Recent studies have shown that cars pollute more on the road thanthey do on paper. Computer models based on federal laboratory tests and analyses seem to underestimate on-road emissions, yet those models help guide air-quality regulations affecting transportation. Researchers around the country are attempting to improve the accuracy of these computerized emission models. UC Davis researchers have developed a new way of looking at one piece of the model -- speed-related emissions. Many air-quality plans rely on improved traffic flow and increased average operating speeds in normally slow areas for emissions control. Within the current system of federal testing and analysis, the UC Davis researchers discovered that speed-related emissions cannot be predicted well by the computer models now used by regulators. "While we know that improving vehicle flow will reduce emissions," says researcher Randall Guensler, "the question now is how much and at what speeds." In future studies, Guensler and his colleagues hope to further refine the computerized model to include other factors that influence emissions -- acceleration, deceleration, idle and cruise speed -- which may be better indicators of on-road emissions than average speed. Guenslerr will present the paper in session No. 73 at 7:30 p.m., Monday, Jan. 11, at the Hilton.