The efficiency of the CMB model for source contribution estimate is determined by the availability of accurate source profiles and therefore its use is limited when source profiles are not known. In this study the source contribution estimate of ambient VOC was carried out using the CMB model with the source profiles obtained from the PMF model. The PMF model fitted the monitored dataset very well with five factors. The source profiles generated by the PMF model resembled most closely the profiles of petroleum refineries, automobiles, wastewater treatment plants, solvent using industries, and solid waste burning. The derived source types match well with the prevailing source types in the study area. Using the source profiles developed by the PMF model, the CMB model could estimate source contributions for more than 80% of samples with
³ 0.9. The integrated method of developing source profiles with the PMF model and using them in the CMB model for estimating source contributions for each measured dataset proved to be successful and can be adopted in any location where source profiles are not known. This method also eliminates the need for ageing the source profiles to take care of atmospheric photochemical reactions.
Keywords: source profile, source apportionment, PMF, VOC, chemical mass balance, CMB, source type