Air monitoring
Air monitoring
We can measure dust and other particles in the atmosphere. To do this, monitoring devices take advantage of a physical property of the pollutant of concern, known as the analyte. An example of such a physical property is that dust and other particles have mass; another example is that particles interact with light. The monitor controls an interaction with the air based on the physical property to quantify the amount of dust or other particles in the air or to collect other information such as composition––are there heavy metals in the particles?
What to consider when interpreting air monitoring data
There is no air monitoring that will tell you everything you may want to know about airborne dust, or that can be used without considering the unique attributes of each technique in terms of the monitoring technology and where/when/how the measurements were made, known collectively as sampling.
As a result, air monitoring data can reveal––but they can also obscure––critical variability in any analyte. Air monitoring data are often presented without some of the context needed to interpret them. This is important when air monitoring data are the basis for justifying regulatory action or inaction. A lack of details may not be mendacious or malicious, as air monitoring can be done without fully considering the implications of all monitoring technology and sampling decisions, even by the person doing it. Now, the onus is on impacted communities to understand and be vocal about what is revealed and hidden by air monitoring in their neighborhoods. While the technical nature of air monitoring can be intimidating and exclusionary, people develop deep knowledge of air pollution variability from living with it and should trust that knowledge.
Here are a few questions to ask scientists and regulators about their air monitoring. These questions can also guide community-led air monitoring project design and interpretation.
Where is air monitoring, and will it “see” the pollutant of concern?
Monitoring location is among the most important factors influencing pollutant data, especially for pollutants that remain in the atmosphere only briefly like dust. For directly emitted pollutants, expect pollutant concentrations to decline exponentially as you move further from the source. As a result, air monitoring that is 1–2 miles from a dust source will tend to measure PM10 concentrations that are dramatically lower than monitoring located within a few hundred yards. Also, be aware of the winds. A monitor located close to an industrial facility, but on the upwind side, will not achieve goals of measuring air pollution impacts from that facility.
What is the timing of air monitoring data?
Air pollution is highly variable in time, and different air pollutants vary in time differently. This temporal variability can be key information for identifying the source and impacts of pollutants. Airborne dust is often episodic, with atmospheric concentrations that are low most of the time punctuated by periodic high dust events. This is because dust settles out of the atmosphere relatively quickly. Dust events often last a few minutes to a few hours. Air monitoring of dust should span these events and collect measurements frequently enough to capture short-duration events. Regulatory monitoring of PM10 does not meet these criteria, but has other aims, see below for discussion.
How is accuracy determined?
Accurate pollutant data reflect the true state of the atmosphere. For example, when you measure a dust concentration of 10 μg/m3 PM10, if that data point is accurate, then there are, in fact, 10 μg/m3 PM10 in the air. Accuracy is typically determined by adding a known amount of the analyte to the monitoring device under controlled conditions. This process is referred to as calibration, although other terms may be used. Calibration needs to be as frequent as the monitor’s accuracy is variable. If calibration is infrequent, there must be some evidence provided that accuracy does not change or drift over time. Keep in mind, high accuracy may not be required for air monitoring to be valuable. For example, to show pollutant concentrations are elevated near an industrial facility, we would need to measure higher pollutant concentrations in the vicinity of that facility compared to further away. This is different from knowing the amount of the pollutant in the air accurately.
Is the measurement precision sufficient to discern pollutant impacts?
Precision describes the certainty with which we know each data point. All monitoring data include some uncertainty. As a result, a dust concentration data point is never just, for example, 10 μg/m3 PM10. It is instead 10.0 ± 0.2 μg/m3 or, put another way, a concentration value between 9.8 and 10.2 μg/m3. For a high-precision measurement, this range is small. For a low-precision measurement, this range is large. Air monitoring data can have any combination of accuracy and precision. It is essential that the precision of the monitoring data be sufficient to distinguish between the conditions of interest. To show pollutant concentrations are elevated near an industrial facility, we would compare pollutant data in the vicinity of that facility and further away, and, for there to be an effect, measured concentration differences must be larger than their uncertainties. For example, compare case (a) 10.0 ± 0.2 μg/m3 PM10 near the facility versus 5.0 ± 0.1 μg/m3 PM10 further away and (b) 10.0 ± 3.0 μg/m3 PM10 near the facility versus 5.0 ± 3.1 μg/m3 PM10 further away. The monitoring in case (a) but not (b) was precise enough to show differences between the facility vicinity and further away. Precision typically improves when you average many data points together, scaling as one divided by the square root of the number of data points in the average (N), or 1/√(N).
What is the physical property of the analyte being detected, and how is that converted to metric reported?
Knowing the physical property of the analyte being used to make the measurement is the first step to understanding the air monitoring data. This is never obvious, and it is not a silly question to ask someone tasked with air monitoring to explain how the monitoring device physically works. They should be able to explain this clearly so it is understood by all, even nonexperts. If they cannot, which is possible, they should admit they do not know and get back to you with that information. If you perceive there to be a lack of transparency around the air monitoring––take note––that lack of transparency can propagate through data interpretation, analysis, and ultimately regulatory decision-making.
The metric reported by an air monitor may not be directly measured and may instead be estimated from a tangentially-related property. For example, some popular low-cost particle sensors used in community air monitoring quantify the interaction of particles with light using a device called a nephelometer. Smaller particles, especially PM1, are preferentially detected because of the positioning of the light path and detector and how the airstream moves within the device. As a result, the monitors rely on a series of mathematical equations derived from calibration experiments conducted in other locations to convert PM1 concentrations to PM10 concentrations. This works where and when PM1 and PM10 have similar sources, as changes in PM1 concentrations are indicative of changes in PM10; however, it does not work for airblown dust sources such as coal piles, which emit PM10 but not PM1.
Are airborne pollutants reaching the part of the monitor where the measurement occurs? Put another way, are their potential sampling losses and how are their impacts minimized?
To make the measurement, the analyte must be delivered to the area within the monitor where the controlled interaction occurs. For pollutants that deposit out of the air quickly, are reactive, that stick to surfaces, or are large particles, a major portion of the analyte may be lost as air moves through any sampling tubing and the monitoring device itself before the measurement is made. These are called sampling losses. Accurate air monitoring minimizes or otherwise accounts for any sampling losses.
Types of monitoring
Regulatory air monitoring
Regulatory air monitoring has the specific aim of determining state compliance with the Clean Air Act. States must show that the concentrations of PM2.5 and PM10, as well as other criteria pollutants, do not exceed the U.S. Environmental Protection Agency (EPA) National Ambient Air Quality Standards (NAAQS). This leads to a very specific air monitoring challenge: air monitoring data collected in any location must indicate the same pollutant concentrations if measured by any monitor located across the U.S. As a result, significant effort is made to standardize air monitoring technology, sampling, and other procedures such as calibration. EPA-approved standardized procedures are referred to as Federal Reference Methods (FRMs). Note the ‘M’ stands for method not monitor, as it encompasses both the monitoring technology itself and a set of approved protocols on monitor location, installation and operation, and data processing and quality evaluation. You may also encounter Federal Equivalent Methods (FEMs), which are methods that are well characterized by the EPA and may produce air monitoring data deemed equivalent to an FRM. FRM or FEM measurements are not necessarily distinguished from other air monitoring techniques by accuracy or precision, rather by their standardized and regimented use and that their data can be used as proof of NAAQS compliance/noncompliance.
The Virginia Department of Environmental Quality (VA DEQ) measures PM10 mass concentrations at two locations in Hampton Roads: one monitor at NASA Langley Research Center in Hampton and one monitor at the NOAA Storage Facility in Norfolk. Both monitors are too far away from coal export terminals to routinely record evidence of airborne coal dust. These monitors are FRM technique 062, which uses a high-volume sampler to pull a large and known quantity of air through a filter, collecting PM10 and smaller particles on a filter surface that is then returned to a laboratory to determine the PM10 mass using a high-precision, high-accuracy balance. Because the volume of air is known, the mass concentration––mass per volume air (μg/m3)––can be computed. Samplers operate continuously in 24 hour periods. This is because the PM10 NAAQS is based on a 24-hour average concentration. Samplers operate on a one-in-six cycle, collecting PM10 data only one day (a 24-hour period) every six days. This means that on most days they collect no data at all, potentially missing many dust events entirely.
Community air monitoring
In the past decade, low and lower-cost air monitoring devices have become widely available, especially for measuring airborne particle concentrations. These devices cost a few hundred to a few thousand dollars each and are often straightforward to install and operate, with minimal data processing required. The devices facilitate community-led air monitoring, especially where residents have air pollution-related concerns and where there is no regulatory monitoring. Because these devices are low cost, many can potentially be purchased to build a monitoring network across a neighborhood. Such spatial information is useful for showing, for example, elevated particle concentrations in the vicinity of suspected sources. While these air monitors cannot technically be used as evidence of NAAQS noncompliance, they can produce data that challenge official narratives, apply pressure on regulators and/or industries to intervene, and function as tools for community organizing and mobilization. As a note: the VA DEQ has referred to the low-cost monitoring devices as “sensors” and the devices used in regulatory monitoring as “monitors”, although this is not a standard nomenclature.
Scientific air monitoring
There are a wide variety of scientific instruments that measure concentrations and other properties of dust and other particles, with new methods still being developed. These monitoring devices may be commercially available or custom-made and used in only one research laboratory. These devices often focus on unresolved scientific and engineering issues, with questions including: What is the composition of particles? How are particles shaped (their morphology) or how are they mixed? Do particles have coatings and what is in those coatings? Do particles cause a response in human cells? How do particles change through chemistry or physics? Can monitoring devices be miniaturized, and can we reduce the power and costs of novel particle monitoring to collect data, for example, across a city or from a drone? Scientific air monitoring is considered the gold standard for monitoring and will likely require an academic or other research organization partner.