Gino–
Yes I think the EPA conversion would be best, particularly on days of high humidity. Is there an equation including an RH factor available for the Xiaomi Mi? It would be interesting to compare the EPA and Xiaomi Mi equations.
1 Like
I DONT KNOW ABOUT XIAOMI MI IS CHINESSE… But i think that its a good air quality monitor. His marks are THE SAME with CONVERSION US EPA (purple air)
1 Like
Hi Gino–
I looked up the official website for XIAOMI (https://www.mi.com/global/ ) but could not find the air monitor that you are using. Perhaps they do not offer it any more?
1 Like
Hi, sorry for the delay in responding. I was checking and Xiaomi has changed the hardware on its meters. The old meter I have measures exactly the US EPA, and the new Xiaomi hardware (partnered with Qinping) leaves much to be desired, providing a very low PM2.5 measurement.
Thank you very much for responding. My Twitter account is @ClimaPeruMundo
Best regards.
Gino.
1 Like
Hey @Lance! Firstly, thank you for your work; having an experienced scientist doing work like this as an independent researcher is such a wonderful boon! Seeing such valuable data generated from such a low-cost setup really encourages me—and, I hope, others—to pursue passion projects and share their information (presuming it’s properly documented, gathered, and presented, of course).
Anyway, I’m writing because I had a read through of your helper doc after being surprised by the RH levels and variation I saw on the map of my local area (Mountain View, CA). I’ve been using custom microcontroller circuits with things like the grove DHT 22 and a cheap AliExpress KY-015 which I think uses the DHT11, to try to gather data on the levels in my poorly-ventilated apartment for medical purposes. I thought, “well, I can try to use the near-ish purpleair and official weather stations to compare to my data when I read at the window and what my weather app is saying, to try to get a good idea of what’s happening.” Of course when I saw the readings, I realized it was not to be a simple comparison, so I’ve started to dive deep, and now have all sorts of questions about sensor technology, variability of humidity across an area, heat transfer, and plenty more that I will be exploring, BUT, for you in particular, I have a few pointed questions, if you’d allow me to take even more of your time than I already have with my verbosity…
- Have you done any work outside what you published in these documents on calibration/characterization of T/RH? I’m particularly wondering if you’ve done any saturated salt solution tests. (I would do some myself, but I am currently unable to afford any equipment
) - Given the extreme variance in the RH in particular that you documented, would you recommend just not using it from the purpleair sensors? If so, have you any recommendations for other local monitoring of RH? Are these PWS’s that WeatherUnderground uses, for instance (they seem to be by AmbientWeather and seem like they use similar enclosures, so I don’t have high hopes) a bit more consistent?
- Or, is humidity just the kind of thing that is so localized that you’re not going to get a good idea of the local outdoor condition from a monitor a few blocks away?
- As an aside, in regard to your discussion on the RH influence on AQI, particularly the typical indoor RH levels, I think they may be higher than you mentioned expecting in many situations, though I currently only have my own data to show for it. I’m not sure if this has been well-studied, but in this area and others where AC is rare, I certainly see a clear path to high indoor humidity. Whether or not that actually makes a meaningful difference to algorithmic design, of course, is complicated, as you have discussed.
Sorry for the exceedingly long post! If you think this very off topic, feel free to let me know and I can move to email [staffa at alumni dot stanford dot edu]! Again, thank you for all of your work and being so available to the community!