meteohub.de

Posted: **Wed Sep 23, 2020 10:15 pm**

There were some fix requests addressing that air quality for a day/month/year shows "0" as minimum. I sniffed the data if it is something within Meteobridge or if the sensor is sometimes reporting this. Now I can say, it is the senor.

I logged this data snippet from an Airlink.

Code: Select all

{ "data": 
  { "did": "001D0A1000B1", "name": "001D0A1000B1", "ts": 1600890284, "conditions": 
    [{ "lsid": 346392, "data_structure_type": 5,  "temp": 78.1, "hum": 44.8, "dew_point": 54.9, "wet_bulb": 60.6, "heat_index": 77.4, 
       "pm_1_last": 0, 
       "pm_2p5_last": 1, 
       "pm_10_last": 1, 
       "pm_1": 1.14, 
       "pm_2p5": 1.41, 
       "pm_2p5_last_1_hour": 1.98, 
       "pm_2p5_last_3_hours": 2.10, 
       "pm_2p5_last_24_hours": 4.25, 
       "pm_2p5_nowcast": 2.20, 
       "pm_10p0": 1.41, 
       "pm_10p0_last_1_hour": 2.33, 
       "pm_10p0_last_3_hours": 2.63, 
       "pm_10p0_last_24_hours": 5.29, 
       "pm_10p0_nowcast": 2.77, 
       "last_report_time": 1600890283, 
       "pct_pm_data_last_1_hour": 100, 
       "pct_pm_data_last_3_hours": 100, 
       "pct_pm_data_nowcast": 100, 
       "pct_pm_data_last_24_hours": 50 }] }, 
     "error": null...

Currently, Meteobridge reads the "pm_XXX_last" values, which can be zero ("pm_1_last": 0).
I could switch to "pm_XXX" instead. Does anyone have more background which is the more appropriate data?
From the Davis documentation then "pm_XXX_last" values looked most suited to me as raw data (averaging etc is done inside Meteobridge).
https://weatherlink.github.io/airlink-local-api/

I think I will switch to "pm_XXX" as "pm_XXX_last" looks unreliable to me although Davis documentaiton says the Airlink only report valid readings via "pm_XXX_last".

Posted: **Thu Sep 24, 2020 3:44 pm**

This is going to be one of the challenges. Some countries indicate AQI by current PMxx values. Some countries use the 24hr average. Some are not well specified. There is even a variance on the friendly labeling of each AQI level: Polluted/Unhealthy/Poor/High etc. https://en.wikipedia.org/wiki/Air_quality_index

A possible solution, when adding the Davis Airlink as a subsequent sensor, is to offer the user the choice of using the Actual, 1-hr, 3-hr or 24-hr values for each sensor. Is that possible?

None of the major countries/regions seem too bothered about PM_1.0. It's all about PM_2.5 and PM_10, but as the Airlink provides the data I'm sure someone will want it.

Posted: **Thu Sep 24, 2020 4:50 pm**

The averaging on different time frames is done by Meteobridge or any other postprocessing program; no one needs a sensor to report this. I think I will stay with the three "pm_XXX" als the "pm_XXX_last" data has proven unreliable resp. returns misleading data.

P.S.: Davis will need the sensor to average over the most popular time frames because they want to keep data uploads to their server minimal and by that can't compute this in good quality on their servers. But with Meteobridge we don't have a need for this poor mens shortcut as we read data locally and frequently and can do all postprocessing where it belongs to.

meteohub.de

Davis Airlink - "0" shown as minimum for a day/month/year

Davis Airlink - "0" shown as minimum for a day/month/year

Re: Davis Airlink - "0" shown as minimum for a day/month/year

Re: Davis Airlink - "0" shown as minimum for a day/month/year