I’m trying figure out why the Canadian scale seems calibrated so differently from the rest of the world. I understand that it uses a different algorithm, but there’s not a lot of transparency on how it’s different.
Two big questions:
Comparing the Purple Air Canadian AQHI and the Environment Canada AQHI is concerning. Purple Air suggests it’s a 3 (“low risk”) outside, and EC says 7. The EC data is a couple hours old, but two hours ago it was marginally worse out there, not better. What’s happening here?
Purple Air is bright red when in AQI (130-150), but blue / low risk when in AQHI (3). What is happening here?
Hello Lee, I had posted this in response to a similar question from a fellow Canadian. I will repeat it here.
I had the same question living in Regina, SK. There are a couple of reasons the PA map calculation of the Cdn AQHI does not seem to match up with what our official agency Environment Cda reports. The AQHI is calculated based on the combined effect of 3 components (PM2.5, Ground Level Ozone, Nitrogen Dioxide). See Air Quality Health Index (Canada) - Wikipedia . The PA sensors only measure PM and that is all the PA map is using to calculate the AQHI so it will always appear too low compared to what ECCC reports. The second difference is that many provinces including BC/SK and maybe ALTA also use a variation of the AQHI they call AQHI-Plus during the summer months when forest fire smoke is the dominate AQ concern. That formula is much simpler. Wildfire Season AQHI-Plus = ceiling value (the closest integer greater than or equal to a given number) of the 1-hour PM2.5 concentration divided by 10. I BC like SK use the AQHI+ when the PM2.5 reaches 51 ug/m3. Alberta uses a different threshold (maybe 81ug/m3 or none at all?) Therefore in BC/SK PM2.5=51 will be reported as AQHI 6…61 would be 7 and so on. See https://www.tandfonline.com/doi/full/10.1080/10962247.2020.1797927 .
I prefer to switch the map to show me Raw PM2.5 concentration and convert using the AQHI+ formula.
Again for Alberta I have found the threshold to switch to the AQHI+ is higher or not at all. I also use the ALT CF=3 conversion factor as the PA devices mostly over estimate PM and I have found that setting more closely matches official sources. Play around with the different settings. Lots of good data to play with.
One other good way to compare the PA devices to official sensors in Canada is to use this AQmap (EN). Originates with UNBC. It shows PM2.5 only but a good resources to see how PA stacks up against official (FEM) sensors.
Sounds good Andrew. Any timelines? Available for this fire season? Also will there be any notification as to when that change happens on the map? I am interested to see how well it works.
We want to release the new map version soon, although we do not have a strict timeline. If you would like, we can contact you when it is released. Does that sound agreeable?
Thanks so much for the detailed information. Just to clarify… if I live in Alberta, have asthma, and want to determine when I need to wear a mask outside in the summer months, I need to use the ALT CF=3 conversion factor and then check the Raw PM2.5 concentration. I can then calculate the AQHI-Plus value and use this number to determine if the risk is low, moderate, high, or very high. Is this correct? This would be a better measurement than using the Canadian AQHI because it takes into account the wildfire smoke?
Now that Alberta has joined most of the rest of Canada, yes using the Alt CF=3 conversion/10 and round up to the nearest integer will certainly get you close to an accurate AQHI+ value for your location. A conversion/correction factor developed in and for Canada specifically can be found at AQmap. The UNBC map displays the PM2.5 count only… Just make sure you select to use the corrected data not the raw data. I believe corrected data is the default.
It’s me again. I just wanted to make sure I understand which raw data you’re talking about. It is the Raw PM2.5 at the bottom of the drop down menu that I should be using to make my calculation, correct?
Kieran, Thanks for the update. I see you utilized the paper by Nilson et.al for the ECCC conversion. That’s great. This will be most helpful to the users of your data here in Canada.
Monique, Apologies for the late reply But now that PA has updated the AQHI layer there is less of a need to use the raw PM2.5 data you were referring to. You should now be able to just select the Canadian AQHI+ and apply the Canada ECCC conversion. The conversion they are using is well known and trusted to convert the PA2.5 values to AQHI+. Keep in mind the AQHI+ value is not taking into account Ozone and Nitrogen Dioxide which the full AQHI uses. However we know that PM2.5 is what most people concern themselves with lately.
