EN Correction for humidity

Language DE

Depending on both humidity and the particle composition, a certain amount of water condenses on the particulate matter. As a result, the particle's effective cross section increases as well as the likelihood of scattering of the laser in the detector. This can lead to the effect that particles which wouldn't have usually been detected during low humidity, might well be measured in high humidity conditions.

This effect cannot be universally represented with a simple formula, because its mainly determined by the composition of the particle which is usually unknown. The efforts to find a formula describing this effect empirically are ongoing^[1],[2]., which is not linear and is especially noticeable above 70% humidity. This explains the fact that the data sheet of the SDS011 specifies the work environment for the sensor at max 70%.

This means that all measurements taken at a humidity level of more than 70% sometimes have very high readings ​​and, strictly speaking, should be discarded. One can try and attempt to partially neutralise this effect with a correction formula to make use of those readings which were gathered outside of the specifications valid work environment.

Following formulas were used so far:

• HackAir, Poland (sds011_nodemcu.ino)
  • gPM10 (h) = 1 + 0.816 ∙ h ^ 5.83
  • gPM2.5 (h) = 1 + 0.488 ∙ h ^ 8.60
• Norbert^[4] recommends Köhler formula, h normalised to 0 <h <1
  • g_Soneja^[3] = 1 + α ∙ h² / (1-h)
  • g_Hänel = (1-h)^-β
  • g_Köhler = 1 + γ / (1 / h-1)
  • with α = 0.25, β = 0.47, γ = 0.22
• RIVM, Netherlands. Formula based on Hänel with multiplier:
  • g_RIVM = δ* (1-h)^-β
    
  • with different values for β and δ. Values for humidity are sourced from KNMI^[5].
    • for PM2.5 (Amsterdam: β=0.38, δ=2.3; Amersfoort : β=0.40, δ=3.4; Venlo: β=0.43, δ=3.9)
    • for PM10 β=0.65, δ=4.56 (with DHT22 as humidity sensor)

List of sensors located directly at official stations:

Bulgaria

10001 10002 (42.669798,23.26841) Station Pavlovo
10003 10004 (42.655486,23.383278) Station Mladost

Austria

Linz:
11149, 11150 Luftprüfstation S341 Römerberg (http://www.land-oberoesterreich.gv.at/125879.htm)
17654, 17655 Luftprüfstation S341 Römerberg (http://www.land-oberoesterreich.gv.at/125879.htm)

Switzerland

Zürich:
574, 575 Station Stampfenbachstr.
601, 602 Gerhardstrasse 1, no station???

Belgium

10975, 10976 Ukkel (Brüssel), Station 41R012 (http://www.irceline.be/de/luftqualitat/messungen/feinstaub/pro-messstation)
21466 official VMM reference station R802
22585 official VMM reference station R805
22587 official VMM reference station R817
22589 official VMM reference station R701
22591 official VMM reference station R702
22593 official VMM reference station R750
22595 official VMM reference station R834

Australia

Sydney:
17733, 17734 Located at the Environment Protection Authority's Rozelle Air Monitoring Station.
17735, 17736 Located at the Environment Protection Authority's Rozelle Air Monitoring Station.

References

1. Laulainen, N. S. (1993). Summary of Conclusions and Recommendations from a Visibility Science Workshop; Technical Basis and Issues for a National Assessment for Visibility Impairment, Prepared for US DOE, Pacific Northwest Laboratory, PNL-8606.

2. Day, D. E., Malm, W. C., Aerosol light scattering measurements as a function of relativehumidity: a comparison between measurements made at threedifferent sites, Atmospheric Environment 35 (2001) 5169–5176

3. Ramachandran, G., Adgate, J. L., Pratt, G.C., Sexton, K. (2003). Characterizing Indoor and Outdoor 15 Minute Average PM2.5 Concentrations in Urban Neighborhoods; Aerosol Science and Technology 37:1, 33 - 45

4. Streibl Norbert Influence of Humidity on the Accuracy of Low-Cost Particulate Matter Sensors; ResearchGate

5. RIVM Ervaringen RIVM SDS011 sensor 12-7-2018