Data from: Vegetation cover in relation to socioeconomic factors in a tropical city assessed from sub-meter resolution imagery

Fine-scale information about urban vegetation and social-ecological relationships is crucial to inform both urban planning and ecological research, and high spatial resolution imagery is a valuable tool for assessing urban areas. However, urban ecology and remote sensing have largely focused on cities in temperate zones. Our goal was to characterize urban vegetation cover with sub-meter resolution aerial imagery, and identify social-ecological relationships of urban vegetation patterns in a tropical city, the San Juan Metropolitan Area, Puerto Rico. Our specific objectives were to: i) map vegetation cover using sub-meter spatial resolution (0.3 m) imagery; ii) quantify the amount of residential and non-residential vegetation; and iii) investigate the relationship between patterns of urban vegetation versus socioeconomic and environmental factors. We found that 61% of the San Juan Metropolitan Area was green, and that our combination of high spatial resolution imagery and object-based classification was highly successful for extracting vegetation cover in a moist tropical city (97% accuracy). In addition, simple spatial pattern analysis allowed us to separate residential from non-residential vegetation with 76% accuracy, and patterns of residential and non-residential vegetation varied greatly across the city. Both socioeconomic (e.g., population density, building age, detached homes) and environmental variables (e.g., topography) were important in explaining variations in vegetation cover in our spatial regression models. However, important socioeconomic drivers found in cities in temperate zones, such as income and home value, were not important in San Juan. Climatic and cultural differences between tropical and temperate cities may result in different social-ecological relationships. Our study provides novel information for local land use planners, highlights the value of high spatial resolution remote sensing data to advance ecological research and urban planning in tropical cities, and emphasizes the need for more studies in tropical cities.

城市植被与社会-生态关系（social-ecological relationships）的精细尺度信息，可为城市规划与生态研究提供关键支撑；高空间分辨率影像（high spatial resolution imagery）则是评估城市区域的重要工具。然而，当前城市生态学与遥感研究大多聚焦于温带城市。本研究旨在利用亚米级分辨率航空影像（sub-meter resolution aerial imagery）刻画城市植被覆盖特征，并识别热带城市——波多黎各圣胡安都会区的城市植被格局社会-生态关系。本研究的具体目标包括：① 利用亚米级（0.3米）空间分辨率影像绘制植被覆盖图；② 量化住宅与非住宅用地的植被占比；③ 探究城市植被格局与社会经济、环境因素之间的关联。研究发现，圣胡安都会区61%的区域为绿地；结合高空间分辨率影像与面向对象分类法（object-based classification），可在湿润热带城市中高效提取植被覆盖信息，分类精度达97%。此外，通过简单的空间格局分析，我们能够以76%的精度区分住宅与非住宅用地植被，且全市范围内住宅与非住宅植被的格局差异显著。在空间回归模型（spatial regression models）中，社会经济因素（如人口密度、建筑年代、独立住宅占比）与环境变量（如地形）均能有效解释植被覆盖的空间差异。但温带城市中已被证实的关键社会经济驱动因子（如收入与住房价值），在圣胡安都会区并不显著。热带与温带城市在气候与文化上的差异，可能导致二者的社会-生态关系有所不同。本研究为当地土地利用规划者提供了全新的参考信息，凸显了高空间分辨率遥感数据在推动热带城市生态研究与城市规划方面的价值，并强调了在热带城市开展更多相关研究的必要性。