This is a testing site only. See the live Public Lab site here »

# Question:What is the correct AQI measure ?

by sagarpreet |

I have been working on showing data from PurpleAir Layer onto a map in the form of a HeatMap .

I have the value of PM2.5 , should i convert it to US-Based value AQI ? Will that make more sense ?

My work on it so far : https://github.com/publiclab/leaflet-environmental-layers/issues/14

Heat Map looks like this if i am using PM2.5 Value :

The conversion can be done using this formula [source: https://en.wikipedia.org/wiki/Air_quality_index]:

This is an awesome post by the way @sagarpreet!!!

Thank you :)

I used the formula given on the pic above ðŸ˜€ . Now the heat map looks like this :

Hi @sagarpreet nice work. I think is better to leave the PM2.5 value because the AQI refers to a combination of pollutants. They https://www.facebook.com/InversionDoc/ are using PurpleAir and show an AQI that is the combination of O3 and PM2.5

Hi @sagarpreet nice work. I think is better to leave the PM2.5 value because the AQI refers to a combination of pollutants. They https://www.facebook.com/InversionDoc/ are using PurpleAir and show an AQI that is the combination of O3 and PM2.5

hi @sagarpreet , could you provide a link to your source for the image of computing the AQI? thanks!

Is this a question? Click here to post it to the Questions page.

Hi @liz ,

The image source of the above formulae is : https://en.wikipedia.org/wiki/Air_quality_index

The map made by purpleair community is here : https://www.purpleair.com/map#2/37.7/8.54 , which i think uses the same formula . Thank you !

Thank you!

I wonder if @marlokeno, @nanocastro, @imvec, @xose, @zengirl2, @JamesWarren, @the_butler, @guolivar, or @DrJessie may be interested in this post or have some input.

Folks, this is related to @sagarpreet's work to develop a common environmental layers mapping library: https://publiclab.org/notes/sagarpreet/06-06-2018/leaflet-environmental-layer-library -- i'm sure he'd love input!

Also, this looks like an interesting related post! https://publiclab.org/notes/Zengirl2/04-20-2017/epa-s-beta-c-tools-community-tools-for-modeling-near-road-or-port-air-quality

This is interesting. What unit is PM2.5 in? If it's ug/m3 then I think it's best to not change it because ug/m3 is one of the most widespread measures used around the world, I believe.

Is this a question? Click here to post it to the Questions page.

Hi @sagarpreet nice work. I think is better to leave the PM2.5 value because the AQI refers to a combination of pollutants. They https://www.facebook.com/InversionDoc/ are using PurpleAir and show an AQI that is the combination of O3 and PM2.5

Yes it makes sense...PM2.5 value seems to be more appropriate . Thank you :)

Hi , I have 2 options for making heatmap :

1.) Either use global maxima and global minima value (of PM2.5 in ug/m3) to make heat map .

2.) Or to use only the values which are in the current viewport of the map .

Which one makes more sense ?

Is this a question? Click here to post it to the Questions page.

I just reviewed this conversation and would like to add to it and would like comments from @sagarpreet and @guolivar if possible. About the USEPA AQI: if you look again at the table and formula @sagarpreet posted above you will notice the column for pm2.5 refers to a 24 hr average. If the point is to inform about the current air quality someone might breath in, a 24 hour average based upon the past 24 hours is not useful.Particulate in the past is not what will enter your breathing now. The pm2.5 standard is based upon a 24 hour average and to keep within the context of the standards, against which all the "health" research has been based, the EPA AQI uses an average of the past 12 hours with an estimation (prediction) of the next 12 hours to assess the air NOW. If you are not able to do that prediction part, you will be misappropriating the risk (health) aspect of the formula and possibly misinforming those you want to communicate with. If you are not able to do the prediction part I maintain you should provide information stating you are using a modified and unassessed version of the USEPA AQI. My suspicion is that PurpleAir does not do this prediction part and, as many people do, takes the most current reading or average over a short period of time, and plugs it into the formula in place of the correct 24 hour past and future average. Whether you use the AQI or particulate density this addresses the color you use and its risk assessment (or actually misappropriation of risk assessment!)

Thanks for reply :) Yes the unit is in ug/m3 .

Can someone also help me with this ?

I have 2 options for making heatmap : 1.) Either use global maxima and global minima value (of PM2.5 in ug/m3) to make heat map .

2.) Or to use only the values which are in the current viewport of the map .

Is this a question? Click here to post it to the Questions page.

Great work @sagarpreet!

This might get rambly because there is no right answer but I'll try to explain why and what then.

The first thing you need to understand is that all AQ indices are designed for COMMUNICATION purposes so the question is not "should I report concentrations or indices?" but rather "what am I trying to communicate?"

There are many indices and all of them serve different purposes. Some are "regulation" related so they refer to the corresponding regulation limits. These are pollutant dependent so you can have a high index for PM2.5 and a low index for Ozone at the same time which will mean that concentrations of PM2.5 are near or above the standard (whatever that is) while Ozone concentrations are well below the standard.

Now, there is an argument that the standards have very little meaning for the general public as their meaning is unclear and they are defined on many different time bases. So, that's where the "health effect based" indices come in. They try to communicate "risk to people" rather than "state of the air" so a high index value would mean that there is a high risk of a health impact (whatever that may be) with these conditions... that's what the AQI tries to do

In general, if you want people to do somehting, show them something they can understand so AQI's are much more useful for that BUT you need to be VERY careful with the averaging times you're using and you need to be very transparent on where you got the calculations from because if you tell people that their air is OK but it isn't, you're in trouble.

So ... decide what you want to communicate and run with it documenting your sources.

Is this a question? Click here to post it to the Questions page.