This webpage comprises

• Introduction for measurement of Gas
• Forecast by MeteoBlue and VentuSky: Dust & Gasses
• Actual read-out for Gas elements
• Derived information
• 24Hour Graphs
• Longterm Graphs
• System Layout
• Why local measurement of gasses?
• Description of the setup

• As start a basic configuration which is rather broadband, with gas sensors which in (limited) quality and (limited) sensitivity are compatible with the dust-sensor GP2Y10
  SGS in it's catalogue has sensors MiCS2714 and MiCS6814, already implemented by others in break-outs to be installed in suitable housing providing good gas flow.
  Type MiCS2714 only reports NO₂-level, and might serve as quickstart 'gasset' in combination with the ADC of an ESP8266.
  Type MiCS6814 reports gas level divided in 3 signals for CO, NH₃ and NO₂, as a kind of spectrum-survey aimed at
  "detection of pollution from automobile exhausts and for agricultural/industrial odors" [quote from datasheet].
• Next level with a 'better' configuration for more accurate and more 'deep'investigation of specific segments
  Alphasense has in it's catalogue the sensors NO2-B43F and OX-B431 for measurement of NO₂ and O₃
  Besides the sensors they can provide compatible interfaces, both mechanical and electronical.
• Speed of the airflow over the sensors must be constant, because otherwise the measurement is affected.
  This requirement implies that the fan must have stabilized/controlled power supply.
• As reference for the gas measurements, the air flowing over the sensors also must be well measured for temperature and humidity.
  This requirement implies that the T&H-sensors must be in the air flow close before or close behind the gas-sensors, preferably in a common tube/canal.
• The characteristics of the MiCS6814 are defined for an environment of 25C&50%RH.
  The definition for the Alphasense' sensors has a wider operational T&H-range, but RIVM's testreport hints at sensitivity for moisture,
  For measurement of gas-content the use of Dry/hot air seems therefore a better basis than the direct, wet/cold outdoors air:
  requiring additional attention because a setup for heating/drying should have good stability and simple reproduction for alignment between the measuring setups, enabling honest comparison between comparable stations ......
• The processing of the sensorsignals anyway requires additional interfaces (incl. A/D-converters) as well as functions for calibration.
  
• As a 1st processing stage and for communication an ESP8266 or a Raspberry may be suitable.
  
• Related to calibration especially an easy availability of good references is a 'hot' issue.

1. the experience for outdoor gas measurement in the perspective of Citizen Science is still limited, and with the experimental basic configuration
   • a physical setup can be developed which has the sensors well located in the air flow
   • experience can be collected related to occuring outdoor phenomena
     [which may confirm the basic configuration in a role as 'gasalarm-alerter']
   • experiments may lead to a functioning approach for local or remote calibration
   • initial investment can be limited
2. also the 'better' configuration needs further developement, because
   • implementation of it's sensors requires a rather extensive infrastructure [= compatible, gas-neutral!! housing with controlled air flow, interfaces and software]
   • periodic calibration is critical and cannot be realised at user's location.
     This document explains why it is difficult/impractical to expect calibration outside a laboratory.
   • variation between sensors, degradation/shift of characteristics and limited life cycle of the sensors
   • components & realisation still are rather expensive (at least for Citizen Science)

Until a setup has been developed and tested, which has high quality&reliability at an acceptable price-level, simple(r) placeholders and some information from 'elsewhere' have to be applied, like Copernicus Atmosphere Monitoring System.
E.g. the presently online Ozone-information from DarkSky/NASA in this page is only valid for higher layers (~stratosphere),
with meaning in relation to attenuation in those layers for UV-radiation from higher-up.
UV-radiation indirectly has effects on air quality and life on ground level, but more on other aspects than gas.
Also positive, such as for the creation of Vitamine-D.
Because of the local effects and the relation to other air components therefore also for for O₃-measurement at ground level a local measuring setup is required.
Information of RIVM bases on models & estimations, at surface level supported by info from a limited number of qualified measuring stations.
Nearest RIVM measuring stations with gas-info are in Eibergen and in Hellendoorn.

Actual status/ Remarks:

Ozone-info (stratosphere) from DarkSky [= calculated Dobson Units] CO-info ground-level measurement [= ADC-values] NO₂-info ground-level measurement [= ADC-values] NH₃-info ground-level measurement [= ADC-values]

The colour classifcation for gas-related air quality is according to the table below.
For O₃ and NO₂ derived from the RIVM-table for Air Quality Index.
The Ozone-information from DarkSky and from NASA is an indication for the thickness of the ozone layer protecting the earth, not a local, atmospheric O₃-measurement with indication in µg/m³.
In absence of local O₃-measurements the info from DarkSky (as Dobson Units = DU) has been roughly (reversely) translated from the map colours of NASA:
'green'= thickest Ozone-layer = best protection, while 'red' = thinnest Ozone-layer = least protection (e.g. in 'Ozone-hole').
The selected classification for the ADC-signals is an arbitrary default, until more experience has been collected:
- for CO applying the O₃-info of RIVM
- for NO₂&NH₃ 'aligning' with the NO₂-scaling of RIVM.

Quality = Colour	Good = Green	Average = Blank	Insufficient = Yellow	Bad = Red
Ozone [DU]	600 ~ 450	450 ~ 350	350 ~ 200	200 ~ 0
O3 [µg/m3] / ADC_CO [Units]	0 ~ 40	40 ~ 100	100 ~ 180	> 180
NO2 [µg/m3] / ADC_NO2_NH3 [Units]	0 ~ 30/ 0 ~ 60	30 ~ 75/ 60 ~ 150	75 ~ 125/ 150 ~ 250	> 125/ > 250

This site borrows dynamic prognosis for gas-oriented Air Quality from the RIVM-site to which this Meteo-System indirectly uploads data through Samenmeten_Dataportaal.
The prognosis for Ozone has a multi-day bracket around a selected (Dutch) postal code.

For fitting combinations the graphs G1A and G1B show:
# Temperature, Humidity and Pressure are real, measured values determined by the dust-related sensors.
- Temperatures, Humidity and Levels are 'as-is'
- Pressure is the difference relative to 1000hPa
# Gas1-implementation = measurement by MiCS6814 for 3 gas types
+ measurement by LM35 for 1 gas/air temperature
# Gas2-implementation = measurement assumed by 2 'peaked' sensors for 2 gas segments, such as from Alphasense.
Till real sensors arrive, 'placeholder'-constants.
The placeholder constants serve as reference lines in the graph.

Graph G1A in this first edition has:
# Gas1-implementation = readout on dynamic data from ESP8266_GP2Y10 with ca. 1mV/Unit
Gas1-levels: gaslevels are raw ADC-values.
Gas1Temperature is a translation of the ADC-value to degrees Celsius.

Graph G1B experimentally translates the ADC-values for gasses into µg/m³,
coarsely aligning with RIVM's NO₂-forecast
# Gas1-leveltranslation:
1 Unit ~ 0.15 µg/m³ for NO₂&NH₃
1 Unit ~ 1 µg/m³ for CO

For comparison the dust-graph for the same 24hours.











The components in the Dust-graphs not always are singularly generated, but often 'combined'.

The 'more accurate' set with Dust sensor type SDS011 + meteosensors BME280&DHT22 in the present housing probably has no space left to accomodate the Alphasense sensors and their interfaces:
investigation required how to configure, how to house and how to link.
Best in a setup which can be shared with organisations like Luftdaten.org or RIVM or Sensebox.de

The uncertainties require a plan of attack for the next phases.

Copyright © 2013-2022 T4S
Samenvatting voor Rechten & Verantwoordelijkheden / Summary for Rights & Liabilities