Though government is enforcing various policies to reduce greenhouse gases and air pollutants emitted by vehicles, it is required to be preceded to understand the emission source in order to enforce better policies and improve evaluation of effect. Emission is calculated by activity data and emission factor. Korea is currently calculating the emission based on the total consumption of fuel by vehicle and top-down model of IPCC default emission factor. Such calculation method has advantages on understanding the aspect of nation wide emission. However, it has its limitation of reading the characteristics of on-road emission in space and time dimension. This revitalized studies on bottom up approach (BUA) which is calculating the emission from the activity data such as average speed, driving distance, traffic volume, number of vehicle registered, etc.
The objective of this research is to analyze emission of greenhouse gases and index based on newly developed unit of activity data that is based on 1 trip driving information from ignition key on to key off. The unit of activity data which is required to calculate the emission of greenhouse gases and air pollutants is analyzed by travel characteristics of actual driving data of every second gathered from OBD II.

The actual driving data is gathered from 10 recreational vehicles (RV) from July 1st of 2012 until May 31st of 2013 via mobile communication network. The vehicles ran 9,679 times in total which averaging in 17.01 km distance and 29 minutes and 12 seconds in duration resulting in fuel efficiency of 11.74 km/L. In particular, the fraction of driving distance driven with cold engines was 14.18 %. 50 % of the trips are found to have less than 10.829 km in distance and 25 % are found to have less than 3.028 km.

10 vehicles have emitted 37,419 kg CO2 eq. in total, calculated from fuel consumption, resulting in 98.29 % of total CO2 emission, 0.11 % of total CH4 emission and 1.60 % of total N2O emission. 10 vehicles have emitted 31,310 kg CO2 eq. in total, calculated from average speed and driving distances, resulting in 99.81 % of total CO2 emission, 0.03 % of total CH4 and 0.16 % of total N2O emission.

In comparison of emission calculation method, the method based on the fuel consumption was 120.22 % of the method based on the average speed. Especially CH4 and N2O had the most wide difference of 405.02 % and 1,222.17 %. Correlation of coefficient of determination (R2) of emission CO2 eq. was 0.9782 which shows high correlation. CO2, CH4 and N2O had also high correlation value of 0.9764, 0.9645, 0.9775. However, the regression slope of CH4 and N2O had 0.2817 and 0.0734 which was resulted in uncertainty of IPCC default emission factor of 1.6 ～ 9.5 kg/TJ of CH4 and 1.3 ～ 12 kg/TJ of N2O. It is because the emission factor represents the average of the worldwide.

Greenhouse gas emission index based on fuel consumption was calculated to be 227.24 g CO2 eq./km and 2.21 g CO2 eq./sec. Emission index calculated from average speed and driving distance was 189.01 g CO2 eq./km and 1.84 g CO2 eq./sec.

Accurate emission should be calculated in prior to evaluation of effect and efficient policy enforcement for adaptation of climate change and reduction of air pollutants. Activity data and emission calculation method developed as the result of the research on mobile emission on roads reduces the uncertainty of national statistics and improves calculation accuracy which can make better evaluation of effect and policy enforcement.