Have the EPA PM2.5 Breakpoints Changed?

Uncategorized
1

Previous posts for the EPA PM 2.5 show these breakpoints:

Good 0 - 50 | 0.0 – 12.0
Moderate 51 - 100 | 12.1 – 35.4
Unhealthy for Sensitive Groups 101 – 150 | 35.5 – 55.4
Unhealthy 151 – 200 | 55.5 – 150.4
Very Unhealthy 201 – 300 | 150.5 – 250.4
Hazardous 301 – 400 | 250.5 – 350.4
Hazardous 401 – 500 | 350.5 – 500.4

And reference AQI Breakpoints | Air Quality System | US EPA as the source. AQI Breakpoints | Air Quality System | US EPA

That page now seems to have the breakpoints at (raw values) 0 - 9.0, 9.1 - 35.4, 35.5 - 55.4, etc. However, the EPA page only shows breakpoints for “24 HOUR” and “24-HR BLK AVG” (whatever that is).

I am using the posted (old?) ‘pm2.5’ sensor value for my calculations, and mine are little lower than the Purple Air PM2.5 map.

Have the breakpoints changed recently?

2

Hi Wiley,

Yes! We did update these breakpoints in a recent map update, so they now reflect the current U.S. EPA breakpoints.

Let us know if you have any additional questions about this.

3

So is Purple using something different than that shown in this post? If so, would someone at Purple update that post and reference it here as well?

And does this imply that the firmware in my Flex has been updated so that it displays changed breakpoints/AQI information when referenced via the LAN? Mine shows V7.04.

Thanks

4

What equation is purpleair using for AQI when pm2.5 > 325.4? If I continue with the same form as for the lower new breakpoints, I’d use AQI = ((999-501)/(99999.9-325.5))*(pm2.5-325.5)+501 for 325.5 and above. Due to the large value of 99999.9, this equation has a very low slope above 325.4, essentially capping AQI a little above 501. However, I’m seeing some sensors reporting AQI’s in the 700s and even higher.

5

I was able to find a couple of sensors on the map with very high AQIs, well above 501. If I compared these high AQIs with their corresponding pm2.5 values, they clearly followed AQI = (500 - 301)/(325.4 - 225.5)*(pm2.5 - 225.5) + 301. That is, the slope from 225.5 to 325.4 continued beyond 325.4. A breakpoint at 325.4/5 was not apparent for these sensors. Do others agree?