Potential of big data approach for remote sensing of vehicle exhaust emissions

Abstract

Abstract At present, remote sensing (RS) is applied in detecting vehicle exhaust emissions, and usually the RS emission results in a definite vehicle specific power (VSP) range are used to evaluate vehicle emissions and identify high-emitting vehicles. When the VSP exceeds this range, the corresponding vehicle emission RS data will not be used to assess vehicle emissions. This method is equivalent to setting only one VSP Bin qualified for vehicle emission evaluation, and generally only one threshold limit is given for each emission pollutant without considering the fluctuation characteristics of vehicle emissions with VSP. Therefore, it is easy to cause misjudgment in identifying high-emitting vehicles and is not conducive to scientific management of vehicle emissions. In addition, the vehicle emissions outside the selected VSP Bin are more serious and should be included in the scope of supervision. This research proposed the methods of vehicle classifications and VSP Binning in order to categorize the driving conditions of each kind of vehicles, and a big data approach was proposed to analyze the vehicle emission RS data in each VSP Bin for vehicle emission evaluation.