Analysis Of Physical Urban Spatial Morphology And Driving Factors In Beijing-Tianjin-Hebei Based On Multi-source Dat

As the urbanisation process of the Beijing-Tianjin-Hebei urban agglomeration faces uneven development,this study analyses the spatial morphology of the cities in Beijing-TianjinHebei and their internal urban areas to provide a scientific reference for actively promoting the national strategy of "Beijing-Tianjin-Hebei Cooperative Development".This study uses Beijing-Tianjin-Hebei as the study area,and analyses the spatial patterns and drivers of the cities in the Beijing-Tianjin-Hebei physical region from 2012 to 2021 based on multiple sources of data.Firstly,an enhanced nighttime light city index is constructed by combining Landsat and NPP/VIIRS images to identify the extent of cities in the physical region;secondly,a deep neural network model is constructed based on multi-source data such as population,nighttime light,POI and road network density to obtain the spatialisation data of population at grid scale from 2012 to 2021;using the population change rate as a measurement index,the spatial pattern of cities is divided into The spatial pattern of cities is divided into heavy expansion,medium expansion,light expansion,heavy contraction,medium contraction and light contraction;the spatial pattern of cities and urban areas is analysed from the perspective of the number of spatial patterns of cities and urban areas,their geographical distribution and the proportion of development in their provinces.The main findings are as follows.(1)The distribution of the extracted cities and built-up areas in Beijing,Tianjin and Hebei is consistent and the area error does not exceed 10%,and the distribution pattern is mainly "peripheral dispersal and central agglomeration",with significant spatial agglomeration characteristics.(2)The overall fitting accuracy of the population spatialization model at the district and county scales constructed based on multi-source feature data and deep neural networks is 97%,which can be used for the retrospective simulation of population at the grid scale;the fitting accuracy of the population spatialization results at the grid scale is 92%,which is closer to the population census data compared with the three types of population data sets in the international arena.(3)From 2012 to 2021,the physical cities in Beijing,Tianjin and Hebei showed an overall trend of expansion;after the optimization of urban structure from 2015,there was a local contraction within the resource-based cities,and the spatial pattern changed from heavy contraction to medium or light contraction,indicating that the economic development of these cities is still dominated by growth,and there is no large-scale and sustained economic recession after the transformation of urban structure;at the same time,some At the same time,some of the old urban areas are close to saturation and have begun to adjust their development patterns internally,leading to a contraction of spatial patterns in urban areas.(4)The three drivers of urban economy,urban scale and urban facilities all play a positive role in the development of cities,but the contribution of different drivers to urban development varies,with the most significant changes in the coefficients of urban economy and urban scale,indicating that increasing the economic level and scale of cities can accelerate urban development.