Spatial Heterogeneity Of Influencing Factors Of Population Flow In China Based On Spatiotemporal Data

Population mobility is one of the most far-reaching geographical processes in China.Since the reform and opening up,population mobility has become a significant phenomenon in the field of population in China's economic and social development.It is the inevitable product of market economy and in turn affects the process of economic and social development.At the same time,the blowout development of large spatial and temporal data provides a certain research basis for geography,especially for human geography.Geographers have been doing a lot of research on population mobility and its influencing factors.However,limited to the research data,most of the research scales are national or provincial scale,and most of the influencing factors index system are constructed using composite quality index system.Therefore,this paper is based on the spatial-temporal data of population flow between provinces in 337 administrative divisions of China provided by Baidu Migration Map.Firstly,through literature review and mathematical formula deduction,the foundation and reason of establishing composite mass index system in gravity model are clarified from the point of view of model construction.And study The relationship and difference between composite quality index system and single quality index.Secondly,study the difference of fitting results of gravity model between single quality index system and composite quality index system on the national scale.Finally,on the basis of the former,using singlemass gravity model,to explore the spatial heterogeneity of the explanatory power of socio-economic indicators to the intensity of inter-provincial population flow in China.It is concluded that:(1)Different quality setting methods will result in obvious differences in the explanatory power of the same quality index to the intensity of population flow in the fitting results of gravity model.The gravity model of single mass and composite mass has obvious differences in the use of the results because of the different way of setting the mass.In previous studies,the composite mass gravity model has been used more widely,but this kind of mass setting method inevitably affects each other in the model.Although the equivalent function expression in product form can represent the total mass scale under the action of each mass synthesis,it can not evade the interaction between quality indicators.Therefore,in exploring how much the quality indicators can explain the intensity of population flow,we need to abandon the situation that the quality indicators can not be excluded because of the function expression.In other words,by fitting and regressing the gravity model with single quality index,the interference between the quality indexes will be weakened essentially,which is more conducive to discussing the difference of the explanatory power of each quality index to the intensity of population flow.(2)Pop,Employee,GDP,Passenger,Wages have obvious spatial heterogeneity in explaining the strength of population mobility.The nugget values of goodness of fit of regression of five index models were respectively 0.268,0.499,0.473,0.331 and 0.245.That is to say,the explanatory power of the five indicators to the intensity of population flow itself has a positive basement effect caused by various random factors.Base values were respectively 0.69,0.611,0.528,0.607 and 0.779.The nugget coefficients of Employee and GDP are 0.817 and 896 respectively,which shows that the spatial heterogeneity in the study area is mainly controlled by random factors and weakly controlled by spatial autocorrelation.At the same time,because the ratio is close to 1,the regionalized variables have constant variation in space.The nugget coefficients of Pop,Passenger and Wages are respectively 0.388,0.545 and 0.315.That is to say,the spatial heterogeneity of the explanatory power of the three influencing indicators is affected by the random part and spatial autocorrelation,and both of them are moderate.Furthermore,the spatial heterogeneity of Pop and Wages is mainly controlled by a small range of special factors,and the degree of spatial heterogeneity caused by spatial autocorrelation is 61.2% and 68.5% respectively,which indicates that the spatial heterogeneity of the explanatory power of the average annual population and the average wage of workers to the intensity of population mobility is 61.2% and 68.5% respectively from spatial autocorrelation.The fitting of the semi-variance function in this paper proves that the complexity and uncertainty of the impact of the same index on the intensity of regional population flow are gradually deepened with the deepening of the research accuracy at the prefectural scale.(3)Quality indicators have obvious spatial differences in the explanatory power to the population flow intensity of administrative divisions.Firstly,considering the spatial damping effect,Employee,GDP,Passenger,Wages and Pop can independently and independently explain 34.72%,29.67%,11.87%,7.72% and 7.12% of China's regional population mobility at prefecture level with a goodness of fit of more than 0.6.Secondly,from the perspective of spatial differences in explanatory power,the lowvalue areas are mainly concentrated in Xinjiang,Tibet,Qinghai,western Sichuan and Northeast China.The high-value areas are mainly located in Beijing,Tianjin and Hebei,the Yangtze River Delta and the Pearl River Delta,and around the Bohai Sea.The reason is that the low-value concentration areas are mostly located in the border areas of China,and the average path of population flow is longer.At the same time,influenced by the level of social and economic development,natural environment,regional culture and other factors,the initiative,frequency and intensity of longdistance inter-provincial population flow are lower,which leads to the low goodness of fit of model regression.In high-value areas,the activities of various social and economic material flows are strong,and they form a mutually reinforcing balance system with the intensity of population flows,which is highly correlated with each other.Therefore,social and economic indicators can explain the intensity of population flows in this region to a higher degree.