Near-road dispersion modeling with CAL3QHC has traditionally been accomplished by assuming vehicles are either idling in queue links or flowing freely in cruise links. With the introduction of the new mobile-source emissions model MOVES, second-by-second activity patterns can be used to produce emission factors (EFs) that vary by vehicular modal activity, that is, acceleration, deceleration, idle, and cruise. By using these EFs in unique modal links in CAL3QHC input files, the predicted concentration of pollutants near roadways can be modeled with greater precision in regard to real-world intersection vehicle behavior. It is noted that this work does not include any comparisons with real-world monitored data, and thus only the precision and not the accuracy of the proposed method is addressed. This work poses the question of how best to include modal links into near-road dispersion modeling. Specifically, it examines dividing acceleration and deceleration segments into multiple sublinks for greater resolution. It is shown that such an approach can produce much higher CO predictions at an intersection (up to 400% higher) compared with the current cruise-and-idle-links modeling approach. A method of dividing links by increments of speed change is suggested. The method relies upon obtaining EFs from standstill to various cruise speeds (or from cruise speed to stopped) and using those results to obtain position-specific acceleration (or deceleration) EFs needed for dispersion modeling inputs. Acceleration EFs (in g/mile) are an order of magnitude larger than cruise EFs; deceleration EFs are smaller than cruise EFs. The number of sublinks used to model one acceleration link makes a difference in the predicted concentrations. MOVES can produce erratic EFs when longer links are broken into smaller sublinks. Implications: MOVES now affords modelers the ability to account for the varying vehicle emissions that occur under various modal operations including acceleration, deceleration, idle, and cruise. These modal activities are dominant near large intersections, and can greatly affect the results of dispersion modeling to predict near-road concentrations of various pollutants. More accurate near-road dispersion modeling techniques are important to engineers and planners for determining which projects likely may cause exceedances of National Ambient Air Quality Standards (NAAQS). The method described herein is also applicable to the modeling of PM2.5 and NOx, which is likely to be of more interest than CO in the near future.
