Integrating Low-Cost Sensors with the New York State Mesonet for Continuous, Dense PM2.5 Monitoring in New York City

ELLIE HOJEILY, Jason Covert, Scott Miller, Kit Moore, Md. Aynul Bari, Cheng-Hsuan Lu, Janie Schwab, Atmospheric Sciences Research Center, University at Albany

     Abstract Number: 551
     Working Group: Instrumentation and Methods

Abstract
A low-cost air quality sensor package to measure particulate matter (PM2.5), ozone (O3), carbon monoxide (CO), nitric oxide (NO), and nitrogen dioxide (NO2) was designed and integrated with 38 New York State Mesonet (NYSM) sites in the New York City Metropolitan Area. A new calibration approach, the Network Calibration Algorithm (NCA), was developed, whereby a single low-cost sensor package (the “keystone” package) was co-located alongside regulatory-grade instruments at the New York State Department of Environmental Conservation Queens College monitoring site for 16 months. For each pollutant, hourly data from the keystone package and reference instruments were used to train a single calibration model that was subsequently applied to all packages at field sites across the network. The calibration model for the PM2.5 sensor (Plantower PMS5003) was a hybrid approach that combined multiple linear regression with machine learning. The PM2.5 model was evaluated using data from 41 sensor packages (over 25000 hours) and showed high correlation with the reference monitor (R²>0.88) and low error (RMSE = 2 µgm⁻³). Hourly PM2.5 data from the 38 NYSM sites from May 2023 through August 2024 are used to characterize temporal pollutant patterns at daily, weekly, and seasonal time scales, to contrast pollutants in urban and rural environments, and to evaluate spatial correlations across the network. This study illustrates that the data from the low-cost sensor augmented Mesonet sites, along with the existing FRM/FEM monitors, can be used to identify and understand emission sources and improve our understanding of urban air pollution.