Calibration of PMS7003

FAQ
1

Hi I’ve just set up an new sensor outside my house. It’s using a PMS7003 particulate sensor.

I’ve noticed that it is the highest reading in my city. Currently reading PM2.5 of 56 ug/m^3, when the 3 nearest nodes average 34 ug/m^3. I suspect that this is because I am using a PMS 7003 and the others (probably) have different sensors.

Reading up on this subject:

This study tested 3 PMS7003 and concluded that they were very consistant with each other (high reproducability), but that they over estimated PM 2.5 (by quite a lot, possibly ~43%).
Which does seem to fit my current experience.

I might be able to persuade a local university to help me calibrate, but until then:

1 Does anyone have a “standard” code or method to post process the raw data from PMS7003 or other Particulate sensors to apply generic calibration to improve accuracy before uploading?
2 Is doing this even a good idea?
3 Would it be smarter to buy and SDS011?

For info - The rest of the set up is a Pi Zero, BME280 and the code is a simplified version of the Pimoroni Enviro+ code.

Thank you in advance for you help,
Drew

2

I’m interested in this as well.
I have just bought a PMS7003 and freaked out that PM2.5 is around 37 ug/m3, which is about AQI 110
I also have a handheld SmartMi PM2.5 sensor for years, and it was showing AQI 25.
image
Not sure how to get a correct value now.

3

Welcome!
No one answered on the first message actually…
So in my opinion, there is a better solution than the individual calibration: in the Netherlans, the RIVM developed a spreading correction algorithm, which change the raw measurements.

4

Can you elaborate on this RVIM algorithm?

5

This is explained at Kalibratie van fijnstofsensoren | Samen meten aan luchtkwaliteit (in Dutch, Google translate does a decent job translating to English).

A scientific paper is available at Atmosphere | Free Full-Text | Development and Implementation of a Platform for Public Information on Air Quality, Sensor Measurements, and Citizen Science

1 Like