A low-cost sensor for high density urban CO2 monitoring

Abstract:

The high spatial-termporal variability of greenhouse gases and other pollution sources in an urban environment can not be easily resolved with current high-accuracy but expensive instruments. We have tested a small, low-cost NDIR CO2 sensor designed for potential use. It has a manufacturer’s specified accuracy of +- 30 parts per million (ppm). However, initial results running parallel with a research-grade greenhouse gas analyzer have shown that the absolute accuracy of the sensor is within +-5ppm, suggesting their utility for sensing ambient air variations in carbon dioxide. Through a multivariate analysis, we have determined a correction procedure that when accounting for environmental temperature, humidity, air pressure, and the device’s span and offset, we can further increase the accuracy of the collected data. We will show results from rooftop measurements over a period of one year and CO2 tracking data in the Washington-Baltimore Metropolitan area.