Dear editor,rnThe existing methods for regional vehicle emission prediction can be roughly categorized into the classes of classical dispersion models and satellite remote sensing models.Gaussian plume models, operational street canyon models and computational fluid dynamics are the classical dispersion models.The mobile source emission factor (MOBILE) and computer programme to calculate emissions from road transport (COPERT) models that have been developed in USA and Europe respectively are the most frequently used emission factor models [1].These models [2, 3] are usually complex models that consider the meteorology, road network geometry, geographical locations, traffic volumes, and emission factors based on a several empirical assumptions and parameters that may not be applicable to all the regions in a city [4].The aforementioned parameters are usually considered to be difficult and expensive to obtain, and the results that are generated using these parameters may be inaccurate.Further, satellite remote sensing of the surface air pollution has been extensively investigated during previous decades [5], and can be considered to be a top-down method.However, such approaches are extremely influenced by the presence of clouds and are considered to be sensitive to other environmental factors such as humidity, temperature, pressure and geographical location [6, 7].