Purple Air – Understanding Your Data

One of the most popular air sensors is Purple Air. Many people use Purple Air sensors to learn about fine particle pollution in their area.

Raw Purple Air data are about two times higher than the real values. We know this from placing Purple Air sensors at our regulatory monitors and comparing measurements.

There are two important considerations when looking at Purple Air data to ensure the data is accurate:

• The data need to be calibrated.
• The data need to be in the right scale and time average.

To improve the accuracy of your Purple Air sensor, follow these instructions:

1. Go to www.purpleair.com, select View Map, and zoom in on your location.
2. On the Purple Air website, in the bottom left corner, click on the Map Data Layer box and select Raw PM2.5 in ug/m3. This will show values for fine particles less than 2.5 microns in micrograms per cubic meter [ug/m3].
3. Then, in the same box, click the Conversion dropdown and select LRAPA. Calibration, also called conversion, adjusts the raw Purple Air values so that they match regulatory air monitors more closely.

   LRAPA is a conversion equation developed by Lane Regional Air Pollution Agency in Eugene, Oregon, and is best suited for our area.

Purple Air Data & Your Health

The color scheme on the Purple Air website – which corresponds to the EPA’s ‘air quality index’ or AQI – is only valid for 24-hour data. If you want to use Purple Air’s data for health recommendations, click on a monitor and look at the 1-day average (second from the right in time options).

Other Considerations

When you click on a monitor a timeseries graph will pop up. The timeseries graph is uncalibrated even if you select the LRAPA conversion. We have let Purple Air know about this issue and hope they fix it soon.

The values in the timeseries graph are raw data and therefore are not accurate. The graph is only useful to look at general patterns over time and to compare one monitor to another.

Purple Air has two sensors, which are both graphed on the timeseries. If the sensors are reading differently, very low, or very high, then one or both may be broken.

You can see current air quality data on our website. Our air monitors display air pollution levels for PM2.5 in µg/m3 over a 24-hour period value – an amount of time that is linked to air quality health effects. If comparing the two websites, remember that the Purple Air timeseries is uncalibrated and the raw values will not match the correct regulatory data. What can be useful is to see how the air quality changes over time. If one monitor sees extended high concentrations while another doesn’t there could be a pollution source nearby or one of the monitors could be reading incorrectly.