近年来城市快速发展所导致的局地气候变化，成为人们日益关注的问题。城市化进程会对天气和气候造成一定的影响，它改变城市的气温、湿度、下垫面能量水汽输送等过程，造成居住舒适性下降、空气污染加剧，并且增加灾害性天气气候事件发生的几率。中国作为全球未来几十年城市化增幅潜力最大的国家，研究其城市化进程对区域气候的影响具有重要意义。在全球变暖的背景下，中国北方和南方地区的城市群规模扩张以及单个城市的扩张是否会导致明显的区域气候效应？未来不同温室气体排放情景下如何精细化描述城市群的局地气候效应？

基于此，本文选取中国新型城市（雄安新区）、北方（京津冀）城市群和南方（粤港澳）城市群作为研究对象，探究全球变暖背景下城市化的空间尺度扩张与区域气候变化之间的联系，分析了大规模城市化与全球变暖背景对区域气候变化影响的共同作用。本研究通过使用中尺度区域气象模式WRF（Weather Research and Forecasting Model），对雄安新区、京津冀城市群和粤港澳大湾区城市群的城市化过程进行数值模拟。得到主要结论如下：

（1）本研究选取雄安新区作为中国新型城市，基于WRF模式评估了不同规模的城市化进程对2016年夏季该地区气象要素的影响。雄安新区在不同城区面积下的增温幅度有所不同：处于建设初期的雄安新区平均气温增加0.25℃；而建设中期（200 km²的城区面积）其平均气温增幅达0.87℃。不同规模的雄安新区对整个京津冀地区的增温影响较小，雄安新区在建设启动期在城区气温增加，但达到中期和远期规模时会导致城市区域及其西南方向部分区域增温，出现城市规模快速扩大导致区域气温大幅度增加的现象。由于在WRF模式中修改了城市区域范围，其相对应的地表反照率从0.18降低为0.15，叶面积指数从默认的3 m²/m²降为1 m²/m²。地表反照率和叶面积指数下降使该区域净辐射能量增加，较低的下垫面植被覆盖度导致地表向大气输送的潜热减少约60W/m²，同时感热通量增加20W/m²。地表相对湿度降低约10%，感热增加引发了微弱的局地次级环流，导致风速微弱增加。进一步分析月平均气温的空间分布得出，雄安新区在建设初期和中期在其行政区内和西南部有较明显的增温，且主要出现在 7、8 月份，在城市规模明显扩大后（200 km²）下增温尤为显著，导致夏季周边地区增温高达1℃。

（2）本文利用英国东英吉利（East Anglia）大学气候研究中心（Climatic Research Unit，CRU）的气温和降水格点资料与WRF模式进行历史情景的对比验证，然后以Community Earth System Model (CESM)?在RCP4.5浓度路径下输出结果作为驱动数据，利用WRF模式对RCP4.5排放情景下未来京津冀地区2020、2030年的气象场进行模拟。并设计京津冀地区城市群扩张约9600 km²的实验，以此探究京津冀地区城市群扩张后对区域气候的影响。结果表明：WRF模式历史实验较好地再现了京津冀地区的平均气温由南到北递减的空间分布特征，并且在地形复杂区域（如燕山山脉、太行山脉）能模拟出气温的极大、极小值中心，与CRU数据的空间相关系数高于0.9。然而，WRF模式对历史降水的模拟效果不佳，表现为京津冀区域内降水模拟偏高。RCP4.5排放情景下，京津冀地区未来年平均气温呈现逐年上升的趋势，在2030年，城市保持原有规模的实验（S1）比历史实验（HIST）高0.5℃左右。城市大面积扩张的情景（S2）下的气温较HIST实验在城市区域高0.5-1℃，S2实验比S1实验在城市区域高0.5℃左右。城市用地大规模扩张，将会导致夏季平均气温明显升高，在2020年和2030年S2实验较S1实验夏季平均气温分别升高1.01℃和0.87℃，但是冬季相反，S2情景下2020和2030年冬季气温分别下降了1.81℃和 1.65℃。

（3）对粤港澳大湾区历史时期（1980、1990和2000）和未来不同 RCP 典型浓度路径排放情景下和粤港澳大湾区未来城市扩张约9000 km²后的未来天气气候状况进行预估（2030、2040和2050），设计了5组实验，分别代表历史实验（HIST）、RCP4.5情景（S1）、RCP8.5情景（S2）、RCP4.5情景下城市扩张（S1_EXP）和RCP8.5情景下城市扩张（S2_EXP）。结果表明：WRF模式能够准确模拟出粤港澳大湾区年平均气温的空间分布特征，空间相关系数在0.80以上，但对粤北丘陵和山地地区气温的模拟偏低，在珠江三角洲地区模拟的气温偏高。HIST实验对区域内极端高温（T90p）和极端低温（T10p）的模拟结果与CRU数据的一致性较好，空间相关系数在0.68至0.90之间。在S1情景下，粤港澳大湾区城市区域在2030、2040和2050年的年平均气温较比HIST实验分别上升0.33、0.66和1.12℃左右，S1_EXP情景下城市区域在2030、2040、2050的年平均气温分别较HIST实验上升了0.51℃、0.83℃和1.26℃，但从华南地区看，温室气体排放带来的升温效应比城市规模增加带来的增温效应更为显著。模拟结果表明，2040年与2050年在佛山与广州北部交界处出现极端高温中心，达32℃以上；极端低温在粤港澳地区也会有增加趋势，但上升幅度不大，在2050年达到0.78℃。粤港澳大湾区的城市区域扩张后，城市下垫面的年平均感热通量基本保持在70-90W/m²左右，而潜热通量在10-20W/m²左右，表明地表向大气输送的热量主要以感热传输为主。扩张的城市区域蒸发量减少，水汽含量降低，由之前的潜热传输为主转变为感热传输为主。S2_EXP情景下的模拟结果与该现象一致，但2050年城市区域的年平均气温增幅比HIST实验更加显著，在2030、2040和2050年较HIST实验分别升高0.53、1和1.58℃左右。

本研究基于 WRF模式，结合 MODIS 土地利用/植被覆盖数据，分析了中国北方京津冀城市群和华南粤港澳城市群城市下垫面扩张情景下区域气温、降水和相关通量的变化特征，并初步探讨了气温变化的机理。本研究有助于区分未来气候变暖和城市规模扩大导致的城市增温效应，为预估城市扩张带来的极端气候事件风险，以及未来的城市规划、环境污染治理和公共健康保障等提供有效的决策参考。

In recent years, the rapid urbanization-induced regional climate change has become a pressing concern and attract considerable attention. Urbanization impacts the weather and climate through changing temperature, humidity, surface energy budget, and water vapor transmission, thus leading to declined living comfort and degraded air pollution, even enhancing the possibility of extreme weather and climate events. China is considered to have the highest potential for urbanization growth in the world in the next few decades. Thus, the assessment of urbanization's impact on regional climate under various scenarios has great significance. Under the context of global warming, will the expansion city cluster in the north and south China or single city expansion cause different regional climate change? How to accurately assess the regional climate change on city clusters under various greenhouse gas emission scenarios in future years? 

In this study, three types of city clusters are selected as study objects, including Xiongan New District as a typical new city, Beijing-Tianjin-Hebei city cluster as northern city cluster, Guangdong-Hong Kong-Macao Bay Area city cluster as southern city cluster. Under the context of global warming, we explore the relation and combined effect of urban expansion scale and regional climate change. This study uses WRF (Weather Research and Forecasting) model to simulate the impacts of urbanization process on meteorological conditions and regional climate over the Xiongan New District, Beijing-Tianjin-Hebei city cluster, and Guangdong-Hong Kong-Macao Bay Area city cluster. The main conclusions are as follows:

(1) This study selects Xionan New District, as a typical new kind of city, using the WRF model to simulate the impact of the Xiongan New District’s start-up and mid-term urban scale on the regional climate in the summer of 2016. Different stages of the urbanization process in Xiongan New District leads to elevated temperature with various magnitudes. The Xiongan New District during the start-up period with 100 km² urban area will result in the average summer temperature increasing by 0.25°C, while the mid-term stage with an urban area reaching 200 km² will lead the temperature enhancement to reach 0.87°C. However, these two stages of the Xiongan New Area have minor impacts on the temperature over the whole Beijing-Tianjin-Hebei region. During the start-up period, the Xiongan New Area mainly affects the internal region. When the Xiongan New Area reaches the mid-term stage, it will cause a temperature increase in the urban and its southwestern area. The urban parameters covered Xiongan New area are modified in the WRF model with the albedo set to 0.15 instead of 0.18, and LAI reduced from 3 m²/ m² to 1 m²/ m². The decrease in surface albedo and leaf area index increases the net radiant energy in the region, and the lower vegetation cover leads to a decrease in latent heat transport from the surface to the atmosphere (about 60 W/m²) and an increase in sensible heat flux of 20 W/m². The relative humidity at the surface decreases by about 10%, and the increase in sensible heat triggers a weak local secondary circulation, leading to a weak increase in wind speed. Further analysis of the spatial distribution of the monthly mean temperature shows that the Xiongan New Area has significant warming during the early and middle phases of construction, mainly in July and August, within the administrative region and in the southwest of the city. The temperature enhancement becomes much intense when the urban area reaching 200 km², leading to a warming of up to 1°C in the surrounding areas in summer.

(2) This study adopted CRU temperature and precipitation gridded dataset for validating the model outputs and then employed CESM RCP4.5 bias correct data as initial and boundary conditions to drive the WRF model for simulating the meteorological conditions over the Beijing-Tianjin-Hebei area in 2020 and 2030 with the urban region assumed to expand 9600 km². The results indicate that the spatial pattern of average temperature across the Beijing-Tianjin-Hebei area and the maximum and minimum temperature in complex terrain within the domain (Yanshan Mountains and Taihang Mountains) were both well-captured by the historical experiment, with the spatial correlation coefficient between model outputs and CRU higher than 0.9. However, the WRF simulation shows weak capability in reproducing precipitation featured by overestimation on the historical experiment. The annual average temperature in the Beijing-Tianjin-Hebei area shows an increasing trend in RCP4.5 scenarios. Specifically, temperature in the S1 experiment (default urban area) and the S2 experiment (expanded urban area) in the year of 2030 are 0.5℃ and 1.0℃ higher than HIST's, respectively. The increase of temperature due to the expansion of urban areas is particularly evident in summer. In summer 2020 and 2030, the temperature in the S2 experiment is predicted 1.01°C and 0.87°C higher than the S1 experiment, while the temperature in the S2 experiment reduces by 1.81°C and 1.65℃ in winter.

(3) In Guangdong-Hong Kong-Macao Bay Area, five WRF simulations were performed in the historical period (1980,1990,2000) and future years (2030,2040,2050) under various representative concentration pathways with the default urban area and projected 9000 km2 urban expansion, including historical experiment (HIST), RCP4.5 scenario experiment (S1), RCP8.5 scenario experiment (S2), RCP4.5 scenario experiment with the expanded urban area (S1_EXP) and RCP8.5 scenario experiment with the expanded urban area (S2_EXP). The results indicate that WRF simulations could capture the characteristics of the spatial pattern of the annual average temperature, with the spatial correlation coefficient between model outputs and CRU data higher than 0.80. However, the temperature in hilly and mountains situated on northern Guangdong was underestimated by the WRF simulations, while the temperature in the Pearl River Delta was overestimated. The simulated extremely high temperature (T90p) and extremely low temperature (T10p) in the HIST experiment show a strong agreement with the CRU data with the spatial correlation coefficient ranged from 0.68 to 0.90. In S1 experiments, the average temperature in the urban area of the Guangdong-Hong Kong-Macao Greater Bay Area will increase by 0.33℃, 0.66℃, and 1.12°C in 2030, 2040, and 2050, respectively. In the S1_EXP experiments, the annual average temperature has increased by 0.51℃, 0.83℃, and 1.26℃ compared with the HIST experiment, respectively. However, for the region of South China, the heating effect attributed to greenhouse gas emissions is much intense than the heating effect brought by the urban area expansion. In 2040 and 2050, there will be an extreme high-temperature center (higher than 32°C) at the junction of Foshan and northern Guangzhou. The extremely low temperature will moderately increase over the Guangdong-Hong Kong-Macao Greater Bay Area, reaching 0.78°C in 2050. From the simulated sensible heat flux and latent heat flux results, when the urban area expanded in, With the expansion of urban areas over the Guangdong-Hong Kong-Macao Greater Bay Area, the annual average sensible heat flux is around 70-90W/m², while the latent heat flux is ranged 10-20W/m², indicating that the heat transferred from the surface to the atmosphere is dominated by sensible heat. The expansion of urban areas caused less evaporation and lower water vapor content, further leading to the dominant heat transport changed from latent heat to sensible heat. The model outputs in the S2_EXP scenario are consistent with this phenomenon, but the annual mean temperature increase in the urban area in 2050 is more significant than the HIST experiment, increasing by 0.53℃, 1.00℃ and 1.58℃ in 2030, 2040 and 2050, respectively, compared to the HIST experiment.

This study used the WRF model combined with MODIS land use/vegetation cover data for analyzing the features and relevant changes in regional temperature, precipitation, and radiation flux under various expansion magnitudes of the Beijing-Tianjin-Hebei city cluster in northern China and the Guangdong-Hong Kong-Macao Greater Bay Area city cluster in southern China. Furthermore, the mechanism of urbanization-induced temperature enhancement was investigated based on sensitivity experiments. This study could help distinguish the warming effects of future climate change and urban area expansion, benefit the risk estimation of extreme climate events, provide insights to future urban planning, environmental pollution management, and public health protection in developing city clusters.