Journal of Pollution Effects & Control
Open Access
ISSN: 2375-4397
ISSN: 2375-4397
Qingqing He
The Chinese University of Hong Kong, Hong Kong
Posters & Accepted Abstracts: J Pollut Eff Cont
Statement of the Problem: Using satellite-retrieved aerosol optical depth (AOD) and statistical model is a potential approach to estimate exposure to PM2.5 for regional studies. However, studies of assessment of ground-level PM2.5 for China at a high spatial resolution have been limited due to the lack of high resolution AOD product. The purpose of this study is to estimate daily highresolution distribution of ground-level PM2.5 using satellite remote sensing. Methodology & Theoretical Orientation: The newly released MODIS AOD data at 3 km resolution were processed as the main predictor. A geographically and temporally weighted regression (GTWR) model was developed to estimate daily PM2.5 concentrations over Beijing-Tianjin-Hebei region from January 1, 2013 to December 31, 2015. The surface PM2.5 measurements were the dependent variable and combined AOD data, land use and meteorological data were used as the independent variables. The GTWR model is able to simultaneously accounts for spatial non-stationarity and temporal variability of the relationship between PM2.5 and AOD, which can enhance the PM2.5 estimation accuracy. Findings & Conclusion: The overall model R2 value generated by GTWR model was 0.84 in model validating process, which was significantly better than those from geographically weighted regression (R2 of 0.51) and temporally weighted regression (R2 of 0.58) models. The annual mean of satellite-derived PM2.5 for China was 70.80 �¼g/m3 over the study period, 100% higher than the national ambient PM2.5 standard of 35 �¼g/m3. The ground PM2.5 predictions shows significant seasonality and winter was the most polluted season. There was virtually no ascending or descending trend for ground PM2.5 concentrations (-0.0002 day-1) from Jan 1, 2013 to Dec 31, 2015. In addition, predicted PM2.5 maps at high-resolution grid are useful to present the detailed particle gradients and investigate PM2.5 hotspots. The findings from the study demonstrated the promising potential of GTWR model for air pollution mapping.
Email: qqhe@link.cuhk.edu.hk