【作者】 徐波；

【导师】 宫鹏； 李艳；

【作者基本信息】 南京大学 ， 地图学与地理信息系统， 2012， 博士

【摘要】 本文首先对中国市域尺度火灾(不含森林、草原、军队、矿山火灾,以下同)及相关因子的时空变化进行探索性分析；然后以重庆、江苏、全国为例,基十向量误差修止模型,研究了经济发展、气候变化与火灾变化的因果关系、长期均衡关系,及经济因子、气候因子对火灾的脉冲影响和贡献度；根据火灾时空变化特征,进行火灾区划,对比分析了火灾的地域分异,提出火灾Kuznets假说并对其进行实证研究：基于空间动态面板数据模型,分析了火灾同化效应、火灾惯性效应、火灾警示效应。论文主要结论如下：1)中国火灾空间聚集显著,近年来火灾态势呈改善趋势通过对中国市域火灾分布研究发现,2002年以前,中国整体的火灾发生率逐年升高,火灾空间聚集趋势也随之加强,较高火灾发生率单元主要分布于东北地区、京津鲁、长三角、新疆、内蒙、宁夏等地,低火灾发生率单元则主要分布于华南、西南等地,区域差异明显；2003年以后,随着“和谐社会”、“执政为民”的理念深入人心,各级政府加强了火灾安全管理和投入,火灾状况逐渐趋向缓和,火灾的空间聚集趋势也随之减弱并趋向分散。改善区共186个,主要分布于东北、华北、华东、华南等地区；恶化区共74个,大部分分布于黑河-腾冲线以西地区及陕西、湖北等地；波动区共77个,分布于西南地区及其它零星地区。这导致中国火灾重心逐渐朝西部转移。2)经济发展、气候变化是促使火灾变化的宏观原因对中国全国、省域(以江苏为代表)、市域(以重庆为代表)研究发现,从长期来看,经济发展、气候变干与火灾态势变化之间存在因果关系。经济因子、气候因子与火灾因子之间存在长期均衡关系,短时间之内火灾变化可能偏离经济发展与气候变化的影响,长久来看火灾态势会回归到由经济水平、气候条件所决定的均衡状态。火灾态势对气候因子的影响敏感度强于对经济因子的影响敏感度。气候变化对火灾的贡献度高于经济发展对火灾的贡献度,湿度正向冲击对火灾有抑制作用,但不同地区经济正向冲击对火灾有不同影响,在江苏体现为负向响应,在重庆则体现为止向响应。考虑到未来气候仍将继续变干,这对未来火灾防灾减灾的压力和挑战将越来越大。3)火灾存在显著的地域分异全国整体上火灾随经济发展呈改善趋势,人均GDP每增加1%,火灾发生率降低0.43%；东北地区由于长期为火灾高发区,对火灾安全较为重视,火灾随经济发展改善幅度最明显,人均GDP每增加1%,火灾发生率降低1.11%；东南地区次之,其经济较为发达,对火灾安全的投入较多,人均GDP每增加1%,火灾发生率降低0.86%；华北、西南火灾也随经济发展而改善,人均GDP每增加1%,火灾发生率分别降低0.69%和0.62%；中南地区人均GDP每增加1%,火灾发生率平均降低0.23%；西部地区随着经济发展,火灾发生率整体早上升趋势,人均GDP每增加1%,火灾发生率上升0.16%。全国各区域气候变干均会促使火灾恶化。西南地区火灾对气候变化的敏感度最高,年平均相对湿度每降低1%,火灾发生率上升4.06%；东北、中南地区火灾对气候变化的敏感度次之,年平均相对湿度每降低1%,火灾发生率分别上升3.62%、2.14%；东南、华北、西部地区火灾对气候变化的敏感度较低,年平均湿度每降低1%,火灾发生率将分别升高1.99%、1.03%、0.86%。气候季节变化越剧烈、经济越发达的地区,火灾对气候变化的敏感度越低。全国总体上经济发展与火灾变化呈“倒U型曲线”,称之为火灾Kuznets曲线(FKC)：东北、华北、东南、西南地区存在FKC；而中南、西部地区经济发展与火灾变化呈“正U型曲线”,不存在FKC。4)时空因素对中国市域火灾态势变化影响显著时空因素对中国火灾态势变化具有很显著的影响,这种影响可以分别引申为火灾同化效应、火灾惯性效应、火灾警示效应。由于存在火灾同化效应,各地区的火灾态势会相互影响,周边地区火灾态势的改善或恶化会促使本地火灾态势改善或恶化；相邻地区火灾发生率若增长1%,则该地区火灾发生率受到连带影响而增长0.57%。火灾受到火灾惯性的影响较大,前期火灾高发的地区未来火灾仍会相对高发,前期火灾已经改善的地区未来火灾仍会继续改善；往期火灾发生率若降低1%,则本期火灾发生率可能降低0.54%。相邻地区的往期火灾,特别是重特大火灾会对本地区产生警示作用,相邻地区的火灾安全投入会对本地产生示范作用并刺激本地区加大投入,进而降低火灾发生率；周边往期火灾每上升1%,则该单元本期火灾将降低0.20%。火灾警示效应与火灾同化效应、火灾惯性作用相比相对较低。更多还原

【Abstract】 Firstly, fire (except fires of forest, grassland, Mine and army troops) spatial-temporal variations in urban area of China had been studied by exploration analysis. Then, VEC (Vector Error Correction) model was used, proved that economic development and climate change do Granger cause fire variations, concluded that the three factors have long-run equilibrium relationships, and analyzed the impulse response and contribution of economic development and climate change to fire variations. Further more, compared to the regional diversity of the fire zonings with panel data model, put forward Fire Kuznets Curve (FKC) hypothesis and confirmed it in different scales. Finally, fire assimilation effect, fire inertia effect, and fire caution effect had been studied with spatial-temporal dynamic panel data model.The article’s main conclusions are listed as below:1) Chinese fire spatial distributions have obvious clustering, and fire situations have mitigated in recent years.Chinese urban fire rate enhanced up and fire spatial clustering strengthened before2002year by year. The higher fire rate units mainly distributed in East-north region, around Bohai Area, the Yangze River Delta,Xinjiang, Neimeng, Ningxia etc.; lower fire rate units mainly distributed in South of China, West-south China etc. Along with the idea of "harmonious society" and "government for the people", more attention to fire safety administration and fire investments, fire situations had mitigated and fire spatial clustering had weakened gradually after2003. The ameliorated units count up to186, distribution in East-north Region, North of China, East of China, and South of China mainly; deteriorated units count up to74, location west of Heihe-tengchong line and Shanxi, Hubei provinces mostly; fluctuated units count up to77, distribution in West-south region and other scattered zones. All of these induced that Chinese fire center of gravity moved to the West gradually.2) Economic development and climate change do cause fire variationEconomic development and climate change both are Granger cause of fire variations; and the three factors have long-run equilibrium relationships. Response sensitivity of climate factor in fires is stronger than that of economic factor in fires. The contributions of climate change to fire variations are higher than that of economic development to fire variations besides fire inertia’s contribution. Humidity positive impulses have restrained effect to fires, but the impulses of economic factor in fires are different in different regions.3) Fires have obvious regional diversityThe whole nation’s fire rates were downward with economic development, per capita GDP increased by1%, while fire rate reduced by0.43%. Fire rates in western regions were upward with economic development, per capita GDP increased by1%, while fire rate increased by0.16%. Other regions’fire rates were downward with economic development. The response sensitivity of economic development to fire variations in East-north region is highest, in East-south is relatively higher, in North of China and West-south region is middle, in Middle-south region is relatively lower, per capita GDP increased by1%, while fire rate reduced by1.11%,0.86%,0.69%,0.62%, and0.23%respectively.Arid climate makes fire situation worse in all regions. The response sensitivity of climate change to fire variations in West-south region is highest, in East-north and Middle-south region is middle, in East-south, north of China, and western region is relatively lower; annual average relative humidity reduced by1%, while fire rate increased by4.06%,3.62%,2.14%,1.99%,1.03%, and0.86%respectively. Where seasonal climate is changed acutely or economy is developed, where response sensitivity of climate change to fire variation is lower.The whole nation’s fire situation figured as "inverted U-curve" with economic development, named as fire Kuznets curve (FKC). There are FKC in East-north region, north of China, East-south region, and West-south region, but not in Middle-south region and West region.4) Fires have significant spatial-temporal lag effects.There is significant fire assimilation effect in China. Fire situations in different regions will influence each other, and neighboring regions have similar fire trends usually; while neighboring regions’fire rate increased1%, local region’s fire rate increased0.57%accordingly. Fire inertia effect has great positive influence to fire, where fire situations are serious or amelioration previous will still are serious or amelioration; while previous period’s fire rate reduced1%, this period’s fire rate reduced0.54%accordingly. Fire has caution effect so that neighboring regions’previous serious fire situations, especially conflagrations, will alarm the local region to strengthen fire preventions or increase safety investments; while neighboring regions previous period’ fire rate increased1%, local region this period’s fire rate reduced0.20%accordingly. The fire caution effect is relatively lower than fire assimilation effect and fire inertia effect.更多还原