Publications

You, W; Zang, ZL; Zhang, LF; Zhang, M; Pan, XB; Li, Y (2016). A nonlinear model for estimating ground-level PM10 concentration in Xi'an using MODIS aerosol optical depth retrieval. ATMOSPHERIC RESEARCH, 168, 169-179.

Abstract
Satellite measurements have been widely used to estimate particulate matters (PMs) on the ground and their effects on human health. However, such estimation is susceptible to meteorological conditions and may result in large errors. In this study, we developed a nonlinear empirical model for seasonal ground-level PM10 prediction in Xi'an, Shaanxi province of northwestern China. The nonlinear model is based on 3 years (2011-2013) of daily PM10 concentration data from 13 PM10 monitoring stations in Xi'an, aerosol optical depth (AOD) data derived from the Moderate Resolution Imaging Spectroradiometer (MODIS), surface meteorological measurements, and NCEP/NCAR reanalysis data. The nonlinear model corrects the AOD data using the height of plenary boundary layer and surface relative humidity, and further adjusts the corrected AOD according to visibility, surface temperature and surface wind speed. Our results show that there is almost a threefold improvement from 0.28 to 0.78 in the correlation coefficient when using the nonlinear model compared to using a linear regression model of AOD and PM10. The root-mean-square error (RMSE) is reduced from 34.42 to 21.33 mu g/m(3) using the nonlinear model over the linear model. Further analysis about meteorological variables shows that relative humidity and visibility are important factors to improve the relationship between AOD and PM10. The relationship between the predicted PM10 concentration from the nonlinear model and observed PM10, concentration is the best in winter, moderate in autumn and spring, and poor in summer. Further validation has shown that the nonlinear model is able to explain approximately 79% (R-2 = 0.79, n = 270, p < 0.01) of the variability in the monthly-mean PM10 concentration with an RMSE of 11.7 mu g/m(3) and mean absolute percentage error of 14.2% based on monthly-mean data set. These results are useful for accessing surface PM10 concentration and monitoring regional air pollution. (C) 2015 Elsevier B.V. All rights reserved.

DOI:
10.1016/j.atmosres.2015.09.008

ISSN:
0169-8095