欢迎访问林业科学,今天是

林业科学 ›› 2021, Vol. 57 ›› Issue (4): 142-152.doi: 10.11707/j.1001-7488.20210415

• 综合评述 • 上一篇    下一篇

森林地表死可燃物含水率预测模型研究进展

孙龙,刘祺,胡同欣*   

  1. 东北林业大学林学院 哈尔滨 150040
  • 收稿日期:2019-11-18 出版日期:2021-04-01 发布日期:2021-05-21
  • 通讯作者: 胡同欣
  • 基金资助:
    国家"十三五"重点研发计划(2018YFD0600205);中央高校基本科研业务费专项资金(2572017PZ05)

Advances in Research on Prediction Model of Moisture Content of Surface Dead Fuel in Forests

Long Sun,Qi Liu,Tongxin Hu*   

  1. College of Forestry, Northeast Forestry University Harbin 150040
  • Received:2019-11-18 Online:2021-04-01 Published:2021-05-21
  • Contact: Tongxin Hu

摘要:

林火是影响森林生态系统的重要因子之一,林火蔓延和发展深受森林可燃物含水率的影响,尤其是林火的发生直接受地表死可燃物含水率的影响。因此,准确预测森林地表死可燃物含水率是预报森林火险和火行为的关键,加强森林死可燃物含水率预测模型研究尤为重要。从森林可燃物含水率的研究方法、研究模型及模型精度3方面综述研究现状,并对比评价现有模型。针对目前研究的诸多问题,提出5点展望:1)加强研究重点火险区野外含水率动态。利用已有的森林火险因子采集站和森林火险监测站获取不同环境因子和可燃物含水率及气象因子监测数据,构建重点火险区基于气象参数的森林可燃物含水率预测模型。2)加强森林可燃物的基础数据监测和收集。这可为全面构建森林火险等级系统奠定坚实的数据基础,同时还应建立精准的森林可燃物类型划分体系。3)加强研究可燃物含水率的空间异质性。应考虑不同影响因子下可燃物含水率动态,特别是了解小尺度内森林可燃物含水率的空间异质性,才能更准确进行林火预测预报。4)结合应用增强回归树(BRT)方法来提高模型精度。在可燃物含水率模型精度影响因子的研究中,运用BRT方法多次随机抽取一定量的数据,量化分析不同因子对模型精度的影响程度。5)结合GIS进行大尺度火险预警研究。综合应用RS和GIS技术,建立可燃物含水率的遥感反演模型,在准确模拟森林可燃物含水率空间分布的基础上,建立基于可燃物含水率的不同火险等级的预测模型。

关键词: 森林可燃物, 地表死可燃物含水率, 预测模型, 模型精度, 火险预测预报

Abstract:

Forest fire is one of the important factors that affect the forest ecosystem. The spread and development of forest fire are deeply affected by the moisture content of forest fuels, especially the occurrence of forest fire is directly affected by the moisture content of dead fuels on the surface. Therefore, accurate prediction of the forest surface dead fuels moisture content is the key to predict forest fire risk and fire behavior, and it is particularly important to strengthen the study of forest dead fuels moisture content prediction model. This article summarizes the research status of forest fuels moisture content in terms of the research methods, research models and model accuracy, and comparatively evaluates the existing models. In view of problems in the current research, five prospects for future research are proposed: 1) Strengthen research on the dynamic of fuel moisture content in key fire risk zone. The existing forest fire danger factor collection stations and forest fire danger monitoring stations are used to obtain the monitoring data of forest fuel moisture content and meteorological factors under different environmental condition. The prediction model of forest fuel moisture content based on meteorological parameters in key fire danger zone is constructed. 2) Strengthen basic data monitoring and collection of forest fuels. In order to build a comprehensive forest fire risk rating system, the basic data monitoring and collection of forest fuels should be strengthened, a solid data foundation should be laid, and an accurate forest fuel type classification system should be established. 3) Strengthen the study on the spatial heterogeneity of the fuel moisture content. In the future research, the dynamic changes of fuel moisture content under different impact factors should be considered, especially the spatial heterogeneity of fuel moisture content of small-scale forests, so that the prediction of forest fire danger can be made more accurately. 4) Improve the accuracy of the models combined with boosted regression tree(BRT). In the study of the influencing factors of the accuracy of the fuel moisture content model, the BRT method should be used to randomly extract a certain amount of data multiple times to analyze the degree of influence of different influencing factors on the accuracy of the model. 5) Conduct research on large-scale fire risk early alarm combined with GIS. Based on RS and GIS technology, the remote sensing inversion model of the fuel moisture content is established. On the basis of accurate simulation of the spatial distribution of forest fuel moisture content, the fuel moisture content prediction models of different fire risk classes is established.

Key words: forest fuel, surface dead fuel moisture content, forecast model, model accuracy, fire danger prediction

中图分类号: