北京典型功能区绿地格局演变及其对地表温度的影响

In the context of rapid urbanization of China, large amounts of greenspace have been converted to urban land. These changes affect the structure and function of urban ecosystems. A well-documented consequence caused by the land conversion is urban heat island (UHI), which increases water consumption and energy use, and impair human health. Urban greenspace can effectively mitigate the UHI. Previous studies have documented that both the composition and configuration (such as largest patch index, mean patch size) of the greenspace can affect the land surface temperature (LST). However, in the complex urban ecosystem, the relationship between greenspace pattern and LST may differ from that of the whole urban area. This is because the social economic activities of urban dwellers may affect the cooling effect of greenspace. In the aim of mitigating UHI through urban greenspace planning and management, it is essential to reveal the pattern, dynamic and cooling effect of greenspace under different human influences. Therefore, the objectives of this study are to (1) explore the pattern and dynamic of greenspace affected by different human activities; and (2) quantify the relationship between greenspace and LST under different human interruption..To understand the quantitative relationship between greenspace pattern and LST under different human activities, it is fundamentally important to (1) accurately map and quantify urban greenspace and its dynamics; and (2) select proper variables to represent different human activities. In this study, we will first develop a new classification approach to accurately map urban greenspace and its dynamics. Then, we will choose meaningful analyze unit to reveal the impact of human activities. Greenspace in urban area are heterogeneous and fragmented, to map the fine greenspace and its changes require high resolution image and proper classification method. Here we combine the object based image analysis technical and backdate/update framework for landscape classification and change detection. The new method is suitable for classifying multi-temporal high resolution images. It will improve the classification accuracy, especially for detecting real landscape change. For the analyze unit, we will use urban functional zone (UFZ) to represent the influence of human activity. People in different UFZ has distinct livelihood and lifestyle, and belong to unified management, which makes it an ideal variable to link the impact of human activities to greenspace pattern and process. This study will focus on three types of UFZ, namely residential district, commercial district and cultural and educational district. .Based on the new classification approach, we will map the land cover within the 5th ring road of Beijing in 2002, 2008 and 2014. High resolution images such as quickbird (with 0.6m spatial resolution) and Geoeye (with 0.6m spatial resolution) will be engaged in the classification procedure. Then, we will delineate the boundary of UFZs based on high resolution images, and retrieve LST using the infrared band of ASTER and Landsat images. Based on the boundary of UFZs, we will quantify and compare the greenspace pattern and changes in different UFZs using landscape metrics, then analyze human impact on this different. In addition, the relationship between greenspace pattern and LST within different type of UFZs will also be quantified and analyzed. The results from this study will enhance our understanding on the relationship between greenspace and LST within different human impact. The quantified relationship between landscape pattern and LST can provide valuable information for greenspace planning, design, and management, especially for specific functional zones. In addition, the developed new classification approach is theoretically and practically important for urban landscape quantification.

绿地具有调节城市微气候，缓解城市热岛的功能。绿地的降温功能，不仅受到绿地的覆盖比例与空间配置的影响，还可能受到人类活动强度的影响。揭示在不同人类活动强度下，绿地降温能力的差异，对改善城市热环境的绿地规划设计具有重要意义。本研究旨在探索在人类活动密集，但强度不同的城市典型功能区（居民区、商业区和文教区）， 绿地的空间格局与变化特征，及其对地表温度的影响。本研究将首先发展一套基于多期高分辨率遥感影像的景观分类制图方法，对2002、2008和2014年北京5环内绿地景观开展分类制图，准确量化绿地的变化。然后，利用目视解译的方法开展居民区、商业区、文教区三类城市典型功能区的制图，并以功能区为统计分析单元，研究三类功能区中绿地格局及其变化的规律、绿地的覆盖比例和配置与地表温度的定量关系。本研究有助于深化城市绿地格局、过程、功能相互作用的科学认知，并对改善城市热环境的绿地规划设计具有重要实践意义。

本项目围绕 “城市典型功能区绿地降温效应”这个关键科学问题，选取城市快速扩张的北京作为研究对象，重点开展了三个方面的研究：（1）城市精细景观格局量化技术方法；（2）不同城市功能区的绿地景观格局分析；（3）居民区绿地格局对其热环境的影响。取得了如下研究进展。.（1）构建了城市精细景观格局的量化方法，用于准确量化城市内部的绿地格局和演变特征。利用树、草的光谱和纹理差异构建了区分树和草的遥感分类指数，实现了树和草的有效区分。对于多期影像的分类，发展了结合回溯与迁移学习的多期高分影像自动分类方法。实现了多期高分影像的准确、高效解译，突破了高分遥感海量数据分类的瓶颈。.（2）定量分析了城市绿地的格局和演变特征，揭示了不同城市功能区绿地格局和演变的差异。研究发现北京城市功能区中，工业区的绿地比例最高；居住区、文教区的绿地比例次之，且绿地比例相似；而商业区的绿地比例最低。分析绿地空间配置发现工业区绿地破碎度最低，居住区、文教区的绿地比例较高，且比较相似；而商业区的绿地破碎度最高。对比深圳的功能区绿地格局，也发现了类似的规律，表现为工业区绿地比例高于居民区绿地，商业区绿地比例最低。此外，深圳的研究还发现了越新的建成区绿地比例越高、配置越合理；城市内部填充式的发展具有更多的绿地空间。.（3）以居民区这类城市典型功能区为对象，揭示了城市绿地格局对气温和地温的影响。研究揭示了居民区较大的气温差异，并证实了城市绿地格局对气温的显著影响：增加绿地覆盖比例，增大绿地斑块面积，以及增加绿地的边缘密度，可显著降低温度。揭示了影响气温的有效范围：在观测站点100m范围内绿地比例和配置会显著影响气温。在此基础上，进一步比较建筑高度和绿地对小区热环境的影响。研究发现：高层小区的地表温度要低于低层小区，建筑高度与地表温度有显著的负相关关系，且建筑高度对地表温度的影响要高于绿地。