| Currently,the world is experiencing an unprecedented large-scale urbanization process.Urbanization is a complicated and multi-scale spatiotemporal geographical process involving the evolutions of multiple factors,such as land use,population,and socioeconomic activities.Among them,the continuous expansion of high-density and high-intensity human activities in space is an intuitive representation of urbanization.Through the land use/cover and landscape changes,urbanization inevitably intensifies urban heat island,increases anthropogenic greenhouse gas emissions and building energy use,and influences the ecosystem at different degrees.Thus,a series of ecological and environmental issues may arise,which are likely to endanger the public health.Meanwhile,irrational urbanization can exacerbate the regional polarization between the rich and the poor,and intensify the imbalance of industrial structure,and thereby induce various socio-economic issues.As human-dominated habitats,urban areas undoubtedly play a crucial role in global change and regional development.Thus,timely monitoring and quantitatively perceiving the dynamics of urban areas and high-intensity human activities over large regions and long time series are urgent.This is of great significance for deeply understanding the urbanization process and its development laws,optimizing regional pattern,realizing sustainable development of resources and environment,and assisting in solving various social problems.With advantages of wide coverage,small storage,continuous time-spanning,and effective characterization of human activities,multi-source nighttime light remote sensing data provide a new opportunity for comprehensively perceiving and quantifying urbanization and socio-economic activities at the large scales.The reconstruction of the urbanization process is on the basis of long-term and continuous monitoring of urbanization dynamics,while single nighttime light remote sensing data,with limited temporal coverage,are unable to realize such long-term monitoring.Therefore,integrating multi-source nighttime light remote sensing data together will be a new trend for the future nighttime light remote sensing studies on urbanization monitoring.However,there are many difficulties and challenges in monitoring urbanization dynamics regionally and globally using multi-source nighttime light remote sensing data.Firstly,nighttime light remote sensing data from different sensors are incomparable,how to integrate these nighttime light signals from different sensors effectively,to conduct the inter-calibration of multi-source nighttime light data,to build longer time-series,comparable and continuous nighttime light dataset,for providing precious data basis for monitoring the long-term urbanization dynamics? Secondly,considering the complexity and diversity of urbanization over different regions and periods,as well as the limited availability of auxiliary data in large scale and long time-series,how to make full use of the spatialstructure characteristics of nighttime light signals in the urban-rural transition zones,to quantify and identify the patterns of urban-rural nighttime light gradient variations over regions with different development levels and patterns in the premise of reducing the dependence on auxiliary data as much as possible,to realize the rapid identification and extraction of urban extents experiencing different urbanization backgrounds,for providing a universal and transferable urban extraction method for deeply perceiving and understanding the spatiotemporal dynamic characteristics of large scale and long time-series urbanization?As a necessary channel for promoting the Maritime Silk Road,as well as a key hub for developing China's foreign trade and economic development,Southeast Asia has important strategic position.Due to historical and political factors,and industrial structure and natural conditions,urban development and regional environment issues in different Southeast Asian countries are severe.However,the spatiotemporal dynamics of urbanization,the trend of environmental quality change,and the potential impact of urbanization on environmental change,in this region,are still lack of in-depth studies in long time series.Considering the limitations and challenges in current studies on urbanization monitoring using multi-source nighttime light remote sensing data,as well as the urgency and importance of both urbanization monitoring and environmental change assessment in Southeast Asia,a series of methods and models were proposed in this study for integrating multi-source nighttime light signals,monitoring and perceiving regional urbanization and comprehensive environmental change through collaborating multi-source nighttime light data and combining other remote sensing time-series products including PM2.5 concentration data,vegetation index data,and land surface temperature data.Based on these methods and models,the spatiotemporal characteristics of urbanization in Southeast Asia over the past 30 years were clarified,the trends and characteristics of the comprehensive environmental quality change in this region were quantitatively assessed,and the potential impacts of the urbanization process on its comprehensive environmental quality change were then discussed.The methods and models developed in this study could provide a robust and general paradigm for studies on the spatiotemporal monitoring of urbanization dynamics and the assessment of comprehensive environmental change,spanning large areas and long time series.The results and conclusions obtained in this study are of great significance and reference value for deeply understanding the urbanization process and its development laws in Southeast Asia,scientifically recognizing the trends and characteristics of the comprehensive environmental change in this region,as well as further exploring the coupling mechanism between urbanization and environmental change.The major results and conclusions of this study are as follows.(1)Developed a consistency calibration model of nighttime light data from DMSP and VIIRS.First,the VIIRS annual-averaged stable nighttime light data from2012 to 2018 were synthesized.Second,an aggregation method of VIIRS nighttime light data based on both space and brightness was proposed.Based on these two key steps,this study developed a consistency calibration model of nighttime light data from DMSP and VIIRS,and then constructed the temporally consistent nighttime light dataset from 1992 to 2018 in Southeast Asia.The model proposed in this study can effectively reduce the inconsistency and incomparability of the nighttime light signals from multiple sources for achieving the temporal extension of the stable annual-averaged DMSP nighttime light data.Using this model,the simulated DMSP nighttime light data show a good agreement with the original DMSP nighttime light data in the same year,in terms of their spatial pattern,profile curve,scatter density,and histogram distribution.These comparisons and evaluations confirm that the consistency calibration model developed in this study is reliable.(2)Proposed a new mapping framework of urban sequence based on the spatial variation of the nighttime light gradient(SVNG).First,the potential urban clusters in Southeast Asia were identified using the cluster-based segmentation algorithm based on nighttime lights.Second,an automatic urban extraction method based on the spatial variation of the nighttime light gradient at local scales was proposed,and then the original urban extents from 1992 to 2018 in Southeast Asia were delineated.Finally,a temporal consistency check was performed on the initial urban extents to update and obtain the final urban sequence.The urban extents obtained in this study perform well in the spatiotemporal consistency with those obtained in other multisource urban remote sensing products.This method can well capture urban extents in areas with different development levels and patterns without relying on auxiliary data,and thereby enable to open a new way to monitoring and understanding the long-term urban dynamics at the regional and global scales.(3)Quantitatively revealed the spatiotemporal dynamics of urbanization in Southeast Asia over the past 30 years.Three aspects,including the spatiotemporal dynamics of nighttime lights,the spatiotemporal dynamics of urban expansion,and the spatiotemporal dynamics of urban development,were performed in this study.In general,the urbanization process in Southeast Asia has developed from an initial stage to a medium-term accelerated stage.The overall urbanization levels of different countries are quite different,and their relative speeds are also different.The urban development in each country is extremely unbalanced and the phenomenon of polarization is prominent.Within different urban clusters,the spatial distribution,expansion scale,speed,and direction,of the urban extents,are a bit different.Urban areas in each city are continuously expanded and tend to be connected as larger aggregated areas.The overall urban development in Southeast Asia exhibits a spatial pattern of the north by northwest to south by southeast.For each Southeast Asian country,the spatial agglomeration characteristics of urban development are weakened,the primacy indices of urban development tend to decrease,and the multi-center development patterns are increasingly prominent.(4)Constructed a comprehensive evaluation model of multi-dimensional environmental change and analyzed the trends and characteristics of comprehensive environmental quality change in Southeast Asia under the background of urbanization.Based on the evaluation model and quantitative assessment method,the spatiotemporal dynamics of multi-scale comprehensive environmental change in Southeast Asia during 2000-2016 were analyzed from three aspects: air quality change,vegetation coverage change,and surface temperature change.In general,the situation of environmental degradation in Southeast Asia is severe.During the transition from urban cores to their peripheral areas,the degree of environmental degradation is gradually deepening.Environmental degradation in urban areas is more common and serious than that in non-urban areas.With the continuous urban expansion,the proportion of degraded areas in urban areas increases rapidly,and the environmental degradation process could slow down when urbanization reaches a certain stage.In summary,a consistency calibration model of nighttime light data from DMSP and VIIRS was proposed for generating the temporally consistent nighttime light data in Southeast Asia from 1992 to 2018;a new mapping framework of urban sequence based on the spatial variation of the nighttime light gradient was developed for obtaining temporally consistent urban sequence(1992-2018)in Southeast Asia;the spatiotemporal dynamics of urbanization in Southeast Asia over the past 30 years were quantitatively clarified and revealed;a comprehensive evaluation model of multi-dimensional environmental change was constructed for analyzing the spatiotemporal dynamics of multi-scale comprehensive environmental change in Southeast Asia,and revealing the potential impacts of urbanization on environmental degradation.This study is of great significance for fully understanding the main stages and characteristics of urbanization in Southeast Asian countries,guiding their sustainable regional development,and promoting the construction of Maritime Silk Road.Besides,this study can also provide references for remote sensing monitoring of urbanization and environmental change spanning large areas and long time series. |
PDF Full Text Request