The Use of Small Sensors to Augment Future Air Quality Services within the Meteorological Service of Canada

摘要

This study demonstrated the performance of the Purple Air II in three locations in Canada over a wide range of meteorological conditions and ambient concentrations. Although the sensors showed data completeness comparable the FEM, reliability is a concern with 4 out of 7 sensors showing significant degradation of performance within time frames of a few to several months. If data from these sensors are to meet even a modest level of data quality, protocols must be developed to monitor and address sensor performance issues such as drift and failure. The inter-sensor precision displayed by the Plantower sensor was very good, while accuracy was found to be variable across the seasons and concentration ranges captured. In particular, the sensors showed a higher absolute error during periods of elevated PM_(2.5) concentrations during wildfire smoke episodes, however, the use of a simple data correction algorithm greatly improved the accuracy. More advanced data correction algorithms should be investigated to further improve sensor accuracy. Despite the reliability and accuracy issues identified here, with the implementation of protocols for maintenance and data post-processing, these sensors could be a very useful tool to aid the MSC in achieving its objectives.