The traditional methodology for estimating vehicle emissions based on vehicle miles traveled and average speed is not reliable because it does not consider the effects of congestion, control devices, and driving mode (cruise, acceleration, deceleration, and idle). The authors developed an analytical model to predict vehicle activity on signalized arterials with emphasis on oversaturated traffic conditions. The model depends only on loop detector data and signal settings as inputs and provides estimates of the time spent in each driving mode, which consequently leads to more accurate...