Satellite-based estimates of outdoor particulate pollution (PM10) for Agra City in northern India

Air quality of north Indian cities has worsened over the last few decades which has been posing a great risk to consequential health-related issues. Ground-based monitoring of particulate matter smaller than 10 mu m (PM10) in Indian cities has been limited to few selective sites at local hot spots, and thus, related health studies at regional scale were constrained. To overcome this issue, we utilized the aerosol optical depth (AOD) from Moderate Resolution Imaging Spectroradiometer (MODIS) onboard EOS Terra and Aqua satellites to estimate the regional PM10 concentration in Agra City located in the northern part of India. The approach envisaged the developments of linear, log-linear, and multi-linear regression models to estimate PM10 using AOD(MODIS) and in situ measured meteorological parameters by utilizing the data of years 2010 and 2011. The results indicated that both hourly and 24-h average PM10 had a weak correlation with AOD(MODIS) when chosen as the only regressor. However, hourly PM10 showed better correlation with AOD(MODIS) (R similar to 0.45) than 24-h average PM10 (R similar to 0.24). The log-linear estimation of PM10 utilizing AOD(MODIS), relative humidity, wind speed, and atmospheric temperature as regressors had the highest correlation coefficient (R = 0.81) and a minimum relative standard error as 8.93 %, and thus, it was able to provide the best estimates of PM10 among all the models considered in this study. However, the model adequacy checks suggested the further scope of strengthening of these linear and log-linear models by adopting some suitable transformations in them.