Abstract View
Trends, Performance, and In-Field Calibration for Low Cost Particulate Matter Sensors in West Bengal, India
V. FAYE MCNEILL, Anindita Dutta, Sandhya Sethuraman, Siddharth Nobell, Rakhi Basu, Sarbani Palit, Daniel Westervelt, Debasis Dasgupta, Shairik Dasgupta, Columbia University
Abstract Number: 414
Working Group: Urban Aerosols
Abstract
We present results from monitoring efforts using low-cost ambient particulate matter sensors (LCS) operating in West Bengal, India (Kolkata, Bandel, and environs). Two datasets are examined: a network of Purple Air sensors operating from August 2019-present and a deployment of locally developed sensors in 2020. LCS were co-located with reference-grade PM monitors (RGM) operated by the West Bengal Pollution Control Board for in-field calibration using random forest and multilinear regression. The Purple Air network performed well without additional correction for PM10 but exhibited a linear bias for PM2.5 which was corrected effectively using the calibration. After calibration, the network correctly diagnosed days as being in or out of attainment with India NAAQS 93% of the time (for both PM10 and PM2.5). Opportunities exist to complement existing government monitoring networks in India with LCS for increased spatial resolution and coverage.