基于空间动态面板数据模型,对2000～2009年中国市域城市火灾数据进行分析,探讨经济发展与气候变化共同作用下时空因素对中国城市火灾态势变化的宏观影响。结果显示:气候变干促使火灾恶化,而经济发展扭转了这种趋势并促使火灾态势整体改善。时空因素对中国城市火灾态势变化具有显著影响,并可被引申为火灾同化效应、火灾惯性效应、火灾警示效应。火灾安全管理部门应该充分利用这些效应,采取积极措施改善火灾态势。本文补充了城市地理学在城市火灾领域的研究不足;将抽象的时空因素引申为具有实际物理含义的虚拟变量,对相关研究具有借鉴意义。

Based on Spatial-temporal Dynamic panel data Model(SDM),with fire statistical data of China's 337 cities in 2000-2009,the influence of spatial-temporal factors on Chinese urban fire variation was analyzed,and the integrated influence of economic development and climate change on urban fire occurrence was considered.By Granger causality tests,per capita GDP and annual average relative humidity were used to represent economic development and climate change respectively.Three factors(fire rate,per capita GDP,annual average relative humidity)have long-run equilibrium relationships,so the Fire-Economy-Climate Model with the three variables is suitable.By making up spatial-temporal factors and some transformation,fire SDM(FSDM)was constructed.The results showed that,arid climate makes the fire situation worse,while economic development turns this trend back and makes the fire situation better.Response sensitivity of climate factors in fires is stronger than that of economic factors in fires,so one should make great efforts to mitigate fire variation because of future arid climate.Spatial-temporal factors have significant influence on Chinese urban fire situation,namely fire assimilation effect,fire inertia effect,and fire caution effect.Under fire assimilation effect,fire situations in different regions influence each other,and neighboring regions usually have similar fire trends,so fire administrative departments should pay more attention to cross-region cooperation;under fire inertia effect,the trend at the places where the fire situation was serious or mitigative in the earlier stage will be persisted in the future,so all regions should adjust fire administrative measures to local conditions;under fire caution effect,neighboring regions' previous serious fire situations will alarm local region to strengthen fire prevention or increase safety investment,and reduce local fire occurrence rate,so fire prevention should learn lessons from neighboring regions' serious fire situations and clear up hidden troubles.Furthermore,1st spatial lag variable is turned to an average value of neighboring units,this method simplifies the estimate process of spatial panel data model;and nonobjective spatial-temporal factors are expressed as meaningful dummy variables,which can provide reference for other studies.