夜间灯光数据替代GDP指标时需关注的问题  

Issues Needing Attention in the Substitution of Night Light Data for GDP

作　　者：秦永[1,2] 方迎风[3] 刘凯敏[2] Qin Yong;Fang Yingfeng;Liu Kaimin(School of Economics,Nanjing University,Nanjing 210093,China;School of Economics,Nanjing Audit University,Nanjing 211815,China;School of Economics and Management,Wuhan University,Wuhan 430072,China)

机构地区：[1]南京大学经济学院,南京210093 [2]南京审计大学经济学院,南京211815 [3]武汉大学经济与管理学院,武汉430072

出　　处：《统计与决策》2023年第2期5-10,共6页Statistics & Decision

基　　金：国家社会科学基金资助项目(20BJL144);国家自然科学青年基金项目(72003143)。

摘　　要：客观的夜间灯光数据被频繁地作为GDP的替代指标,出现在越来越多的经济学文献中。然而随着SNPP卫星替代DMSP卫星,研究领域从城市扩展到农村,有三个新问题常常被学者们忽略:DMSP/OLS和SNPP/VIIRS夜间灯光数据是否可比?夜间灯光数据能否替代农村地区的GDP?哪一个夜间灯光指标最优?忽视这三个问题很可能导致得出错误的结论。文章通过对DMSP/OLS和SNPP/VIIRS夜间灯光数据的比较发现,分辨率、成像时间、传感器的差异使二者不具可比性。我国县级农村层面的数据证明,夜间灯光数据仍然可以作为GDP的替代指标;农业占GDP比例较高的子样本证实,夜光强度为GDP的最优替代指标。Objective nighttime light data are frequently used as an alternative indicator of GDP and appear in more and more economic literature.However,three issues are usually ignored by the economists while the satellite SNPP replace satellite DMSP and the field of research is expanded from urban to rural areas.Are DMSP/OLS and SNPP/VIIRS nighttime light data comparable?Can we substitute nighttime light for GDP in rural areas?Which nighttime light indicator is the best?Neglecting the three issues is likely to result in wrong conclusions.By comparing the night light data of DMSP/OLS and SNPP/VIIRS,this paper finds that the differences in resolution,imaging time and sensor make them incomparable.The data at the county level proves that the night light data could still be used as an alternative indicator of GDP,and that the subsample with a higher proportion of agriculture to GDP confirms that luminous intensity is the best proxy for GDP.

关 键 词：夜间灯光数据 农村 GDP 夜光强度 夜光总值 

分 类 号：F221[经济管理—国民经济]