This study aims to analyze the regional variation in the source of air pollution, identify the percentage contribution of each pollutant, and distribute the mass contribution of each source category using multivariate analysis. The nine air monitoring sites were successfully divided into three groups using hierarchical agglomerative cluster analysis (HACA) (clusters 1, 2, and 3). The collected meteorological data is non-parametric data for the years 2020–2021 which includes PM2.5, PM10, SO2, NO2, NO, NOx, CO, wind speed, humidity, wind direction, temperature, cloud cover, and surface radiation. The most major air pollution sources were identified using FA. Multiple linear regression (MLR) and principal component regression (PCR) were utilized to create an equation model explaining contaminants' impact in each cluster. However, it was shown that the most important pollutants impacting the value of the air pollutant index (API) are gaseous pollutants (NOx and SO2) and particulate matter (PM10 and PM2.5). Gas and non-gas pollutants have a 65% influence on cluster 1 and meteorological conditions have a 35% effect. Cluster 3 is influenced by 65% particle and non-gas pollutants and 35% weather conditions, compared to Cluster 2 which is 100% affected by gas and particulate pollutants because of its spatial location. This study shows the value of the multivariate modeling technique in minimizing the time and expense associated with monitoring redundant stations and parameters.