哈长城市群县域碳排放空间溢出效应及影响因素研究——基于NPP-VIIRS夜间灯光数据的实证

基于2012-2016年NPP-VIIRS夜间灯光数据测算哈长城市群共计68个县级单元碳排放数据，采用空间自相关、位序规模法则、空间计量模型以及空间马尔科夫链模型，对县级尺度城市碳排放空间特征、影响因素和动态空间溢出作用进行分析.结果表明：①城市碳排放呈现逐年下降的趋势，空间Moran''s I指数表明研究区城市碳排放存在高度的空间自相关，碳排放的高值集聚性呈降低的趋势.②位序规模法则表明，研究区全部城市的碳排放属于次位型分布，高位次城市的碳排放表现突出；在前十位城市的碳排放规模先减少再增加，由分散向集中演变.③多种空间面板模型对比分析表明，城市经济水平和人口密度因素呈现显著的正相关关系；固定投资和外商投资因素仅在时间固定模型中起到正向影响作用；技术进步以及路网密度因素则起到显著的抑制作用.④空间马尔科夫链分析结果表明，城市碳排放空间溢出效应明显，位于低水平区域与高水平区域的城市在转移过程中保持稳定；位于中低水平区域的城市与高碳排放的城市相邻会降低转移概率，反之则会提升转移概率.

Study on spatial spillover effects and influencing factors of carbon emissions in county areas of Ha-Chang city group: Evidence from NPP-VIIRS nightlight data

Based on the carbon emission data, which were calculated from the 2012-2016 NPP-VIIRS nightlight data of a total of 68 county-level units in the Ha-Chang urban agglomeration, the spatial characteristics, influencing factors and dynamic spatial spillover effects of urban carbon emission were analyzed at county level with spatial autocorrelation, rank-size rule, spatial econometric model and spatial Markov-chain model. The results indicate that, firstly, city carbon emission shows a downward trend year by year. The spatial Moran''s I index indicates that there is a high spatial autocorrelation of urban carbon emission in the study area, and the high value aggregation of carbon emission is decreasing. Secondly, according to rank-size rule, all the urban carbon emission in the study area shows a sub-type distribution, and high-level cities witnesses a huge amount of carbon emission. The amount of carbon emission from the top ten cities reduced initially and then increased, from decentralization to concentration. Thirdly, having compared using various spatial panel models, the factors of the development level of urban economic and population density show a significant positive correlation; only in the time fixed model, play the fixed investment and foreign investment factors a positive role; technological progress and road network density factors play a significant inhibitory role. At last, spatial Markov chain analysis results show that urban carbon emission space spillover effect is obvious, and low-level cities and high-level ones remain stable during the transfer process; if low- and medium-level cities are adjacent to those with high carbon emission, the probability of transition will be reduced; and if not, it will be promoted.