Question:How long do you need to co-locate an air quality tool to get what you need to calibrate or "ground truth" another tool?

stevie is asking a question about calibration: Follow this topic

by stevie | December 07, 2021 20:50 | #28339

This came up in Open Call today!!


We were in an medical environment, usually class 100,000. These instruments are different from the ones used by public lab, but based on the same principles. Normally, the instruments would get about 10 minutes warm up time. Then we would do a "sanity check". We'd put a 0.25 micron air filter on the air input and see what the instrument read. It should read zero. If not, there was a calibration problem. Now there are differences between the instruments we used and the instruments used here. In the medical field, the instruments normally have pumps that pull the air through. Also, calibration was required every six months to guarantee the particles were sized correctly. This is not cheap. Hope this helps answer the question. Regards!

Reply to this comment...

I think it all depends on the sampling rate and whether the data converge, ha.

For electronic monitors sampling every few seconds, two weeks isn't bad. I think a month won't hurt. But it also depends on what happens in that month.

Ideally, the "real" PM2.5 readings cover the full range of what is possible during your sampling period, so, in Louisiana, for example, I would run a co-location during Sugar Cane harvest season, to make sure the meters read at similar rates during cane burns, when PM swings are large and dramatic.

With Particulate Matter, there are usually issues with non-convergence at the lower readings. It's either good or bad that different units seem disagree more at lower PM levels. Consult your expert.

Reply to this comment...

From this Guidebook for Developing a Community Air Monitoring Network: If possible, permanently install a community monitor at the colocation site so your conversion equation can be kept up to date. Particle composition in an area may change over time, so the conversion equation should be updated regularly to account for this. If ongoing colocation isn’t feasible, the colocation should last for a year if possible so that seasonal differences can be accounted for. If this is not possible, then the community air monitor should be colocated for one month during each season.

The EPA Colocation Instruction Guide doesn't specify how long to colocate the sensors, but they recommend setting the reading frequency of the low-cost sensor to at least as frequent as the reference sensor.

Great info! thank you!

Reply to this comment...

@fongvania @stevie Thank you for the EPA Colocation Instruction Guide slide deck. @TravisLondon The info linked above in the slide deck does a great job of explaining precision and accuracy in data for air quality monitors, and how you can use existing high-quality monitors owned and maintained (typically) by existing federal or state agencies as reference monitors; you compare data from the reference AQ monitors to data from the DYI or other monitor you want to use for field use by collocating them: putting them together in the same place at the same time so they're both analyzing the same air. In an ideal situation, the DYI or other gives the same data as the reference.

Reply to this comment...

Log in to comment