Cumulus Support

https://en.wikipedia.org/wiki/Segal's_law

I am a person with more than one Air Quality sensor and trying to determine if the air is Excellent, Good, Moderate, or whatever
My Davis AirLink Weatherlink.com says: AQI = 0.06; PM1 = 1.8 µg/m³; PM2.5 = 1.3 µg/m³; PM10 = 1.3 µg/m³; and using the Index: Canada AQH1
One Ecowitt WH41 says: AQI = 38 Good; PM2.5 = 9 µg/m³; Don't know what Index Ecowitt uses...
Another WH41 says: AQI = 63 Moderate; PM2.5 = 18 µg/m³

All 3 are with a foot of each other.


Good breathing,
Paul

Respect that you have three AQ sensors!
And I do agree that air quality is a confusing field of research.

Especially the AQIs by the different countries with their different interpretations and make belief it is hard science. Which it isn't. The actual concentration of the particulate matter concentrations can be measured and easily represented, focussing on the variation between devices. Calibration of the devices can be done but is a lot of work. sensor.community has taken the approach of the theory of large numbers and in Western Europe there is no a high density of low cost sensors giving a good idea of where the pollution is (I wrote two posts on the subject).

But I agree that the variation between your three devices is big. Too big. One issue may be the maintenance. Cleaning the devices every 6 month is useful. Blowing through them is one method (compressed air, or blow yourself. Not too high pressure). Opening them and cleaning piece by piece sometimes can be done. Some handy ability required.

A good example of the sensor.community approach could be seen yesterday morning 10 September in the area where I live. Two images, one zoomed in, one the context of western Europe. Now search for the origin of the pollution. And believe me: I could feel it on my lungs and my nose. This was real. No talk about cheap worthless equipment necessary. Just concentrations. Source: https://sensor.community, click on a sensor to see its recent history.

Thanks Hans, you are much more aware and understanding than I.
You may be right on the sensor deteriorating over time. I inherited the WH41 from testing in 2018 (one 433MHz and the other 915MHz). One was outside for a couple of years until the battery died... the other used for less than a month. Recently got them reconnected to two separate GW1000s. Taking electronics apart is not my foray...

So why AirLink AQI mostly does not go over 1.0 but the WH41 is in the area of 50 plus or minus. The WH41 PM2.5 is around 10 +/- while AirLink is usually much lower µg/m³ http://www.komokaweather.com/cumulusmx/ ... airquality

@Hans I've just opened your November 2020 article and reading that should help to better understand.

Enjoy,
Paul

No offense to Ecowitt, but their WH41 is dreadful, despite the 'honeywell sensor' claim. If the outside air has high moisture content, forget it! The WH41 will give you a reading that's less useful than an ashtray on a motorbike.

High (and very low) moisture is a killer for all cheap PM sensors. That is why calibration through MLR for sensors (types?) helps (see here). But that is a lot of work and at the extremes still does not really work. But what to expect from € 20 sensors where scientific acceptBle sensors are in the range of 100 times that amount?

You may remember Thread / post that ruffled a few feathers at the time, but I think now perhaps is better accepted ?

@beteljuice: ah yes, that was one of the better threads on this.

HansR wrote: ↑Sat 11 Sep 2021 6:44 pm One issue may be the maintenance. Cleaning the devices every 6 month is useful. Blowing through them is one method (compressed air, or blow yourself. Not too high pressure). Opening them and cleaning piece by piece sometimes can be done. Some handy ability required.

In addition to the above what I wrote about cleaning AQ sensors i forgot to show you what I use to blow them through every now and then. And don't wait too long because dust may form a crust in the laser beam and may require more difficult operation to clean.

As a 'bump' to Air Quality measurement equipment in general ....

Some info I gleaned from another forum

Gaia Air Quality Module.

Interesting concept, "Buy as a hardware service" - Everything is ready to go (THREE sensors).
The data gets sent to them, sanitised and available to you.
You pay monthly or yearly (10% discount) and the unit will be replaced every TWO years.
Said to be as accurate as profession kit, although it doesn't mention heaters. Perhaps they have algorithm for RH > 70%

Some useful info on the site, include distance placement from roads according to traffic density.

Netigo pro kit - there is also a pre-soldered version (assembly only required)

Has a heater, and the pro has an external antenna for greater wifi connectivity. but only seems to be a single sensor - approx 80 Euro (base price)

Ha, you're just ahead of me and repaired the link.

Nice though: about the same cost as an AirLink (every two years, which I thought has two sensors, at least the PurpleAir has) and with some kind of triple differential measurement. Nice spot.

But those chinese simply take the sensor readings for their map from the sensor.community site as my sensor is on it and the only place I drop the data is sensor.community. So don't think they are all that big. Data stolen! So

Perhaps they have algorithm for RH > 70%

No I don't think they will, it is just raw data from the sensor.community.

The NAM (Nettigo) is I believe a much improved version of the sensor.community DIY kit - using the same sensor - and this one has tried to overcome the 70+% humidity situation.

The GAIA (? not sure) and the NAM automatically send data to sensor.community. It would be nice if there would be some generic method in CMX to read/log/display the data of this type of sensors but I think that is a step or two too far as there is no direct interfacing to these. You would have to make that yourself. The only way I know now is the webserver interface of the AirLink but maybe the new http://... interface might be a possibility for the future. I did not look at it that way.

With the http interface you might be able to get the data from sensor.community and store it locally as a device belonging to your weatherstation. A bit like taking the weather from a METAR which is done by some. It could be a CMX feature, it could be a template feature.

Anyway.... I get carried away a bit. GAIA and NAM: nice alternatives. NAM I like best.

Just bought the NAM sensor btw: € 88,- really cheap with a nice device to compensate humidity (essential for where I live where HUM did not fall below 70% this year) and a nice box to pack it and mount it outside (in my Stevenson screen ). I hope to finish it all coming spring and report back when all done.

Re. GAIA

The pages are a bit disjointed, but it looks like their data is pushed to air-net and that is where all the multiple sources are coming from.

They have their own APIs, but they seem to interrogate air-net,
ie' Any source station

Here is Hans station - scroll down the page for extra info / accreditation.

I am not sure the data is pushed to air-net. Pull is more likely I think. Chinese polling the worlds air quality
Anyway, thats my station all right. I'll look i to that air-net but i do not mind. One reason for the sensor is to use it for the aggregtion like this (or sensor.community itself). It does have added value.

They must be pulling is or sensor.community is pushing every sensor forward because the API of air-net is different from the sensor.community.

Another minor thing is that air-net is only using the AQI of the USA for pm2.5 without any reference to the actual concentration.

HansR wrote: ↑Sat 18 Sep 2021 1:44 pm

HansR wrote: ↑Sat 11 Sep 2021 6:44 pm One issue may be the maintenance. Cleaning the devices every 6 month is useful. Blowing through them is one method (compressed air, or blow yourself. Not too high pressure). Opening them and cleaning piece by piece sometimes can be done. Some handy ability required.

In addition to the above what I wrote about cleaning AQ sensors i forgot to show you what I use to blow them through every now and then. And don't wait too long because dust may form a crust in the laser beam and may require more difficult operation to clean.

I took your advice after my neighbour had a particularly smoky chimera running a week or two ago and I was getting strange high readings afterwards. Blew it through with compressed air cleaner and hey presto readings returned to reality!!