Spatial Distribution Analysis Of Fine Scale Daytime And Nighttime Population Based On The Internet Mining Information And Building Data Estimation

Gridded spatial information of population can effectively overcome shortcomings of the statistical population data like poor spatial resolution,low spatial analysis feasibility,difficult combination with other non statistical data,and so on.Furthermore,fine temporal and spatial resolution daytime and nighttime population distribution information can reflects more rich and accurate characteristics of real population situation,and consequently benefiting significantly to city living space research,planning and construction,ecological environment construction,disaster prevention and control and evaluation,and so on.In this study,a daytime and nighttime gridded population spatial distribution estimation model was proposed based on the internet platform mining information and building vector data,by took the six districts of Beijing,namely Dongcheng,Xicheng,Chaoyang,Fengtai,Shijingshan and Haidian Districts as research area.The main research results are as follows:1.A nighttime population spatial estimation model was establishedBased on the "down-top method" in population census,the relationship of the number of single-storey houses,the size of households and the number of floors in each "dwelling" building was used to build the nighttime population estimation model.Based on the household Internet mining data,building vector data and statistical data,the three variables in each "dwelling"building was achieved,and accordingly the spatial distribution of population at night was estimated.The results of precision verification showed that the model achieves good a estimation result,with the proportion of each household population(1～4 persons)estimated by Dongcheng District is not much different from the statistical data,and its total absolute error is only 1.97%.2.A daytime population spatial estimation model was establishedAccording to the different relationships between architectural function and daytime population accommodation,different estimation models were used for "dwelling" buildings,"working" buildings,"medical" buildings and "commercial" buildings.In terms of "dwelling"buildings,the number structure and spatial distribution of 12 main family structures are estimated based on a number of statistical data,so as to obtain the spatial distribution of elderly population,adult population and juvenile population.Then the daytime population in "residential" buildings is estimated according to the characteristics of daytime travel of people in different age groups.For the rest four types building,the area weight interpolation method was adopted to estimate the daytime population distribution.The accuracy verification results showed that the total absolute error of the estimated population of each age group and the statistical data was between 9 and 25%.3.The different between daytime and nighttime population spatial distribution in the study area is obvious.The nighttime population is concentrated between the Second Ring Road to the Fourth Ring Road of Beijing,while the daytime population is concentrated in the industrial bases within the Second Ring Road and outside the Fourth Ring Road of BeijingOn the whole,spatial distribution of nighttime population presents a feature of "less distribution inside and outside,more distribution in the middle",that is,the nighttime population is concentrated between the Second Ring Road and the Fourth Ring Road of Beijing,and less distributes within the Second Ring Road and outside the Fourth Ring Road,while the spatial distribution characteristics of daytime population are the opposite.It is characterized by "more internal and external distribution,less middle distribution",that is,the daytime population is concentrated in the area within the Second Ring Road and the Fourth Ring Road and outside the Fourth Ring Road,while the high-value area from the Second Ring Road to the Fourth Ring Road is less distributed and more dispersed.