top of page
  • Triple Helix

Citizen Science: How is PurpleAir Different from EPA Air Quality Monitoring?

Written by Meg Fay '23

Edited by Jasmine Shum '24

Ever wonder what’s in the air you breathe? It may include particulate matter…

As an air pollutant, particulate matter (PM) has known effects on human health, with both short-term and long-term exposure having the potential to be quite damaging. Exacerbation of existing asthma and new-onset asthma, especially in children, has been consistently correlated to the inhalation of air particulates [1]. Additionally, the impact of PM exposure on genetic material suggests greater implications that PM is a carcinogen. In fact, areas near railways and train platforms that burn diesel which releases PM have been identified as “higher-risk cancer zones”. Drawing on public health concerns, the Environmental Protection Agency (EPA) and locally the state-managed Department of Environmental Management (RIDEM) are constantly monitoring levels of fine particulate matter. But individual citizens and community groups also want to know what’s really in the air they breathe daily, and it's becoming increasingly simple.

Companies like PurpleAir sell monitors that allow anyone to measure air quality. Their data is open-source and updated in real-time for all citizen scientists to access here [2]. But with low-cost monitors like these becoming popular, it’s all the more consequential to understand the data collection process and limitations of these devices. PurpleAir sensors operate using a completely different technology from the compliance monitors used by the EPA. Lower-quality data outcomes can serve as a barrier for people hoping to use their observations for social or environmental change.

So what technology is the EPA using?

The EPA uses official protocols, known as Federal Reference Methods (FRMs), and all other measurements are judged in correspondence to FRMs [3]. FRMs consist of 3 main components: an impactor, a filter, and an air pump. The impactor sorts the particles by size so that ultrafine particles can be analyzed differently than coarse particles. The pre-weighed filter collects the sorted particles over the course of a specified period of time, typically 24 hours. The pump draws in a precise volume of air at a consistent speed. All of these parts allow PM to be studied through gravimetric analysis. By dividing the weight of the PM (dirty filter minus clean filter) by the volume of air pulled through the instrument, PM concentrations are calculated and expressed in micrograms of particles per cubic meter of air (μm/m3). These methods are extremely controlled and accurate but are inaccessible to the public for three main reasons: cost, training, and attention. FRMs can cost upwards of $20,000, need to be analyzed carefully in a lab setting, and require someone to remove the filters from the locations which can be in dangerous locations like on rooftops.

There are other options. Technologies with quality deemed similar enough to FRMs are classified as Federal Equivalent Methods (FEMs). They must meet standards in accuracy, precision, range, detection limit, pollution specificity, freedom from co-pollutant interference, noise, drift, and multi-site measurement performance [3]. These typically include innovative technologies that reduce costs or increase real-time reporting capabilities. But even this designation does not include PurpleAir sensors, so what technology is inside low-cost monitoring systems?

The operating device inside the low-cost PurpleAir sensors is called a Plantower and it is a laser dust sensor [4]. These devices quantify light that is scattered by the particles as they pass through the sensor and then output an electrical signal, corresponding to the quantity of scattered light. This is then related to the mass concentration in μm/m3 [5]. The amount of light scattered can vary greatly depending on the size and distribution of the particles. One specific source of error is that PM absorbs water, which means that particles can scatter more light than they would in dry conditions [6]. But PurpleAir’s air quality monitor is only charging a cool $250 for personal monitoring. While this means that the PurpleAir devices demonstrate a distinct trade-off between price and accuracy, recent studies show major improvements to low-cost monitoring networks.

To see one example of research that is geared toward shrinking the data quality gap between government-operated monitors and community-based efforts look no further than this study done at the University of Northern British Columbia [7]. They provide a robust framework for the evaluation of low-cost air quality sensor models and present an optimized correction model for North American PurpleAir sensors using humidity to calibrate raw data. Efforts like these increase the utility of PurpleAir data, with even the U.S. EPA AirNow team and the U.S. Forest Service looking to use them to assist in wildfire smoke impacts [8].



[1] AAFA Home Page [Internet]. Asthma & Allergy Foundation of America. [cited 2022 Dec 11]. Available from:

[2] Download Sensor Data - Data - PurpleAir Community [Internet]. [cited 2022 Dec 11]. Available from:

[3] US EPA O. PM 2.5 Federal Reference Method | Ambient Monitoring Technology Information Center | US EPA [Internet]. [cited 2022 Nov 19]. Available from:

[4] Plantower PMS5003 Air Quality Sensor [Internet]. [cited 2022 Nov 28]. Available from:

[5] contributors PL. Optical Monitoring of Particulate Matter [Internet]. Public Lab. [cited 2022 Nov 19]. Available from:

[6] Zou Y, Clark JD, May AA. Laboratory evaluation of the effects of particle size and composition on the performance of integrated devices containing Plantower particle sensors. Aerosol Science and Technology. 2021 Jul 3;55(7):848–58.

[7] Nilson B, Jackson PL, Schiller CL, Parsons MT. Development and evaluation of correction models for a low-cost fine particulate matter monitor. Atmospheric Measurement Techniques. 2022 Jun 3;15(11):3315–28.

[8] Development O of R&. PurpleAir PM2.5 performance across the U.S.#2 [Internet]. [cited 2022 Dec 11]. Available from:


7 views0 comments
bottom of page