近年来我国的城市化进程加速，在城市化建设过程中，地表覆盖类型由土壤或植被覆盖变成建筑或道路等城市用地，这种下垫面性质的改变导致城市与周围农村的温差加大，从而产生城市热岛效应（UHI）。杭州作为长三角城市群中发展快速的“新一线”城市，近年来成为经济发达、人口众多的中心城，从而导致了它的热岛效应日益严重。当前遥感卫星能够对地表进行大范围、长时间的重复观测，因此利用长时间序列的遥感卫星数据可以对杭州市的热岛效应强度及其驱动因子进行深入研究和分析，弥补杭州热岛效应研究的空缺，为缓解杭州热岛效应提供决策支持。另外对我国其他发达城市地区的热岛效应变化研究提供借鉴。 本文以杭州市为研究区，使用长时间序列的多源遥感数据产品对杭州的UHI进行相关研究。使用MODIS NDVI、DMSP/OLS等产品提取不透水面（ISA）、水体等热岛驱动因子，展开UHI与其驱动因子的相关分析；利用ISA和MODIS LST产品计算地表城市热岛强度（SUHII），分析SUHII的时空分布特征及变化趋势。在杭州地区开展利用长时间序列多源遥感数据提取多年地表热岛强度（SUHII）的研究并进行时空变化的分析，有助于了解杭州UHI的发展与演变机制，对城市的规划建设具有重要的指导意义。论文的主要研究内容和结论如下： 1、热岛效应驱动因子的提取和研究 使用多源遥感数据产品分别得到三种热岛效应驱动因子数据：（1）使用数据融合的方法，由2000-2013年MODIS NDVI和DMSP/OLS数据提取人类居住指数（HSI），用于表征不透水面（ISA）；（2）利用MODIS第4、6波段反射率数据（MOD09A1）提取修正的归一化差值水体指数（MNDWI），用于识别、表征水体；（3）使用MODIS植被指数产品（MOD13A2）中的NDVI表征植被。提取的驱动因子与验证数据进行对比发现精度较高，HSI与MODIS土地覆盖类型数据的平均匹配率为90.7%，用30m土地覆盖数据验证得到的RMSE为0.0762；MNDWI与30m土地覆盖数据的匹配率为76.3%，证明用HSI表征ISA以及用MNDWI表征水体在粗空间分辨率和大尺度范围内表现较好。热岛效应与其驱动因子的相关性研究表明，LST与不透水面有较强的正相关性，平均R2为0.62，HSI每上升0.1，LST上升0.4K；与植被覆盖有较强的负相关性，平均R2为0.61，NDVI每上升0.1，LST下降0.58K；与水体有一定的负相关性，平均R2为0.51，MNDWI每上升0.1，LST下降0.63K。LST与不透水面的相关性最强。另外大型水体对周边的气候和环境有较强的影响，会削弱热岛效应强度。 2、热岛效应定量分析 通过拟合2000-2013年LST和HSI之间的线性关系，得到相应的城市地表热岛强度（SUHII），基于SUHII、HSI和LST数据，对杭州的UHI进行定量化的分析。利用核密度分析（KDE）方法对HSI数据进行空间处理后，消除邻接效应带来的影响，增强HSI和LST的相关性，实验发现在KDE搜索半径为2500m时得到的结果最好。使用LST和KDE分析后的HSIKDE数据进行线性回归，以拟合方程的斜率作为SUHII的值，拟合方程的R2除冬季夜晚外均大于0.8，表明拟合结果精度较高。杭州市的SUHII提取结果表明，白天热岛效应显著：全年、夏季和冬季的SUHII均大于1.5K，夏季尤其剧烈，达到4K以上；夜晚热岛效应有所减弱，全年和冬季结果大部分小于1K，夏季在2K左右。针对杭州各个区域的研究结果表明：杭州市SUHII体现出北高南低，东高西低的特征。杭州东北部的SUHII高于杭州西南地区，并且杭州市区（HZ），余杭区（YH）和萧山区（XS）等最发达的地区UHI增强现象显著，UHI效应的增强主要是由于城市扩张，城市人口活动和建成区热量排放的扩大造成的。以上结果说明杭州具有强烈的UHI效应，并且UHI具有当地的区域特征，该地区未来需要采取相关的措施来降低SUHII。

In recent years, China's urbanization process has accelerated, the type of land cover is changing from soil or vegetation to urban land such as buidlings or roads during the process. The change in the nature of the underlying surface rose an increase in the temperature difference between the city and the suburbs, resulting in the urban heat island (UHI) phenomenon. The Yangtze River Delta urban agglomeration is the fastest growing region in China. Among them, Hangzhou is the “new first-line” city, and has developed rapidly in recent years and has become a city with developed economy and large population, which leading to the particularly prominent UHI effect. Remote sensing satellites can observe large-scale and long time-series observations of the earth's surface. Therefore, long-term remote sensing satellite data can be used to study and analyze the heat island effect intensity and driving factors in Hangzhou, which helps to study the spatio-temporal change of UHI in large area and fill the blank in Hangzhou's UHI research. In addition, it provides guidance for the study of the UHI change in other developed urban areas of China. This study takes Hangzhou as a research area and uses long-term multi-source remote sensing data products for analysis. The MODIS NDVI, DMS/OLS and other products were used to extract Humen settlement index(HSI) to represent the impervious surface area(ISA), water body and other UHI driving factors, and conduct the correlation analysis between UHI and these factors. The ISA(HSI) and MODIS LST were used to calculate the surface urban heat island intensity(SUHII), and to analyze its temporal and spatila distribution characteristics. The multi-source remote sensing data are used to analyze the spatio-temporal changes at the regional scale, which helps to understand the development and evolution mechanism of UHI, and is of great significance for guiding the construction of the city. The main research contents and conclusions of the thesis are as listed follows: 1.Research on driving factors of UHI The multi-source remote sensing products were used to obtain three UHI drivers:(1)the human settlement index (HSI) extracted using data fusion method within time-series MODIS-NDVI and DMSP-OLS nighttime light data; (2) the modified normalized water body index (MNDWI) extracted using MODIS reflectance data (MOD09A1); (3) the NDVI in the MODIS vegetation index product (MOD13A2) is used to characterize the vegetation. The extracted driving factors are compared with the verification data toevaluate their accuracy. The average matching rate of extracted HSI and MODIS land cover type data is 90.7%, and the RMSE verified by 30m land cover data is 0.0762; the matching ratio of MNDWI to 30m land cover data is 76.3%, which proves that HSI and MNDWI performs well in coarse spatial resolution and large scale. The correlation between the UHI effect and its driving factors shows that LST has a strong positive correlation with the ISA(HSI) and a strong negative correlation with vegetation coverage. LST has a certain negative correlation with the water body. LST has the strongest correlation with the ISA(HSI). In addition, large water bodies have a strong influence on the surrounding environment, which will weaken the SUHII. 2.Quantitative study of heat island effect The HSI data is spatially processed by the Kernel Density Analysis (KDE) method before analysis, the influence of the adjacency effect is eliminated, and the correlation between HSI and LST is enhanced. The experimental results show that the best results are obtained when the KDE search radius is set to 2500 m. The time-series urban surface heat island intensity (SUHII) is derived by fitting a linear relationship between LST and HSIKDE. The slope of the regression function was used as the SUHII. The R2 of the fitted functions was greater than 0.8 except for the winter night, indicating that the fitting result was accurate. Then the quantitative analysis of the UHI effect in Hangzhou by time-series SUHII, HSI and LST. The SUHII extraction results in Hangzhou show that the UHI effect is significant during the day, and the SUHII is greater than 1.5K in the annual, summer and winter, especially in the summer, reaching more than 4K; the UHI effect is weakened at night, around 1K in the annual and winter, about 2K in summer. According to the results of various districts in Hangzhou, the SUHII in the northeastern part of Hangzhou is higher than that in the southwestern part of Hangzhou, and the UHI enhancement phenomenon in the most developed areas such as Hangzhou urban area (HZ), Yuhang (YH) and Xiaoshan (XS) is significant. The UHIeffect enhancement was mainly due to urban expansion, urban anthropogenic activities and the heat emissions of the built-up area. The above results show that Hangzhou has a strong UHI effect, and UHI has local regional characteristics, and the relevant measures for reducing the SUHII needs to be taken in the future.