Estimation of long-term population exposure to PM2.5 for dense urban areas using 1-km MODIS data

The lack of long-term PM2.5 measurements in developing countries makes it difficult to quantify the overall PM2.5 pollution exposures and health impacts in these countries where the PM2.5 concentrations are often very high. Moreover, it is also difficult for traditional fixed-site monitoring to capture the substantial spatial variability of PM2.5 over dense urban areas or regions with significant topography. Hence, recent developments in satellite based remote-sensing allowing the reconstruction of long-term, wide-area and high-resolution estimates of current and historical PM2.5 concentration is an important step forward, allowing the quantification of the long-term pollution exposure of PM2.5 in developing cities and in dense urban areas using the satellite-derived PM2.5 data. In this study, instead of just looking at the spatial average PM2.5 concentrations, we have studied the long-term population exposure of PM2.5 by analyzing the population-weighted PM2.5 concentrations at regional, city and district scales by combining 1 km x 1 km satellite-derived PM2.5 and population density data sets. The variation of population exposure to PM2.5 across the urban areas in the Pearl River Delta (PRD) region from 2000 to 2014 was studied. Our result shows that the PM2.5 concentrations over the PRD increased steadily from 2000 to 2004, remained at quite a high level through 2008 and then started to decline after 2009. More importantly, our analysis also shows that, at regional, city and district levels, the population-weighted mean PM2.5 concentrations from data with 1 km resolutions are typically the highest, followed by the population-weighted mean PM2.5 concentrations from data with 10 km resolutions and then the simple spatial PM2.5 averages. This suggests that the use of simple spatial concentrations can lead to systematic underestimation of the overall population exposure and the associated health impacts. This systematic difference is related to the positive correlation between PM2.5 pollutant concentration and population density, and shows the usefulness of using high-resolution satellite-retrieved PM2.5 concentrations to quantify the overall population exposure. The higher population weighted mean PM2.5 concentration in comparison with simple spatial average indicates that, for more effective reduction of overall population exposure and protection of public health, control efforts must be further targeted at high-population high-pollution areas, and land-use and city planning should also encourage population to redistribute away from the highly polluted areas. (C) 2016 Published by Elsevier Inc.National Science Foundation of China [41575106, 41375103]; NSFC/RGC Grant [N_HKUST631/05]; UGC Special Equipment Grant [SEG_HKUST07]; UGC Special Research Fund Initiatives [SRFI11IO01]; ECF project [ECWW09EG04]; NSFC-GD Grant [U1033001]; Fok Ying Tung Graduate School [NRC06/07.SC01]SCI(E)EIARTICLEyingli@ust.hk13-2217