Rugel Emily J, Henderson Sarah B, Carpiano Richard M, Brauer Michael
School of Population & Public Health, 2206 E Mall, Vancouver, BC V6T 1Z9, Canada.
School of Population & Public Health, 2206 E Mall, Vancouver, BC V6T 1Z9, Canada; British Columbia Centre for Disease Control, 655 W 12th Avenue, Vancouver, BC V5Z 4R4, Canada.
Environ Res. 2017 Nov;159:474-483. doi: 10.1016/j.envres.2017.08.033. Epub 2017 Sep 18.
Natural spaces can provide psychological benefits to individuals, but population-level epidemiologic studies have produced conflicting results. Refining current exposure-assessment methods is necessary to advance our understanding of population health and to guide the design of health-promoting urban forms.
The aim of this study was to develop a comprehensive Natural Space Index that robustly models potential exposure based on the presence, form, accessibility, and quality of multiple forms of greenspace (e.g., parks and street trees) and bluespace (e.g., oceans and lakes).
The index was developed for greater Vancouver, Canada. Greenness presence was derived from remote sensing (NDVI/EVI); forms were extracted from municipal and private databases; and accessibility was based on restrictions such as private ownership. Quality appraisals were conducted for 200 randomly sampled parks using the Public Open Space Desktop Appraisal Tool (POSDAT). Integrating these measures in GIS, exposure was assessed for 60,242 postal codes using 100- to 1,600-m buffers based on hypothesized pathways to mental health. A single index was then derived using principal component analysis (PCA).
Comparing NDVI with alternate approaches for assessing natural space resulted in widely divergent results, with quintile rankings shifting for 22-88% of postal codes, depending on the measure. Overall park quality was fairly low (mean of 15 on a scale of 0-45), with no significant difference seen by neighborhood-level household income. The final PCA identified three main sets of variables, with the first two components explaining 68% of the total variance. The first component was dominated by the percentages of public and private greenspace and bluespace and public greenspace within 250m, while the second component was driven by lack of access to bluespace within 1 km.
Many current approaches to modeling natural space may misclassify exposures and have limited specificity. The Natural Space Index represents a novel approach at a regional scale with application to urban planning and policy-making.
自然空间可为个体带来心理益处，但针对人群层面的流行病学研究结果却相互矛盾。改进当前的暴露评估方法对于增进我们对人群健康的理解以及指导促进健康的城市形态设计而言十分必要。
本研究旨在开发一个综合自然空间指数，该指数能基于多种绿地（如公园和行道树）和蓝地（如海洋和湖泊）的存在、形态、可达性和质量，有力地模拟潜在暴露情况。
该指数是为加拿大温哥华地区开发的。绿地存在情况通过遥感（归一化植被指数/增强植被指数）得出；形态从市政和私人数据库中提取；可达性基于诸如私人所有权等限制条件。使用公共开放空间桌面评估工具（POSDAT）对200个随机抽样的公园进行了质量评估。将这些测量值整合到地理信息系统（GIS）中，基于假设的心理健康途径，使用100至1600米的缓冲区对60242个邮政编码区域的暴露情况进行了评估。然后通过主成分分析（PCA）得出单一指数。
将归一化植被指数与评估自然空间的其他方法进行比较，结果差异很大，根据测量方法的不同，22%至88%的邮政编码区域的五分位数排名发生了变化。总体公园质量相当低（0至45分制下平均分为15分），不同邻里层面家庭收入水平之间未见显著差异。最终的主成分分析确定了三组主要变量，前两个成分解释了总方差的68%。第一个成分主要由公共和私人绿地及蓝地以及250米范围内公共绿地的百分比主导，而第二个成分则受1公里范围内无法到达蓝地的影响。
当前许多模拟自然空间的方法可能会对暴露情况进行错误分类，且特异性有限。自然空间指数代表了一种在区域尺度上的新方法，可应用于城市规划和政策制定。
