Participant Detail

Research Title

Bridging the Air Quality Data Gap: Transfer Learning from Korea's Monitoring Network to Predict PM2.5 in Data-Sparse Southeast Asian Cities

Research Abstract

Air pollution is the second leading risk factor for death globally, claiming over 8 million lives annually, with disproportionate impact on low- and middle-income countries in Southeast Asia. Despite this burden, many countries in the region maintain critically sparse monitoring infrastructure. Indonesia, for example, operates only 56 reference-grade air quality stations for a population of 280 million — far below international guidelines. Expanding monitoring networks through conventional infrastructure alone would require tens of millions of dollars, a prohibitive investment for nations with limited environmental budgets. This study explores whether machine learning models trained on data from countries with dense monitoring networks, such as South Korea, can be transferred to predict air quality in data-sparse Southeast Asian cities. By leveraging freely available meteorological, satellite, and air quality datasets, we aimed to evaluate how effectively transfer learning can supplement limited local monitoring and reduce the infrastructure investment required for reliable air quality prediction. Our goal is to demonstrate a scalable, cost-effective pathway toward equitable environmental health protection in developing nations.