A177-0002
Assessing rural vs. urban PM2.5 disparities in the Indo-Gangetic Plain: A low-cost network approach
Tuesday, 15 December 2020
Poster
Saumya Singh1, Mark Joseph Campmier2, Adithi Upadhya3, Sreekanth Vakacherla4, Harshraj Mishra5, Jonathan Gingrich6, Meenakshi Kushwaha3, Ravi Kant P. Pathak7, Julian D Marshall8, Sagnik Dey9 and Joshua Apte10, (1)University of Texas at Austin, Austin, TX, United States, (2)University of California Berkeley, Berkeley, United States, (3)ILK Consultancy, Bangalore, India, (4)CSTEP, Bangalore, India, (5)IGP-CARE, BHARAT UDAY, Hamirpur, India, (6)University of Texas at Austin, Department of Civil, Architectural and Environmental Engineering, Austin, United States, (7)University of Gothenburg, Gothenburg, Sweden, (8)University of Washington, Seattle, United States, (9)IIT Delhi, New Delhi, India, (10)University of California, Berkeley, Berkeley, CA, United States
Abstract:
The Indo-Gangetic Plain (IGP) experiences severe air pollution, with typical annual-average fine particulate matter (PM2.5) concentrations in the range 75 – 150 µg m
-3, leading to average life expectancy for nearly a billion people to be shortened by approximately two years. Air pollution policies often emphasize city-by-city management, yet air pollution in the IGP is a regional challenge, and in-situ observations of rural air quality are limited. Here, we present initial results from our ongoing study to characterize rural vs. urban air pollution gradients in North India, based on a stratified sampling scheme designed to cover a broad range of conditions from villages to towns to megacities and evaluation of sensors performance during the 12 months period including collocation and field deployment.
Our experimental approach employs a carefully maintained network of lower-cost PM2.5 sensors (Purple Air [PA] II, each encompassing two Plantower PMS5003 nephelometers), calibrated on the basis of multiple beta-attenuation reference PM2.5 monitors (MetOne BAM-1022) and hourly-average measurements. During two mass colocation events in urban Bangalore (South India) and rural Hamirpur (IGP), we observed that PA sensors produced highly reproducible results with low unit-to-unit variability. The PA sensors estimated BAM-reported PM2.5 with R2 > 0.8 and normalized root-mean-squared error (NRMSE) of 20-30%, for concentration ranging from 5 to 150 µg m-3.
We deployed PA sensors in a clustered pattern to characterize within-settlement variability along a gradient from remote rural to highly urban conditions in eastern Uttar Pradesh. We observed mass concentration of PM2.5 ranged from 8.5 to 560 with an average of 82.7 µg m-3 in Varanasi city (population ~1.2 million) whereas Mai village (population ~5200) observed a concentration range from 9.1 to 512.7 µg m-3 with an average of 105.2 µg m-3. Overall, we found that PM2.5 concentrations in our pilot sample of villages and small towns were often similar to or higher than the nearest large city. These initial results suggest that additional observations in rural settings may provide valuable new information about the sources, dynamics, and consequences of PM2.5 across this region.
