Dear Dr. Rao,
Thank you for the opportunity to respond to the comments from Dr. Nam regarding our paper. The MOVES model is impressive; it was undoubtedly a massive undertaking by the U.S. Environmental Protection Agency (EPA) in developing such a comprehensive model for U.S. vehicles. It is a noteworthy improvement over the MOBILE6 model, which it replaced, especially regarding project-level air quality analyses, the subject of our paper. We thank Dr. Nam for letting us know that, for some reason, “A design decision was made” to ignore acceleration during the first second of a drive cycle. We completely agree that setting acceleration to zero for 1 second makes an insignificant difference to total emissions over longer drive cycles, those approaching 1 hour. However, for short drive cycles representing dynamic modal vehicle activity on roadway segments near signalized intersections, this is a shortcoming of the MOVES model. Short drive cycles are repeated in the MOVES model to represent an hour simulation, which increases the effect of ignoring acceleration during the first second of the drive cycle. Developing a project-specific operating-mode distribution external to MOVES is a way to avoid the internal conversion of a user-specified drive cycle to an operating mode distribution—provided that the MOVES computation procedure of ignoring acceleration during the first second of the drive cycle is not implemented in the external process.
Further, we disagree that this shortcoming poses largely a theoretical problem. It is not intuitive that different results would be obtained when dividing a vehicle trajectory into smaller segments. As we found during our work, the length of the link does produce significantly different results (very short cycles do occur near intersections). To best describe vehicle activity near an intersection or busy roadway, it seems to us that links should be significantly shorter than a representative 1-hour drive cycle. This yields better precision both spatially and temporally. We believe that this was demonstrated in our paper and that such research will enlighten policy.
Over the past 20 years, EPA mobile source emission models have been quite good for doing emissions inventory work. However, they have not been as good for doing intersection modeling. Intersection (i.e., “hotspot”) modeling is one of the primary uses of a mobile source model. As we know, idle emission factors (EFs) are extremely important in CAL3QHC, the EPA-approved dispersion model for intersections. Neither MOBILE5 nor MOBILE6 would give EFs directly; both versions required one to run the model at 2.5 mph and then multiply that EF by 2.5 mph to obtain the idle EF. MOVES is better in this regard, but still it seems that its prime orientation is towards emission inventory work; hot-spot modeling is still treated as a secondary concern.
We are glad to hear that EPA is considering improving the VSP algorithms to better accommodate very short drive cycles. As for entering second-by-second drive cycles, practitioners often do not have the time to develop specific drive cycles for each project level analysis. Rather, they rely on EPA defaults when analyzing hotspots, and we would humbly suggest that one or more default short-cycle drive cycles be developed that would be appropriate to intersection analyses.
Sincerely,
Mark Ritner, Kurt K. Westerlund, C. David Cooper, and Michael Claggett
University of Central Florida