Has anyone looked into discrepancies between AirNow’s AQI readings and Purple Air’s readings? I have yet to purchase a Purple Air device but there are two near where I live and they both had AQI’s just over 90 at the same time Air Now reported 65.
Is it just that Purple Air is measuring something different? Purple Air devices are very affordable, how do they stack up against what I assume are more costly and more accurate/reliable devices that feed AirNow. Or is it something else?
The discrepancies AirNow and the PurpleAir App are influenced by a couple of factors:
Data resolution
The US EPA conversion formula for PurpleAir data
Data Resolution
AirNow displays hourly averaged PurpleAir data that updates every hour. Conversely, the PurpleAir map displays 10-minute averaged data by default. However, you can change the data resolution on the PurpleAir Map to match the averaging period on the AirNow map, which will bring the data more in line. You can do so by changing the averaging period in the PurpleAir map’s configuration menu to 1-hour.
In order to fully align the data, you’ll also need to use the formula AirNow applies to raw PurpleAir data.
US EPA Conversion Formula
The US EPA has created a formula that alters PurpleAir data and is meant for wildfire conditions. This formula is applied to the PurpleAir data that appears on the previously mentioned AirNow Fire and Smoke map. You can learn more about the US EPA conversion formula here: The “Apply Conversion” Field .
The field “Apply conversion” on the PurpleAir Map is set to “No” by default. This means that no formulas are applied and you are seeing raw PurpleAir data. You can replicate the conversion that AirNow uses by changing the value in the “Apply Conversion” field on the map’s configuration menu to “US EPA.”
I have been wondering about the discrepancies myself between what our local area shows on Purple Air and what AirNow does. Here’s the thing. AirNow only measures HOURLY data. It is NOT real-time like Purple Air is. AirNow is owned by the government. It’s run by the EPA. The EPA is in charge of all the government sensors. The AQI is also an index that’s measured according to whatever the dominant pollutant is that day. A lot of times it’s a lot lower than Purple Air if it’s measuring ozone as the dominant pollutant for your area, and if it’s measuring non-woodsmoke PM2.5 and that ozone is the higher reading than non-woodsmoke PM2.5. At the same time, there’s SOME ozone, the WOOD smoke PM2.5 may be a lot higher than ozone. Sometimes AirNow could be measuring PM10 pollution, and it MAY be dominant over ozone. So, these are possible reasons for discrepancies between AirNow and the real-time Purple Air readings.
I had asked for a phone call - 973-650-4162 — but perhaps you don’t do this as policy. Ok, we can go back and forth a bit via email.
If a Purple Air sensor was placed directly next to an EPA sensor, and each delivered their results at the same moment, would they get the same results?
Purple Air retails for ~$220 and my wild guess is that the EPA sensors are much more costly. I get leapfrog technology and all that, but is there any comparative takeaway? Are the readings equally accurate?
I’m starting to get the AQI calculation and can I ask, does Purple Air deliver an AQI, or something else? "The AQI is also an index that’s measured according to whatever the dominant pollutant is that day. A lot of times it’s a lot lower than Purple Air if it’s measuring ozone as the dominant pollutant for your area,…” Or does Purple Air ONLY measure non-woodsmoke PM2.5? What about the recent Canadian wildfires, which obviously were woodsmoke — is this different?
When you use the Purple Air website, on the left-hand side of the page you can set the readings such that they display non-woodsmoke US EPA AQI, or if you toggle the WOODSMOKE button, the readings are slightly different but it is the wood smoke calculation based on the US EPA reading.
We used to get similar readings on our Purple Air sensors (classic model) as compared to EPA readings on AirNow.gov when we selected US EPA conversion. However, lately the numbers are wildly different - this is something we noticed this year (2024). The discrepancy is far greater than it used to be and this is quite puzzling.