基于MaxEnt模型的森林火险区划等级研究——以烟台黄渤海新区为例

doi:10.13466/j.cnki.lczyyj.2024.02.002

摘要/Abstract

摘要：

森林火灾是一种突发性强、破坏性大的自然灾害,受气象、地形、植被、人类活动等影响。科学制定森林火险区划等级是预防和管控森林火灾的基础。最大熵(MaxEnt)作为机器学习模型的一种,被广泛应用于国内外预测森林火灾发生概率、识别森林火灾风险等级的相关研究,并被证实具有较高准确性。烟台黄渤海新区位于胶东半岛、黄渤海交界处,经济发达、人口密度大,人类活动对森林影响显著,在全国森林防火工作中具有代表性。以烟台黄渤海新区为例,采用MaxEnt模型,结合历史火情数据和气象、地形、地物类型、社会经济等主导环境变量,预测森林火灾发生概率,划定森林火险区划等级。结果表明:1)优势树种(组)、人口密度对森林火灾发生具有重要影响;2)烟台黄渤海新区高、中、一般风险区面积接近,高风险区主要位于西南部山区,一般风险区主要位于东北部沿海。经检验,本研究模型预测结果具有较高准确性,可为全国森林火险区划提供参考。

关键词: 森林火险区划, MaxEnt, 烟台

Abstract:

Forest fires are sudden and destructive natural disasters influenced by various factors such as meteorology,terrain,vegetation,and human activities.The establishment of a forest fire risk zoning is essential for effective prevention and controlling forest fires,which is of great significance for carrying out related work in the future.As a machine learning method,MaxEnt has gained widespread domestic and international usage in predicting probabilities of occurrences and identifying zoning of forest fire risk due to its demonstrated high accuracy.Yantai Yellow and Bohai Seas New Area,located at the junction of Jiaodong Peninsula and Yellow and Bohai Seas has a developed economy,high population density,and significant human impact on forests,making it representative of forest fire prevention efforts in China.This study took Yantai Yellow and Bohai Seas New Area as an example and utilized MaxEnt to predict the probability of forest fire occurrence and delineate forest fire risk zoning based on historical fire incident data,as well as dominant environmental variables such as meteorology,topography,land cover types,and socio-economic factors.The research findings indicate that:1)Dominant tree species(group)and population density have significant impacts on forest fire occurrence;2)In Yantai Yellow and Bohai Seas New Area,the areas with high,medium,and general risk levels are similar in size.High-risk areas are mainly concentrated in the southwestern mountainous region,while general-risk areas are primarily found along the northeastern coast.Through verification,this study's results have shown a high level of accuracy and can provide a methodological exploration for subsequent national forest fire risk zoning.

Key words: forest fire zoning, MaxEnt, Yantai

中图分类号:

S762

李杰, 晁碧霄, 闫钰倩, 姜帆, 张大为, 孙永康, 邹全程. 基于MaxEnt模型的森林火险区划等级研究——以烟台黄渤海新区为例[J]. 林草资源研究, 2024,(2): 8-16.

LI Jie, CHAO Bixiao, YAN Yuqian, JIANG Fan, ZHANG Dawei, SUN Yongkang, ZOU Quancheng. Forest Fire Risk Zoning Based on MaxEnt Model—A Case Study of Yantai Yellow and Bohai Seas New Area[J]. Forest and Grassland Resources Research, 2024,(2): 8-16.

图/表 9

图1

表1

图2

图3

图4

图5

图6

图7

表2

参考文献 32

[1]	SLOBODAN M, NENAD M, DRAGAN P, et al. Forest fire probability mapping in Eastern Serbia:Logistic regression versus random forest method[J]. Forests, 2020, 12(1):5.
[2]	国家林业和草原局. 2021中国林草资源及生态状况[M]. 北京: 中国林业出版社, 2022.
[3]	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, 2019,10:386-403.
[4]	中华人民共和国国务院新闻办公室. 国新办举行春夏森林草原火灾防控新闻发布会[EB/OL].(2020-04-09)[2024-03-25]. http://www.scio.gov.cn/xwfb/gwyxwbgsxwfbh/wqfbh_2284/2020n_4408/2020n04y09r/#1.
[5]	中共中央办公厅, 国务院办公厅. 关于全面加强新形势下森林草原防灭火工作的意见[EB/OL].(2023-04-20)[2024-03-25]. https://www.gov.cn/zhengce/2023-04/20/content_5752410.htm?eqid=e00582e50005a3d300000005647c6f06.
[6]	TAN Chaoxue, FENG Zhongke. Mapping forest fire risk zones using machine learning algorithms in Hunan Province,China[J]. Sustai-nability, 2023, 15(7):6292.
[7]	GUO Futao, SU Zhangwen, WANG Guangyu, et al. Understanding fire drivers and relative impacts in different Chinese forest ecosystems[J]. Science of the Total Environment, 2017,605:411-425.
[8]	AKAY A E, SAHIN H. Forest fire risk mapping by using GIS techniques and AHP method:A case study in Bodrum (Turkey)[J]. European Journal of Forest Engineering, 2019, 5(1):25-35.
[9]	安佳怡, 冯仲科, 马天天, 等. 基于GIS格网的重庆合川区森林火险等级区划[J]. 中南林业科技大学学报, 2022, 42(9):91-101.
[10]	王双, 张贵, 谭三清, 等. 基于空间logistic的湖南省森林火灾风险评价[J]. 中南林业科技大学学报, 2020, 40(9):88-95.
[11]	TIEN Bui D, BUI Q, NGUYEN Q, et al. 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[J]. Agricultural and Forest Meteorology, 2017,233:32-44.
[12]	MA Wenyuan, FENG Zhongke, CHENG Zhuxin, et al. Identifying forest fire driving factors and related impacts in China using random forest algorithm[J]. Forests, 2020,11:507.
[13]	LI Yudong, FENG Zhongke, CHEN Shilin, et al. Application of the artificial neural network and support vector machines in forest fire prediction in the Guangxi Autonomous Region,China[J]. Discrete Dynamics in Nature and Society,2020:1-14.
[14]	SEVINCA V, KUCUKB O, GOLTASC M. A Bayesian network model for prediction and analysis of possible forest fire causes[J]. Forest Ecology and Management, 2020,457:117723.
[15]	PHILLIPS S J. A brief tutorial on Maxent network of conservation educators and practitioners,center for biodiversity and conservation,American Museum of Natural History[J]. Lessons in Conservation, 2009,3:108-135.
[16]	MAKHAYA Z, ODINDI J, MUTANGA O. The influence of bioclimatic and topographic variables on grassland fire occurrence within an urbanized landscape[J]. Scientific African, 2022,15:e01127.
[17]	MISHRA B, PANTHI S, POUDEL S, et al. Forest fire pattern and vulnerability mapping using deep learning in Nepal[J]. Fire Ecology, 2023, 19(1):3.
[18]	TARIQ A, SHU H, SIDDIQUI S, et al. Spatio-temporal analysis of forest fire events in the Margalla Hills,Islamabad,Pakistan using socio-economic and environmental variable data with machine learning methods[J]. Journal of Forestry Research, 2021, 33(1):183-194.
[19]	KIM S J, CHUL-HEE L, KIM G S, et al. Multi-temporal analysis of forest fire probability using socio-economic and environmental variables[J]. Remote Sensing, 2019, 11(1):86.
[20]	VILAR L, GÓMEZ I, MARTÍNEZ-VEGA J, et al. Multitemporal modelling of socio-economic wildfire drivers in central Spain between the 1980s and the 2000s:Comparing generalized linear models to machine learning algorithms[J]. PLos One, 2016,11:1-17.
[21]	国家林业局森林防火指挥部办公室. 全国森林火险区划等级:LY/T 1063—2008[S]. 北京: 国家林业局, 2008.
[22]	BANERJEE P. Maximum entropy-based forest fire likelihood mapping:Analysing the trends,distribution,and drivers of forest fires in Sikkim Himalaya[J]. Scandinavian Journal of Forest Research, 2021, 36(4):275-288.
[23]	CHEN Feng, DU Yongsheng, NIU Shukui, et al. Modeling forest lightning fire occurrence in the Daxinganling Mountains of Northeastern China with MaxEnt[J]. Forests, 2015, 6(5):1422-1438.
[24]	YATHISH H, ATHIRA K V, PREETHI K, et al. A comparative analysis of forest fire risk zone mapping methods with expert knowledge[J]. Journal of the Indian Society of Remote Sensing, 2019, 47(12):2047-2060.
[25]	SARI F. Identifying anthropogenic and natural causes of wildfires by maximum entropy method-based ignition susceptibility distribution models[J]. Journal of Forestry Research, 2022, 34(2):355-371.
[26]	艾秋月. 山东省森林火灾发生时空规律及防控对策研究[D]. 山东: 山东农业大学, 2023.
[27]	胡海清. 林火生态与管理[M]. 北京: 中国林业出版社, 2005.
[28]	吕常笑, 秦乃花, 李萍, 等. 山东省森林火灾时空分布及火源特征研究[J]. 森林防火, 2023, 41(2):1-6.
[29]	GOLTAS M, AYBERK H, KUCUK O. Forest fire occurrence modeling in Southwest Turkey using MaxEnt machine learning technique[J]. IForest-Biogeosciences and Forestry, 2024,17:10-18.
[30]	TRANG P T, ANDREW M E, CHU T, et al. Forest fire and its key drivers in the tropical forests of northern Vietnam[J]. International Journal of Wildland Fire, 2022, 31(3):213-229.
[31]	JANIEC P, GADAL S. A comparison of two machine learning classification methods for remote sensing predictive modeling of the forest fire in the North-eastern Siberia[J]. Remote Sensing, 2020, 12(24):4157.
[32]	ANTONELLA A D, CARLO R, MARCO C, et al. Modelling the meteorological forest fire niche in heterogeneous pyrologic conditions[J]. Plos One, 2015, 10(2):e0116875.

数据类型	环境变量	单位
生物气候	最干季平均温度	℃
	最暖季平均温度	℃
	最干季降水量	mm
	最暖季降水量	mm
地形	高程	m
	坡度	(°)
	坡向
地物类型	优势树种(组)
	土地利用类型
	植被覆盖类型
	归一化植被指数
社会经济	距居民点距离	m
	距道路距离	m
	GDP	万元/km²
	人口密度	人/km²

镇、街道	一般风险区		中风险区		高风险区		总计村数量/个
镇、街道	面积占比/%	村数量/个	面积占比/%	村数量/个	面积占比/%	村数量/个	总计村数量/个
大季家街道	53.43	12	24.13	12	22.44	13		37
潮水镇	27.53	20	46.27	40	26.20	18		78
福莱山街道	100.00	1	0.00	0	0.00	0		1
古现街道	14.52	11	31.60	21	53.88	29		61
烟台黄渤海新区	37.39	44	31.70	73	30.91	60		177