Spatiotemporal Changes Of Thermal Environment Pattern In Changsha In The Recent Past Decade

Urban thermal environment has been a hot topic in the field of the eco-environmental effects of city.Researches to the evolution law of the urban thermal environment can make contributions to release the negative effects of urban heat island(UHI)and promote the sustainable development of cities.Changsha as the"rise of central China plan" strategy supporting point,the urbanization process is moving forward rapidly,the contradiction between urban development and the ecological environment is further intensified.The study of spatiotemporal Changes of thermal environment in Changsha City,and the relationship between the spatial pattern and composition of urban landscape and spatial variation of land surface temperature(LST)on landscape scale,not only ease the UHI effect in Changsha City and improve the city's ecological environment,but also has the reference value to the city in the midwest regional center.The paper analyses the spatiotemporal changes of LST in the central city of Changsha.Twenty-six Thematic Mapper(TM),Enhanced Thematic Mapper Plus(ETM+),Thermal Infrared Sensor(TIRS)thermal images from 2000 to 2013 are used to retrieve the LST of urban center in Changsha.Then it is classified as low temperature region,sub-low temperature region,middle temperature region,sub-high temperature region and high temperature region.The landscape types of urban center in Changsha in 2004,2009 and 2013 are classified based on the land use data.Numerous methods have been applied in the paper,landscape centroid migration,spatial autocorrelation,the moving window and gradient analysis,and spatial regression model.Detailed study thermal landscape centroid evolution,spatial agglomeration pattern and gradient variation characteristics from two aspects of the seasonal and interannual variations in Changsha.Quantify the relationship between the spatial pattern and composition of urban landscape and spatial variation of LST.The main research contents and conclusions are drawn as follows.(1)Spatiotemporal variation analysis of thermal environment.From two aspects of the seasonal and interannual variations,reveal characteristics of spatial changes of LST distribution and thermal landscape,and thermal landscape centroid evolution in Changsha.With the rapid urbanization of Changsha,the districts of UHI expand and become more decentralized from 2000 to 2013.Area of UHI increases by 96.80 km2.Moreover,the newly increased UHI districts are mainly distributed in new industrial parks such as Yuelu,Tianxin,Yuhua and Xinsha,which relates to the government development layout.Centroid metastasis of heat island landscape(high temperature region and sub-high temperature region)is consistent with urban land.They move to the Northeast.However,centroids of low temperature region and sub-low temperature region are in the opposite transfer direction as high temperature region.Centroids of high temperature region and sub-high temperature region have same transfer directions in different seasons in 2013.Move to the Northwest,then move to the Southeast with time.Centroid of low temperature region moves from the West to East with season.(2)Spatial pattern of thermal landscape.The global Moran's I and LISA index are used to analyse general characteristics and local characteristics of spatial agglomeration pattern of thermal environment in Changsha.Analyse the trend of changes of landscape index in different years and seasons in landscape level.On this basis,by combining with the moving window and gradient analysis,the paper explores the spatial changes of urban thermal landscape pattern in Changsha downtown from 2004 to 2013 and analyse thermal landscape evolution from 8 directions.The spatial distribution of LST presents agglomeration in Changsha from 2000 to 2013.Area of high-high temperature region increases with time,while low-low temperature region migrates to Northward with distribution of discretization.There is an obviously spatial variation of thermal landscape pattern on landscape scale in the center of Changsha.From the center to the Northwest,East and South,the thermal landscapes tend to be more fragmented and diversified,and the landscape shapes become more complexity from the downtown to the suburb.Regular thermal landscapes general cover with uniform landscape types.Complicated landscape types with difference thermodynamic properties will lead to thermal landscapes fragmentation,complexity.The changes of thermal landscape pattern indexes can reflect the change of thermal landscapes.(3)Thermal environment effect of landscape pattern.In order to study the relationships between spatial pattern of landscape types and LST,zonal statistic is applied to investigate the changes of various urban landscape patterns under different thermal landscape ranks.Spatial regression models are used to quantify the relationship between landscape compositions and LST at different resolutions.The higher and more concentrated farmland and forest land plaques are in the thermal landscape,the more significant surface cooling effect there is.On the contrary,the higher LST is when the urban land dominance is greater,the degree of cohesion is higher and landscape shape is simpler.And what's more,the heat island effect will become more remarkable in the study area.Results of this study indicate that the resolution of 900m is the most suitable for measuring the relationship between landscape compositions and LST.All coefficients are significant at the 99%confidence level.The value of R2 is 0.8951.Coefficients of farmland,grassland,forest land,urban green land and water are-0.0292,-0.0158,-0.0244,-0.0209,-0.0488 respectively.Which implies that increasing 1%cover of farmland when decreasing 1%cover of urban land will lead to a decrease in LST by 0.0292 ?.Increasing with same proportion,water,farmland and forest land have stronger cooling effect on LST,which play a key role in releasing UHI effect.