Sensitivity of a Chemical Mass Balance model for PM2.5 to source profiles for differing styles of cooking

Use of a Chemical Mass Balance model is one of the two most commonly used approaches to estimating atmospheric concentrations of cooking aerosol. Such models require the input of chemical profiles for each of the main sources contributing to particulate matter mass and there is appreciable evidence from the literature that not only the mass emission but also the chemical composition of particulate matter varies according to the food being prepared and the style of cooking. In this study, aerosol has been sampled in the laboratory from four different styles of cooking, i.e. Indian, Chinese, Western and African cooking. The chemical profiles of molecular markers have been quantified and are used individually within a Chemical Mass Balance model applied to air samples collected in a multi-ethnic area of Birmingham, UK. The model results give a source contribution estimate for cooking aerosol which is consistent with other comparable UK studies, but also shows a very low sensitivity of the model to the cooking aerosol profile utilised. A survey of local restaurants suggested a wide range of cooking styles taking place which may explain why no one profile gives an appreciably better fit in the CMB model.