Characteristic And Impact Factors Of Spatial-temporal Evolution Of Land Surface Temperature In Changsha Based On Landsat Remote Sensing Data

Rapid urbanization has significantly changed the urban thermal environment,produced many ecological and environmental problems,and seriously hindered the sustainable development of the city.An in-depth study of the temporal and spatial evolution pattern of the urban thermal environment in the process of urbanization and revealing the driving mechanism of the dynamic coupling of impervious surface and vegetation cover on the evolution of the urban thermal environment can provide a scientific theoretical basis for improving the urban thermal environment and promoting urban sustainable development.Taking Changsha City as the study area,based on the Landsat land surface temperature(LST)data from 1986 to 2020,this study first compared different methods(i.e.,urban heat island method,min-max normalization method,centralization method,and pseudo-invariant normalization method)to normalize multi-temporal LST to make them temporally comparable,then revealed the Spatio-temporal evolution of LST during urbanization,and finally investigated the dynamic coupling of impervious surface area and vegetation cover and its effects on the thermal environment during urban expansion.The main findings are below.(1)Concerning LST normalization.The pseudo-invariant normalization method assumes the remote sensing index will keep stable when the surface characteristics remain unchanged.It can effectively reflect the regional warming caused by urban expansion and the spatial heterogeneity of the temporal evolution of LST caused by urban expansion.The pseudo-invariant normalization method,compared with the other three methods can more accurately quantify the response of LST to urban expansion and is a scientific and effective normalization method for Landsat LST.(2)Concerning the Spatio-temporal evolution of LST during urban expansion.LST in the study area increased by 1.5 K.The increase rate is slower in the period before 2000(0.07 K per year)than that after 2000(0.073 K per year).There was high spatial heterogeneity in the change of LST with 55% of the study area experiencing LST increase and 15.6% of the study area witnessing LST decrease in the city center.The dynamic evolution process of LST can be summarized into eight temporal patterns,namely,continuous decreasing after first increasing,a slight increase in the early stage,slightly decreasing in the late stage,slow and consistent increasing,increasing in the early stage,and later stage,continuous increasing after stable warming in the early stage(logistic growth pattern),decreasing and increasing,increasing and decreasing.The different pattern of LST dynamic evolution is closely related to the urbanization stage.Generally,the temperature rises in the urbanization stage,but the temperature is stable in the non-urbanization stage.(3)The study area experienced a rapid expansion of impervious surface area and decreasing in vegetation cover during the study period.The impervious surface area expansion heated the city,while urban greening cooled the city.The impacts of impervious surface increase are stronger than the impacts of urban greening.There is a negative feedback effect between LST and vegetation cover.The negative feedback effect is the strongest in the urban center and is the weakest in the rural areas.The dynamic coupling of impervious surface and vegetation cover is more apparent in the urbanizing region than that in the urban center,for example,in regions with a logistic growth pattern of LST change(i.e.,pattern 8)and a first stable and continue increase pattern of LST change(i.e.,pattern 5).