Spatio-Temporal analysis of PM2.5 using MODIS, MISR and SeaWiFS imagery over Lahore division

Abstract

Particulate matter (PM2.5) is considered a major cause of mortality worldwide. In Pakistan, specifically in the industrial and capital city of Lahore, the air quality is so deteriorated that the provincial government closed schools and other institutions. A study was designed to (1) investigate the decadal trends of PM2.5, urbanization, and (2) model metrological factors responsible for PM2.5 over the Lahore division for the long period 1998-2019. The PM2.5 temporal change was analyzed using the non-parametric Mann-Kendall trend test. Based on Mann Kendall test results Lahore division was divided into three regions: low, medium, and PM2.5 polluted sites to know urbanization change through non raster Mann Kendall test. The impact of metrological factors was analyzed by multiple linear regression. The results highlighted that all the cities except Nankana sahib experienced a significant (p ≤ 0.05) increase in this period, with highest values of 113 μgm-3 and lowest values of 49 μgm-3. The air quality of the Lahore division was worse in Lahore, Okara and some parts of Sheikhupura and Kasur, showing the highest PM2.5 concentrations of 113 μgm-3. Urbanization in low PM2.5 polluted site showed no significant (p ≥ 0.05) trend with Z=-1.04, tau=0.33, P=0.29, sens slope=809 moderate PM2.5 polluted site showed increasing trend with Z=2.08, tau=1, P=0.03, sens slope=1506 while high PM2.5 polluted site showed increasing trend with Z=2.5, tau=3, P=0.01, sens slope=3236 respectively. The correlation coefficients of PM2.5 concentration with temperature, wind speed, and precipitation are -0.53, -0.41 and -0.65, respectively, showing strong negative correlations. The correlation coefficients of PM2.5 concentration with relative humidity and pressure was 0.69, showing positive correlation and 0.28, respectively. Multiple linear regression showed that metrological factors have 82% tendency to alter the concentrations of PM2.5.