Extracting relationship between air pollution and precipitation using spatio-temporal analysis in Tehran metropolis
Keywords: Precipitation, Air pollution, Pearson correlation coefficient, Polynomial regression, Modelling
Abstract. Precipitation is one of the main stages of the water cycle, and it is required for the organisms to survive on the planet. In contrast, air pollution is a phenomenon that has greatly affected the human life nowadays. Population growth, development of factories and increasing number of fossil fuel vehicles are the most influencing factors on air pollution. In addition to understand nature of precipitation and air pollution, finding relationship between these two phenomena is necessary to make appropriate policies for reducing air pollution. Furthermore, studying trends of precipitation and air pollution in the past, is helpful to forecast the times and places with less precipitation and more air pollution for a better urban management. In this study, we tried to extract any probable relationship between these two parameters by investigating their monthly measured amounts in 22 municipal districts of Tehran in three epochs of time (2009, 2013 and 2017). Carbon Monoxide (CO) was considered as the indicator of air pollution. Results of the study show that the parameters have a significant relationship with each other. By using Pearson Correlation Coefficient and One-Way Variance (ANOVA) test, relationship between the data for each month and for each district of Tehran were studied separately. As the time has passed and the air pollution has increased, the correlation between the parameters in districts has decreased. In addition, during the cold months of the year, the correlations decrease since the fact that precipitation is not the only influencing factor on the air pollution due to the rise of air “Inversion”. Finally, the polynomial regression model of carbon monoxide based on precipitation was extracted for each of the three years. The model suggests a degree three polynomial equation. The obtained coefficients from the regression model show that the relationship between parameters was stronger in the years with more rainfalls. This can be due to the more significant impact of other influencing factors on air pollution, such as population density, wind direction, vehicles and factories in the areas or conditions with a less rainfall.