Health & Environmental Research Online (HERO)


Print Feedback Export to File
2461672 
Journal Article 
Estimation of atmospheric particulate matter based on MODIS haze optimized transformation 
Wang, Q; Zha, Y; Gao, Jay; Shen, Dan 
2013 
Yes 
International Journal of Remote Sensing
ISSN: 0143-1161
EISSN: 1366-5901 
34 
5 
1855-1865 
WOS:000323247900019 
This research is an attempt to simulate the relationship between haze optimized transformation (HOT) and aerosol optical thickness (AOT), and explore the influence of typical ground covers on this relationship using the 6S atmospheric radiative transfer model for the Chinese city of Nanjing. The HOT data were derived from moderate resolution imaging spectroradiometer (MODIS) satellite images recorded in the winter and spring seasons of December 2007-May 2009. They were analysed in conjunction with ground observed atmospheric particulate matter (PM) data so as to establish their quantitative relationship. Such a relationship may open a new avenue for remotely estimating atmospheric PM based on HOT. The results obtained indicate that HOT is related positively to AOT. This relationship is most accurately depicted by a second-order polynomial equation. Although built-up areas, waterbodies, and vegetation have differing HOT values, all of them bear a close and consistent correlation with AOT. HOT of built-up areas, waterbodies, and vegetative surfaces derived from MODIS images is also positively correlated with PM10 (PM with diameter <10m), which was measured near the surface. The second-order polynomial equation has a coefficient of determination (R-2) value of 0.375 (built-up), 0.344 (water), and 0.362 (vegetation) and a root mean squared error (RMSE) of 0.0258, 0.0264, and 0.0261, respectively. The closeness in R-2 value and RMSE for different ground covers suggests that correlation is marginally affected by the ground cover. It is thus concluded that HOT can be used as a reliable alternative for estimating PM10 from MODIS data.