Modeling Mobile Source Emissions Based Upon In-Use and Second-by-Second Data : Development of Conceptual Approaches for EPA ’ s New MOVES Model

EPA is undertaking an effort to develop a new set of modeling tools for the estimation of emissions produced by on-road and off-road mobile sources. The product of this effort will be the Multi-scale mOtor Vehicle & equipment Emission System, hereby known as MOVES. The design of MOVES is guided by the following four considerations: (1) the model should encompass all pollutants (e.g., HC, CO, NOx, particulate matter, air toxics, and greenhouse gases) and all mobile sources at the levels of resolution needed for the diverse applications of the system; (2) the model should be developed according to principles of sound science; (3) the software design of the model should be efficient and flexible; and (4) the model should be implemented in a coordinated, clear, and consistent manner. The key purpose of this paper is to demonstrate methods for developing modal emission rates from on-board data and laboratory second-by-second data for a 'pilot' dataset of light duty gasoline vehicles (LDGV). This activity builds upon previous work at NCSU and elsewhere in which similar modeling approaches have been demonstrated for LDGVs, heavy duty diesel vehicles and non-road diesel vehicles (e.g., construction equipment). The conceptual model development is focused upon HC, CO, NOx, and CO2 hot stabilized LDGV tailpipe emissions. The methods demonstrated here include the development of 'modes' based upon vehicle specific power (VSP). The modes can be weighted to represent different speed traces and, therefore, can be used to support both mesoscale and macroscale estimation of emission inventories. Methods for characterization of variability and uncertainty for the conceptual models, and validation of the conceptual models versus independent data, are discussed. Based upon the results of the conceptual model development, recommendations are made regarding future development of MOVES.