I have also posted this on the wxforum, but lets see if anyone here has the answer...
My question is specifically about the UK and their COMEAP Air Quality Index (AQI) scale, but it probably applies more widely - though the answers may be different.
AQI values for the UK are a 24 hour measurement, and for any moment in time we can quote a rolling 24 hour number for the AQI.
But when the AQI is either forecast for future days, or recorded for past days what is the value that is quoted?
The highest rolling 24 hour AQI value of the day?
The mean of the rolling AQI values for the day?
The AQI for the 24 hours of that day - i.e. the value at 23:59?
The AQI at midday?
or...?
Welcome to the Cumulus Support forum.
Latest Cumulus MX V3 release 3.28.6 (build 3283) - 21 March 2024
Cumulus MX V4 beta test release 4.0.0 (build 4018) - 28 March 2024
Legacy Cumulus 1 release v1.9.4 (build 1099) - 28 November 2014 (a patch is available for 1.9.4 build 1099 that extends the date range of drop-down menus to 2030)
Download the Software (Cumulus MX / Cumulus 1 and other related items) from the Wiki
Latest Cumulus MX V3 release 3.28.6 (build 3283) - 21 March 2024
Cumulus MX V4 beta test release 4.0.0 (build 4018) - 28 March 2024
Legacy Cumulus 1 release v1.9.4 (build 1099) - 28 November 2014 (a patch is available for 1.9.4 build 1099 that extends the date range of drop-down menus to 2030)
Download the Software (Cumulus MX / Cumulus 1 and other related items) from the Wiki
Air Quality Measurements
-
- Posts: 1183
- Joined: Fri 27 Jul 2012 11:29 am
- Weather Station: Chas O, Maplin N96FY, N25FR
- Operating System: rPi 3B+ with Buster (full)
Re: Air Quality Measurements
Should be...
https://assets.publishing.service.gov.u ... _index.pdf
(page 17) that 24 hour mean at end of day is what should be reported in UK for past days. I have not seen any mention of the time at which a day ends, so it is reasonable to assume that past days are reported by calendar days. Again, the same reference does not say explicitly what is used for future forecasts, but it is reasonable to assume these are also calendar days.
In the late 1970s, when I worked in air pollution, we used filters for measuring particulate pollution, and the colour of the filter gave the reading (against a standard scale). Some sites, if my memory is correct, had mechanisms for controlling which set of filters was exposed, so the figures probably were by calendar day. Some relied on manual action to change the filter, so these were actually left alone overnight and the change time might be 9am or 10am, so the figures were not by calendar day. Obviously back then we could not produce real-time numbers!
[In case it is of interest to anybody, back then, for gaseous pollutants we reported the 98th percentile (selected to eliminate the maximum just in case it was an outlier, i.e. false) of all reported measurements arranged not by time of day but in ascending order in calendar day. This was new at the time, having previously reported the median, which also required sorting by magnitude.
I did not actually work in the team reporting daily values, my team reported monthly and annual values (except for Pb where in those days before lead-free petrol we reported weekly values); our computers back then were not powerful enough to sort all reported measurements for monthly or annual periods, so for these we had a finite number of ranges that measurements were assigned to, so the 98th percentile for these longer periods was an approximation]
It is made very clear in
https://assets.publishing.service.gov.u ... _index.pdf
(page 17) that 24 hour mean at end of day is what should be reported in UK for past days. I have not seen any mention of the time at which a day ends, so it is reasonable to assume that past days are reported by calendar days. Again, the same reference does not say explicitly what is used for future forecasts, but it is reasonable to assume these are also calendar days.
In the late 1970s, when I worked in air pollution, we used filters for measuring particulate pollution, and the colour of the filter gave the reading (against a standard scale). Some sites, if my memory is correct, had mechanisms for controlling which set of filters was exposed, so the figures probably were by calendar day. Some relied on manual action to change the filter, so these were actually left alone overnight and the change time might be 9am or 10am, so the figures were not by calendar day. Obviously back then we could not produce real-time numbers!
[In case it is of interest to anybody, back then, for gaseous pollutants we reported the 98th percentile (selected to eliminate the maximum just in case it was an outlier, i.e. false) of all reported measurements arranged not by time of day but in ascending order in calendar day. This was new at the time, having previously reported the median, which also required sorting by magnitude.
I did not actually work in the team reporting daily values, my team reported monthly and annual values (except for Pb where in those days before lead-free petrol we reported weekly values); our computers back then were not powerful enough to sort all reported measurements for monthly or annual periods, so for these we had a finite number of ranges that measurements were assigned to, so the 98th percentile for these longer periods was an approximation]
- HansR
- Posts: 5874
- Joined: Sat 20 Oct 2012 6:53 am
- Weather Station: GW1100 (WS80/WH40)
- Operating System: Raspberry OS/Bullseye
- Location: Wagenborgen (NL)
- Contact:
Re: Air Quality Measurements
I am not sure it would be useful to use the 24hr average as a prediction.
Most predictions nowadays come from satellite and models and there is a huge science around it. So making even the suggestion that a 24hr av value or a derivative from there would be a prediction I think is not a correct signal.
I will give some links below for a view on the problem and techniques used
- https://www.hindawi.com/journals/amete/2019/1279565/
- https://www.sciencedirect.com/science/a ... 9719334813
- https://www.sciencedirect.com/science/a ... 0920307249
- https://aqicn.org/forecast/world/
Anyway, single station values have no predictive value is my point. In that sense the 24 hr value is locally of little interest it is the 1 or 3 hr value which dampen the actual values a bit which help to evaluate when you got to run or stay and close windows and doors. I assume that answers your question a bit?
Hans
https://meteo-wagenborgen.nl
CMX build 4017+ ● RPi 3B+ ● Raspbian Linux 6.1.21-v7+ armv7l ● dotnet 8.0.3
https://meteo-wagenborgen.nl
CMX build 4017+ ● RPi 3B+ ● Raspbian Linux 6.1.21-v7+ armv7l ● dotnet 8.0.3
- mcrossley
- Posts: 12697
- Joined: Thu 07 Jan 2010 9:44 pm
- Weather Station: Davis VP2/WLL
- Operating System: Bullseye Lite rPi
- Location: Wilmslow, Cheshire, UK
- Contact:
Re: Air Quality Measurements
Well, yes, but the UK AQI scale like many of the others is based on a 24 hour average, so a 3 hour figure whilst of interest is of no value to compare against official figures. The premise being that exposure to short term pollution is less detrimental to health than the exposure over the longer 24 hour period. The paper referenced above recognises that using a 24 hour value has some short comings, like their example of the evening of 5th November, but that is what has been chosen as the standard so it is what I will use for my daily figures.
I do also graph the 3 hour value for what it is worth on my web site alongside the raw readings, so I can see both short term and 24 hour values.
I do also graph the 3 hour value for what it is worth on my web site alongside the raw readings, so I can see both short term and 24 hour values.
- HansR
- Posts: 5874
- Joined: Sat 20 Oct 2012 6:53 am
- Weather Station: GW1100 (WS80/WH40)
- Operating System: Raspberry OS/Bullseye
- Location: Wagenborgen (NL)
- Contact:
Re: Air Quality Measurements
True, but I was more on the predictive value of the 24 hr av.mcrossley wrote: ↑Tue 08 Dec 2020 11:45 pm Well, yes, but the UK AQI scale like many of the others is based on a 24 hour average, so a 3 hour figure whilst of interest is of no value to compare against official figures. The premise being that exposure to short term pollution is less detrimental to health than the exposure over the longer 24 hour period. The paper referenced above recognises that using a 24 hour value has some short comings, like their example of the evening of 5th November, but that is what has been chosen as the standard so it is what I will use for my daily figures.
I do also graph the 3 hour value for what it is worth on my web site alongside the raw readings, so I can see both short term and 24 hour values.
No matter what the standard of the country is, the 24 hr av. does not contain predictive value.
Hans
https://meteo-wagenborgen.nl
CMX build 4017+ ● RPi 3B+ ● Raspbian Linux 6.1.21-v7+ armv7l ● dotnet 8.0.3
https://meteo-wagenborgen.nl
CMX build 4017+ ● RPi 3B+ ● Raspbian Linux 6.1.21-v7+ armv7l ● dotnet 8.0.3
- mcrossley
- Posts: 12697
- Joined: Thu 07 Jan 2010 9:44 pm
- Weather Station: Davis VP2/WLL
- Operating System: Bullseye Lite rPi
- Location: Wilmslow, Cheshire, UK
- Contact:
Re: Air Quality Measurements
Agreed. But I do not think I implied that did I? I just asked how they arrive at the AQI value for a day - and the predicted value for a day presumably must be the same calculation. How they arrive at predicted values I have no idea but obvious things like the expected weather conditions will be the major factor in high values (no wind, high pressure system, cold etc)
- HansR
- Posts: 5874
- Joined: Sat 20 Oct 2012 6:53 am
- Weather Station: GW1100 (WS80/WH40)
- Operating System: Raspberry OS/Bullseye
- Location: Wagenborgen (NL)
- Contact:
Re: Air Quality Measurements
Well you started with
But again, non-native English speaker etc...
that looks pretty predictive to me.But when the AQI is either forecast for future days, ...
But again, non-native English speaker etc...
Hans
https://meteo-wagenborgen.nl
CMX build 4017+ ● RPi 3B+ ● Raspbian Linux 6.1.21-v7+ armv7l ● dotnet 8.0.3
https://meteo-wagenborgen.nl
CMX build 4017+ ● RPi 3B+ ● Raspbian Linux 6.1.21-v7+ armv7l ● dotnet 8.0.3
-
- Posts: 884
- Joined: Sat 16 Jul 2016 11:59 pm
- Weather Station: Davis VP2+ & GW1000 (Standalone)
- Operating System: Win10 Pro / rPi Buster
- Location: Australia
Re: Air Quality Measurements
Something I've pondered before, regards air quality graphs is the inclusion of some sort of wind information.
Here I know two basic sources of particle matter.
One to the West @ 5km, the other around North @ under 1km.
The sort of thing that would be useful would be a wind indicator on the top border of the graph....
Bit like this, which is also colour coded for speed.
Now that would make another great @Beteljuice Play Thing......
Cheers.
Here I know two basic sources of particle matter.
One to the West @ 5km, the other around North @ under 1km.
The sort of thing that would be useful would be a wind indicator on the top border of the graph....
Bit like this, which is also colour coded for speed.
Now that would make another great @Beteljuice Play Thing......
Cheers.
You do not have the required permissions to view the files attached to this post.
:Now: :Today/Yesterday:
Main Station Davis VP2+ Running Via Win10 Pro.
Secondary Stations, Ecowitt HP2551/GW1000 Via rPi 3 & 4 Running Buster GUI.
:Local Inverell Ecowitt Station: :Remote Ashford Ecowitt Station:
Main Station Davis VP2+ Running Via Win10 Pro.
Secondary Stations, Ecowitt HP2551/GW1000 Via rPi 3 & 4 Running Buster GUI.
:Local Inverell Ecowitt Station: :Remote Ashford Ecowitt Station:
-
- Posts: 884
- Joined: Sat 16 Jul 2016 11:59 pm
- Weather Station: Davis VP2+ & GW1000 (Standalone)
- Operating System: Win10 Pro / rPi Buster
- Location: Australia
Re: Air Quality Measurements
Just found the Legends for the above.
You do not have the required permissions to view the files attached to this post.
:Now: :Today/Yesterday:
Main Station Davis VP2+ Running Via Win10 Pro.
Secondary Stations, Ecowitt HP2551/GW1000 Via rPi 3 & 4 Running Buster GUI.
:Local Inverell Ecowitt Station: :Remote Ashford Ecowitt Station:
Main Station Davis VP2+ Running Via Win10 Pro.
Secondary Stations, Ecowitt HP2551/GW1000 Via rPi 3 & 4 Running Buster GUI.
:Local Inverell Ecowitt Station: :Remote Ashford Ecowitt Station: