火灾风险与共存能力的整合减灾框架:计划烧除的空间优化策略OACHSSCD
An integrated disaster risk reduction framework for fire risk and coexistence:spatial Optimization Strategies for Prescribed Burning
面对气候变化加剧带来的异常与极端火险事件频发,火管理范式正由传统被动控制转向以灾害风险韧性为导向的"与火共存".自然生态系统中的动植物与传统生态智慧为适应低强度火灾提供了宝贵经验,揭示了可持续共存的潜力.尽管计划烧除是公认有效的火灾风险管理工具,但在中国具体国情下,"与火共存"的系统化理念及其与计划烧除的整合研究和应用仍显不足.在研究国内外"与火共存"的基础上,创建"灾害风险预测-与火共存能力评估-计划烧除战略优先区(Risk and Coexistence-Prescribed Burning,RCO-PB)"框架.该框架旨在从理论层面系统化梳理火灾风险与生态系统共存能力间的复杂关系,并构建一套面向区域尺度火管理空间优化的普适性方法论.核心目标在于通过此框架,有效降低高强度火灾风险,科学规划火管理优先区,并为维护生物多样性提供适宜的生态环境.在RCO-PB框架指导下,运用机器学习和深度学习模型,对研究区域(重庆市嘉陵江流域)的火灾风险进行精准预测,并对生态系统的与火共存能力进行了深入评估.应用空间优化模型识别并确定了计划烧除的战略优先区,重点关注那些共存能力强但火灾风险高的区域.结果表明:土地覆盖因子在火灾风险模型预测中起到关键作用,研究区内 3%的地区极易发生火灾(风险值≥0.659),7%的地区发生火灾的风险很高(风险值≥0.411),北碚缙云山火险等级极高.与火共存能力高价值区集中在植被复杂、茂密的生态系统,合川南部、北碚西部和北部具有较高的共存能力(前 10%,值>0.9).计划烧除优先区分布在以马尾松纯林、针阔混交林(马尾松-青冈栎/香樟等)为主的林区,主要对这些区域的成熟林木以及灌木进行计划烧除处理.该框架为重庆应用计划烧除管理火灾风险提供有利依据,并为我国西南地区面临高强度火灾的山地城市提供了参考.
Confronting intensifying climate change and the heightened frequency of anomalous and extremely damaging fire events.There is a pressing need to shift fire management from"command and control"approach towards resilience-oriented strategies to taking action that coexists with fire disaster risk reduction.Natural ecosystems,their flora and fauna,and associated traditional ecological practices offer valuable insights into adaptation following low-intensity fire events,demonstrating significant potential for sustainable coexistence within certain fire regimes.While prescribed burning is widely acknowledged as an effective risk mitigation tool,the systematic conceptualization of"coexistence with fire"and its practical integration with prescribed fire management remain unexplored within China's specific national context.By examining both theoretical and practical frameworks for"coexistence with fire",both domestically and internationally,we propose the"Risk and Coexistence-Prescribed Burning(RCO-PB)"framework.This framework aims to systematically examine the complex relationship between wildfire risk and ecosystem coexistence capacity at a theoretical level and to develop a generalizable methodology for spatially optimizing fire management interventions at regional scales.Its core objectives are to effectively reduce high-intensity wildfire risk,scientifically identify priority areas for fire management,and foster ecological conditions conducive to biodiversity conservation.Guided by the RCO-PB framework,we employed machine learning and deep learning models to achieve high-precision wildfire risk prediction within the study area(Jialing River Basin,Chongqing)and conducted an in-depth assessment of ecosystem coexistence capacity with fire.Subsequently,a spatial optimization approach was applied to identify and delineate strategic priority zones for prescribed burning implementation,specifically targeting areas exhibiting high coexistence capacity concurrently with high wildfire risk.The results indicate that:land cover was the dominant predictive factor in the wildfire risk model.Within the basin,3%of the area was classified as extremely high risk(risk value≥0.659),and 7%as high risk(risk value≥0.411),with Beibei District's Jinyun Mountain identified as an extreme risk hotspot.Areas of high fire coexistence capacity were predominantly concentrated within ecosystems characterized by complex and dense vegetation structure.Notably,southern Hechuan District,western Beibei District,and northern Beibei District exhibited the highest coexistence capacity(top 10%,value>0.9).Spatial optimization identified priority areas for prescribed burning primarily within forest stands dominated by pure Pinus massoniana(Masson pine)and mixed coniferous-broadleaf forests(e.g.,Pinus massoniana-Cyclobalanopsis glauca/Cinnamomum camphora).Prescribed burning treatments within these priority zones would focus primarily on managing fuel loads associated with mature trees and shrubs.The RCO-PB framework provides a robust,spatially explicit foundation for implementing prescribed burning to manage wildfire risk in Chongqing.Furthermore,it offers a valuable reference model for developing sustainable fire management strategies in other mountainous urban regions of Southwest China facing significant threats from high-intensity wildfires.
郗婕;傅微;唐清源;Luca Maria Francesco Fabris
北京建筑大学建筑与城市规划学院,北京 100044||复旦大学生物多样性与生态工程教育部重点实验室,上海 200433北京建筑大学建筑与城市规划学院,北京 100044||复旦大学生物多样性与生态工程教育部重点实验室,上海 200433北京建筑大学建筑与城市规划学院,北京 100044北京建筑大学建筑与城市规划学院,北京 100044||米兰理工大学建筑规划与工程学院,米兰 20133
与火共存计划烧除火灾管理机器学习深度学习
coexistence with fireprescribed burningfire managementmachine learningdeep learning
《生态学报》 2026 (3)
1385-1396,12
生物多样性与生态工程教育部重点实验室开放课题(2025-FDU-KF-01)国家自然科学基金项目(41901220)
评论