导　　师： 夏丽华

学科专业： 120405

授予学位： 硕士

作　　者： ;

机构地区： 广州大学

摘　　要： 绿地作为城市土地重要的组成部分，科学合理的绿地规划是发挥其生态服务功能，营造可持续的城市生态环境的重要措施。加强新技术的研究在城市绿地规划中的应用是必然的趋势，为探讨该领域的遥感技术机理和应用，提高绿地规划管理水平，文中以城市绿地养护监测及绿地规划中治理热岛为主要研究内容，主要工作、研究方法和应用实践如下： （1）城市植被胁迫监测 城市植被养护该工作是城市绿化建设的基础工作之一，如植被长势动态监测、落叶枯萎病清查等，开展城市植被胁迫研究在提升城市林业和园林管理、提高植被生态服务功能有重要作用，本文研究采用EO-1遥感卫星Hyperion传感器过境采集的广州市东边建成区高光谱影像，在大气纠正处理后，依据植物光谱响应原理进行以下方法试验：选取合适的植被指数在ENVI-Forest Health中分类；通过混合像元分解获得植被丰度，结合纯净像元的识别进行植被胁迫的分类；对比两种试验结果表明了在植被胁迫识别中混合像元分解--植被纯净像元提取的方法要比指数法更好地表示植被胁迫的特征；提取影像中万亩果园的植被端元光谱形态作对比验证，同时使用了地物光谱辐射计对不同健康状况的植物冠层光谱测量以作地面参考验证，实验结果表明植被潜在胁迫的指示标志，光谱反射率近红外区间（680-1300nm）“红边斜率”变化较为吻合。本研究前期通过对影像指示的地点实地调查和资料回顾，总结出植被胁迫的城市空间现象，研究的结果希望该监测方法能为节约土地资源为目的、对保育养护型的城市林业管理提供定性和定量的研究。 （2）城市地表温度反演 城市热岛格局遥感监测是城市绿地规划的工作内容之一，而城市绿化工作也是致力于缓解热岛效应。在城市小尺度环境的热红外遥感影像反演地表温度的研究中，本文采用EOS//TERRA卫星的ASTER数据，采用国内学者毛克彪针对ASTER数据推导出来其反演地表温度的劈窗算法（Split-window Algorithm），经如下预处理：辐射定标—大气纠正—几何纠正后。按照该算法反演地表温度的设计体系进行如下步骤：普朗克定律计算亮温、基于NDVI地表性质的辐射率赋值、MOTRAN辐射传输在线计算获取实时大气透射率，来实现该算法反演地表温度。通过对地表温度演算结果分布频率的评价（296.7~304.1K）及与当时气温相比较，对照高分辨率影像中的地表性质后，认为该算法用于ASTER热红外数据反演地表温度能提供丰富的城市热环境信息，热岛监测能为城市建设管理方面提供有益的启示。 （3）以疏风散热为目标的城市绿地系统规划 通过解释在太阳辐射下地表升温成因和植被降温的机理，用Hyperion数据获得的植被丰度和ASTER数据反演的地表温度进行回归统计分析，说明了植被丰度与地表温度明显负相关，从数量关系上说明了植被越茂密降温作用越明显。参考《广州市城市绿地系统规划2001-2020》，以广州市自然、人文特征为依据，在已有城市绿地系统规划的各种理论与方法的基础上，文中引入植被胁迫遥感监测以及城市热岛监测，论述本文的研究成果能为城市园林绿化规划设计等方面提供有益的建议，以GIS制图分析技术为支撑，以图片和文本形式展示出切实可行、科学合理的绿地布局，以营造良好的城市环境为目标。 Urban green space comprises a critical part of urban land use. A scientific and rationalurban green space planning is to react a full use of greenery’s ecological services, and as animportant strategy to build up a sustainable urban ecology environment. To strengthen theresearch of new technology and its application is definitely a trend in urban green spaceplanning. In this dissertation, research on mechanism and application of remote sensingtechnology in urban space planning was conducted, for the goal of planning and managementpromotion. As a proposal, our research mainly based on urban green space conservationmonitoring and urban heat island eliminating-oriented planning. Our works, researchmethods, and applications are listed below: （1） Urban vegetation stress monitoring Urban vegetation conservation should be one of basic tasks of urban greenery construction,such as vegetation growing condition monitoring, leaf falling-off and withering check-out. Tostart the urban vegetation stress research plays an important part of increasing urban forestrymanagement and its ecological services. In this thesis we used hyperspectral images in theeastern Guangzhou City, China, which is captured by the EO-1satellite-borne Hyperionsensor. With the atmospheric correction preprocessing, according to plant spectra responsetheory, we wanted to conduct the vegetation stress identification, through these two methods:on one hand we selecting the appropriate vegetation indices for image classification viaENVI-Forest Health tool; on the other hand we through “Spectral Mixture Analyst” toobtain vegetation abundance, and combined with purest pixel indentifying to extractvegetation stress. Comparing these experimental results, it shows that in the feature extractionof vegetation, the vegetation abundance analysis methods has a higher precision than thevegetation index method, and a better characterization of vegetation stress, for the regard thatnon vegetation feature spectral is excluded. In respect of validation, we extract the spectrumshape of vegetation endmembers in Wanmu orchard for comparison; and by usingspectroradiometer ASD, we through on-site spectrum measurement of vegetation canopy indifferent health status. The result supports that the image spectra and field measurement havecertain similarity. The result shows that spectrum reflectance in near-infrared section /(680-1300nm/), is comparable with the “red-edge slop” change, which is the symbol ofvegetation under-stress. In the previous days, indicating by the image, we had on-site survey,queried those high spatial resolution images in near-by days, and reviewed materials about theurban construction yearbook via internet, then the cause of vegetation stress’s urban spatialphenomenon had came to a conclusion. We concluded and also hope that our study onmonitoring methods can be applied on preservative-oriented land use, andconservative-oriented urban forestry management, by qualitatively and quantitativelyresearch. （2） Urban surface temperature retrieval Urban heat island pattern remote sensing monitoring should be one of basic tasks ofurban green space planning, and urban greening’s task is to migrate heat island effect. Withrespect of a urban-scale circumstance, we investigated the thermal infrared remote sensingwhich is of higher resolution image, in this thesis we used ASTER data from EOS//TERRAsatellite, adopted the “Split-window Algorithm” developed by domestic researcher, MaoKe-biao, to retrieve land surface temperature. With the processes: radiationcalibration—atmospheric correction—geometric correction, and then conformed to itsretrieving design, we explicated the parameters acquisition in detail that conducting thealgorithm for retrieving: calculating the bright temperature by Plunk’s law, assigning surfaceemissivity based on NDVI value’s classification, and acquiring the real time atmospheretransparence via MOTRAN radiation-transmission online calculation. The retrieving resultshows value distributing frequency located between296.7~304.1K, which was comparablewith the exact date. After comparing to the surface characteristic in high spatial resolutionimage, the result demonstrated this algorithm applied at ASTER data for land surfacetemperature retrieval can providing plentiful information of land surface temperature, by suchurban heat island monitoring can provide significant instruction for urban construction andmanagement. （3） An orientation for heat island migrating urban green space system planning By explaining the mechanism of ground surface’ heating up by solar radiation andvegetation cooling effect, we used vegetation abundance acquired from Hyperion datacoupled with land surface temperature from ASTER data, the result demonstrated a apparent negative correlation between vegetation abundance and land surface temperature, whichquantitatively shows the cooling effect depends on the thickness of vegetation. Based onUrban Green space Comprehensive Planning in Guangzhou,2001-2020, this dissertationintroduces vegetation stress and urban heat island monitoring methods, combine with ecology,geography, and landscape planning for multiple disciplinary research, elaborated that thisresearch can offer instructive suggestion for urban greenery design and planning. Under theguidance of green space planning practice, with the support of GIS cartographic techniqueand analysis, we presented our research achievement in application, design the green spaceallocation scientifically and rationally, for the goal of favorable urban ecology andenvironment construction.

关 键 词： 城市绿地规划 高光谱遥感 植被胁迫 热红外遥感 地温反演

分 类 号： [TP79 TU985 P237]

领　　域： [自动化与计算机技术] [自动化与计算机技术] [建筑科学] [天文地球] [天文地球]