21世纪以来，随着社会经济的不断发展，人口的快速增长打破了粮食、资源与环境之间的平衡关系，成为全世界最为关注的问题之一。通过对人口分布进行研究，不仅可以有效地缓解人口激增带来的压力，同时还能够为经济发展、政策制定、产业布局和城市规划提供重要的参考依据。但目前对人口分布的认识大多是基于国家、省市和县域层面，如何精确地获取更为详细的人口空间分布信息，成为各国家学者们致力于探究的重要任务。

本文以郑州市为研究区，对该区域2013-2020年间的人口空间分布演变特征进行分析。研究内容包括以下几方面：首先对NPP/VIIRS夜间灯光数据背景噪声、极高值及年际数据不连续的问题进行校正。接下来基于校正后的NPP/VIIRS灯光数据分别与郑州市及各区县常住人口数据，采用皮尔逊相关性分析验证了两者之间的相关性，通过夜间灯光数据与郑州市各区县人口数据分区建模，实现人口数据的空间化。然后以空间化结果为基础，通过时空动态分析郑州市人口分布演变特征变化。最后利用地理探测器分析影响人口分布的驱动力因子，并针对驱动力因子分析结果对郑州市人口分布提出一些建议。主要的研究结论如下：

（1）利用郑州市夜间灯光亮度值与常住人口统计数据进行相关性分析，结果表明：郑州市常住人口数、各区县常住人口数与对应的夜间灯光值呈现出较强的正相关性。

（2）利用夜光遥感数据的灯光强度DN值和常住人口数据进行回归分析，并基于回归分析结果进行分区建模。最后将2020年空间化统计结果与第七次人口普查数据进行拟合，拟合结果R²=0.9339，精度可以满足研究要求。

（3）从时间上看，郑州市常住人口数量在2013-2020年呈现出增长趋势。从空间上看，2019-2020年期间人口高值区域范围扩大，郑州市主城区人口数从2013-2019年呈现出缓慢稳定增长趋势，但在2019-2020年呈现出大幅度的增加。利用SLOPE趋势分析法对郑州市人口分布结果进行分析，发现郑州市的人口动态变化规律为三个人口密集区域为人口快速增长；人口中等区围绕着人口密集区域；人口稀少区的人口数量变化较小。人口重心呈现“先东南，再东北”方向移动的趋势，这表明郑州市的人口分布不均衡。从标准差椭圆方面可以看出郑州市人口分布呈现空间紧缩的趋势，在移动的主要方向上“先分散，后极化”，在次要方向上呈现出“先分散、后极化、再分散”的趋势。通过对郑州市人口密度进行空间自相关分析，发现其具有较强的正向空间自相关性，空间聚集程度呈现出不断增强的趋势，人口分布则为高度聚集和低度聚集两种情况。高度聚集的区域主要位于市中心城区，而低度聚集的区域则分布在远离市中心城区的周围区域。

（4）利用地理探测器分析郑州市人口分布的驱动力，发现2014-2020年社会因子的平均影响力明显高于经济因子。社会因子中科教文化、医疗保健服务、城镇化率保持着较高的影响力，经济因子中的社会消费品零售额对人口分布的影响力随着时间的推移逐渐增强。根据影响力可以发现，人们会优先选择拥有更好服务和城镇化率高的地区。并且，这些因素对人口分布的影响非单一作用的效果，而是通过多种因素的交互作用产生相互增强或者非线性增强的效果。随着时间的推移，整体上来看各因子之间各类因子之前的交互作用都在增强，社会因子之间的相互作用整体上比经济因子之间的相互作用对人口分布影响力强。驱动因子差异性分析结果表明：随着时间的推移，高程与其他因子对人口分布没有显著差别，而大多数的影响因子两两之间对人口分布具有明显差别。

Since the 21st century, with the continuous development of society and economy, the rapid growth of population has broken the balance between food, resources, and the environment, becoming one of the most concerned issues in the world. By studying population distribution, not only can it effectively alleviate the pressure brought by population surge, but it can also provide important reference basis for economic development, policy formulation, industrial layout, and urban planning. However, the current understanding of population distribution is mostly based on the national, provincial, and county levels. How to accurately obtain more detailed population spatial distribution information has become an important task that scholars from various countries are committed to exploring.

This article takes Zhengzhou City as the research area and analyzes the spatial distribution and evolution characteristics of population in the region from 2013 to 2020. The research content includes the following aspects: Firstly, correcting for background noise, extremely high values, and discontinuity in inter annual data of NPP/VIIRS nighttime lighting data. Next, based on the corrected NPP/VIIRS lighting data and the permanent population data of Zhengzhou City and various districts and counties, Pearson correlation analysis was used to verify the correlation between the two. By modeling the nighttime lighting data and the population data of various districts and counties in Zhengzhou City, the spatialization of population data was achieved. Then, based on the spatialization results, analyze the changes in population distribution and evolution characteristics of Zhengzhou City through spatiotemporal dynamics. Finally, geographic detectors are used to analyze the driving factors that affect population distribution, and some suggestions are made for the population distribution of Zhengzhou based on the driving factor analysis results. The main research conclusions are as follows:

(1) The correlation analysis between the nighttime light brightness value and the statistical data of permanent population in Zhengzhou shows that there is a strong correlation between the number of permanent residents in Zhengzhou, the number of permanent residents in each district and county, and the corresponding nighttime light value.

(2) Regression analysis was conducted using the light intensity DN value of night light remote sensing data and the resident population data, and zoning modeling was conducted based on the regression analysis results. Finally, the spatial statistical results of 2020 were fitted with the data of the seventh population census, and the fitting result R2=0.9339, with an accuracy that meets the research requirements.

(3) From a time perspective, the number of permanent residents in Zhengzhou showed an increasing trend from 2013 to 2020. From a spatial perspective, during the period of 2019 to 2020, the range of areas with high population values expanded. The population of the main urban area of Zhengzhou showed a slow and stable growth trend from 2013 to 2019, and showed a significant increase from 2019 to 2020. By using the SLOPE trend analysis method to analyze the population distribution results of Zhengzhou City, it was found that the population dynamic change pattern of Zhengzhou City is characterized by rapid population growth in three densely populated areas, medium population areas surrounding densely populated areas, and small population changes in sparsely populated areas. The population center of gravity shows a trend of moving in the direction of "first southeast, then northeast", indicating that the population distribution in Zhengzhou City is uneven. From the aspect of standard deviation ellipse, it can be seen that the population distribution in Zhengzhou shows a trend of spatial contraction, with a trend of "dispersion first, polarization later" in the main direction of movement and a trend of "dispersion first, polarization later, and dispersion later" in the secondary direction. Through spatial autocorrelation analysis of population density in Zhengzhou City, it was found that it has a strong positive spatial autocorrelation, and the degree of spatial aggregation shows a continuously increasing trend. The population distribution is divided into two types: high aggregation and low aggregation. Highly concentrated areas are mainly located in the city center, while low concentrated areas are distributed in the surrounding areas far from the city center.

(4) Using geographic detectors to analyze the driving forces of population distribution in Zhengzhou City, it was found that the average influence of social factors was significantly higher than that of economic factors from 2014 to 2020. The social factors such as science and education culture, healthcare services, and urbanization rate maintain a high influence, while the influence of social consumer goods retail sales on population distribution in economic factors gradually increases over time. Based on their influence, it can be found that people prioritize areas with better services and high urbanization rates. At the same time, the impact of these factors on population distribution is not a single effect, but rather a mutually reinforcing or non-linear enhancing effect through the interaction of multiple factors. Over time, overall, the interactions between various factors have increased, and the interactions between social factors have a stronger impact on population distribution than those between economic factors. The results of differential analysis of driving factors indicate that over time, there is no significant difference in population distribution between elevation and other factors, while most influencing factors have significant differences in population distribution between pairs.