Abstract: The study of temporal and spatial distribution of population has great practical significance in urban planning management, land use layout optimization and ecological environment assessment. According to the spatial displacement law of people’s daily activities, this research established a relationship model for three components which are urban population, time (daytime and nighttime) and land use with the help of GIS modeling and spatial analysis technology, then applied this model to explore the temporal and spatial characteristics of different types of population. Taking Qinhuai District of Nanjing, China as the study area, this research used the remote sensing image with the spatial resolution of 0.27 m combined with the statistical data and field investigation to quantitatively estimate urban population during the daytime and nighttime, then analyzed its spatial distribution characteristics at 100 m grid scale. Furthermore, the results of urban population spatialization were verified under the scale of streets and buildings. The research results are as follows: (1) The use of multi-source data including high spatial resolution remote sensing images and streetscape maps can effectively overcome the difficulties of obtaining the land use types and building data under the complex urban underlying surface, which can improve the spatial resolution of urban population estimation. (2) The spatial distribution of population on the street and the building scale can be reasonably estimated by using the land use type and the spatial attribute information of buildings. (3) Because of the influence of many factors, such as the function layout of urban buildings and the difference between urban and rural development, there are significant distribution differences between the daytime and nighttime population in spatial structure. During the daytime, the spatial distribution of population is much wider, and it has more obvious characteristics of spatial agglomeration compared to that of nighttime. During the nighttime, the region of high value distribution of population is relatively dispersed, and the agglomeration areas transfer to the resident place.
Key words：population at daytime and nighttime; spatial and temporal distribution; remote sensing(RS); geographic information analysis; Nanjing; Qinhuai District