Development of PM2.5 concentration maps using multi-source datasets

The project aims to create a national-scale dataset of daily average PM2.5 concentration maps from 2019 to 2021, utilizing a Mixed Effect Model (MEM) that incorporates various data sources. The study area for this project includes Vietnam, with a focus on monitoring air pollution levels in the country. According to the World Health Organization (WHO), air pollution poses a significant threat to the environment and human health globally, with 7 million premature deaths attributed to both outdoor and indoor air pollution each year. In Vietnam, air pollution levels are among the highest in the world, with many provinces and cities exceeding national standards for annual average PM2.5 concentrations. The methodology employed in this project involves the use of a MEM that incorporates ground station monitoring data of PM2.5 dust concentration, satellite image products (Aerosol Optical Depth, AOD), maps of meteorological parameters such as humidity and Planetary boundary layer height (PBLH), as well as land use maps indicating traffic density, vegetation index, and topography. The steps involved in constructing the PM2.5 dust concentration map include preprocessing the station and map data, improving the quality of the map data, integrating the map data with the station data, building and validating the MEM to estimate daily PM2.5 maps, synthesizing monthly/yearly maps of PM2.5, evaluating the quality of the map data against station data, and comparing the results with global PM2.5 mapping products. The results of this project include the estimation of daily PM2.5 maps using two MEM models (one main and one auxiliary model). The main model was developed using a combination of ground station monitoring data and satellite image products, while the auxiliary model was used to improve the accuracy of the PM2.5 maps in areas with limited ground station data. The results show that the daily PM2.5 maps were able to capture the spatial distribution of PM2.5 dust concentration in Vietnam, with a high degree of accuracy. The project also includes the development of a WebGIS system that displays near-real-time air quality information, as well as a video for educational and promotional purposes. The project's deliverables include a PM2.5 dataset from 2019-2021, a report on the status of PM2.5 and health impact assessment in Vietnam in 2021, and a video for educational and promotional purposes. The Mixed Effect Model (MEM) used in this project is a statistical model that incorporates various data sources to estimate daily PM2.5 maps. The MEM was developed using a combination of ground station monitoring data and satellite image products, and was validated using a combination of station data and global PM2.5 mapping products. The results show that the MEM was able to capture the spatial distribution of PM2.5 dust concentration in Vietnam, with a high degree of accuracy. The project's findings have important implications for air pollution monitoring and management in Vietnam. The development of a national-scale dataset of daily average PM2.5 concentration maps from 2019 to 2021 provides a valuable resource for policymakers and researchers, and can be used to inform the development of effective air pollution management policies. The project's results also highlight the importance of incorporating multiple data sources, including ground station monitoring data and satellite image products, to estimate daily PM2.5 maps.