城市热岛效应即城区温度比四周郊区高的现象，是城市化发展过程中所产生的最明显的局地气候变化。在全球变暖的气候背景下，夏季热浪事件频发，城市热岛与热浪的协同效应使城市区域面临日益严峻的高温压力。此外，由于城市化进程的不断加快，城市热岛效应持续存在，其对城市区域长期增温速率的影响也日益明显。随着城市持续扩张、城市人口不断增加，城市热岛效应及其对热浪和增温速率的影响越来越受到各国政府和有关部门的重视，其对城市居民健康、城市生态系统、城市环境和气候的影响已经不容忽视，尤其在城市化发展迅速、人口高度密集的城市和地区。

虽然针对城市热岛效应、热浪和增温速率的相关研究已经取得了巨大的进展，但由于研究数据、研究区域和评估方法不同，相关结论存在一定的差异和争议，有待进一步分析和完善。近年来，地面气象观测系统、卫星遥感观测系统和地表覆盖基础产品不断发展和完善，为城市热岛效应、热浪和增温速率的进一步研究提供了充足和可靠的数据支撑，相关研究将为政府和有关部门的决策提供更可靠的科学参考，以更好地改善城市热环境、提升城市功能。

鉴于此，首先，以北京这个人口密集的超大城市为例，本文第2章基于城市密集气象网络观测的气温数据和卫星遥感观测的地温数据，着重研究气温城市热岛和地温城市热岛的日变化、季节变化和站间变化，并从中梳理出二者的共性和差异，得出二者的主控因子。其次，以北京、上海和广州这三个不同气候背景下的超大城市为例，本文第3章基于气温、太阳辐射等观测数据，进一步定量分析了不同气候背景下，热浪和城市热岛的协同效应表现及其可能的影响因素。最后，以城市化发展迅速的中国地区为例，本文第4章基于最新的地表覆盖年变化数据和气温数据，分析了中国气象站点周围的城市化变化，对比了日最高气温（T_max）和日最低气温（T_min）在城市和郊区的增温速率差异，并定量评估了城市热岛效应对中国地区增温速率的影响。

主要研究结果归纳如下：

（1）以北京为例，基于2013-2015年间的气温数据和地温数据，分析了气温城市热岛和地温城市热岛的日变化、季节变化和站间差异。结果表明，典型城市站的夜间气温城市热岛在冬季比夏季要晚2-3小时才开始降低，这主要是因为冬季晚上的人为热排放更强（供暖）。其次，气温城市热岛在晴天晚上随时间增加而减弱，在阴天晚上随时间增加而增强。典型城市站的气温城市热岛在晴天的22:30比01:30（晚上）高0.70 °C，而在阴天的22:30却比01:30低1.05 °C（郊区为农田时）。这主要是因为在阴天晚上，云层吸收地表长波辐射，并发射长波辐射，导致大气逆辐射作用增强，补偿了地表发射长波辐射而散失的热量。在季节变化特征上，当郊区类型由农田改为山区森林后，夜间城市热岛的季节差异明显减小，典型城市站的夜间气温城市热岛在夏季和冬季的差异由-1.48 °C减小到了-0.49 °C，夜间地温城市热岛（01:30时刻）的季节差异从-2.94 °C减小到了-0.11°C。这主要是因为山区森林结构的季节变化不大，而农田在夏季有作物覆盖、在冬季几乎都是裸地。这一差异使得山区森林在冬季的长波辐射冷却效率低于农田。此外，归因分析表明，站点周围地表的不透水面覆盖率差异可以解释夜间气温城市热岛42%-55%的站间差异，可以解释夏季白天地温城市热岛40%-43%的站间差异。

（2）以北京、上海和广州为例，基于2013-2015年气温、风速、风向、相对湿度和日总太阳辐射的观测数据，对比分析了不同气候背景下热浪和城市热岛协同效应的表现差异。结果表明，在热浪天，北京和广州城市热岛的增强主要发生在晚上，而上海城市热岛的增强主要发生在白天，并且这些变化与热浪天太阳辐射的增强显著相关。此外，在上海的热浪天，风向的变化对白天城市热岛的增强有重要的贡献。和普通天相比，在热浪天，北京和广州的夜间城市热岛分别增强了0.93 °C和0.79 °C，上海的白天城市热岛增强了0.85 °C。在热浪期间，北京和广州的夜间城市热岛和日总太阳辐射呈显著正相关（北京：相关系数r=0.76，p<0.01；广州：r=0.30，p<0.01），上海的白天城市热岛和日总太阳辐射呈显著正相关（r=0.62，p<0.01）。同时，在上海的热浪天白天，超过63%的风来自西南方向同样处于高温压力的内陆；而在普通天白天，超过70%的风来自东南方向温度相对较低的东海。此外，亚热带湿润季风气候背景使得上海在热浪天白天处于高温高湿的条件下，上海城市区域在白天的体感温度最高值（49.15 °C）远高于北京（36.05 °C）和广州（33.68 °C）。

（3）将研究区域扩展至全中国，基于最新的30m空间分辨率的土地覆盖年变化数据以及2454个站点的T_max、T_min和T_mean（日平均气温）数据，定量评估了1985-2017年间城市热岛效应对观测增温速率的影响。结果表明，在城市站，T_min的增温速率明显高于T_max。这一差异在城市站大约是全国所有站平均值的三倍，但在郊区站的差异却非常小。分析得出，这主要是由于1985年来我国气象站点周围环境的普遍城市化，影响了观测结果的空间代表性，导致了T_max和T_min的增温速率差异。从1985年到2017年，周围地表不透水面覆盖率低于20%的站点数量从1912个减少到了780个。城市化导致的城市热岛效应对T_min的增温速率起到了重要的贡献。在1985-2017年，T_min和T_max的年平均增温速率在城市站分别为0.575 °C/10yr（即摄氏度/10年）和0.378 °C/10yr。其次，城市热岛和城郊不透水面覆盖率差呈显著正相关，基于这一相关性和全国所有站的平均气温和平均不透水面覆盖率，评估了城市热岛对全国平均增温速率的贡献。结果表明，城市热岛效应对T_max、T_min和T_mean全国平均增温速率的贡献分别为7.2%（0.026 °C/10yr）、30.2%（0.127 °C/10yr）和24.4%（0.085 °C/10yr）。

The urban heat island (UHI) effect refers to the phenomenon that urban areas being warmer than surrounding suburbs. It is the most obvious local climate change generated during the process of urbanization. Under the background of global warming, the occurrence of summer heat waves is more frequent, and the synergistic effect of UHIs and heat waves makes urban areas face increasingly severe heat pressure. In addition, due to the accelerating process of urbanization, the UHI effect has continuously existed, and its influence on the long-term warming rate over urban areas is becoming more and more obvious. With the continuous urbanization expansion, and the continuously increasing urban population, the UHI effect and its impact on heat waves and warming rates are getting more attention from the government and relevant departments, their impacts on the health of urban residents, the urban ecological system, the urban environment and local climate should not be ignored, especially in cities and regions with a rapid urbanization process and a highly dense population.

Considerable progress has been made in the research on UHIs, heat waves and warming rates. However, due to different research data, study areas and assessment methods, there are some differences and controversies in relevant conclusions, and further studies need to be performed and improved. In recent years, the ground meteorological observation system, the satellite remote sensing observation system and land cover products are continuously developing and improving, which provide researches on the UHI effect, heat waves and warming rates with a sufficient and reliable data support. The relevant research conclusions will greatly help the decision-making of the government and relevant departments with a more reliable scientific basis, which will help better improve the urban thermal environment and enhance urban functions.

Therefore, first of all, taking the densely populated megacity of Beijing as an example, Chapter 2 of this paper studied diurnal variabilities, seasonal variabilities and interstation variabilities of atmospheric UHIs and surface UHIs using different observation data, i.e., air temperatures observed by the urban dense meteorological network and land surface temperatures observed by the satellite remote sensing system, respectively. The generalities and differences between the two types of UHIs were sorted out, and the main control factors of the two were then obtained. Secondly, taking the three cities under different climatic backgrounds, i.e., megacities of Beijing, Shanghai and Guangzhou, as examples, Chapter 3 of this paper quantitatively analyzed the synergistic performance of heat waves and UHIs under different climatic backgrounds and their possible influencing factors based on air temperatures, solar radiation and other observation data. Finally, taking the country with a rapid urbanization process, i.e., China, as an example, Chapter 4 of this paper analyzed the urbanization changes in the environment around meteorological stations based on the latest land cover products and air temperature observations. The difference between warming rates of the daily maximum temperature (T_max) and the daily minimum temperature (T_min) at both urban stations and rural stations were compared. The quantitative influence of the UHI effect on warming rates in China was also evaluated.

We drew the following conclusions:

(1) Taking Beijing as an example, we analyzed the diurnal, seasonal and interstation variabilities of atmospheric UHIs and surface UHIs based on the air temperature (T_a) data and the land surface temperature (T_s), respectively, which were collected from 2013 to 2015. The results show that nighttime atmospheric UHIs at typical urban stations in winter began to decrease 2~3 hours later than those in summer. This phenomenon is mainly related to the increased anthropogenic heat release (the heating) in winter. Secondly, atmospheric UHIs decreased with time under clear nights but increased with time under cloudy nights. On clear nights, atmospheric UHIs at typical urban stations were 0.70 °C stronger at 22:30 than those at 01:30 (at night) when croplands were used as the rural reference. Whereas on cloudy nights, for typical urban stations, atmospheric UHIs at 22:30 were 1.05 °C weaker than that at 01:30. This is mainly because on cloudy nights, clouds absorb long-wave radiation from the surface and also emit long-wave radiation, resulting in enhanced downward long-wave radiation, which compensates for the heat lost by long-wave radiation emitted from the ground. For characteristics in seasonal variations of UHIs, the seasonal variation of nighttime UHIs decreased significantly when the rural reference was changed from croplands to mountainous forests. For nighttime atmospheric UHIs, their seasonal differences at typical urban stations decreased from -1.48 °C to -0.49 °C. For nighttime surface UHIs, their seasonal differences at typical urban stations at 01:30 decreased from -2.94 °C to -0.11 °C. The main reason for this phenomenon is that there is little seasonal change in the structure of mountain forests, whereas croplands are covered with crops in summer but are almost bare in winter. This difference makes the long-wave radiation cooling efficiency of mountain forests being lower than that of croplands in winter. In addition, the analysis shows that the difference in the percentage of impervious surfaces around stations could explain 42%-55% of the inter-station variation in nighttime atmospheric UHIs and 40%-43% of the inter-station variation in summer daytime surface UHIs.

(2) Taking Beijing, Shanghai and Guangzhou as examples, this paper studied the synergistic effect of heat waves and UHIs under different climate backgrounds based on air temperatures, wind speed, wind direction, relative humidity and daily total solar radiation observations collected from 2013 to 2015. The results showed that the intensified UHIs under heat waves in Beijing and Guangzhou mainly occurred during nighttime, whereas for Shanghai, the intensified UHI under heat waves mainly occurred during daytime. These changes were significantly related to increased solar radiation under heat waves. In addition, the wind direction change also contributed to the intensified daytime UHI in Shanghai under heat waves. Compared to nighttime UHIs under normal conditions, nighttime UHIs under heat waves in Beijing and Guangzhou were amplified by 0.93 °C and 0.79 °C, respectively, the daytime UHI in Shanghai under heat waves was amplified by 0.85 °C. During the heat wave period, the daily total solar radiation was significantly related to nighttime UHIs in Beijing and Guangzhou (Beijing: r=0.76, p<0.01; Guangzhou: r=0.3, p<0.01), the daily total solar radiation was significantly correlated with daytime atmospheric UHIs in Shanghai under heat waves (r=0.62, p<0.01). In the meantime, for Shanghai, more than 63% of the wind came from neighboring inland hot cities in the southwest during daytime under heat waves, whereas under normal conditions, more than 70% of the wind originated from the East China Sea with a lower temperature in the southeast during daytime. Additionally, the background of the Marine monsoon climate makes a high relative humidity and high air temperatures occurred at the same time during the day in Shanghai under heat waves, which led to a higher daily maximum apparent temperature in urban areas of Shanghai (49.15 °C) than that in urban areas of Beijing (36.05 °C) and Guangzhou (33.68 °C).

(3) Expanding the study area to the whole country, i.e., China, we quantitatively analyzed the impact of the UHI effect on observed warming rates during the period of 1985-2017 based on the latest land cover data with a 30-meter spatial resolution and an annual temporal resolution, and the T_max, T_min and T_mean (the daily mean air temperature) observations collected at 2454 stations. The results show that the warming rate of T_min was much higher than that of T_max at urban stations, and this difference was approximately three times larger over urban stations compared to the average difference over all of the stations in China. However, this difference disappeared at rural stations. According to the analysis, this result is mainly due to the widespread urbanization of the environment around meteorological stations in China since 1985, which affected the spatial representation of observation results and led to the difference in warming rates between T_max and T_min. The number of stations with a percentage of impervious surfaces under 20% around them decreased from 1912 to 780 during the period 1985-2017. The UHI effect caused by urbanization plays an important role in the warming rate of T_min. During the period of 1985-2017, the annual warming rates of T_min and T_max over urban stations were 0.575 °C/10yr (i.e., °C/10 years) and 0.378 °C/10yr, respectively. Secondly, the UHI and the urban-rural difference in the percentage of impervious surfaces were significantly correlated. Based on this correlation relationship, along with the average percentage of impervious surfaces and the average warming rate at all of the stations, the urbanization effect on national averaged warming rates was also evaluated. The result shows that UHI contributions to the national mean warming trend in T_max, T_min and T_mean were 7.2% (0.026 °C/10yr), 30.2% (0.127 °C/10yr) and 24.4% (0.085 °C/10yr), respectively.