PurpleAir sensor readings are virtually indistinguishable from government monitors according to new peer reviewed study

In August of 2016, the Gabriola Island Clear Air Society brought online the first PurpleAir sensor network in Canada. Our sensor network has documented extensively the staggering amount of air pollution coming from wood stoves and fireplaces on our tiny island.

As the number of PurpleAir sensors expanded across the province and elsewhere in Canada, many began to raise questions about the accuracy of these low-cost devices. There have been several studies to-date that all show the same thing: these devices work, are accurate, reliable, and fill in gaps in air quality monitoring.

The newest study from Australia by Dr. Dorothy Robinson provides definitive evidence that should quash any concerns about using PurpleAir sensors. See the study here. 

Gabriola Island Clean Air Society director Dr. Michael Mehta has the following to say about this study.

This paper provides empirical evidence to show that the use of low-cost PM2.5 monitors for measuring air quality on a distributed basis has merit. By benchmarking the accuracy of PurpleAir monitors in particular through a comparison with data from co-located TEOM measurements from government monitors placed in the New South Wales region of Australia, the author convincingly demonstrates that residential wood burning practices are far more polluting than is often recognized.

 

The paper reviews how household wood burning for home heating purposes creates substantial pollution at the PM2.5 scale, and shows how monitoring air quality at a more granular level using low-cost IOT (internet of things) devices can improve how exposure-response relationships may be refined to increase the accuracy of estimates of health costs from such practices.

 

It is clear from the analysis in this paper that standard government approaches for monitoring air quality are inadequate when it comes to capturing the spatial and temporal variation associated with wood heating. As the author points out, lives are at-risk from inadequate monitoring regimes.

 

In this empirical section of this paper, the author show that PurpleAir monitors track extremely closely in their readings when compared with FEM instruments that cost substantially more. By developing calibration equations, the author showed that PurpleAir monitors can be a reliable and accurate instrument, and moreover that these monitors are relatively unaffected by temperature or humidity inside the device.

 

PurpleAir monitors have been deployed globally and are now used by local governments, air quality boards, citizen scientists, and others to document the prevalence of wood smoke in communities that are underserved by government monitoring stations. In this paper, the author compares various communities including some that have open pit mines and power stations which have been the focus of research on air pollution in those communities. The author concludes, based on the use of PurpleAir technology that wood smoke is still the dominant source of winter air pollution even in such communities.

 

To estimate health costs, the author reviews several studies including some comprehensive meta analyses, and discusses how low-cost devices like PurpleAir have several advantages including the provision of real-time, localized information to help at-risk individuals to proactively take steps to reduce exposure. An extensive discussion on policy implications is also provided in this paper. 

 

The paper builds on several existing studies, and provides additional support for the case that low-cost sensors are revealing a hidden problem resulting from the hyperlocal effects of wood smoke exposure from residential burning practices.