Addressing Disparities in Air Quality Monitoring: Using Machine Learning and Remote Sensing to Estimate the Distribution of PM2.5 in Mexico
ISEF · 2023 Earth and Environmental Sciences Third Award
Overview
Air pollution is the single greatest environmental health threat, according to the WHO, and disproportionately impacts developing nations, which suffer 90% of the 5 million premature deaths globally per year due to air pollution. This is in part due to the "outsourcing" of pollution to countries with lesser air quality regulation and research on regional air pollution monitoring when compared to the United States, Europe, and East Asia, as most international treaties hold countries responsible only for emissions produced within their borders. Air pollution, however, is a global matter, and must be addressed with research at the regional level. To combat the disparity in the geographic focus of research, this project details the process of creating the first continuous map of fine particulate matter (PM2.5), the most dangerous pollutant, over Mexico, using data sources and methods that may be easily applied to other regions of the world as a baseline for more specialized regional studies. A nationwide model to estimate PM2.5 in regions with few ground stations is developed using satellite data, historical ground pollutant monitoring, and land use variables. This is compared to NASA’s MERRA-2 hourly surface particulate matter estimates for 2018. To provide continuous estimates of air pollution in regions without monitoring stations, a machine-learning model that incorporates historical ground monitoring, meteorological parameters, and physical model simulations is developed and tested, with cross-validated R scores up to 0.88 - the first of its kind covering all of Mexico.
Awards (3)
- Third Award of $1,000 $1,000
- Fondazione Bruno Kessler: Award to participate in summer school "Web Valley"
- U.S. Agency for International Development: Third Award Climate and Environmental Protection
Competition history
- ISEF 2023
Resources
Related projects
JSHS · 2023
SOUTHWEST An Integrated Approach for Immediate and Long-Term Air Quality Regulation and Monitoring in Mexico
ISEF · 2021
Using Machine Learning to Combat Air Pollution by Forecasting Tropospheric Ozone Levels
ISEF · 2022
Performance of Machine Learning Algorithms for Predicting Air Pollution Parameters
ISEF · 2022
Urban Ozone Forecasting and Policy Recommendations Through Photochemical Modeling and Machine Learning
ISEF · 2019
A Comprehensive Spatiotemporal Model for Interpolation of Tropospheric Fine Particulate Matter Concentration
ISEF · 2022
Comparison of Ambient Air Pollutants in Varying Income Tracts Using a Constructed, Portable Device That Precisely Measures Ambient Air Concentrations (PAQMS)
ISEF · 2018
The Air We Breathe- Reducing Health Risks by Improving IAQ: An Innovative, Smart, and Responsive Ductless System Optimized by Stochastic Simulation and Machine Learning
ISEF · 2024
National Ground-Level NO2 Predictions via Satellite Imagery Driven Hybrid Neural Networks
Closest projects by meaning, across every fair and year in the corpus.
Browse more like this
Source: Regeneron International Science and Engineering Fair