Background

Fine particulate matter (PM_2.5) is associated with various adverse health outcomes. The MicroPEM (RTI, NC), a miniaturized real-time portable particulate sensor with an integrated filter for collecting particles, has been widely used for personal PM_2.5 exposure assessment. Five-day deployments were targeted on a total of 142 deployments (personal or residential) to obtain real-time PM_2.5 levels from children living in New York City and Baltimore. Among these 142 deployments, 79 applied high-efficiency particulate air (HEPA) filters in the field at the beginning and end of each deployment to adjust the zero level of the nephelometer. However, unacceptable baseline drift was observed in a large fraction (> 40%) of acquisitions in this study even after HEPA correction. This drift issue has been observed in several other studies as well. The purpose of the present study is to develop an algorithm to correct the baseline drift in MicroPEM based on central site ambient data during inactive time periods.

Method

A running baseline & gravimetric correction (RBGC) method was developed based on the comparison of MicroPEM readings during inactive periods to ambient PM_2.5 levels provided by fixed monitoring sites and the gravimetric weight of PM_2.5 collected on the MicroPEM filters. The results after RBGC correction were compared with those using HEPA approach and gravimetric correction alone. Seven pairs of duplicate acquisitions were used to validate the RBGC method.

Results

The percentages of acquisitions with baseline drift problems were 42%, 53% and 10% for raw, HEPA corrected, and RBGC corrected data, respectively. Pearson correlation analysis of duplicates showed an increase in the coefficient of determination from 0.75 for raw data to 0.97 after RBGC correction. In addition, the slope of the regression line increased from 0.60 for raw data to 1.00 after RBGC correction.

Conclusions

The RBGC approach corrected the baseline drift issue associated with MicroPEM data. The algorithm developed has the potential for use with data generated from other types of PM sensors that contain a filter for weighing as well. In addition, this approach can be applied in many other regions, given widely available ambient PM data from monitoring networks, especially in urban areas.

中文翻译：

---

开发一种校正 MicroPEM 基线漂移的方法

 背景

细颗粒物 (PM _2.5 ) 与各种不良健康结果相关。 MicroPEM (RTI, NC) 是一种微型实时便携式颗粒传感器，带有用于收集颗粒的集成过滤器，已广泛用于个人 PM _2.5暴露评估。五天部署的目标是总共 142 次部署（个人或住宅），以获取纽约市和巴尔的摩儿童的实时 PM _2.5水平。在这 142 次部署中，有 79 次在每次部署开始和结束时在现场应用了高效颗粒空气 (HEPA) 过滤器，以调整浊度计的零位。然而，即使在 HEPA 校正之后，在本研究的大部分 (> 40%) 采集中仍观察到不可接受的基线漂移。在其他几项研究中也观察到了这种漂移问题。本研究的目的是开发一种算法，根据非活动时间段内的中心站点环境数据来校正 MicroPEM 中的基线漂移。

 方法

基于非活动期间 MicroPEM 读数与固定监测点提供的环境 PM _2.5水平以及 MicroPEM 过滤器上收集的 PM _2.5重量的比较，开发了运行基线和重量校正 (RBGC) 方法。将 RBGC 校正后的结果与单独使用 HEPA 方法和重量校正的结果进行比较。使用七对重复采集来验证 RBGC 方法。

 结果

对于原始数据、HEPA 校正数据和 RBGC 校正数据，存在基线漂移问题的采集百分比分别为 42%、53% 和 10%。重复的 Pearson 相关分析显示，RBGC 校正后，决定系数从原始数据的 0.75 增加到 0.97。此外，回归线的斜率从原始数据的0.60增加到RBGC校正后的1.00。

 结论

RBGC 方法纠正了与 MicroPEM 数据相关的基线漂移问题。开发的算法有可能与其他类型的颗粒物传感器生成的数据一起使用，这些传感器也包含称重过滤器。此外，鉴于监测网络广泛提供环境 PM 数据，这种方法可以应用于许多其他地区，特别是在城市地区。