Abstract In 2017, Assembly Bill 617 was approved in state of California, which mandated the allocation of resources for addressing air pollutant exposure disparities in underserved communities across the state. The bill stipulated the implementation of community scale monitoring and the development of local emissions reductions plans. We aimed to develop a streamlined, robust, and accessible PM2.5 exposure assessment approach to support environmental justice analyses. We sought to characterize individual PM2.5 exposure over multiple 24-hr periods in the inland Southern California region, which includes the underserved community of San Bernardino, CA. Personal sampling took place over five weeks in Spring of 2019, and personal PM2.5 exposure was monitored for 18 adult participants for multiple, consecutive 24-hr periods. Exposure and location data were available at 5-s resolution, and participant data recovery was 50.8% on average. A spatial clustering algorithm was used to classify data points as one of seven microenvironments. Mean and median personal-ambient PM2.5 ratios were aggregated along SES lines for eligible datasets. GIS-based spatial clustering facilitated efficient microenvironment classification for more than 900,000 data points. Mean (median) personal-ambient ratios ranged from 0.26 (0.14) to 2.78 (0.65) for each microenvironment when aggregated along SES-lines. Aggregated ratios indicated that participants from the lowest SES community experienced higher home exposures compared to participants of all other communities over consecutive 24-hr monitoring periods, despite high participant mobility and relatively low variability in ambient PM2.5 during the study. The methods described here highlight the robust and accessible nature of the personal sampling campaign, which was specifically designed to reduce participant fatigue and engage members of the inland Southern California community who may experience barriers when engaging with the scientific community. This approach is promising for larger-scale, community-focused, personal exposure campaigns for direct and accurate analysis of environmental justice.

中文翻译：

---

一种表征南加州内陆社区规模空气污染暴露差异的数据驱动方法

摘要 2017 年，加利福尼亚州批准了第 617 号议会法案，该法案要求分配资源以解决全州服务欠缺社区的空气污染物暴露差异问题。该法案规定了社区规模监测的实施和地方减排计划的制定。我们旨在开发一种简化、稳健且易于使用的 PM2.5 暴露评估方法，以支持环境正义分析。我们试图描述南加州内陆地区在多个 24 小时内暴露于个人 PM2.5 的特征，其中包括加利福尼亚州圣贝纳迪诺服务不足的社区。个人采样在 2019 年春季进行了超过五周的时间，并且对 18 名成年参与者的个人 PM2.5 暴露进行了多次连续 24 小时监测。曝光和位置数据以 5 秒的分辨率提供，参与者数据平均恢复率为 50.8%。使用空间聚类算法将数据点分类为七个微环境之一。对于符合条件的数据集，个人环境 PM2.5 比率的平均值和中值沿 SES 线汇总。基于 GIS 的空间聚类促进了对超过 900,000 个数据点的有效微环境分类。当沿着 SES 线聚合时，每个微环境的平均（中值）个人环境比率范围为 0.26 (0.14) 到 2.78 (0.65)。汇总比率表明，尽管参与者流动性高且环境 PM2.5 的变异性相对较低，但在连续 24 小时监测期间，来自最低 SES 社区的参与者与所有其他社区的参与者相比经历了更高的家庭暴露。5 在学习期间。此处描述的方法突出了个人抽样活动的稳健性和可访问性，该活动旨在减少参与者的疲劳并吸引南加州内陆社区的成员，他们在与科学界接触时可能会遇到障碍。这种方法有望用于更大规模、以社区为中心的个人暴露活动，以直接和准确地分析环境正义。