基于夜间灯光的城市居民直接碳排放及影响因素——以中原经济区为例

城市居民直接能源消耗及其碳排放对区域碳减排政策的制定具有重要影响。受统计数据缺乏与研究方法的限制,当前的研究不仅较少探讨精细空间尺度上的城市居民直接碳排放,同时也缺乏在县级尺度上对人均居民直接碳排放的影响因素进行深入分析。鉴于此,论文以中原经济区为例,通过引入夜间灯光数据,利用增强型饱和校正模型估算了网格尺度上的城市居民碳排放,并采用地理加权回归模型对其影响因素进行了分析。主要研究发现：中原经济区的碳排放总体空间特征是西北高、东南低。郑州市市辖区的城市居民直接碳排放总量位于首位,而邢台县、辉县市和襄垣县的人均碳排放较高。此外,就其影响因素来看,人均GDP、碳排放强度、第二产业比重和HDD（Heating Degree Days,热度日）均表现为正效应,城镇化率为负效应,而CDD（Cooling Degree Days,冷度日）的系数有正有负。城市居民直接碳排放的影响因素分析为中原经济区制定切实可行的区域碳排放政策提供了重要的基础理论依据。

Urban Residential CO2 Emissions and Its Determinants: A Case Study of Central Plains Economic Region

Urban residential energy consumption and CO₂ emissions have a major impact on regional carbon reduction policy. Due to lack of data and limitation of methods, spatial characteristics of urban residential CO₂ emissions and its influencing factors at county scale have rarely been discussed. Therefore, taking the Central Plains Economic Region as a case, this paper used the DMSP-OLS nighttime light imageries corrected by enhanced saturation correction model to estimate the spatial distribution of carbon emissions at the 1 km resolution and analyze its influencing factors at the county scale with geographical weighted regression model. The rationality of our method was confirmed in the process of estimating carbon emissions. The linear regression model between the estimated CO₂ emissions and the statistical CO₂ emissions of urban residents with R² of 0.837 1 demonstrated that the inversion model based on the nighttime light imageries has strong feasibility and suitability. Based on the carbon emissions of urban residents at 1 km resolution, we can calculate carbon emissions across administrative boundaries. In terms of spatial characteristics, CO₂ emissions in the northwest Central Plains Economic Region were significantly higher than those in the southeast. Moreover, Zhengzhou District held the first place of CO₂ emissions with total amount of 2.47 ×10⁶ t, accounting for 6.58% of total CO₂ emissions. However, Xingtai, Huixian and Xiangyuan were the regions with high CO₂ emissions per capita, and these regions with higher carbon emissions per capita should be paid more attention. With respect to influencing factors, per capita GDP, carbon intensity, proportion of secondary industry and HDD (Heating Degree Days) all have the positive effects on urban residential carbon emissions at the county scale. However, urbanization rate presented a negative effect. Furthermore, it's important to note that CDD (Cooling Degree Days) has positive impact on urban residential carbon emissions in some cities while has negative impact in other cities. However, the maximum coefficient of its negative impact was 0.046 6, which could be ignored compared with the coefficient of positive impact. Therefore, CDD would be regarded as a positive influencing factor as a whole. Overall, the analysis of influencing factors in this paper provides an important theory basis for policy-makers to carry out more feasible policy on regional carbon emissions in the Central Plains Economic Region.