Time Series-based Information Extraction Of Impervious Surface And Analysis Of Urban Expansion In Beijing

As urbanization is accelerated,there are growing impervious surface with less surface of other types,which poses severe challenges to the urban ecological environment and causes security problems such as waterlogging,urban heat island effect and groundwater shortage in the urban area.To support urban development while protecting the ecological environment has become a priority in sustainable development in China and even in the world.Remote sensing technology can help to solve this problem by making it possible to research on the transformation of long time series of impervious surface and urban expansion.Taking Beijing as an example,this paper employs Landsat time series satellite images to extract information on impervious surface and conduct a spatiotemporal measurement analysis of urban expansion based on time series.It aims to provide a new solution to extracting the time series information on impervious surface and analyzing urban expansion so as to realize the extraction and dynamic monitoring of highprecision distribution feature of impervious surface for a long-time series.The findings are as follows:(1)To deal with the missing values and outliers in some Landsat data,the CFMask algorithm is employed to extract the cloud image information and compensate the missing values with the median fusion method.the paper also presents an improved linear regression pixel smoothing method to eliminate noise and enhance the temporal correlation of images.Moreover,a cloudless and highly spatiotemporal Beijing dataset(BJLS)is established.(2)A new method of extracting impervious surface(TSIS)is presented in this paper.An impervious surface augmentation index(ISAI)is proposed to improve the impervious surface information,but it includes water and shadow information.In order to solve this problem,the paper suggests a modified urban water index(MUWI)with improved threshold to collect water information in the research area.It is combined with the urban shadow index(USI)and the digital elevation model to collect information about shadow created by buildings and hills respectively in the research area,and the mask processing is also carried out respectively.After the test on time consistency,distribution characteristics of impervious surface are obtained in the research area from 2000 to 2020.The average accuracy of the impervious surface obtained in this way reaches90.33%,and the Kappa coefficient is between 0.6215 and 0.8934,with an average value of 0.76.Therefore,it can be considered that this method can work effectively on the extraction of impervious surface time series information.After analyzing the distribution characteristics of impervious surface in Beijing,it can be found that the impervious surface was mainly concentrated in the main urban area from 2001 to 2004,and expanded outwards from 2005 to 2008,mainly in the northern part of Haidian District and the outer suburbs such as Miyun District,Changping District and Daxing District.(3)A spatiotemporal measurement analysis of urban expansion in Beijing is made in this paper,including quantitative and spatial characteristics of urban expansion in Beijing.The research shows that Beijing has experienced two urban sprawl peaks from 2000 to 2005 and 2009 to 2011.During the urban development period,the urban construction is advanced in a reasonable way with clearly defined boundaries and compact urban land use.