Vehicle emission is a major source of air pollution in Dhaka. Old fleet, lack of maintenance, improper traffic and parking management, overloading, fuel adulteration etc. are responsible for high emissions from the vehicle sector. In this study, vehicle emissions have been measured on-road in Dhaka using an Automotive Gas Analyzer and Smoke Opacity Meter to determine the existing vehicle emission scenario in the city. Concentrations of carbon monoxide (CO) and hydrocarbons (HC) in the emissions from CNG/gasoline vehicles, and opacity of the emissions from diesel vehicles were measured. The results were compared with the corresponding national limit values. It was found that all types of CNG vehicles performed very well with more than 80% satisfying the corresponding limit values. Private cars ranked at the top in performance among the CNG/gasoline vehicles. Diesel vehicles were found as the worst polluters in the vehicle sector; emissions from about 75% of the diesel vehicles had opacity more than 65 HSU, the national limit value for emissions from diesel vehicles. Motor cycles were also highly polluting; 60% of the motor cycles emitted CO and HC concentrations higher than the respective national emission limit values.

Decomposition and transformation algorithms are important tools in data analyses, including air pollution data processing, which has been a principal environmental concern for many decades. One of the main sources of air pollution is vehicular emissions, which contain harmful greenhouse gases. In addition to such factors as speed and acceleration affecting the quantity of emissions and fuel consumption, pavement roughness is an indirect factor. The portable emissions measurement system (PEMS) is used to collect on-road emissions data every second, and smartphone apps such as Roadroid are used to estimate road roughness. These data can be combined and correlated with vehicular activities such as speed, acceleration, and vehicle specific power and analyzed in detail to reveal the actual impact of roughness on emissions. However, analytical modeling can provide only an average estimation and does not retain local roughness and how it affects emissions. Such detailed information is necessary for pavement diagnosis and maintenance and emissions control. This paper presents a discrete wavelet transform (DWT) procedure that can offer local and supplementary relationships between pavement roughness and emissions that other simplified techniques cannot. PEMS data of four road segments in greater Houston, Texas, were considered to be decomposed by one-dimensional DWT for analysis of the wavelet subbands and their energies. The results of this study can be used in planning to minimize emissions with consideration for pavement conditions.