Comparing Two Filtering Methods to Differentiate between on-Road Vehicle Plumes and Background Measurements.

Comparing two data filtering methods to differentiate between on-road vehicle plumes and background measurements.

Maria Eugenia Velasco Cunningham, Blake Actkinson, Fangzhou Guo, Subin Yoon, Sergio Alvarez, Rebecca Sheesley, Sascha Usenko, Robert Griffin, and James Flynn.

Mobile measurements cover large areas of interest and provide a comprehensive spatial evaluation of local air pollution. However, the data sampled can also be biased by direct on-road vehicle emissions, especially for primary pollutants. The ability to distinguish quickly and precisely between a plume and background levels is crucial for source identification of key species of interest. Baylor University’s Mobile Air Quality Laboratory 2 (MAQL 2) was equipped with an extensive suite of air quality instrumentation and deployed to San Antonio, TX, during the spring 2021 San Antonio Field Study (SAFS). During this campaign, the MAQL 2 provided a month of trace gas, aerosol, and Volatile Organic Compounds (VOC) data measurements made in San Antonio’s urban core and its surroundings. This study utilizes two data filtering algorithms, Density-Based Spatial Clustering of Applications with Noise (DBSCAN) Plume Detection Tool and State-Informed Background Removal (SIBaR) on trace gas measurements (carbon dioxide, nitrogen oxide, and nitrogen dioxide) to systematically distinguish on-road emission sources from background/regional levels. Both filtering methods identify periods when local sources' contributions were high during the time MAQL 2 was on the road. This study determines and compares the consistency and accuracy of each tool for the trace gases being analyzed. The quick identification of plumes and their sources helps to interpret on-road measurements, identify on-road emissions, allow for a better comparison to models, and aid in the evaluation of on-road chemistry.