Evaluation of mobile emissions contributions to Mexico City ' s emissions inventory using on-road and crossroad emission measurements and ambient data

Mobile emissions represent a significant fraction of the total anthropogenic emissions burden in the Mexico City Metropolitan Area (MCMA) and, therefore, it is crucial to use top-down techniques informed by on-road exhaust measurements to evaluate and improve traditional bottom-up official emissions inventory (EI) for the city. We present 5 the measurements of on-road fleet-average emission factors obtained using the Aerodyne mobile laboratory in the MCMA in March 2006 as part of the MILAGRO/MCMA2006 field campaign. A comparison of our on-road emission measurements with those obtained in 2003 using essentially the same measurement techniques and analysis methods indicates that, in the three year span, NO emission factors remain within 10 the measured variability ranges whereas emission factors of aldehydes and aromatics species were reduced for all sampled driving conditions. We use a top-down fuel-based approach to evaluate the mobile emissions from the gasoline fleet estimated in the bottom-up official 2006 MCMA mobile sources. Within the range of measurement uncertainties, we found probable slight overpredictions of 15 mean EI estimates on the order of 20–28% for CO and 14–20% for NO. However, we identify a probable EI underprediction of VOC mobile emissions between 1.4 and 1.9; although estimated benzene and toluene mobile emissions in the inventory seem to be well within the uncertainties of the corresponding emissions estimates. Aldehydes mobile emissions in the inventory, however, seem to be under predicted by factors of 20 3 for HCHO and 2 for CH3CHO. Our on-road measurement based estimate of annual emissions of organic mass from PM1 particles suggests a severe underprediction (larger than a factor of 4) of PM2.5 mobile emissions in the inventory. Analyses of ambient CO, NOx and CO/NOx concentration trends in the MCMA indicate that the early morning ambient CO/NOx ratio has decreased at a rate of about 25 1.9 ppm/ppm/year over the last two decades and that the decrease has been driven by reductions in CO levels rather than by NOx concentration changes, suggesting that the relative contribution of diesel vehicles to overall NOx levels has increased over time

Inventories. Based on these facts it should be accepted in the ACP journal. The authors seem to be experienced and well versed in different aspects of atmospheric chemistry, emissions measurements and car emissions, thus providing insightful conclusions. Nevertheless the following general comments should be considered: 1. -I think that the decrease in percentage of vehicles without emission controls thanks to the introduction of newer cars is as relevant as the rate of removal of older cars. It should be a two thronged policy. I disagree with the conclusion that the introduction of new cars with better emissions controls is not as relevant.
[Response] The decrease in the percentage of vehicles without emission control technologies occurs by the two reasons mentioned: the introduction of new cars and the removal of older cars. Our point is that, in terms of reducing mass emissions, a policy strategy aimed for a faster-than-normal removal of old cars would be highly efficient. Both effects are somewhat related because the removal of old cars implies faster introduction of newer cars andchanges in transportation mode. 2. -Although the increase of Diesel sales relative to gasoline is a hard number, other reasons could add to the explanation on the increase of NOx levels such as: Proliferation of air conditioning availability in private cars in Mexico City. This adds to the burden (about 2KW) of internal combustion engines increasing NO emissions. NO removal efficiency of catalytic converters tend to decrease faster than for HC's and CO. Related to this is the contents of sulfur in the gasoline of Mexico that may change the behavior and reduce the overall efficiency of catalytic converters. These issues are not commented in the text.
[Response] We thank the reviewer for suggesting these two points, we have added C2850 ACPD 9, C2849-C2853, 2009 Interactive Comment Full Screen / Esc

Printer-friendly Version
Interactive Discussion Discussion Paper them to the discussion of NOx variability in the revised manuscript.
3. -The geographical distribution of brand and age of cars in Mexico City (as in many cities) tends to be inhomogeneous: affluent neighborhoods will have newer less polluting cars, with different CO, CO's and other ratios than in poor neighborhoods. This fact adds to the uncertainty of conclusions based on-road and cross-road measurements and in general to all efforts of obtaining EI. In the text it should be mentioned if the authors considered this aspect in their field experiments and if so how.
[Response] That is absolutely true. Our study does not specifically explore that effect and a statement on that regard has been added. From a series of remote sensing measurements at four sites in Mexico (Schifter et al, 2003;2005) we can observe that the measured emission ratios are consistently between the 95 percentile CI among the sites. It would be interesting to design an experiment to explore this topic. Some specific comments are: In the abstract paragraph 15 p.6364 ovepredictions are expressed as a percentage and underpredictions as a factor. Same in paragraph 15 p.6381. Authors should be consistent.
[Response] Very often studies refer to over-predictions in that way, so we prefer to keep it as it is.
In paragraph 10 p.6365, and paragraph 5 p.6367 the authors should give numeric data when referring to contribution of pollutants in Mexico City by mobile sources.
[Response] We have included these emissions contributions in the revised manuscript. In paragraph 5 or 10 p. 6366 the following reference on the topic of motor stress and emissions should be included: A Vehicle Emissions System Using a Car Simulator and a Geographical Information System: Part 1âËŸAËGTSystem Description and Testing Aron D. Jazcilevich, Alejandro García-Fragoso, Agustín García Reynoso, Michel Grutter, Ulises Diego-Ayala, Jim Lents and Nicole Davis Journal of the Air Waste Management Association, Volume 57, October 2007 Also a reference on this topic is the International Vehicle Emission Model of ISSRC, http: www.issrc.org currently being used to obtain vehicle emissions in Mexico City taking into account motor stress.
[Response] Thanks for the references.
In paragraph 5 p. 6371 are there or are there not sufficient data to compare emissions between 2003 and 2006? At some points the authors say there are and here they hesitate.
[Response] We have clarified that statement indicating that it is referring to the observed changes (not data) of on-road measurements of NO only.
In paragraph 10 p.6372 it is stated that VOCs increased 4.6 percent a year in Mexico City. This contradicts previous statements in the paper.
[Response] Please see our response for reviewer 1 on the same issue.
In paragraph 15 p. 6375 it should be revised: The sentence does not give a context of the aircraft measurements. Interactive comment on Atmos. Chem. Phys. Discuss., 9, 6363, 2009.