(CNA, Taipei, June 21) A research team from National Cheng Kung University (NCKU) has used AI to estimate Taiwan's air pollution risk for the next century for the first time, with their findings published in an international journal. The study indicates that under extreme warming scenarios, while overall air pollution in Taiwan may decrease, the northeast, the border of Taichung and Changhua, and Kaohsiung-Pingtung will paradoxically become new air pollution hotspots by the end of the century.
Climate change and air pollution are considered the two most significant environmental risks affecting human health today. The NCKU team's research, titled "How Climate Change Itself Will Alter the Distribution of Air Pollution Risk in Taiwan," has been published in the international journal Journal of Environmental Management.
Wu Chih-ta, a professor in NCKU's Department of Geomatics and Spatial Information, and the corresponding author of the project, told CNA that the research primarily compiled 20 years of climate and air pollution data in Taiwan to simulate possible changes over the next 100 years.
This project integrated long-term air quality monitoring data from 1994 to 2019 in Taiwan, geographical environmental factors, land use information, population data, and climate model data used in the Intergovernmental Panel on Climate Change (IPCC) Sixth Assessment Report (AR6). Through a Geospatial Artificial Intelligence (Geo-AI) model, it assessed changes in future fine particulate matter (PM2.5) concentrations under different climate change scenarios.
According to the study, climate change-induced warming and increased rainfall generally help reduce PM2.5 concentrations, leading to a downward trend in average PM2.5 levels across Taiwan in most scenarios. Wu Chih-ta explained that it's similar to summer when air pollution is rarely discussed, as high temperatures increase the height of the atmospheric mixed layer, allowing pollutants to disperse and dilute easily without accumulating near the surface. Additionally, increased rainfall in some future scenarios also aids in removing PM2.5 from the air through wet deposition.
However, Wu Chih-ta pointed out that the impact of climate change on air quality is not evenly distributed. Under extreme warming scenarios, it was found that from 2081 to 2100, Taiwan's northeast, the border of Taichung and Changhua, and the Kaohsiung-Pingtung region will become hotspots where air pollution paradoxically worsens.
Wu Chih-ta explained that because the AI model estimates weather patterns 100 years into the future, it is sometimes difficult to perfectly explain why PM2.5 increases using traditional climate models. The research speculates that it might be related to changes in local meteorological conditions, such as accumulated high temperatures, altered temporal and spatial distribution of rainfall, or changes in atmospheric diffusion conditions, which could disrupt the original air quality balance. However, further research is needed to verify these mechanisms.
Wu Chih-ta added that the premise of this project assumes that all human industrial activities, socioeconomic conditions, land use, and anthropogenic emission sources remain constant in the future, with only temperature and rainfall changing. Conversely, by adjusting land use, increasing urban greening to reduce the urban heat island effect, promoting energy transition, and regulating human activities, it is possible to proactively mitigate pollution.
Wu Chih-ta stated that this is the first time a high-resolution Geo-AI model has been integrated with IPCC climate change scenarios to estimate future climate risks. The goal is to recognize the differentiated risks that different regions may face under climate change, enabling the formulation of more precise and forward-looking adaptation strategies. (Editor: Chang Ya-ching) 1150621
Stand with facts, your every donation is a force to protect press freedom
Download the CNA "One-Stop News" APP for real-time updates
The text, images, and videos on this website may not be reproduced, publicly broadcast, publicly transmitted, or used without authorization.
FACT BOX
- Source: CNA (Central News Agency)
- Category: 研究
- Organizations: Journal of Environmental Management