Measurements of trace gases were taken onboard the NASA DC-8 during the Deep Convective Clouds and Chemistry (DC3) field project with the goal of understanding the role that midlatitude deep convection plays in altering the vertical distribution of atmospherically-relevant species. Measurements of VOCs were obtained via UC Irvine’s whole air sampler (WAS) instrument, while measurements of CH4, O3, NOx, N2O, water vapor, CO, and meteorological variables were performed by a variety of other instruments operated by collaborators onboard the DC-8.

Using known VOC atmospheric lifetimes and measured VOC mixing ratios in the PBL, a tracer for rapid vertical lofting of air from the planetary boundary layer (PBL) to the upper troposphere/lower stratosphere region (UT/LS) by convection was created. In this study, it was found that light hydrocarbons associated with oil and natural gas (O&NG) and vehicular sources were widespread throughout the PBL of the DC3 study regions. In the UT/LS, enhanced levels of these light hydrocarbons were strongly correlated with water vapor, indicating a convective source. On the other hand, decreases in the measured mixing ratios of CFCs, HCFCs, and other long-lived halocarbons (LLHCs) in the UT were used as tracers for stratosphere-to-troposphere transport (STT). These two sets of tracers were used to divide the DC3 WAS merge into many subsets of data corresponding to: the PBL, convective outflow in the UT, convective outflow in

the LS (i.e. overshooting tops), STT-influenced air in the troposphere, background UT air, and background LS air.

Using these derived subsets of data, interactions and mixing between stratospheric intrusions and tropospheric air masses was investigated. A large number of stratospherically-influenced samples were found to have reduced levels of O3 and elevated levels of CO (both relative to background stratospheric air); indicative of mixing with anthropogenically-influenced air. Using n-butane and propane as tracers of anthropogenically-influenced air, it is shown that this type of mixing was present both at low altitudes and in the UT. At low altitudes, this mixing resulted in O3 enhancements consistent with those reported at surface sites during deep stratospheric intrusions, while in the UT, two case studies were performed to identify the process by which this mixing occurs. In the first case study, stratospheric air was found to be mixed with aged outflow from a convective storm, while in the second case study, stratospheric air was found to have mixed with outflow from an active storm occurring in the vicinity of a stratospheric intrusion. From these analyses, it was concluded that deep convective events may facilitate the mixing between stratospheric air and polluted boundary layer air in the UT. Throughout the entire DC3 study region, this mixing was found to be prevalent: 72% of all samples that involve stratosphere-troposphere mixing show influence of polluted air. Applying a simple chemical kinetics analysis to these data, it is shown that the instantaneous production of OH in these mixed stratospheric-polluted air masses was 11 ± 8 (± 1σ) times higher than that of stratospheric air, and 4.2 ± 1.8 times higher than that of background upper tropospheric air, which could result in a quick, high magnitude pulse of O3 production and reduced lifetimes of OH-controlled species in the UT.

These derived subsets of data were used to investigate the effects of deep convection on mixing ratios of organic chlorine and organic bromine in the UT and LS over the DC3 study region. In the LS, it was found that mixing ratios of organic chlorine in overshooting tops were higher than mixing ratios of organic chlorine in the background LS by an average of 217 ± 179 pptv (6.3 ± 5.2% enhancement), while the total organic bromine mixing ratio in overshooting tops was higher than that of the background LS by an average of 2.8 ± 3.2 pptv (17.4 ± 19.9% enhancement). In both cases, short-lived halocarbons made up a large portion of this enhancement. In the UT, convection was found to play a much more complicated role on the organic halogen content of the region. Model back trajectories and analysis of the chemical composition of the background UT revealed that long-range transport of outflow from East Asia and from the central Pacific affected the background UT of the DC3 study region to varying degrees on different days. When the background UT was affected by East Asian outflow, mixing ratios of organic chlorine in convective outflow were lower than those of the local background UT by up to 150 ± 115 pptv (3.7 ± 2.9% enhancement). On the other hand, when the background UT was affected by clean outflow from the central Pacific, mixing ratios of organic chlorine in convective outflow were higher than the local background UT by up to 115 ± 98 pptv (3.2 ± 2.6% enhancement). Mixing ratios of organic bromine in the background UT were unaffected by these long-range transport processes. However, mixing ratios of organic bromine in convective outflow were highly variable and were affected by the transport of brominated very short-lived halocarbons (VSLH) from the Gulf of Mexico to the surface of the DC3 study region. When organic bromine enhancements in convective outflow were calculated on a flight-by-flight basis, organic bromine mixing ratios in convective outflow were higher than those in the background UT by an average of 1.7 ± 1.6 pptv (8.5 ± 8.1%). Based on these results, it is speculated that

deep convection may play an indirect role in climate change by introducing pulses of short-lived halocarbons into the UT/LS, which in turn result in relatively quick (on the order of a few months) pulses of O3 loss in the region.

ABSTRACT OF THE DISSERTATION

Global Observations of Oceanic Alkyl Nitrates and Dimethyl Sulfide during the NASA Atmospheric Tomography Mission (ATom)

By

Christopher Thomas Woods

Doctor of Philosophy in Chemistry

University of California, Irvine, 2020

Professor Donald R. Blake, Chair

Oceanic trace gases were measured onboard the NASA DC-8 flying laboratory during the NASA Atmospheric Tomography field campaign with the intent to examine the spatial and temporal fluctuations to the global distribution of these species, as well as establish a dataset of global background levels for future use in the validation of satellites and the constraint of chemical transport models. Airborne in situ measurements of VOCs were obtained via the UC Irvine whole air sampler (WAS) system and analyzed with gas chromatography/mass spectrometry.

The ATom mission was comprised of 4 deployments, each corresponding to the seasons. Each deployment, the NASA DC-8 flew a near circumnavigation of the earth, from 82 °N to 86 °S, while performing vertical soundings from 0.5-12.5 km in altitude. Flight paths over the Pacific, Atlantic and Southern Oceans allowed for the investigation of seasonal effects on anthropogenic and biogenic emissions in the remote atmosphere.

Alkyl nitrates (RONO2) are an important source of reactive nitrogen to the remote atmosphere, where they can constitute the majority of the NOy budget. In the Northern Hemisphere, C2-C5 nitrates are mostly generated photochemically. However, smaller carbon number alkyl nitrates (C1-C4) have a significant ocean source with spatial and temporal variability that are not fully understood. These ‘light’ alkyl nitrates were found to contribute up to 88% of measured NOy in the Southern Hemisphere marine boundary layer. Global measurements of RONO2 showed enhancements at the equator and in the Southern Ocean, with seasonal maximums coinciding with conditions of increased marine biological productivity. MeONO2 was the most abundant, with enhancements > 100 ppt at the equator, contributing a up to 77% of measured NOy mixing ratios alone. Observations of 2-Butyl nitrate in the southern hemisphere marine boundary layer were shown to correlate with MeONO2, suggesting an oceanic source unconsidered in recent literature.

Dimethyl Sulfide (DMS), another oceanic trace gas, is a by-product of marine biological production with impacts to short-wave forcing through incorporation into aerosol, causing direct and indirect radiative effects. In remote atmospheres, DMS may condense onto existing particles or participate in new particle formation: becoming a significant contributor to cloud condensation nuclei (CCN). Global observations of DMS in the boundary layer showed enhancements in the tropics (> 150 ppt), the Southern Ocean (>100 ppt) and the northern midlatitudes in the Atlantic (>100 ppt). Although, these increases in DMS production were not found to follow a consistent latitudinal gradient.

Many California regulations try to decelerate climate change by mitigating emissions from various industries. Greenhouse gases (GHGs) such as methane, carbon dioxide, and nitrous oxide must be fully understood to meet these strict reduction goals. Landfills release GHGs during decomposition and active dumping, while dairy farms release GHGs enterically (i.e., from cows) and from manure management. This study aims to better understand these GHGs at active and closed Orange County landfills as well as at a Visalia dairy farm in California. All samples were collected using whole air sampling techniques and analyzed using gas chromatography to identify and quantify many trace gases. Select oxygenates (i.e., methanol, ethanol, and acetaldehyde), dimethyl sulfide, and carbonyl sulfide were also a focus at these locations. Unexpected and novel sources of these gases were revealed, which have possible implications for pollution and the global sulfur budget.

Samples were collected at the landfills seasonally during four campaigns: Spring 2018, Summer 2018, Fall 2018, and Winter 2019 and at the dairy farm during five campaigns: September 2018, March 2019, June 2019, September 2019, and January 2020. Samples from both industries were compared to airborne and remote air samples to determine their enhancements relative to background concentrations. This research establishes previously unexplored or misrepresented sources of various gases, which is important for the success of the state’s reduction efforts, the environment, and the health of surrounding communities.

California’s San Joaquin Valley, an extremely productive agricultural area, also contains many disadvantaged communities. Residents typically experience low socioeconomic status and a disproportionate amount of air pollution, which can lead to health problems. In addition to GHGs, this study also explores how direct emissions from dairy farms may affect these communities living downwind throughout the San Joaquin Valley. Orange County is more affluent but is often a nonattainment area for several pollutants including ozone and particulate matter, which affect the people living there. This study explores the contribution of trace gases from dairy farms and landfills in California to the formation of pollution and odor in these surrounding communities. Solutions for decreasing trace gas emissions from these sources are also proposed.

This thesis will explore the diagnostic and monitoring potential of exhaled breath for various infectious diseases. The specific infections studied include Borrelia hermsii, salmonella enterica, and two forms of endotoxin. Samples of breath were collected from mice both individually and in groups as well as from rats. The animals were awake and restrained in a breath sample tower for a majority of the studies. One study was also conducted using awake unrestrained rats in a glass bulb that was designed for breath sample collection by the Rowland Blake lab. In addition to the breath samples, feces samples were obtained from mice infected with salmonella enterica and incubated in bioreactors from which the headspace was collected for analysis. Samples were collected in evacuated stainless steel canisters. Blank samples were also collected to determine concentrations in the air used during sampling.

All samples, excluding the feces samples, were analyzed for carbon monoxide and carbon dioxide concentrations using gas chromatography. Several samples were also analyzed for various volatile organic compounds using the Rowland Blake non-methane hydrocarbon gas chromatography system. The results of the gas chromatography measurements were analyzed to determine if differences existed between infected and uninfected samples throughout each study. This was accomplished statistically, as well as visually using plots.

The results of the Borrelia and endotoxin studies revealed a significant increase in the ratio of CO/CO2 in the breath of the infected mice and rats, with a dose response being observed for the endotoxin study. Upon treatment, the ratio of CO/CO2 returned to normal within several hours of antibiotic administration. For the salmonella study, carbonyl sulfide, dimethyl disulfide, carbon disulfide, heptanal, i-propylbenzene, and isoprene were identified in the feces; and acetone, isopropanol, and a gas of unknown identity in the breath. The endotoxin study identified i-pentane in the breath of mice. The study with rats identified two gases, acetone and an unknown gas. The results of these studies suggest that exhaled breath has potential for monitoring of the infections studied, however further studies are needed to explore the diagnostic potential.

Wildfires have been increasing in frequency and severity over the past few decades. Wildfires, agricultural fires, land use burns, and other forms of biomass burning are sources of particulate matter and trace gases, including carbon monoxide, to the atmosphere. These gases are detrimental to air quality and the Earth’s climate, and can react to form tropospheric ozone. Many of these trace gases, like benzene, are also hazardous to human health. Here the trace gases from biomass burning in the United States are examined, and the effects on human health, air quality, and climate are reported.Several large-scale field campaigns have been mounted to study the emissions of fires. These campaigns are usually focused on collecting data from very fresh and aged fire smoke with little influence from background emissions. Therefore, the interactions between fire smoke and urban air have not been well studied. Positive matrix factorization is used here to analyze mixed smoke and urban air masses to tease out the relative emissions contributions from each source. The emission ratios of trace gas compounds to carbon monoxide are also studied to see if the types of fuels, environment, or burn conditions affect the relative emissions of particular trace gases between different fires. While most trace gases are emitted at relatively consistent rates between different fires and fuels, ethyne, methyl chloride, and methyl bromide exhibited significantly different emission ratios between western wildfires and smaller southeastern agricultural fires. Finally, chemical box modeling is used to examine the effects of biomass burning emissions on the chemistry of urban air quality. Using inputs for trace gas emissions based on observed data a box model was run for three conditions: urban emissions only, smoke mixed with urban emissions, and heavy smoke mixed with urban emissions. The results indicate that smoke emissions can augment ozone production in urban areas. These results are compared to observations from Pasadena, CA during the 2020 Bobcat Fire.

Short lived greenhouse gases are an immediate target for reductions in climate policy planning. To understand to what extent legislation to limit greenhouse gas emissions is working, monitoring projects are necessary. In this work, whole air samples were collected downwind of the Los Angeles megacity over the course of two years between June 2014 and May 2016. Samples were analyzed via gas chromatography instruments to quantify over 70 volatile organic compounds with limits of detection on the single part per trillion level.

Unlike larger hydrocarbons, the daytime average mixing ratio for both methane and ethane did not decrease in the region since a previous study in 2007 suggestive of continued natural gas leakage. Using monthly ethane to methane enhancement ratios and natural gas composition of gas delivered to the region showed that 72 ± 16% of all excess methane emitted in the SoCAB was due to natural gas leakage. This result is significantly higher than the fraction due to natural gas in previous bottom-up inventories in the region.

Positive matrix factorization was applied to the data set to aid in apportionment of hydrocarbons to various sources in the region. A seven factor result was obtained in which natural gas, oil and gas refining, biogenic, industrial sources, BTEX, gasoline liquids and tailpipe emissions were matched to source profiles. Via this analysis, 59% (57-65%) of the methane seen in the SOCAB was due to the leakage of unburned natural gas. This agrees well with the result obtained using the composition of delivered gas and two recent studies of the region using other data sets.

Finally, several classes of refrigerants which are potent greenhouse gases were studied. The emission of banned chlorofluorocarbons was noted to have completely stopped in the region. Hydrofluorochlorocarbons are currently being phased out and so emissions estimates were calculated and compared to older data sets to show that regulations are working to limit their emission. HCFC-142b emissions were noted to have dramatically stopped during the sampling period as the result of a phase down rule that went into place on January 1, 2015.

Emissions of hydrocarbons associated with oil and natural gas infrastructure are of current significant interest to atmospheric scientists. Studies aim to address the local and regional air quality impacts of oil and gas operations throughout North America. In Fort McMurray, multiple facilities mine and upgrade crude bitumen. The “Industrial Heartland” in Fort Saskatchewan, Alberta is Canada’s largest hydrocarbon processing center, home to over 40 chemical and petrochemical facilities. Ambient samples collected in both regions during the summer of 2010 revealed significant enhancements in over 40 VOCs. Many of these measured values were similar to or greater than some of the world’s largest urban and industrial centers. The primary sources of these VOCs included propene fractionation, diluent separation, and bitumen processing facilities. Hydroxyl radical reactivity (ROH) was calculated in each region as a measure of ozone formation potential. On average, background samples had a ROH of 4.4 s-1, while industrial plumes averaged 9.2 s-1, with reactivity reaching up to 60 s-1 in the most concentrated plumes. Acetaldehyde, propene, and 1,3-butadiene had the largest contributions to the ROH values.

The Barnett Shale of northern Texas is one of the most developed and active natural gas shale plays in the United States. Emissions from the many oil and gas system components in the region have not been fully characterized. Whole air samples were collected throughout the region in October 2013, targeting known methane sources. Hydrocarbon mixing ratios, correlation plots, and stable isotopes were used to discern emission signatures for thermogenic (oil and gas) versus biogenic sources of methane. Ratios of ethane and propane to methane were distinct for each source type, although highly variable, and used to characterize natural gas as either wet or dry. Integration of these ratios into a bottom-up methane emissions inventory for the region predicted a median ethane flux of 5500 kg C2H6 hr-1, which was consistent with top-down estimates. Lastly, the impact of emissions on local photochemistry and a statistical source apportionment suggested that thermogenic sources are responsible for nearly 70% of hydroxyl reactivity, matching the 64-72% predicted by the CH4 inventory. Overall, oil and natural gas activities were the dominant source of CH4 in the Barnett Shale region.

In addition to their influence on air quality and climate, VOCs emitted from oil and natural gas operations are of concern because of their potential health risks. Compounds such as 1,3-butadiene and benzene are hazardous air pollutants and known cancer-causing agents. In the communities closest to the Industrial Heartland, a 13-year record revealed a higher rate of male hematopoietic cancers than other communities in Alberta. The industrial emissions cannot directly be linked to cancer rates, but the elevated VOC concentrations in the region warrant further monitoring and research. In the Barnett Shale, oil and wet natural gas emit increased amounts of potentially harmful gases, including benzene. Using two methods, benzene emission estimates were calculated for the region, and ranged from 48 ± 16 to 84 ± 26 kg C6H6 hr-1.

In recent years, hydrocarbon emissions from oil and natural gas (ONG) industries has been an area of extensive research in an effort to reduce air quality impacts from ONG fugitive emissions. In particular, the Northern Front Range Metropolitan Area (NFRMA) in Colorado, an area with extensive ONG infrastructure, has been designated as an area of ozone National Ambient Air Quality Standard (NAAQS) non-attainment area. There are various factors, from multiple pollution sources to the local meteorology, which exacerbate ozone pollution events in the NFRMA.

The Front Range Air Pollution and Photochemistry Experiment (FRAPPÉ) field campaign was designed to answer the scientific question: “What are the factors controlling NFRMA surface ozone and are current emission controls sufficient to reduce ozone levels below the NAAQS?” This campaign took place from 26 July – 17 August, 2014, and consisted of a network of ground and aircraft measurements. Flight plans and sampling were designed to measure background conditions, source specific emissions, and meteorological effects by assessing local and regional photochemistry. The project aimed to create a comprehensive data set that will be used to elucidate the complex mechanisms that lead to enhanced ozone production.

Specifically, this work characterized emissions from the NFRMA and the Colorado western slope, and determined emissions from the ONG industry are the dominant source of hydrocarbons throughout the study area. Furthermore, from the examination of local meteorology events and spatial hydrocarbon distribution, it was determined that ozone pollution events in the NFRMA are a result of both local emissions and long-range transport.

In response to the posed question, despite strict emission controls in Colorado, emissions from the ONG industry are still pervasive throughout the research study area. To improve local air quality and achieve NAAQS ozone attainment, the effect of local meteorology needs to be strongly considered in future emission standards.