作为城市绿地系统的重要组成部分，城市公园绿地对于城市居民具有高度的开放性和公共性，除了提供生态系统调节服务外，还为城市居民提供了社交娱乐的公共空间，增强了居民的归属感和参与感。在城市化进程中，城市绿地面临着诸多挑战和变化。一方面，城市绿地面积受到不同程度的压缩和削减。另一方面，城市绿地空间分布和质量不均衡现象依然存在，城市绿地对人口暴露的不公平性问题也在增加。虽然通过严格执行国家城市土地使用规范，我国城市绿地建设取得显著成效，但多年来，我国城市绿地系统规划将提高绿地率、绿化率、人均绿地面积作为目标，缺少对绿地布局公平性的重视。因此，有必要对公园绿地资源在空间布局的公平性进行评估，为城市公园绿地的空间布局优化和使用效率提升提供支持。

本研究以北京市六环内区域为例，探索各级公园的空间分布格局和热点区域。其次，构建了各级公园绿地复杂网络和整体公园绿地复杂网络，并利用复杂网络分析方法，获取了各网络的宏观模式、微观特征和中观结构相关信息。进一步，基于北京市六环内居民小区POI数量，得到公园绿地可获得性的空间分布情况。最后，以公园节点重要性数据为基础，量化了居民小区公园绿地可获得性的公平程度。主要结果如下：

（1）北京市六环内各级公园在空间上分布不均，具有明显的空间集聚和多中心特征。此外，每个公园等级都观察到多个空间聚集区域，其数量和规模各不相同。一级公园从中心区域向西北形成一个规模较大的空间聚集区域。二级公园分布相对均匀，空间聚集区域分散，规模不大，各环线中均可观察到存在高密度或较高密度中心的聚集区域。三级公园形成了西南方向规模较大的三个空间聚集区域和东北方向规模较小的多个空间聚集区域。四级的空间格局与三级公园相似，空间聚集区域集中于西南方向。

（2）在宏观层面，一级公园网络在各级网络中规模最大，其节点倾向于聚集成群，形成高度连通的子网络；二级公园网络具有破碎性的特征；三级公园网络的局部聚集程度高；四级公园网络的节点之间连接程度较高。在微观层面，各级公园网络中拥有高度中心度和高接近中心度节点的空间分布格局相似，具有高接近中心度的节点数量更多；重要性程度高的公园节点位于西城区、东城区、海淀区、丰台区、石景山区和大兴区。在中观层面，各级公园网络之间存在社群数量差异，大多数地理空间上接近的节点倾向于属于同一社群，社群边界与现有的行政边界之间存在空间差异。

（3）公园绿地在居民小区缓冲区内的可获得性呈现出不均匀的空间分布，并且各级公园展现出不同的空间特征。从可视化分析和评价结果来看，居民小区一级公园可获得性的空间分布主要呈现出“核心—边缘”结构；二级公园的可获得性高的居民小区主要位于北京市六环内区域的中心和东部；三级公园可获得性高的居民小区分散分布于北京市六环内西南方向的区域；四级公园可获得性高的居民小区位于北京市六环内区域的中心和西部；居民小区的整体公园可获得性大致呈现出“核心—边缘”结构。

（4）北京市六环内居民小区各级公园可获得性的基尼系数分别为0.861、0.858、0.893和0.919，表明在北京市六环内的区域，居民小区在获取各级公园上具有高不公平性，各级公园绿地布局不均衡。西城区的一级公园可获得性的基尼系数最低（Gini=0.357），西城区内居民小区在获得一级公园资源方面比较公平。虽然二级公园的可获得性高的居民小区主要位于西城区和通州区，但这两个区的二级公园可获得性的基尼系数并非最低。虽然丰台区南四环附近和石景山区西五环附近存在三级公园可获得性最高的居民小区聚集区，但这两个区的三级公园可获得性的基尼系数并非最低。基尼系数最低的东城区内存在四级公园可获得性高的居民小区聚集区。

（5）北京市六环内居民小区的整体公园可获得性的基尼系数为0.647，不公平程度较重。根据基尼系数，东城区（Gini=0.292）、西城区（Gini=0.312）和石景山区（Gini=0.397）的居民小区在公园绿地可获得性上比较公平；昌平区（Gini=0.737）、顺义区（Gini=0.745）、朝阳区（Gini=0.761）表现出明显的不公平。

综上，本研究证实北京市六环内公园绿地在空间上分布不均，各级公园具有明显的空间集聚和多中心特征，且居民小区在获取各级公园绿地资源上存在较高的不公平性。在城市公园绿地复杂网络的优化策略上，以宏观、微观和中观三个层面为切入点，建议提高公园绿地连通性、提升重要公园绿地覆盖率和实现社群公园差异化。本研究加深了对城市公园绿地体系的认识，促进了对城市公园绿地空间公平的理解，有助于评价现行政策和为公园绿地规划提供指导意见，以促进城市公园绿地的空间布局公平性和城市可持续性发展。

As an integral component of urban green space systems, urban parks play a crucial role in providing highly accessible and public spaces for residents. Apart from offering ecosystem regulation services, they serve as social and recreational hubs, enhancing residents' sense of belonging and participation. However, amidst the process of urbanization, urban green spaces face various challenges and changes. On one hand, their areas are subject to varying degrees of compression and reduction. On the other hand, disparities in the spatial distribution and quality of urban green spaces persist, leading to increased inequities in residents' exposure to such spaces.

While significant progress has been made in urban green space construction in China through strict adherence to national urban land use standards, there has been a lack of emphasis on the fairness of green space layout in urban green space system planning over the years. Therefore, there is a need to assess the spatial equity of park green space resources to support the optimization of urban park green space layout and enhance usage efficiency.

Taking the area within the Sixth Ring Road of Beijing as an example, this study explores the spatial distribution patterns and hotspots of parks at various levels. Subsequently, it constructs complex networks of park green spaces at different levels and utilizes complex network analysis methods to obtain macroscopic patterns, microscopic characteristics, and mesoscopic structural information of each network. Furthermore, based on the number of Points of Interest (POIs) in residential neighborhoods within Beijing's Sixth Ring Road, the spatial distribution of park green space accessibility is determined. Finally, based on park node importance data, the fairness of park green space accessibility in residential neighborhoods is quantified. The main results are as follows:

(1) Parks at various levels within the Sixth Ring Road of Beijing exhibit uneven spatial distribution, with evident spatial agglomeration and multi-center features. Each level of parks observes multiple spatial agglomeration areas, varying in number and scale.

(2) At the macroscopic level, first-level park networks exhibit the largest scale among all levels, with nodes tending to cluster into groups, forming highly connected sub-networks. Second-level park networks have the lowest network demonstrate fragmented characteristics. Third-level park networks exhibit a high degree of local clustering. Fourth-level park networks display a relatively high level of connectivity between nodes. At the microscopic level, there is a similar spatial distribution pattern of nodes with high centrality and closeness centrality across all levels of park networks, with a greater number of nodes exhibiting high closeness centrality. Parks with high levels of importance are located in Xicheng District, Dongcheng District, Haidian District, Fengtai District, Shijingshan District, and Daxing District. At the mesoscopic level, there are differences in the number of communities between park networks of different levels, with most geographically close nodes tending to belong to the same community, and spatial differences existing between community boundaries and existing administrative boundaries.

 (3) Park green space accessibility within residential neighborhood buffer zones exhibits uneven spatial distribution, with different levels of parks demonstrating distinct spatial characteristics. From the visual analysis and evaluation results, the spatial distribution of first-level park accessibility in residential neighborhoods mainly exhibits a "core-periphery" structure. Residential neighborhoods with high accessibility to second-level parks are mainly located in the central and eastern areas of the Sixth Ring Road area of Beijing. Residential neighborhoods with high accessibility to third-level parks are scattered in the southwestern direction of the Sixth Ring Road area of Beijing. Residential neighborhoods with high accessibility to fourth-level parks are located in the central and western areas of the Sixth Ring Road area of Beijing. And the overall park accessibility in residential neighborhoods roughly exhibits a "core-periphery" structure.

(4) The Gini coefficients for accessibility to parks of various levels in residential communities within the Sixth Ring Road of Beijing are 0.861, 0.858, 0.893, and 0.919, indicating high inequality in access to parks of all levels within this area, reflecting an uneven distribution of green spaces. The Gini coefficient for accessibility to first-level parks in Xicheng District is the lowest (Gini = 0.357), suggesting relatively fair access to first-level park resources for residential communities within this district. Although residential communities with high accessibility to second-level parks are mainly located in Xicheng District and Tongzhou District, these two districts do not necessarily have the lowest Gini coefficients for accessibility to second-level parks. While there are clusters of residential communities with the highest accessibility to third-level parks near the southern Fourth Ring Road in Fengtai District and near the western Fifth Ring Road in Shijingshan District, these two districts do not necessarily have the lowest Gini coefficients for accessibility to third-level parks. In Dongcheng District, which has the lowest Gini coefficient, there are clusters of residential communities with high accessibility to fourth-level parks.

(5) The overall Gini coefficient for park accessibility in residential communities within the Sixth Ring Road of Beijing is 0.647, indicating a significant degree of inequality. According to the Gini coefficients, residential communities in Dongcheng District (Gini = 0.292), Xicheng District (Gini = 0.312), and Shijingshan District (Gini = 0.397) demonstrate relatively fair access to park green spaces. Conversely, Changping District (Gini = 0.737), Shunyi District (Gini = 0.745), and Chaoyang District (Gini = 0.761) exhibit pronounced disparities in park accessibility.

In summary, this study confirms that park green spaces within the Sixth Ring Road of Beijing exhibit uneven spatial distribution, with parks at various levels demonstrating distinct spatial agglomeration and multi-center features. Furthermore, residential neighborhoods display high levels of inequality in accessing park green space resources at different levels.