Air quality monitoring with low-cost sensors

ODINs (outdoor dust information nodes) were designed to be used in a network (typically 20 to 30 nodes) over an area impacted by high emissions of particulate matter. A typical network deployed for a minimum of a month was sufficient to robustly map air quality over a study area.

ODINs provided valuable insights in airsheds impacted by wood-burning for domestic heating, including: 

• Improved understanding of the representativeness of current and future permanent monitoring sites. 
• Identifying groups at higher risk from poor air quality. 
• Informing options for more targeted mitigation. 
• Providing before and after evaluation of air quality interventions. 

Visualising and analysing data

ODINs measured particulate matter (PM2.5 and PM10) every 5 minutes or faster (see features below) and sent it to the cloud via the mobile phone network.

Every morning at 6am these data were automatically converted into an animation of air quality over the past 24 hours. These animations accumulated as a playlist hosted on an online video sharing site.

The animations demonstrated the development of ‘sweet’ and ‘sour’ spots of pollution as pollution was transported from some locations and accumulated in others. After a few weeks aggregate maps were created showing average pollution overall and at different times of the day. These maps often showed that the areas experiencing the biggest impacts do not always correspond to the areas with the highest emissions. To view an example – see our research in Arrowtown in winter 2019.