Study on Spatio-Temporal Distribution Characteristics and Susceptibility Analysis of Forest Fire

doi:10.13466/j.cnki.lczyyj.2023.05.006

Abstract

Abstract:

The situation of forest fire prevention in China is severe.Analyzing and studying the spatial-temporal characteristics and susceptibility of forest fire can provide scientific basis for forest fire prevention.Based on the data of the first national forest and grassland fire risk survey,the temporal and spatial distribution characteristics of forest fire in China's 31 provinces during 2011—2020 were analyzed,and the forest fire susceptibility model in China's 31 provinces was established by using the random forest algorithm through the construction of forest fire driving factors such as fuel,meteorological conditions and terrain.The results were as follows:1)The interannual change of the frequency and burned areas of forest fire in 31 provinces showed a downward trend from 2011 to 2020.The difference was significant in different geographical regions.Forest fires in winter and spring accounted for 85.48%.2)The fuel load per unit area is the most important driving factor of forest fire susceptibility,followed by monthly mean temperature,monthly minimum relative humidity and monthly mean precipitation.3)Receiver operating characteristic curve(ROC),area under curve(AUC)and accuracy(ACC)were used to analyze the accuracy of the prediction model.The values of AUC and ACC were 0.87 and 0.84,respectively,indicating a high accuracy of the susceptibility model.4)The forest fire susceptibility in China's 31 provinces had obvious regional differences.Northeast,Southwest and East China were dominated by high and medium-high susceptibility levels,Central China and South China were dominated by medium-low susceptibility levels,and North and Northwest China were dominated by low and very low susceptibility levels.

Key words: forest fire, random forest algorithm, spatial-temporal characteristic, susceptibility analysis

CLC Number:

S762

ZHANG Guoli, CI Xuelun, YANG Xueqing, JIANG Chunying, SUN Zhichao, MENG Haiding. Study on Spatio-Temporal Distribution Characteristics and Susceptibility Analysis of Forest Fire[J]. Forest and Grassland Resources Research, 2023, (5): 48-55.

Figures/Tables 8

Tab.1

Fig.1

Fig.2

Fig.3

Fig.4

Fig.5

Fig.6

Fig.7

References 24

[1]	岳超, 罗彩访, 舒立福, 等. 全球变化背景下野火研究进展[J]. 生态学报, 2020, 40(2):385-401.
[2]	Jain P, Castellanos-Acuna D, Coogan S C P, et al. Observed increases in extreme fire weather driven by atmospheric humidity and temperature[J]. Nature Climate Change, 2022, 12(1):63-70. doi: 10.1038/s41558-021-01224-1
[3]	Senande-Rivera M, Insua-Costa D, Miguez-Macho G. Spatial and temporal expansion of global wildland fire activity in response to climate change[J]. Nature Communications, 2022, 13(1):1-9. doi: 10.1038/s41467-021-27699-2
[4]	Wu Zhiwei, He H S, Keane R E, et al. Current and future patterns of forest fire occurrence in China[J]. International Journal of Wildland Fire, 2020, 29(2):104-119. doi: 10.1071/WF19039
[5]	田晓瑞, 代玄, 王明玉, 等. 多气候情景下中国森林火灾风险评估[J]. 应用生态学报, 2016, 27(3):769-776.
[6]	史培军, 袁艺. 重特大自然灾害综合评估[J]. 地理科学进展, 2014, 33(9):1145-1151. doi: 10.11820/dlkxjz.2014.09.001
[7]	高博, 单仔赫, 曹丽丽, 等. 大兴安岭地区森林火灾月动态变化及发生预测研究[J]. 中南林业科技大学学报, 2021, 41(9):53-62.
[8]	梁慧玲, 王文辉, 郭福涛, 等. 比较逻辑斯蒂与地理加权逻辑斯蒂回归模型在福建林火发生的适用性[J]. 生态学报, 2017, 37(12):4128-4141.
[9]	谢绍锋, 欧阳君祥, 肖化顺. 基于泰森多边形与条件熵的林火易发性空间分布研究[J]. 林业资源管理, 2017(4):50-58.
[10]	Zhuang Zijun, Yuan Xiaobing, Pei Jun, et al. An unsupervised representation learning approach for modelling forest landform characteristics and fire susceptibility assessment[J]. Journal of University of Chinese Academy of Sciences, 2023, 40(2):227-239.
[11]	高超, 林红蕾, 胡海清, 等. 我国林火发生预测模型研究进展[J]. 应用生态学报, 2020, 31(9):3227-3240. doi: 10.13287/j.1001-9332.202009.014
[12]	Zhang Guoli, Wang Ming, Liu Kai. Forest fire susceptibility modeling using a convolutional neural network for Yunnan Province of China[J]. International Journal of Disaster Risk Science, 2019, 10(3):386-403. doi: 10.1007/s13753-019-00233-1
[13]	Jain P, Coogan S C P, Subramanian S G, et al. A review of machine learning applications in wildfire science and management[J]. Environmental Reviews, 2020, 28(4):478-505. doi: 10.1139/er-2020-0019
[14]	Zhang Guoli, Wang Ming, Liu Kai. Deep neural networks for global wildfire susceptibility modelling[J]. Ecological Indicators, 2021, 127(2):107735. doi: 10.1016/j.ecolind.2021.107735
[15]	Xie Ying, Peng Minggang. Forest fire forecasting using ensemble learning approaches[J]. Springer London:Neural Computing and Applications, 2019, 31(9):4541-4550.
[16]	潘登, 郁培义, 吴强. 基于气象因子的随机森林算法在湘中丘陵区林火预测中的应用[J]. 西北林学院学报, 2018, 33(3):169-177.
[17]	马文苑, 冯仲科, 成竺欣, 等. 山西省林火驱动因子及分布格局研究[J]. 中南林业科技大学学报, 2020, 40(9):57-69.
[18]	苏佳佳, 刘志华, 焦珂伟, 等. 气候变化对中国林火干扰空间格局的影响[J]. 生态学杂志, 2021, 2(1):1-12.
[19]	蒋春颖, 杨雪清, 张国丽, 等. 森林火灾风险评估技术体系探讨[J]. 林业资源管理, 2023(2):17-26.
[20]	Breiman L. Random forests[J]. Machine Learning, 2001, 45(1):5-32. doi: 10.1023/A:1010933404324
[21]	方匡南, 吴见彬, 朱建平, 等. 随机森林方法研究综述[J]. 统计与信息论坛, 2011, 26(3):32-38.
[22]	Satir O, Berberoglu S, Donmez C. Mapping regional forest fire probability using artificial neural network model in a Mediterranean forest ecosystem[J]. Geomatics,Natural Hazards and Risk, 2016, 7(5):1645-1658. doi: 10.1080/19475705.2015.1084541
[23]	Greiner M, Pfeiffer D, Smith R D. Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests[J]. Preventive Veterinary Medicine, 2000, 45(1):23-41. doi: 10.1016/S0167-5877(00)00115-X
[24]	中央政府门户网站. 森林防火条例[A/OL].(2008-12-05) [2023-08-28]. https://www.gov.cn/flfg/2008-12/05/content_1171407.htm.

编号	指标	驱动因素	缩写	分辨率
1	可燃物	单位面积总可燃物载量/(t/hm²)	Fuelload	30″
2	可燃物	可燃性	Flamm	30″
3	气象数据	月大风日数/(d/月)	Windmax	0.01°
4		月平均风速/(m/s)	Windmean	0.01°
5		月平均降雨量/mm	Pre	0.01°
6		月平均气温/℃	Tas	0.01°
7		月最高气温/℃	Tasmax	0.01°
8		月最小相对湿度/%	Rhumin	0.01°
9	地形	高程/m	DEM	15m
10		坡度/(°)	Slope	15m
11		坡向/(°)	Aspect	15m