Trusting PurpleAir: Air Quality Sensor Evaluations

PurpleAir sensors are arguably the most trusted and widely used low-cost air quality monitors in the world. Researchers have studied their accuracy, long-term performance, and data reliability in different locations and conditions, proving the effectiveness of PurpleAir.

Sensor Accuracy Evaluations

To measure their accuracy, PurpleAir sensors have been tested against high-quality, government-grade air monitors. One study found up to a 97% correlation with federal equipment (look at the summary report), while another showed that adjusting PurpleAir data using local reference monitors can improve reliability.

Data Adjustments and Conversions

The air changes depending on factors like location, weather, and time of day. For example, the air in Colorado is different from the air in North Carolina because the particles in the air vary. PurpleAir sensors measure these tiny particles, called particulate matter (PM). The type and amount of PM in the air depend on these same factors, which is why air quality differs between places.

Think of a golf ball and a ping pong ball—they look similar, but without picking them up, you wouldn’t know which one is heavier. PurpleAir sensors, specifically the laser counters inside them, work in a similar way. They count the number and size of particles in the air but can’t directly measure how heavy they are. Instead, they estimate it—just like you might assume a golf ball is heavier than a ping pong ball based on what you know. The laser counters then use particle volume and assumed mass to calculate density, which is how PM levels are measured.

 

 

The laser counters make different assumptions for outdoor and indoor data, although we don't know what those assumptions are, as Plantower, the company that makes the laser counters, does not share them publicly.

Since air quality varies by location and conditions, researchers have created conversion equations to adjust PurpleAir data. These adjustments account for factors like different locations, wildfire smoke, and environmental conditions. By applying these conversions, PurpleAir sensors can better match federal monitors and provide relevant data to you.

Algorithm Evaluation and Development

The conversions mentioned above rely on the density measurements produced by PurpleAir sensors. However, one researcher wanted to better understand how these values are created. Instead of relying on the existing method, he developed his own algorithm. Like the laser counters, this algorithm assumes a mass and then calculates its own particle density values. You can find this algorithm on the PurpleAir map as ALT cf=3.4, just go to the map's settings and update the "Apply Conversion" field.

Long-Term Sensor Performance

Some of the earliest PurpleAir sensors, installed in 2016, are still active today. While many of those early models aren't being used anymore, the network has expanded globally, providing a rich history of data. This network has allowed researchers to study how PurpleAir sensors perform over time in real-world conditions.

 

Want to know more? View Specific Evaluations Here