ID: 1888
Presenting Author: YUXIAO JIANG
Session: 642 - Enhancing the credibility and impact of climate change and health impact assessments
Status: pending
This study develops a climate-sensitive framework integrating machine learning–based air pollution retrieval and β-based health assessment to improve climate-health risk evaluation.
Climate change is intensifying air pollution episodes and exacerbating associated health risks across Southeast Asia. In Thailand, recurrent haze events and long-term exposure to fine particulate matter (PM2.5) and nitrogen dioxide (NO2) have emerged as critical public health concerns. To address these challenges, this study introduces a novel, climate-sensitive framework that integrates high-resolution exposure modeling with β-based short-term health impact assessment. We employ advanced machine learning algorithms that combine ground-based observations, satellite reanalysis products, meteorological data, and other auxiliary variables to develop a robust, high-resolution air pollution retrieval model. The resulting dataset provides spatially continuous estimates of PM2.5 and NO2 at a 1-kilometer resolution, for the period 2019–2023. These refined exposure estimates are then linked with daily hospital admission records to quantify short-term morbidity risks using established exposure–response coefficients (β). Trend analyses and regional comparisons across six areas of Thailand reveal significant spatial and temporal disparities in both pollution exposure and health outcomes. Integrating machine learning-based retrieval methods with epidemiologically grounded health risk estimation, this study demonstrates novel methodological pathways that enhance the robustness, credibility, and policy relevance of climate-sensitive health impact assessments, thereby strengthening population resilience under future climate scenarios.
JIANG Yuxiao is a PhD candidate at The Chinese University of Hong Kong, focusing on satellite remote sensing, atmospheric pollution, and climate-sensitive health impact assessment.
Coauthor 1: XINGCHENG LU
Coauthor 2: Manlika Sukitpaneenit