Comparing indoor and outdoor monitors

jaj · February 16, 2025, 6:05pm

I have two indoor monitors (I-9PSL) and one outdoor monitor (O-1PST) which I have been putting together over the past few weeks. Now I have all three set up and running right next to each other indoors. The two indoor monitors seem to have good agreement within expected uncertainties for PM2.5, temp, RH, CO2, TVOC and NOx. The outdoor monitor, the last to be set up just a couple of days ago, however, does not. All three are within a few cm of each other, with the two indoor monitors on a table mounted with the supplied feet to ensure airflow around it, and the outdoor one just above them on the wall mounted on the screw holes with the vent facing downwards as it should be. I still need to set up my outdoor installation, so I thought for the time being I could compare all their readings indoors. All three are on firmware version 3.1.21

I know that the temperature and RH values for the outdoor monitor have corrections applied assuming outdoor use, so I am not particularly concerned about those. And I do not have an outdoors reference to check it against later anyway so I will just have to accept whatever values it gives. The CO2 value also will take a few days for the 8-day rolling average to be fully evaluated. But the PM2.5 value, although clearly within the ±10µg/m3 at 0-100µg/m3, I was expecting to show closer agreement. Same goes for the TVOC.

Is there some similar outdoor correction applied to PM2.5 measurements too that I am unaware of? Or am I expecting too much and is this kind of deviation normal?

Just in case it is relevant:
I am aware that airgradient have developed corrections for certain batches of the PMS sensor. My two indoor monitors have the relevant correction applied for their batch numbers + the EPA correction on top. My outdoor PMS sensor is not one of those affected batches, so it only has the EPA correction applied (though I do not see the option to apply a batch-level correction on the dashboard for the outdoor monitor anyway, not sure why not).

nickf1227 · February 16, 2025, 8:32pm

I know the PM sensor in indoor units have gotten ALOT of love by the AirGradient team over the past 6 months.
My understanding is the sensor is basically the same between the two, pms5003t vs pms5003, with the interface being slightly differant. But I could be wrong.

But based on the work by AirGradient folks for the pms5003, I would not be surprised if there are just some minor variances/differances in the batches of pms5003ts that haven’t yet been identified.

It is intersting that the particle count is so differant between the outdoor/indoor units though. Some of this may just be due to the effects of the airflow patterns being very differant in the different enclosues, but I wouldn’t begin to guess how much that would matter.

TVOC may just need some more time to calibrate. I think the default 7 day one isn’t perfect, and I increased “tvocLearningOffset” to 120 on mine. I think given time they would drift closer together.

jaj · February 16, 2025, 9:57pm

Thanks Nick.
Yes I also understand, though perhaps mistakenly, that the 5003 and the 5003T are essentially the same in terms of characteristics.

Regarding the count, you are right - in fact I was so fixated on the PM2.5 and TVOC numbers that it was only after I hit Post on my original post above that I noticed the 0.3, 1 and 10 numbers were hugely different. Though I remember an article here on the AirGradient site saying that in these sensors, the other PM numbers are more unreliable than the PM2.5 one. And the PM1 and PM2.5 numbers are usually correlated, indeed the PM2.5 numbers usually include the PM1 count if I remember correctly. So, I probably should not give much weight to those other PM numbers other than PM2.5.

Otherwise, I guess you are right, it is a case of waiting to see if any measuring discrepancies are found between the two. I did notice that I had my outdoor monitor set to Local control in the HA integration, but it had the same configuration as my cloud configuration in the AirGradient dashboard so I don’t think that should have made a difference. I think when i set up the integration it automatically changes it to Local. Anyway, I changed it back to Cloud (as my indoor monitors are) shortly after my initial post above, but still see similarly different numbers for PM2.5. But I will keep an eye on that over the coming days.