Study On Population Spatial Distribution Based On Night Light Remote Sensing Image Data Of Luojia No.1

In the context of rapid urbanization,population has always been the focus and hot issue of various departments.The dynamic monitoring of the spatial distribution of urban population is conducive to urban spatial planning and resource allocation.The census data is based on the statistical data of administrative units.The spatial resolution is low,and it consumes manpower and financial resources,which is not conducive to the spatial visualization of population data.The current population spatialization method is mainly based on the establishment of models based on land use types and other population distribution factors and statistical population data.This method requires manual interpretation and requires a large amount of work.With the application of night light remote sensing images in the fields of city information extraction and socio-economic parameter estimation,night light image data can effectively observe the light information inside the city.Therefore,the luminous remote sensing image data can be used to simulate the spatial distribution of population data,thereby predicting the population data.However,most of the currently used NPP /VIIRS or DMSP / OLS night light image data has a low spatial resolution,and is mainly used in countries or economically developed provinces and cities.China launched the Luojia-1 luminous remote sensing image data in June 2018,the spatial resolution of which was increased to 130 m,and few people currently use the Luojia-1 remote sensing satellite image data to simulate population data.Based on the above reasons,this paper uses Luojia-1 night light remote sensing image data,combined with census data,to explore the correlation model of night light data and population data.A linear regression analysis method is used to explore the regression equations of population and night light data at different scales at the provincial,municipal,and county levels,and is used to estimate the population.Due to the complexity of urban structure,we study the spatial distribution models of population at different scales.The main research contents of this article are as follows:(1)Explore the correlation between population data and night light radiation values,and establish regression analysis models for night light remote sensing image data and population data at different administrative scales at the provincial,municipal,and county levels.(2)By discussing the relationship between night light radiation value and population data,this paper divides the provincial regions into economically developed regions and economically underdeveloped regions according to their economic development level,and performs regression analysis and research under different economic development levels.Models for population estimation at different scales and test the model.(3)Discuss the spatialization model of population data based on night light remote sensing images.Taking the urban area of Chengdu as an example,the linear regression analysis method is used to explore the relationship between building data,light radiation values and demographic data in the study area,and an estimation model of night light data in the population and building profile data is obtained.Predict the regional population,and use the grid as the scale to make a population density thematic map.The results show that the night lighting data and the population data have a high correlation at the provincial and municipal levels,and the correlation coefficients are above 0.6.The results are ideal.The combination of night lighting data and building contour data predicts the number of people and the actual The population is very close,with an absolute total error of 27,200 and an average relative error of about 17%.The model accuracy is good.