Researcher's Profile

CHANG-FU WU (吳章甫)

     ORCID Follow

Research Outline      2015-03-11 15:06:56

My research interest involves developing and applying innovative technologies and models for exposure assessment and environmental monitoring. I am also interested in conducting risk assessment on environmental and occupational health issues.

Air quality modeling
Developing effective control strategies to reduce population exposure to certain hazardous air pollutants (HAPs) requires identifying sources and quantifying their contributions to the mixture of HAPs and the associated health risks. One approach is to use receptor-based source apportionment models (e.g. Positive Matrix Factorization or PMF, Multilinear Engine or ME) to distinguish sources. Once the contributions are determined, the associated risk could be estimated from the source profiles. In one of our previous projects, the risk apportionment approach, which is a combination of receptor modeling and risk assessment, was developed to estimate source-specific lifetime excess cancer risks of selected hazardous air pollutants. It was found that the diesel exhaust contributed less than the wood burning on the basis of mass concentration; nevertheless, they presented similar cancer risks. This highlights the value of the risk apportionment approach for prioritizing control strategies to reduce the highest population health risks from exposure to air pollutants. Other than these receptor models, we are also applying dispersion models (e.g. ISC3, AERMOD) for simulation studies.

Source-specific Exposure and Health Risk Assessment
Many epidemiological studies have found adverse health effects from exposure to air pollutants. However, most of these studies relied on air pollution data collected at central sites. Our team conducted a panel study of 17 mail carriers by monitoring their personal ozone and size-fractionated PM exposures, as well as several health indicators of their cardiovascular functions. The study results showed that personal exposure to ozone and PM between 1.0 and 2.5 mm (PM2.5-1) affected the vascular functions significantly. The collected personal PM filters were further analyzed for their elemental concentrations. Through the receptor model of absolute principal component analysis, three sources of PM2.5-1 were identified (urban dust, regional sources, and brake wear). The statistical analysis found a significant linkage between the PM from regional sources and the observed vascular effects.

Source-specific GIS Modeling
Assigning exposure based on fixed-site monitoring results in many epidemiological studies may lead to misclassification. Land use regression models, which were built based on air monitoring data collected at multiple locations and included predictor variables (e.g., land use, population, and traffic data) obtained through the geographic information system (GIS), can be used to capture the intraurban variability in air pollutant concentrations. Our team has extensive experiences in building land use regression models for PM2.5 concentrations and compositions. These models can be used not only for calculating individualized exposures estimates and but also for identifying major emission sources or activities (e.g., types of land usages).

Optical remote sensing techniques
Optical remote sensing (ORS) instruments are useful tools for monitoring air pollutants. However, they usually provide only path-integrated data. By applying the Computed Tomography (CT) algorithm or radial plume mapping (RPM) technique, we can obtain the spatial distribution information of air pollutants in the environment. Combining with meteorological data, we can identify emission sources, and further assess their health impacts on residents in nearby communities. For example, our team applied this technique to estimate the emission flux of greenhouse gases in rural China and of VOCs at several industrial areas in Taiwan. The main ORS instruments applied in our lab include Open-Path Fourier Transform Infrared Spectroscopy (OP-FTIR), Ultraviolet Differential Optical Absorption Spectroscopy (UV-DOAS), and LIght Detection and Ranging (LIDAR).