Changes in ozone production and VOC reactivity in the atmosphere of the Mexico City Metropolitan Area

Highlights

• Changes in atmospheric reactivity and P(O₃) suggest a need for new emission controls.

• Observed decreases in early morning VOC-OH reactivity with sustained high O₃ levels.

• NO₂/NO increases suggest changes in radical budgets for secondary aerosol formation.

• Marked spatial P(O₃) dependency suggests spatially different O₃ sensitivity regimes.

• Emerging sources of VOCs relevant for P(O₃) should be part of air quality management.

Abstract

The introduction of aggressive emission control policies propelled three decades ago has led to significant reductions of ozone (O₃) levels in the Mexico City Metropolitan Area (MCMA). However, in recent years the levels have stalled, and there is evidence that the production of secondary pollutants may have started to rebound. Similar responses to changes in O₃ precursors observed in other urban areas suggest major changes in atmospheric reactivity and O₃ production P(O₃), prompting the introduction of new emission control policies. In this study, we evaluate the changes in VOC-OH reactivity and P(O₃) using data of O₃, volatile organic compounds (VOCs), carbon monoxide (CO), and nitrogen oxides (NO_x) collected from the 1990s to 2019 by the local air quality monitoring network and during two dedicated 6-week field measurement campaigns in the MCMA. The results show a reduction of VOC-OH reactivity from 38 s⁻¹ in 2003–2006 to about 19 s⁻¹ in 2018 within the urban core, a slight increase from around 8 s⁻¹ to 9 s⁻¹ for the northern peripheral urban sites in the same period, and a small decrease from about 18 s⁻¹ to 15 s⁻¹ during 2014–2016 in the south of the city. Alkanes are still key contributors to VOC-OH reactivity, whereas the contribution from aromatic and alkene species has decreased, in agreement with reductions of VOC emissions from mobile sources. Changes in P(O₃) suggest that increases in the relative contributions from highly oxygenated volatile chemical products from solvents consumption, personal care and other products are responsible for sustained high O₃ levels in recent years. We also found that there is a marked spatial dependency of P(O₃) in the northern, central, and southern areas of the city, suggesting spatially different VOC and NO_x sensitivity regimes. Finally, significant increases in NO₂/NO ratios suggest changes in the pathways for the nighttime accumulation of HO_x radicals and NO_y species that could impact the morning photochemistry and radical budgets for secondary aerosol formation. These results strongly suggest a need for new measurements of radical budgets in the MCMA, the expansion of a continuous measurement network of VOCs with the addition of oxygenated species, and the use of these data to support comprehensive modeling studies for the design of effective emission control strategies.

Introduction

Ozone (O₃) concentrations in the Mexico City Metropolitan Area (MCMA) reduced substantially from the early 1990s up to about 2010, but since then the levels have stalled. Recent severe air pollution episodes suggest that the production of secondary pollutants may have started to rebound (Molina et al., 2019). Simultaneously, advancements in vehicle and emissions-control technologies and the introduction of cleaner fuels have led to drastic reductions of particle matter (PM), volatile organic compounds (VOCs), carbon monoxide (CO), and nitrogen oxides (NO_x) emissions from mobile sources (e.g., Schifter et al., 2014; Koupal and Palacios, 2019; EPA, 2019). Current CO and VOC emission rates of new vehicles are orders of magnitude smaller than in the late 1990s (e.g., Bishop, 2019), albeit impacts on total ambient organic reactivity are still not fully documented. Aggressive emission controls for electricity generation along with changes in residential, commercial, and industrial activities due to population and economic growth have also affected the long-term variability of VOC and NO_x emission sources. Nevertheless, after decades of substantial O₃ reductions, there has been little additional progress in recent years in Mexico and many parts of the world, including China (Li et al., 2019), Europe (Yan et al., 2019) and the US (Yan et al., 2018). This stagnation indicates that there are new challenges to address for reducing O₃ levels in urban areas.

The initial O₃ reductions in the late 90s and beginning of this century in the MCMA were related to aggressive emission controls of O₃ precursor species. These included improving fuels (e.g., removal of lead and reformulation of gasoline, reduction of sulfur content in gasoline and diesel), adopting catalytic converters, removing an oil refinery and heavy industries, shifting to natural gas in power plants and other industries, reformulating liquefied petroleum gas (LPG) for cooking and water heating, introducing a vehicle inspection and maintenance program, and a “no driving day” (Hoy no Circula) rule, (SEDEMA, 2019). The question of why O₃ levels have not further decreased in recent years needs to be scientifically addressed to re-design and implement new effective emission control policies. Due to its meteorological conditions and emissions characteristics, the MCMA provides a unique opportunity to investigate the factors that drive changes in O₃ levels in large urban areas, particularly of developing nations.

Formulation of successful emission control policies to reduce O₃ levels requires knowledge of the chemical regime for O₃ production P(O₃). Photochemical P(O₃) is a complex function of NO_x and hydrogen oxide radicals (HO_x = OH + HO₂ + RO₂) abundances and the partitioning of NO_x between nitric oxide (NO) and nitrogen dioxide (NO₂), and of HO_x in OH, HO₂, and RO₂. In turn, ambient levels of HO_x radicals and atmospheric reactivity depend strongly on the VOC abundance and chemical composition. In the MCMA, O₃ formation and its spatial distribution are further impacted by air mass transport patterns that are locally induced by the complex topography, high altitude, and intense solar radiation.

The topography of the MCMA is characterized by a wide-open basin in the north, with high rim mountains at the west, south, and east, and a mountain passage at the southeast that drive the meteorological transport patterns in the city (Fig. 1). The wind circulation patterns associated with air pollution events have been studied in detail (e.g., Fast and Zhong, 1998; Doran and Zhong, 2000; de Foy et al., 2005, 2006; 2008; Fast et al., 2007). These studies show that although there is some interaction with synoptic-scale circulations, most of the day-to-day variability in surface wind speed and direction is affected by mesoscale, thermally driven mountain–valley flows. Thermal flows have significant vertical shear that leads to recirculation within the basin. The high elevation (2240 m.a.s.l) and low latitude (19°N) of the basin induce strong diurnal variation of the boundary layer that increases vertical mixing. However, near-surface convergence zones due to momentum down-mixing of thermally driven winds and light synoptic winds aloft can promote the accumulation of air pollutants in the basin. The studies also suggest that the basin has relatively effective venting through the southeast passage and open northern plateau, leading to short day-to-day accumulation of pollutants.

The interaction of synoptic and locally-driven circulations leads to high seasonal and spatial variability of O₃ within the MCMA. The high O₃ season spans from mid-February to mid-June, and the highest O₃ concentrations are usually observed during the warm-dry period from March to May when high-pressure systems induce days with clear skies and weak winds (SEDEMA, 2018a). Anticyclonic westerly winds bring mostly dry air currents from November to April, whereas the rainy season from May to October is associated with moist easterly trade winds. In terms of spatial variability, since the beginning of O₃ monitoring in the MCMA in the late 80's, the data showed that O₃ concentrations are usually higher southwest of the city and lowest in the northeast due to dominant near-surface northerly winds (e.g., Bossert, 1997; Fast and Zhong, 1998). As shown in Fig. 1, high O₃ levels are still significantly more prevalent in the southwest than in the northeast, further suggesting that air transport thus far has a primary role in determining the spatial distribution of O₃ in the MCMA.

The most comprehensive studies on the P(O₃) regimes in the MCMA to date are from the MCMA-2003 and MILAGRO 2006 field campaigns (Molina et al., 2007, 2010). The MCMA-2003 field campaign involved a multinational team of experts conducting measurements and modeling studies to better understand the air pollution problems in the MCMA (Molina et al., 2007). State-of-the-art instrumentation generated extensive measurements of oxidant precursors, radicals, photochemical products, VOC and PM intermediates, and meteorological parameters. A highly-instrumented mobile laboratory was deployed for vehicle-chase measurements and at multiple locations. An eddy covariance flux tower measured fluxes of VOCs and demonstrated to be a valuable tool for evaluating the emissions inventory (Velasco et al., 2005). Some of the key results included a better understanding of the meteorology and transport of the urban plume (de Foy et al., 2006); an improved speciated emissions inventory from on-road vehicles (Zavala et al., 2006); an improved characterization of the sources and atmospheric loadings of VOCs (Velasco et al., 2007) and the first spectroscopic detection of glyoxal in the atmosphere (Volkamer et al., 2005), which was subsequently used as a tracer to estimate the relative contributions of primary and secondary formaldehyde (HCHO) (Garcia et al., 2006; Lei et al., 2009); a more comprehensive characterization of NO_x and VOCs precursors of O₃ formation (Shirley et al., 2006; Sheehy et al., 2010; Lei et al., 2007); improved knowledge of the composition and mass loadings of primary and secondary fine PM (Salcedo et al., 2006), and the finding that the formation of secondary organic aerosol (SOA) was much more efficient than predicted by traditional models (Volkamer et al., 2006).

MILAGRO (Megacity Initiative: Local And Global Research Observations) was an international scientific collaboration to examine the emissions, transformations, and the export of atmospheric pollutants from a megacity using Mexico City as a case study (Molina et al., 2010). Measurements took place over 6-week period during 2006 and included a wide range of instruments at three supersites, on six research aircraft, and satellites. Additional platforms included mobile laboratories and stations located at urban and boundary sites. The suite of measurements expanded upon those of the MCMA-2003 campaign by including measurements of targeted emission sources, chemical, physical and optical properties of aerosols, boundary layer development, and other key meteorological parameters at multiple ground sites. The measurements provided much more comprehensive information on the characterization of meteorology, emissions, chemistry, and radiation specific for Mexico City and the surrounding region. Emission inventories were re-evaluated with observations, including area-averaged emission fluxes at the ground site, aircraft measurements, and source apportionment analysis (e.g., Aiken et al., 2009; Karl et al., 2009; Velasco et al., 2009; Zavala et al., 2009). Simultaneous measurements of NO_x, VOCs, O₃, HO_x, as well as several radical precursors such as HONO, HCHO, and glyoxal, provided the observations needed to constraint modeling studies for improving our understanding of O₃ formation and its sensitivity to precursors and SOA evolution (e.g., Li et al., 2011; Song et al., 2010). Urban plumes were detected at distances up to 1000 km downwind with continued O₃ production for several days due to the formation of peroxyacetyl nitrates (PANs) and effectively increasing the NO_x atmospheric lifetime (Mena-Carrasco et al., 2009).

Field observations and modeling exercises of radical production rates, as well as NO_x termination reactions, VOC/NO_x and CO/NO_x ratios, and chemical tracers during both dedicated field campaigns, all indicated that P(O₃) was highly VOC-limited (i.e., NO_x-saturated) within the urban core, while the surrounding low-NO_x areas were either NO_x- or VOC-limited depending on the meteorological conditions (Lei et al., 2007, 2008; Song et al., 2010). However, the recent reversal from a downward O₃ trend together with changes in VOC and NO_x emissions indicates the need to re-evaluate the O₃ formation regimes.

In this study, we investigate the reasons for the slow progress in O₃ abatement for the MCMA in recent years by analyzing the changes in P(O₃) regimes and VOC-OH reactivity. The study is based on VOC, CO, and NO_x data collected from the 1990s to 2019 by the local air quality monitoring network and during the two dedicated air quality field campaigns described above. Changes in atmospheric reactivity are assessed by recent and past VOC measurements, while changes in P(O₃) across the city are evaluated under a pseudo-stationary state (PSS) using as reference concurrent NO levels to understand potential impacts of current and future reduction emission policies. In addition, the contributions of locally and regionally produced O₃ are investigated.