Posts

Showing posts from July, 2018

Comparing PurpleAir network accuracy to satellite data from Copernicus

Image
For those who question the value and accuracy of low cost particle sensors made by PurpleAir and other companies, consider this evidence. Here we have PM2.5 satellite predictive modelling from the Copernicusopen access database of surface PM2.5 for North America (note the hot spots likely from forest fires). When we compare a map of real-time sensors from PurpleAir it should be immediately apparent that they agree with the satellite models in general ways.



If you need more specific proof, the satellite estimate is giving a current surface PM2.5 reading of 33 micrograms/m3 and the PurpleAir sensors on Gabriola Island are bang on!




What's puzzling and somewhat disturbing is the inaccuracy of the air quality monitors used by the BC Ministry of Environment. Compare what their PM2.5 levels are at 15 micrograms/m3 in nearby Nanaimo to what the Copernicus satellite indicates at 39 micrograms/m3. This holds even when one takes into consideration the realtime versus one hour averaging peri…