasan RS. Evaluating the added predictive ability of a new marker: from area ussnder the ROC curve to reclassification and beyond. Stat Med, 2008, 27(2): 173-181, DOI
47	Pham BT, Abolfazl J, Mohammadtaghi NA, Tran DD, Hoang PHY, Tran VP . Performance evaluation of machine learning methods for forest fire modeling and prediction. Symmetry, 2020, 12(6): 1022, DOI
48	Prasad VK, Badarinath K, Eaturu A. Biophysical and anthropogenic controls of forest fires in the Deccan Plateau. India J Environ Manage, 2008, 86(1): 1-13
49	Qiu M, Zuo Q, Wu Q, Yang Z, Zhang J. Water ecological security assessment and spatial autocorrelation analysis of prefectural regions involved in the Yellow River Basin. Sci Rep-UK, 2022, 12(1): 5105, DOI
50	Rishickesh R, Shahina A, Khan N (2019) Predicting forest fires using supervised and ensemble machine learning algorithms. Int J Recent Tech Eng 2 8(2):3697–3705.
51	Sachdeva S, Bhatia T, Verma AK. GIS-based evolutionary optimized Gradient Boosted Decision Trees for forest fire susceptibility mapping. Nat Hazards, 2018, 92: 1399-1418, DOI
52	Sayad YO, Mousannif H, Al Moatassime H (2019) Predictive modeling of wildfires: a new dataset and machine learning approach. Fire Safety J 104 (MAR.):130–146.
53	Sebastián-López A, Salvador-Civil R, Gonzalo-Jiménez J, Sanmiguel-Ayanz J. Integration of socio-economic and environmental variables for modelling long-term fire danger in Southern Europe. Eur J Forest Res, 2008, 127(2): 149-163, DOI
54	Sevinc V, Kucuk O, Goltas M. A Bayesian network model for prediction and analysis of possible forest fire causes. Forest Ecol Manag, 2020, 457, DOI
55	Shakesby RA. Post-wildfire soil erosion in the Mediterranean: review and future research directions. Earth-Sci Rev, 2011, 105(3–4): 71-100, DOI
56	Shao Y, Feng Z, Sun L, Yang X, Li Y, Xu B, Chen Y. Mapping China's forest fire risks with machine learning. Forests, 2022, 13(6): 856, DOI
57	Shu LF, Zhang XL, Dai X, Tian XR, Wang MY. Forest fire research (II): fire forecast. World For Res, 2003, 16(4): 34-37 (in Chinese)
58	Sun L, Shang ZC, Hu HQ. Application of a Poisson regression model and anegative binomial regression model in the forest fire forecasting. Scientia Silvae Sinicae, 2012, 48(5): 126-129 (in Chinese)
59	Sun T, Zhang W, Chen W, Tang X, Qin Q. Mountains forest fire spread simulator based on geo-cellular automaton combined with Wang Zhengfei velocity model. IEEE J-Stars, 2013, 6: 1971-1987
61	Sun JX, Zhong CH, He HW, Hugeman G, Li H. Continuous remote sensing monitoring and changes of land desertification in China from 2000 to 2015. J Northeast For Univ, 2021, 49(3): 87-92 (in Chinese)
62	Suryabhagavan KV, Alemu B. GIS-based multi-criteria decision analysis for forest fire susceptibility mapping: a case study in Harenna forest, southwestern Ethiopia. Trop Ecol, 2016, 57(1): 33-43
63	Tien BD, Le KTT, Nguyen VC, Le HD, Revhaug I. Tropical forest fire susceptibility mapping at the Cat Ba National Park area, Hai Phong city, Vietnam, using gis-based kernel lo-gistic regression. Remote Sens, 2016, 8(4): 347-347, DOI
64	Tien BD, Bui QT, Nguyen QP, Pradhan B, Nampak H, Trinh PT (2017) A hybrid artificial intelligence approach using GIS-based neural-fuzzy inference system and particle swarm optimization for forest fire susceptibility modeling at a tropical area. Agr Forest Meteorol 233 (Complete):32–44.
65	Tuyen TT, Jaafari A, Yen HPH, Nguyen-Thoi T, Phong TV, Nguyen HD, Van Le H, Phuong TTM, Nguyen SH, Prakash I . Mapping forest fire susceptibility using spatially explicit ensemble models based on the locally weighted learning algorithm. Ecol Inform, 2021, 63, DOI
66	Verde JC, Zêzere JL. Assessment and validation of wildfire susceptibility and hazard in Portugal. Nat Hazards Earth Sys, 2010, 10(3): 485-497, DOI
67	Wang Y, Fang ZC, Niu RQ, Peng L (2021) Landslide susceptibility analysis based on deep learning. J Geo-Inf Sci 23 (12):2244–2260. (in Chinese). https://doi.org/10.12082/dqxxkx.2021.210057.
68	Wotton BM, Nock CA, Flannigan MD. Forest fire occurrence and climate change in Canada. Int J Wildland Fire, 2010, 19(3): 253-271, DOI
69	Wu ZW, He HS, Keane RE, Zhu ZL, Shan YL. Current and future patterns of forest fire occurrence in China. Int J Wildland Fire, 2020, 29(2): 104, DOI
70	Xie Y, Peng M. Forest fire forecasting using ensemble learning approaches. Neural Comput Appl, 2019, 31(9): 4541-4550, DOI
71	Yi K, Bao Y, Zhang J. Spatial distribution and temporal variability of open fire in China. Int J Wildland Fire, 2017, 26(2): 122-135, DOI
72	Yin BC, Wang WT, Wang LC. Review of Deep Learning. J Bjing Univ Tech, 2015, 41(1): 48-59 (in Chinese)
73	Zeng C, Zeng Z, Cao ZY, Zou Q, Yu CX. Forest fire dynamic monitoring based on time series and multisource satellite images: A case study of the Muli county forest areas in Sichuan province. Remote Sens Tech Appl, 2021, 36(03): 521-532 (in Chinese)
74	Zhang JJ, Fu WJ, Du Q, Zhang GJ, Jiang PK. Spatial variability characteristics of carbon densities in the forest litter in Zhejiang province. Sci Silvae Sinicae, 2014, 50(2): 8-13 (in Chinese)
75	Zhang ZX, Xu MX, Liu J, Li Q. Spatial variation reasonable sampling number of soil organic carbon under different geomorphic types on the loess plateau. J Nat Resour, 2014, 29(12): 2103-2113 (in Chinese)
76	Zhang Y, Lee JD, Wainwright MJ, Jordan M I (2017) On the learnability of fully-connected neural networks. Paper presented at 20th International Conference on Artificial Intelligence and Statistics, AISTATS 2017, Fort Lauderdale, United States.
77	Zhao P, Zhang F, Lin H, Xu S. GIS-Based forest fire risk model: a case study in Laoshan National Forest Park. Nanjing Remote Sens, 2021, 13(18): 3704, DOI
78	Zhong M, Fan W, Liu T, Li P. Statistical analysis on current status of China forest fire safety. Fire Safety J, 2003, 38(3): 257-269, DOI

[1]	Fatih Sari. Identifying anthropogenic and natural causes of wildfires by maximum entropy method-based ignition susceptibility distribution models [J]. JOURNAL OF FORESTRY RESEARCH, 2023, 34(2): 355-371.
[2]	Chuyan Wan, Shouraseni Sen Roy. Geospatial characteristics of fire occurrences in southern hemispheric Africa and Madagascar during 2001–2020 [J]. JOURNAL OF FORESTRY RESEARCH, 2023, 34(2): 553-563.