Developing Quality Checks for Real-time Low-Cost PM_(2.5) Sensors Deployed by Citizen Scientists

摘要

As part of a NASA-funded research project, we have been performing field deployment of low-cost fine particulate matter (PM_(2.5)) sensors through citizen scientists in the South Coast air basin in the Los Angeles area and in the inland Riverside/San Bernardino regions of Southern California. This network is being expanded to the Research Triangle Park, NC and North India regions. The objective of the project is to use low-cost sensors deployed by citizen scientists to generate high-quality data to examine the spatial gradients in surface PM_(2.5) concentrations and satellite observations. In this presentation, we will share our approach to developing quality checks for verification of real-time sensor data and our plans for implementing them as part of our online data framework. The network uses the PurpleAir sensors. Our earlier work showed potential differences in response between the sensors. Here, we are examining ways to identify potentially erroneous responses and malfunctioning sensors, so that the citizen scientists can be notified. Implementation of such a real-time system can help maintain and improve the quality of data collected by the sensor network.