【作者】 董珊；

【导师】 许国昌；

【作者基本信息】 哈尔滨工业大学 ， 物理， 2021， 硕士

【摘要】 城市,作为人类活动及经济发展的核心区域,其变化和发展一直是政府及社会各界所关心的热点问题。城市扩张是人民生活水平提高的结果,显示了城市经济的发展脉络,但随之而来的则是生态环境受到影响、能源被消耗、资源被浪费。为了走好“新型城镇化”道路,使城市发展同生态环境相协调,本文对过去的发展情况进行了调查与评估,希望能够为今后的发展路线提供决策方面的辅助。粤港澳大湾区规划自2016年提出后一直是国家重点关注的区域发展战略,将粤港澳地区作为发展历程研究的整体对象,具有较为重要的现实意义。与此同时,遥感技术的发展与遥感计算云平台的出现也使得对大区域、长时间的城市信息提取成为可能。本文基于Google Earth Engine平台,通过对比实验确定适合该区域提取方法,对粤港澳大湾区1990年-2020年30年间的遥感图像进行处理分类,并对结果进行分析。根据对不同分类器分类效果对比和不同合成方法效果对比确定了CART决策树分类器与中值合成作为提取的基本方法。对同年不同分辨率数据分类结果进行对比,发现在长时间序列遥感图像信息提取中分辨率改变对结果存在影响,因此,在本次研究过程中需保持分辨率不变。在城市信息提取方法确定后,对30年间数据进行处理、合成、分类,得到粤港澳大湾区植被、水体、裸地、建筑的空间变化及面积变化情况,并对其变化趋势进行分析。得到了植被面积整体呈“下降-稳定-下降”三个阶段的变化趋势;水体面积与裸地面积整体在均值上下波动;建筑面积呈上升趋势,且变化速率各阶段有所不同,具体表现为“缓慢-加快-缓慢-加快”的形式。对粤港澳大湾区中各个城市建筑信息提取结果进行展示分析,对各城市扩张空间趋势、面积变化趋势及年均速率进行分析。除了对提取结果的分析,本文还对大湾区整体城市建筑面积的扩张以空间区域扩张及发展和谐程度为指标进行了分析。对粤港澳大湾区整体建筑面积变化进行趋势拟合,确定适配模型为S型曲线模型。通过夜间灯光遥感数据对粤港澳大湾区城市空间分布上的扩张趋势进行拟合分析,发现惠州北部、江门、肇庆相较于其它城市扩张趋势较弱。为进一步探究城市扩张过程与生态环境的相互作用,本文还对粤港澳大湾区城镇化发展与生态环境耦合度及和谐度进行了建模研究。结果显示,耦合度和和谐度都整体呈“上升-平稳-下降”趋势。从整体上看,本文相较于其它论文以高精度完成了研究区域的遥感影像分类与分类信息提取,并针对整个研究区域及区域中各行政范围进行了不同尺度、不同侧重的城市扩张情况分析,在此基础上还加入统计数据对整个研究区域进行城镇化发展与生态环境之间和谐度的分析,具有较好的创新型和现实意义。更多还原

【Abstract】 City,as the core area of human activities and economic development,its change and development has always been a hot issue that the government and all walks of life are concerned about.The expansion of cities has improved people’s living standards and promoted economic development,but what comes with it is the impact on the ecological environment,the consumption of energy,and the use of resources.In order to take the road of "new urbanization",to coordinate urban development with the ecological environment,to understand and evaluate the city’s past development,which will help to provide assistance in the decision-making of future development routes.The concept of the Guangdong-Hong Kong-Macao Greater Bay Area has been the country’s key regional development strategy since it was put forward in 2016.It is of great practical significance to study the development process of the GuangdongHong Kong-Macao region.The development of remote sensing technology and the emergence of remote sensing computing cloud platforms have also made it possible to extract large-region,long-term urban information.Based on the Google Earth Engine platform,this paper determines the suitable extraction method for this region through comparative experiments,processes and classifies the remote sensing images of the Guangdong-Hong Kong-Macao Greater Bay Area from 1990 to 2020,and analyzes the results.According to the comparison of the classification effects of different classifiers and the comparison of the effects of different synthesis methods,the choice of CART decision tree classifier and median synthesis as the basic method of extraction is determined.In the same year,the classification results of different resolution data were compared,and it was found that the resolution change in the long-term remote sensing image classification had an impact on the results.It was determined that the remote sensing data in this study kept the resolution unchanged.After the urban information extraction method is determined,the 30-year data is processed,synthesized,and classified to obtain the spatial and area changes of vegetation,water bodies,bare land,and buildings in the Guangdong-Hong KongMacao Greater Bay Area,and analyze the trend of changes.The vegetation area as a whole shows a three-stage trend of "declining-stabilizing-declining".The area of the water body fluctuates above and below the average value.The area of bare land has an upward trend and fluctuates up and down the average value.The building area is on the rise,and the rate of change is different in each stage,showing a four-stage change result of "slow-fast-slow-fast".Display and analyze the results of building information extraction in each city in the Guangdong-Hong Kong-Macao Greater Bay Area,and analyze the spatial trend,area change trend and annual average rate of each city’s expansion.In addition to the analysis of the extraction results,the expansion of the urban construction area and the expansion of the space in the Bay Area are also analyzed.The trend of building area changes in the Guangdong-Hong Kong-Macao Greater Bay Area is fitted,and the most suitable model is determined to be the S-curve model.Through the night light remote sensing data,the expansion trend of the urban spatial distribution in the Guangdong-Hong Kong-Macao Greater Bay Area is fitted and analyzed,and it is found that the expansion trend of northern Huizhou,Jiangmen,and Zhaoqing is weaker than other cities.In order to explore the interaction between the process of urban expansion and the ecological environment,a modeling study was carried out on the degree of coupling and harmony between urbanization development and the ecological environment in the Guangdong-Hong Kong-Macao Greater Bay Area.It was found that the trend of coupling and harmony was "rising-steady-falling".Compared with other papers,this paper completes the remote sensing image classification and classification information extraction of the study area with high accuracy,and analyzes the urban expansion of different scales and different focuses for the entire study area and each administrative district.On this basis,statistical data is also added to analyze the harmony between urbanization development and ecological environment in the entire study area.On the whole,it has relatively good innovation and practical significance.更多还原