【摘要】 夜间灯光数据在社会经济活动方面得到广泛应用，有必要研究夜间灯光信息的影响因素。文章针对2020年度NPP/VIIRS合成灯光数据，从县域单元尺度，采用回归分析法及区域分析法，研究人口数量、GDP和区域开发强度对县域灯光值的影响。结果表明：三次函数、二次函数、幂函数和线性函数对于灯光均值和总值都有较好的拟合效果，效率上，幂函数最适合模拟灯光值与社会经济数据的关系；三因素模型的拟合效果优于双因素模型的拟合效果，双因素模型的拟合效果优于单因素模型的拟合效果；县域处于开发早期的灯光信息更受不透水面率影响，开发中期同时受不透水率、人口和GDP的影响，开发完成期主要受人口和GDP的影响；人口数对于灯光总值、不透水面率对于灯光均值强相关，GDP对于灯光总值较强相关，仅京津冀考虑三因素时与灯光最大值相关性较好。结果可为应用NPP/VIIRS提供参考。更多还原

【Abstract】 Nighttime light(NTL) data is widely used as a representation of human social and economic activity. However, few studies have analyzed NTL information with its interfering factor. This paper investigates the impacts of main socioeconomic data such as population size, gross domestic product(GDP) and regional development intensity on three NTL statistics values from the county level, from both national and regional perspectives, using regression analysis and regional analysis in the 2020 NPP/VIIRS synthetic lighting data. The results show that: cubic function, quadratic function, power function and linear function fit well for NTL mean and total value of the NTL,while in terms of comprehensive efficiency, power function is most suitable for modeling NTL values and socio-economic data; the three factor model is the best, followed by two factors and single factor; the NTL of the county is more affected by the impervious surface rate in the early stage of development, affected by the impervious rate, population and GDP in the middle stage, and mainly affected by the population and GDP in the completion period; population size is strongly correlated with the total value of NTL,such as development intensity with NTL mean, GDP is highly correlated with the total value of NTL,only Beijing-Tianjin-Hebei region has a passable relevant three factors model with NTL maximum. The research results provide a reference for the refined application of NPP/VIIRS data in estimating population, GDP and land urbanization.更多还原