Coupling Relationship Between Urban Spatial And Socio-economic Development With Multi-source Remote Sensing Data

Cities are where most human beings live now and in the future.The rapid urbanization all over the world brings efficiency and agglomeration in the economy and technology but also brings a series of urban problems related to society,environment,and resources.Since the 21 st century,countries,regions,and organizations around the world have been working on solving urban development problems and seeking scientific approaches to urban development.Goals and strategies such as the "Sustainable Development Goals(SDGs)","New Urbanization" and "High-quality urban development" are proposed.How to measure the development status scientifically is one of the basic issues that turn out to be particularly important.The process of urban development includes two important aspects: physical space development and social-economic development.These two have the same origin and converge,and they are an inseparable organic whole.Therefore,a full consideration of these two aspects in urban development research and clarifying the coupling relationship between the two are important guarantees for the overall evaluation of urban development.With the development of urban science and extensive use of technologies such as remote sensing and geographic information science,the research on the measurement of urban physical space and social-economic development status has been widely carried out.On top of that,research on the relationship between urban physical space development and social-economic development has been constantly emerging.However,there still exist some certain research gaps.First of all,the research on the correlation between urban physical space and social-economic development mostly relies on statistical data.For large-scale long-term research,the data accuracy is easily affected by problems such as statistical dimensions' inconsistency or monotonicity,which narrow the research boundary.At the same time,statistical data lacks continuous spatial attributes,thus most existing studies ignore the influence of urban spatial characteristics on the relationship between the two.Secondly,the existing research is insufficient for the extraction of lagging information between urban physical space and social-economic development.The mechanism of how these two lagging characteristics influence the urban development process requires further exploration.Finally,most of the researchers use simple indicators such as volume and height to describe the vertical expansion of urban space,which omits the details of the development in urban physical space such as the three-dimensional morphology of buildings and landscape structure.The research on the correlation between urban vertical space expansion and social-economic development still needs to be supplemented and improved.By summarizing previous studies,this article uses multisource remote sensing data to study the coupling relationship between urban physical space development and social-economic development in different spatial and temporal dimensions.The main research contents and results include:(1)We built an exponential model to represent the relationship between the artificial surface ratio and nighttime light intensity at the prefecture-level city.The study area covers 330 prefecture-level cities in Mainland China.We proposed a new and simple remote sensing method for urban development status,namely the urban "physical space – social economy" curve(UPSC).This curve can be used to reveal the urban socioeconomic development status,functional type of cities,and urban spatial structure at different scales.Our results show the model accuracy of most cities is good except for some cities in Western China.The coefficient ? of the exponential curve is the average NTL radiance(ANR)without artificial surfaces,which can indicate the base intensity of socioeconomic activity.The coefficient ? represents the concentration of socioeconomic activities in the areas with high urban density and reflects the urbanrural differences in socioeconomic activities.An analysis of their spatial patterns in China shows that not only cities in developed areas but also those in mountainous regions of Southwest China have a higher ? value.Areas with large ? are typically located in developing areas in Central and Western China.Furthermore,UPSC is helpful to identify the urban functional type,especially the industrial type,owing to the sensibility of UPSC coefficients for the urban function.The residuals of UPSC can indicate the demand for urban physical and economic construction in different areas of the city.Moreover,combining the distribution of artificial surface ratio with the corresponding residual can provide further information on urban planning and development strategy.(2)The Granger Causality Test(GCT)method was used to examine the Granger causality between the artificial surface areas and the variations of nighttime light intensity in 260 prefecture-level cities in China from 2000 to 2018.The time-series lag relationship between the urban horizontal expansion and social-economic development was further analyzed based on the artificial surface data and NPP-VIIRS-like nighttime light remote sensing data.The results show that the change of artificial surfaces in 101 cities is the Granger cause of nighttime light change,while the change of nighttime light in 108 cities is the Granger cause of the artificial surface change.Also,88 cities have the two-way directional Granger causality.Most of these cities are located in the east of Hu Line,and the lag period of the two-way directional causality is mostly 4-5years.From the spatial distribution of the lag period,the average growth rate of the nighttime light,and the average growth rate of the artificial surface,the response of nighttime light change to the artificial surface change is not significant or has a long lag period in most developed cities.The cities with a short lag period are mostly located in the areas with relatively weak social-economic development,and the economic development in these cities is most benefited from the contribution of land-rental.On the other hand,the response of the artificial surface changes to the nighttime light changes often needs a long time.The urban expansion of such cities is mostly oriented by their own needs and policies,instead of being strongly dependent on economic development.For some cities that are in the early stage of development,the urban infrastructure and social-economic basis are relatively weak,urban expansion turn out to have a quick response to the increase in social economics.88 cities that have the twoway directional Granger causality formed a development cycle of “land investment –stable economic benefits – urban expansion”,among which cities with a cycle of 10 years being the most.Finally,the coupling curve of urban physical space and social economy is modified according to the lag data of each city.Given a fixed impervious surface of 2010,the UPSC is fitted using corresponding lag nighttime light data of each city.The results show that some cities' fitting results tend to be more accurate.Most of them are common in having a more significant lag effect of nighttime light brightness variation as impervious surface changes.This shows that future urban studies need to take lag issues as well as the direction of lag characteristics into consideration.3)The e Xtreme Gradient Boosting(XGBoost)regression model was used to establish the correlation between 2D and 3D urban landscape structure and night light intensity at block and street scale using multi-source remote sensing data such as highresolution land cover data,laser point cloud data,digital surface model and Luojia-1-01-nighttime light data.The contribution of each landscape structure to night light intensity was analyzed using the Shapley Additive ex Planations(SHAP)model.The results show that the combination of two-dimensional and three-dimensional metrics explains most of the urban social-economic development conditions.Besides,different landscape metrics' contribution varies as the metrics' value varies.Even some interactions between several indices affect the contribution value.From the results of two-and three-dimensional landscape metrics' contributions,it can be concluded that the high efficiency of land use and the diversity of buildings play a positive role in the night light intensity on the block and street scale.They play a more significant role on the street scale where large buildings and the vertical development of blocks play a dominant role.Overall,a complicated urban structure that has a mixed use of land,dense diverse urban functions with high efficiency tend to be beneficial to the social and economic activities of streets and blocks,which can bring light to urban planning.At the same time,these landscape structure characteristics are consistent with the main characteristics of the compact urban sustainable development model,which proves the practicable consistency of sustainable development and socio-economic development.