| The Quaternary is the latest geological period.The emergence and development of mankind is one of the most important events in the Quaternary.The surface landscape deeply records the brand of human activities.Urbanization is proceeding at an unprecedented speed.Urbanization will lead to urban heat island effect,which will increase natural mortality,affect human health and comfort,increase energy consumption.The study of urban heat island and its associated determinants is of great significance for human health,social sustainable development and decision-making of major environmental problems.In the context of rapid urbanization,it is particularly necessary to study the temporal trend of urban heat island.The early research used air temperature data from meteorological stations to study the urban heat island effect,but the meteorological stations are sparsely and unevenly distributed,which may cause some uncertainties.Satellite-based land surface temperature data has continuous spatial coverage and is easy to use.Therefore,it has been widely used to study urban heat island effect in recent decades.However,there are several problems in the research of urban heat island based on remote sensing.In terms of data,there are many land surface temperature products,and the difference of different land surface temperature data in the study of urban heat island is unknown.In addition,the accuracy of air temperature data estimated by remote sensing data is low,and the temporal accuracy of estimated air temperature data has not been verified.When using the estimated air temperature to study the temporal trend of urban heat island,it is necessary to verify the accuracy of interannual variation and long-term trend of the estimated air temperature.In terms of application,the spatial,diurnal and seasonal variations of urban heat island effect have been systematically studied,but there is a lack of research on the long-term trend of urban heat island effect.In addition,the relationships between urban heat island effect and interannual climate variability and warming are still controversial.Combined with satellite,in situ observation and other auxiliary data,using machine learning model,spatial analysis,regression analysis and other methods,this study compares the land surface temperature value,missing value and abnormal value of different land surface temperature products.Subsequently,this study developed a temporally accurate gridded air temperature dataset.Furthermore,this study investigated the temporal trend of urban heat island effect and its associated determinants in global and Chinese major cities.Finally,this study revealed the relationships between urban heat island and interannual climate variability or global warming.The content of this thesis mainly includes the following three parts:(1)In this thesis,the land surface temperature values,missing values and abnormal values in the MYD21 and MYD11 land surface temperature products were compared in detail.The results show that the MYD21 product in the mainland of China was 1 and 0.65℃higher than that of the MYD11 product during the daytime and nighttime,respectively.This is because the two products use different algorithms to retrieve the land surface temperature.There are more missing and abnormal values in MYD21 data than MYD11 data,which may affect the use of data.This thesis studied the influence of different methods to input data on the accuracy of estimated air temperature.It was found that inputting data for the same month into the model can generate more accurate results than inputting all data into the model.Using temporal variables(i.e.,month and year)can significantly increase the accuracy of estimated air temperature.This is because the relationship between air temperature and land surface temperature is different in different years and seasons.The monthly mean air temperature data with 1 km resolution in the mainland of China during2001–2018 were mapped using in situ air temperature,satellite-based land surface temperature and other auxiliary data.The MAEs of the estimated air temperature data ranged from 0.474 to 0.607℃.In addition,the temporal accuracy of the estimated air temperature was verified.The coefficient of determination(R~2)of interannual variation and long-term trend of estimated mean air temperature were 0.731 and 0.848,respectively.(2)This thesis studied the temporal trends of urban heat island and their associated determinants in global and Chinese major cities.The annual mean daytime and nighttime surface urban heat island intensity increases significantly in 42.1%and30.5%of 397 cities around the world,respectively,for the period of 2001–2017.An interesting phenomenon is that the increase in rural enhanced vegetation index(EVI)contributes 22.5%to the increase in daytime surface urban heat island intensity.The urban heat island intensity increased significantly in more than 50%cities in the mainland of China.The increase of the urban impervious surface and the decrease of vegetation coverage due to human activities were the main reasons for the increase of the urban heat island intensity.Meteorological factors affect the surface urban heat island intensity in the urban center by affecting soil moisture,vegetation growth and albedo.The extremely high value in MYD21 land surface temperature product will affect the calculation of surface urban heat island intensity,while MYD11 land surface temperature product will underestimate the temporal trend of surface urban heat island intensity due to the use of fixed emissivity data.(3)This thesis studied the relationships between surface urban heat island and interannual climate variability.During the daytime,the surface urban heat island intensity in northern China decreases in extremely hot summer and increases in extremely cold winter.The interannual standard deviation of land surface temperature,EVI and albedo in urban centers is obviously lower than that of rural areas in northern China.This thesis proposed a series of methods to estimate the contribution of urban heat island to warming by using gridded air temperature data.These methods can avoid the uncertainty of estimating the contribution of urban heat island to warming by using meteorological stations.The contribution of urban heat island to warming was small at the national and regional scales in the mainland of China,because the urban area accounts for a small proportion of the total area.Finally,the contribution of urban heat island to warming was more than 50%at the local scale.In this thesis,the comparison of land surface temperature products and estimation of air temperature data can support the urban heat island and climate change research.The research on the temporal trend of urban heat island and its associated determinants can enhance the understanding of the urban heat island effect.It lays a foundation for predicting the future urban heat island,evaluating the negative effects brought by the change of urban heat island and decision making for major environmental problems. |
