Research in atmospheric chemistry is a basic, discovery-driven field of science that is needed to understand the world in which we live. This fundamental research is also at the core of a number of societally relevant issues such as climate and weather, and the health of humans and ecosystems. The Committee identified six core areas that encapsulate the chemical nature of the atmosphere and provide the building blocks for breakthroughs in the future:
- Human activities and natural processes govern emissions that determine the chemical composition of the atmosphere. Because atmospheric composition responds to changes in emissions as a result of societal choices and a changing Earth system, the development of accurate emission inventories is a necessary foundation for understanding and predicting changes in the atmosphere.
- Chemical transformations affect the spatial and temporal variability of gases and particles. The fundamental study of reactions that occur in the atmosphere is clearly central for predicting the distribution and nature of constituents in the air.
- Atmospheric oxidants control the lifetimes, distribution, and products of emitted species. Highly reactive chemicals like ozone and the hydroxyl radical participate in many chemical reactions, so their sources, sinks, and concentrations, as well as their spatial and temporal variability, are required to understand reactions that drive atmospheric chemistry.
- Chemical processes interact with atmospheric dynamics to control the distribution of trace gases and particles. Motions of the atmosphere, not just emissions and reactions, affect atmospheric composition, and the understanding of both must be coupled to evaluate distributions.
- Particle chemical and physical properties affect cloud and aerosol particle radiative properties, cloud microphysics, and precipitation processes. The distribution and nature of particles in the atmosphere alters the climate as well as behavior of the water (including precipitation) in the Earth system.
- Atmospheric trace gases and aerosol particles impact and alter global biogeochemical cycles. An understanding of atmospheric chemical composition is required to evaluate the present and future health of natural and managed ecosystems, marine and other aquatic environments.
The first topic, emissions, highlights the substantial influence of humans on the atmosphere. Both human emissions and changes in the natural system are accelerating; a predictive capability that provides foresight into the consequences of those rapid changes will aid in preparation and decision making. The second through fourth topics—chemical transformations, oxidant distributions, and relationship with atmospheric dynamics—are fundamental elements of the predictive capability required to determine the distribution, transport, and fate of chemicals throughout the atmosphere. These processes determine how emissions and other human activity lead to societal impacts. The fifth topic connects atmospheric chemistry to other physical systems that affect humans, while the sixth connects the atmosphere to biogeochemical cycles affecting ecosystems and food production.
Support of research in these core topics remains vital for both atmospheric chemistry itself and for understanding its connections to human health and welfare as discussed in the next chapter. The remainder of this chapter provides an overview of each of the six topics described above, followed by examples of recent and ongoing advances within the past decade. These recent advances are described using language common in the atmospheric chemistry field to capture the nature and complexity of topics, rather than directing the presentation to the nonexpert in the field, as we have attempted to do throughout most of the report.
3.1 HUMAN ACTIVITIES AND NATURAL PROCESSES GOVERN EMISSIONS THAT DETERMINE THE CHEMICAL COMPOSITION OF THE ATMOSPHERE
An overarching goal of atmospheric chemistry research is to understand human-induced changes in the atmosphere so that accurate predictive capabilities can be developed to assess future scenarios and aid in the development of effective policies to minimize risk. This goal requires accurate data on the emissions to the atmosphere, including gases and particles from all sources, anthropogenic as well as natural. Although substantial improvements have been made in recent years, the current knowledge of emissions into the atmosphere in both the pre- and post-industrial eras is inadequate.
The amount, location, and timing of emissions to the atmosphere are controlled by both environmental conditions and human activities. Global change and economic development alter these emissions, and the economic growth of human society has changed the distribution of both natural and anthropogenic sources. The develop-
ment of inventories that track emissions from their sources is a scientific endeavor that contributes to the entire discipline of atmospheric chemistry and should be managed in parallel with other core research areas. Emissions link human and natural activity to changes in Earth system behavior, and are required for traceable, attributable, and quantified assessments of how global change and social choices will affect the atmosphere. Improved understanding of the magnitude and the location of emission sources, as well as connections with the technological advances, social decisions, and environmental conditions that determine emissions, is crucial for predicting climate and weather and mitigating health impacts.
The composition of the atmosphere and the controlling chemistry derive from emissions of gases and particles. Emissions include a very large suite of organic and inorganic compounds, including trace metals such as mercury. Quantification of these emissions in an inventory provides data needed for a multitude of atmospheric chemistry applications, such as inputs to chemistry and climate models, evaluation of trends, and explanation of in situ observations. Substantial advances have been made in the development of emission inventories via techniques that combine a variety of scientific approaches, which may include laboratory, in situ and remotely sensed observations, assessment of technology and energy use, and evaluation of human behavior.
There are several approaches to build, interrogate, and evaluate emission estimates. “Bottom-up” approaches typically apply combined knowledge of emission rates and the activities that contribute to them. “Top-down” approaches use models of the atmosphere to infer emissions from measured concentrations of atmospheric constituents. Improvements to quantifying emission data require the integration of these different approaches. Analysis of remote sensing data of the atmosphere near urban centers has produced a compelling constraint on emissions (e.g., Berezin et al., 2013; Hilboll et al., 2013; Konovalov et al., 2008; Pfister et al., 2005). For example, remote sensing of tropospheric NO2 from space-based satellite instruments has revealed a dramatic change in the global distribution of anthropogenic emissions over the past several decades. Emissions strongly increased over China, the Middle East, and India, while simultaneously declining throughout the more developed world (including the United States, Western Europe, and Japan) (Hilboll et al., 2013). Using a variety of observations and chemical modeling tools, bottom-up emission inventories and emission trends
can be more thoroughly constrained (e.g., Brioude et al., 2013; McDonald et al., 2012; Palmer, 2008; Polson et al., 2011). Novel detection methods and tracer techniques have revealed gaps in inventories for emissions and their variability from specific sectors (e.g., Mellqvist et al., 2010; Pétron et al., 2012; Roscioli et al., 2015). Addressing these gaps results in better predictions, for example, linking precursor emissions with impacts on ozone (Ahmadov et al., 2015). Other observational advances, such as improved direct flux estimates from high time resolution measurements from the surface (e.g., Karl et al., 2001) and aircraft (Karl et al., 2009; Warneke et al., 2010), have provided the means to quantify emissions over local and regional scales. These various techniques not only enable the quantification of species emitted to the atmosphere, but also evaluate the conditions that control the emissions. It is important to continue application and development of new approaches to verify and predict emissions and to evaluate their impact on atmospheric chemistry and climate (Bond et al., 2013; Tong et al., 2012). Some world regions are undergoing rapid growth or transition, requiring dynamic methods of emission estimation that can rapidly respond to changes in circumstances.
As emission inventories are developed, key scientific questions are generated through their subsequent use. An unanticipated disagreement between atmospheric measurements and an emission inventory is usually an indication of a lack of sufficient understanding of sources and emissions (see Box 3.1). With additional observations and study, causes of gaps and errors in emission estimates can be identified and improved in subsequent inventories. A recent example involves long-term measurements of ethane at sites in the Northern Hemisphere (see Figure 3.1), which show increases in ambient concentrations. This increase was not predicted by models until the emission inventories for ethane were revised to include those from oil and gas extraction and production activities in North America.1 Continued analysis of these datasets up to 2015 suggests increased and previously unrecognized emissions of ethane and propane from natural gas development in North America are important (Helmig et al., 2016).
Recent discoveries of high wintertime ozone concentrations in rural areas of the United States in Wyoming (Rappengluck et al., 2014; Schnell et al., 2009) and Utah (Warneke et al., 2014) are other examples of unanticipated disagreements between measurements and emission inventories. Traditionally, high O3 concentrations are observed during summertime in urban and suburban areas with high motor vehicle, residential, and industrial emissions; this paradigm has subsequently guided ozone measurement and management strategies. Atmospheric models do not predict winter
1 Ethane emission inventory: https://nar.ucar.edu/2015/acom/d2-ethane-emissions-inventory.
peak ozone episodes in western rural areas, which are now attributed in part to the rapid development of oil and gas activities and the inability of bottom-up emission inventories to sufficiently characterize the rapidly changing emissions from this sector (Ahmadov et al., 2015).
Emissions from Specific Sources
The relative magnitude of emissions from biogenic or anthropogenic sources can play a key role in atmospheric chemistry and the chemical pathway of emissions. Biogenic and anthropogenic emissions typically have fundamentally different governing processes, temporal scales, and spatial scales that require their own inventory methodologies. Additionally, particle phase emissions represent a unique challenge as they can be emitted directly or formed from secondary processes in the atmosphere.
Vegetation is the dominant source of reactive hydrocarbons to the atmosphere (Guenther et al., 2006) in the form of volatile organic compounds (VOCs) such as isoprene and monoterpenes (Guenther et al., 1995). Model descriptions of the emissions
of biogenic VOCs have improved considerably over the past decade (e.g., Arneth et al., 2007; Guenther et al., 2012) and are included in most regional and global models. At the same time new measurements, including for example flux measurements from aircraft (Karl et al., 2013) and over the open ocean (e.g., Kim et al., 2014; Marandino et al., 2007), are providing critically needed observational constraints on these emissions.
These constraints are vital to quantifying the gas-phase emissions that drive atmospheric chemistry processes in many regions of the globe.
In urban areas, anthropogenic emissions drive the chemistry and resulting impacts, such as ozone and particulate pollution. In former decades, ozone mitigation strategies have been informed and validated by combined model/measurement programs. In more recent decades, the field of atmospheric chemistry has been tasked with gaining insight into the difficult problem of urban and anthropogenic aerosol particles (Heald et al., 2005; Odum et al., 1997; Volkamer et al., 2006). Because particulate matter is frequently formed by processes occurring in the atmosphere, it is a dynamic component that does not lend itself to conventional inventory development approaches (Donahue et al., 2006; Zhang et al., 2015). Current and future research examining urban outflow will expand the understanding of the mechanisms and impacts of anthropogenic and biogenic emissions as they are processed in the atmosphere and contribute to particle mass (Camredon et al., 2007; Gentner et al., 2012; Huang et al., 2014). Constraints on the magnitude of each source are vital to this research.
The atmosphere has been a known habitat for microbes for centuries (Womack et al., 2010). Aerobiology studies focusing on determining the concentrations of microbes in the atmosphere remain disproportionally low relative to water and soil (Barberán et al., 2015; Behzad et al., 2015; Kellogg and Griffin, 2006; Womack et al., 2010). The understanding of the sources and impacts of bioparticles has increased in recent years, but the overall understanding of their impacts on atmospheric composition, climate, and human health remains weak. The first attempts to model emissions and transport of fungal spores, bacteria, and pollen (Burrows et al., 2009b; Heald and Spracklen, 2009; Jacobson and Streets, 2009; Mahowald et al., 2005) are now starting to be tested against new observational constraints from fluorescence measurements (Gabey et al., 2010; Huffman et al., 2010; Perring et al., 2015). The oceans have become recognized as a significant source of not only sea salt but also surface active biological species including bacteria, lipids, proteins, sugars, and viruses. When waves break, bubbles burst at the surface ocean releasing sea spray aerosols that can be heavily enriched in biogenic material (Blanchard, 1964; O’Dowd et al., 2004). Because the biota concentrates important nutrients, phosphorus can be a good indicator of primary biogenic particles (Mahowald et al., 2008). Studies have also explored the ability of bioparticles from different sources to act as cloud condensation nuclei or ice nuclei (DeMott et al., 2016; Steiner et al., 2015; Tobo et al., 2013), the importance of which is discussed in Section 3.4.
Biomass burning is another activity that produces substantial amounts of trace gases and particulate matter to the atmosphere. Estimation methods for biomass burning
emissions (e.g., Urbanski et al., 2011; van der Werf et al., 2006; Wiedinmyer et al., 2011) have taken advantage of controlled laboratory experiments (e.g., Stockwell et al., 2015) and ambient observations (e.g., Burling et al., 2012; Christian et al., 2007) to constrain the particle and gas emissions from this important source on time and spatial scales relevant for local, regional, and global applications, although uncertainties in these estimates remain high (e.g., Al-Saadi et al., 2008; Urbanski et al., 2011). Remote sensing observations of land surface characteristics (e.g., Friedl et al., 2002) and burning activities (e.g., Giglio et al., 2006; Wooster et al., 2003) have enabled further constraints on biomass burning emission estimates. Large-scale open biomass burning, including wildfires (see Figure 3.2), emit close to half of the global particulate emissions and are an important contributor to global trace gas emissions.2
In short, quantitatively defining the composition and sources of all emissions to the atmosphere requires taking both natural and anthropogenic sources into account, while using many different observational and modeling tools to constrain the estimates and to provide realistic uncertainties on those estimates. Such data need to be provided on spatial and temporal scales appropriate for each type of source and need to interface with scales used in models and measurements.
Most emitted gases and particles can further react to form new chemical species. Many compounds that are unreactive in the lower atmosphere, such as chlorofluorocarbons (CFCs), are transported to the stratosphere where they break down, leading to dramatic impacts on natural ozone cycles (Burkholder et al., 2015; Crutzen and Lelieveld, 2001; Finlayson-Pitts and Pitts, 2000; Rowland, 2001, 2006). Transformations are at the core of atmospheric chemistry.
Figure 3.3 provides an overview of the gas-phase chemistry of organic compounds, oxides of nitrogen and sulfur that alter the chemical composition of directly emitted species to form a variety of products in both the gas and condensed phases (Atkinson and Arey, 2003; Finlayson-Pitts and Pitts, 2000; Seinfeld and Pandis, 2006). The gas-phase inorganic chemistry of oxides of nitrogen and sulfur are relatively well known, while their interactions with organic compounds—for example to form organosulfates—are less well understood. Oxidation of organic compounds generally produces
compounds that are more polar and less volatile (Goldstein and Galbally, 2007). Products with different vapor pressures and solubility behaviors, combined with direct emissions of VOCs of varying molecular mass, lead to a continuum of organic species that have been designated intermediate, semi-, low, and extremely low volatility organic compounds (Donahue et al., 2012). Such products are major components of secondary organic aerosol (SOA) particles and are likely involved in the earliest stages of particle nucleation and growth (Bianchi et al., 2016; Ehn et al., 2014; Kulmala et al., 2013; Riipinen et al., 2012; Tröstl et al., 2016; Zhang et al., 2011). The condensed phase of particles, fogs, and clouds provides the opportunity for species to partition between phases, to deposit into rivers, lakes, and oceans. Atmospheric transformations of particles containing micronutrients, such as iron, are likely to be of great importance to both the cycling of sulfur and organic acids as well as iron bioavailability (Meskhidze et al., 2005).
Many important advances over the last few decades have contributed to the understanding of the atmospheric transformations of chemical constituents. Following are
a few examples whose importance for understanding atmospheric processes and impacts has recently been recognized, but that currently lack solid understanding.
Kinetics, Mechanisms, and Products of Gas-Phase Reactions
Rate constants for the gas-phase reactions of organic compounds with OH, O3, NO3, and Cl, and their dependence on temperature and pressure have been measured for many reactions (IUPAC, 2015; Sander et al., 2011), and structure-reactivity relationships have been developed for the initial oxidation steps of relatively small VOCs (Calvert et al., 2015). Mechanistic understanding has also substantially improved (Calvert et al.,
2015), an example being the isomerization of some organic peroxyl radicals to form products with high oxygen to carbon (O:C) ratios and low volatility.
Advances in theoretical chemistry have provided an important tool to understand and predict both kinetics and reaction pathways in increasingly complex systems that may not be possible to study experimentally at present (Vereecken et al., 2015). For example, autooxidation reactions are well known in the condensed phase (Sevanian et al., 1979), but were highlighted in theoretical studies as also being possible in the gas phase (Sevanian et al., 1979; Vereecken et al., 2007), which was subsequently confirmed in environmental chamber studies of isoprene oxidation products (Crounse et al., 2011; Krechmer et al., 2015).
Many gas-phase smog chamber reaction studies have been carried out in the absence of water vapor or under low humidity conditions. There is increasing recognition of the role of water vapor in altering reaction pathways of some organic compounds, particularly those that are highly soluble in water such as glyoxal, methylglyoxal, and organic acids. For example, one study reported the observation of glyoxylic acid and its germinal diol counterpart in the gas-phase (Plath et al., 2009), illustrating a gas-phase pathway to convert aldehyde into acid and aerosol particles.
The Importance of Phase and Partitioning
The importance of exchange between the gas and particle phases has become abundantly clear for understanding the composition of the atmosphere. While the phase behavior of inorganic salts in pure form as a function of relative humidity (RH) is well known, that of SOA is not. Models have historically treated SOA as an oily material, where internal diffusion is sufficiently fast that quasi-equilibrium between the gas-phase and particles holds. However, recent laboratory and field studies have shown that SOA can be a highly viscous semi-solid material (Bateman et al., 2015; Kidd et al., 2014; Perraud et al., 2011; Pöschl and Shiraiwa, 2015; Renbaum-Wolff et al., 2013; Shiraiwa et al., 2011; Virtanen et al., 2010, 2011) in which diffusion in the particles is much slower, and the time scales to reach equilibrium can be longer than represented by quasi-equilibrium in atmospheric models (You et al., 2014). A further recent discovery is that liquid-liquid separation can occur in particles (O’Brien et al., 2015; You et al., 2014), and that this separation and the particle morphology depends on the composition and the relative amounts of the different components. These properties can affect gas uptake and partitioning, gas concentrations, optical properties of the particles (You et al., 2012) and their ability to act as cloud condensation nuclei and ice nuclei (Schill and Tolbert, 2013).
Chemistry and Photochemistry in the Condensed Phase and at Interfaces
The presence of all three phases—gas, liquid, and solid—in the atmosphere dictates that partitioning between the phases and chemical/photochemical processes occurs in all three phases. In addition to multiphase chemistry involving gases and liquids, it is known that unique chemistry can occur at interfaces. Reactions on solids and in surface films are generally described in the community as “heterogeneous chemistry.”
Fogs and clouds present a common aqueous medium for reactions in the troposphere, and water and organic species can also be an important component of particles. While oxidation of dissolved SO2 (in all of its forms collectively referred to as S(IV)) in the aqueous phase to sulfuric acid and sulfates is well known, recent studies show that multiphase chemistry involving water may be important in atmospheric transformation for a variety of species (Biswas et al., 2008; Davidovits et al., 2011; Ervens, 2015; Herrmann et al., 2015; McNeill, 2015; Zhang et al., 2015; Zhu et al., 2005). This mechanism has been shown to influence the chemical signature of trace constituents, such as formation of SOA from species vented from the boundary layer to the free troposphere (e.g., Froyd et al., 2010; Sorooshian et al., 2006). These precursors generate particles aloft, which have large radiative impacts because they scatter and absorb incoming solar energy as well as diffuse backscattered radiation from clouds below (Seinfeld, 2008). Ubiquitous and abundant small, water-soluble compounds from gas-phase oxidation of biogenic and anthropogenic emissions can undergo oxidation in the aqueous phase to form SOA (Herrmann et al., 2015; Zhang et al., 2015; Zhu et al., 2005). Reactions with inorganic constituents in wet aerosol particles (Noziere et al., 2009; Perri et al., 2010; Yasmeen et al., 2010), and with oxidants (e.g., ozone [Grgić, 2010]) also contribute to this aqueous-phase oxidation. Evidence for SOA formation is provided by laboratory experiments demonstrating that the aqueous chemistry of small water-soluble compounds forms organic acids/salts (e.g., oxalate) as well as high molecular weight compounds that are also found in atmospheric aerosol particles.
Multiphase chemistry also plays a role in determining the optical properties of particles, for example the formation of visible light-absorbing products known as “brown carbon” (Laskin et al., 2015; Moise et al., 2015), which adds to particulate mass emitted directly from some combustion processes such as biomass burning (Dutkiewicz et al., 2011; Laskin et al., 2015). However, the nature and number of individual chromophores, their formation mechanisms, and atmospheric fates remain open questions (Phillips and Smith, 2014, 2015).
There has been increasing recognition of the importance of heterogeneous chemical processes, which occur not only on the surfaces of particles (Al-Abadleh and Grassian, 2003), but also on surfaces in the boundary layer, including the built environment,
ocean, vegetation, and ice (George et al., 2015). The air–sea interface covers nearly three-quarters of the Earth’s surface and can profoundly impact atmospheric composition (Ryder et al., 2015). Reactions at interfaces can modify surface properties, changing their ability to act as cloud condensation and ice nuclei, and altering the bioavailability of the particle constituents to both ecosystems and humans. In addition, reactions on biological surfaces can cause damage to humans and ecosystems (Pöschl and Shiraiwa, 2015). In an effort to study heterogeneous chemistry on more realistic atmospheric surfaces, new flow tube approaches have been developed (Bertram et al., 2009). These studies offer the potential for bridging the existing gap between laboratory studies and field observations.
On a molecular level, the mechanisms and kinetics of chemical and photochemical reactions on surfaces differ from those in the gas-phase or in aqueous solutions. For example, enhanced thermal and photochemical reactions have been observed at the interface compared to the bulk for organic systems (e.g., Donaldson and Valsaraj, 2010; Enami et al., 2015; Griffith and Vaida, 2012; Heath and Valsaraj, 2015; Kameel et al., 2014), for inorganic ions (Knipping et al., 2000; Tobias and Hemminger, 2008) and for nitric acid/nitrate photochemistry on “urban grime,” and on other environmental surfaces (Baergen and Donaldson, 2013; Du and Zhu, 2011; George et al., 2015; Zhou et al., 2003; Zhu et al., 2008). Similarly, there is increasing evidence that some organic species in particles and/or on surfaces in the atmosphere act as photosensitizers at interfaces (George et al., 2015), initiating a cascade of unique interfacial chemistry.
Water plays an important role in heterogeneous chemistry by reacting with particles and surfaces or changing their reactivity, affecting particle growth, or competing for surface sites for heterogeneous reactions (Du et al., 2014; Kolb et al., 2010; Rubasinghege and Grassian, 2013). Reactions on ice or snow surfaces play an important role, particularly in the chemistry of the polar regions (Anastasio and Jordan, 2004; Grannas et al., 2002; Hamer et al., 2014; McNeill et al., 2006; O’Driscoll et al., 2008; Thomas et al., 2012; Toom-Sauntry and Barrie, 2002). Gas uptake in/on snow and subsequent chemistry and photochemistry in the snow contribute to changing chemical composition above or in the snowpack, oxidant formation, and ozone depletion, showing the important role of surface atmospheric chemistry within the cryosphere component of the Earth system (Abbatt et al., 2012).
Understanding the Molecular Basis for Particle Nucleation and Growth in the Atmosphere
common to have orders-of-magnitude differences between measured rates of new particle formation and those predicted using classical nucleation theory. The discoveries that have dramatically altered the understanding of atmospheric nucleation include the following:
- amines are far more efficient than ammonia in forming particles with sulfuric acid and are ubiquitous in air (Almeida et al., 2013; Ge et al., 2011; Glasoe et al., 2015; Pratt et al., 2009a; Smith et al., 2010);
- assumptions inherent in classical nucleation theory may not be directly applicable to at least some of the important nucleation processes occurring in the atmosphere (Kupiainen-Maatta et al., 2014); and
- organic species likely play a key role in nucleation and growth (Bianchi et al., 2016; Bzdek et al., 2013; Donahue et al., 2011; Kulmala et al., 2013; Metzger et al., 2010; Riccobono et al., 2014; Riipinen et al., 2012; Schobesberger et al., 2013; Tröstl et al., 2016; Zhang et al., 2011).
Theoretical approaches, including quantum chemistry, have played a central role in elucidating mechanisms on a molecular level and providing a predictive capability that was previously lacking. The combination of theoretical and experimental approaches have provided new understanding of how small clusters form and grow via stepwise additions of acid, base, and water, and the decomposition mechanisms for the clusters that occur in competition with their growth (Bzdek et al., 2013; DePalma et al., 2014; Olenius et al., 2014; Ortega et al., 2012; Vehkamaki and Riipinen, 2012).
An area in which less progress has been made is in reconciling measurements and models for the formation and growth of the mass of SOA associated with particles. It was almost a decade ago that it became clear that models could not reproduce the SOA particle mass measured in many locations (de Gouw et al., 2005; Heald et al., 2005; Johnson et al., 2006; Volkamer et al., 2006). Better agreement has been attained with improved emission inventories and identification of additional precursors and semivolatile and intermediate volatility species, as well as updated chemical mechanisms and chemistry in the condensed phase (Hayes et al., 2015; Heald et al., 2011; Shrivastava et al., 2011). However, even when models reproduce measured mass concentrations, more specific properties such as volatility and O:C ratio may not be consistent (Chen et al., 2011). The new ability to measure extremely low volatile organic compounds (Ehn et al., 2014) has highlighted how ubiquitous these compounds are and their likely importance in SOA formation. In addition, as discussed above, multiphase chemistry and photochemistry are likely to play an important role in SOA formation, particularly in high relative humidity environments.
Ability to Measure Composition of Gases and Particles
A number of techniques, particularly those involving mass spectrometry, have been developed to measure gases at sub-part per trillion levels, particles as small as approximately 8 nm in size, as well as clusters. For example, chemical ionization combined with high resolution time-of-flight mass spectrometry has been applied to measure a variety of both inorganic gases such as nitryl chloride (Simpson et al., 2015), organic species with extremely low vapor pressures (Ehn et al., 2014; Nozière et al., 2015), and molecular clusters (Schobesberger et al., 2013). Online real-time analysis of particles using laser desorption/ionization or thermal desorption techniques (Bzdek et al., 2012; Canagaratna et al., 2007; Middlebrook et al., 2003; Murphy, 2007; Murphy et al., 2006; Nozière et al., 2015; Prather et al., 2008; Tobias and Ziemann, 2000; Zelenyuk et al., 2009; Zhang et al., 2014) has provided insight into particle composition and how it changes with the nature of precursors and aging in air. The development and widespread use of commercial aerosol mass spectrometry has enabled characterization of particle composition in many locations throughout the world, providing insights into spatial and temporal variations under a number of conditions (Jimenez et al., 2009) and enabled the observation of the oxidation of low-volatility organic compounds to form organic aerosol particles in the lab (Robinson et al., 2007; Sage et al., 2008). Obtaining molecular structures of complex organic species in particles remains challenging using online particle mass spectrometry techniques. However, advances in off-line techniques applied to collected particle samples—such as multidimensional and combined chromatography-mass spectrometry (Nozière et al., 2015; Zhang et al., 2014) as well as ambient ionization methods (Ifa et al., 2010) and in situ chromatography with online derivatization (Isaacman et al., 2014) for gas and particle phase species—have improved understanding substantially. Advances in instrumentation have been major contributors to propelling the field forward over the last decade and is likely to continue in the coming years as well.
Trace oxidants are present in extremely low concentrations in the atmosphere, yet they determine the chemical fates of many atmospheric species. These oxidants are highly reactive, with the most reactive oxidants having lifetimes on the order of seconds or less, while tropospheric ozone (O3) can last for weeks. The hydroxyl radical
(OH) is the dominant oxidant in the atmosphere. However, a suite of other gas-phase and aqueous-phase oxidants in the oxygen, hydrogen, organic carbon, nitrogen, and halogen chemical families are important drivers of atmospheric chemical reactions. The hydroperoxyl radical (HO2) and organic peroxyl radicals are also key players. Another important atmospheric oxidant is the Criegee intermediate, which is formed in the reaction between ozone and organic compounds that have double bonds. Lastly, a suite of oxidants grouped under “reactive oxygen species” (ROS) in the condensed phase are also of importance, especially for their role affecting human health (see Chapter 4.2).
Over the past few decades, atmospheric chemistry research has transitioned from a dearth to a wealth of oxidant observations of gas-phase OH, HO2, unspeciated peroxy radicals (Heard and Pilling, 2003; Monks, 2005; Stone et al., 2012), NO3 (Brown and Stutz, 2012), and halogens (Simpson et al., 2015). Advances in understanding oxidant chemistry in certain environments have generated a good first-order understanding of this chemistry in the stratosphere and free troposphere. In more complex environments—particularly those with more VOCs or heterogeneous reaction pathways that compete with gas-phase reactions, and in the condensed phase—the understanding is less complete.
Stratospheric Ozone Depletion
As described in Chapter 2, one success story in atmospheric chemistry is discovery of the processes that cause stratospheric ozone depletion (Newman et al., 2009). An integral component of this understanding is in stratospheric oxidant chemistry, where observations and models generally agree within uncertainties for odd-hydrogen chemistry (Wennberg et al., 1994) and halogen chemistry (von Hobe et al., 2013; WMO, 2014). However, despite this success, there are still outstanding questions about oxidant sources in the lower stratosphere. For example, overshooting mid-latitude cumulus convection may inject water vapor into the lower stratosphere, setting the stage for enhanced heterogeneous conversion of inorganic chlorine to forms that can catalytically destroy ozone (Anderson et al., 2012). Continued monitoring of stratospheric composition is essential to ensure that the stratospheric halogen oxidant concentrations continue to decrease and that stratospheric ozone recovers (see Figures 2.5 and 2.6).
Global Atmospheric Gas-Phase Oxidation Potential
A key issue for atmospheric chemistry research is the global atmospheric oxidation capacity, especially associated with OH since it controls the lifetime of methane and other important greenhouse gases. Studies have shown that the interannual variability in global atmospheric OH concentrations is small (<5 percent; Ciais et al., 2013; Montzka et al., 2011). However, the distribution of OH regionally and between hemispheres (Patra et al., 2014) and the abundance and impact of the full range of atmospheric oxidants on atmospheric composition are still uncertain.
Oxidants seem to be best understood in cleaner environments, such as much of the tropical-free troposphere where oxidation chemistry involves primarily OH in the gas-phase and hydrogen peroxide in the aqueous phase. In these regions, measured and modeled OH generally agree (Cantrell et al., 2003; Ren et al., 2008, 2012). However, measured HO2 is less than modeled in the mid-troposphere (Ren et al., 2008; Tan et al., 2001), perhaps because of heterogeneous reactions of HO2 on aerosol particles (Mao et al., 2010).
The nitrate radical (NO3) is an important nighttime oxidant for anthropogenic and biogenic VOCs (Brown and Stutz, 2012). NO3 is coupled to dinitrogen pentoxide, which is lost to the surface where it can react with sea spray to produce volatile nitryl chloride, a daytime Cl oxidant source (Thornton et al., 2010). Several studies in urban areas and power plant plumes indicate that the chemical processes producing NO3 are generally understood, but questions remain about the distribution of NO3 in the free troposphere and its vertical nocturnal distribution in urban areas.
Halogens can also influence oxidant budgets. In the Arctic, bromine drives a catalytic cycle that can lower boundary-layer ozone to near-zero levels (Barrie et al., 1988; Pratt et al., 2013; Simpson et al., 2015). On a global scale, satellite measurements suggest that bromine monoxide (BrO) might be distributed in the troposphere throughout the mid-to-high latitudes, giving it a role in determining global oxidant levels (Wagner et al., 2001). In remote coastal regions, iodine and other emitted biogenic species can play an important role in the local oxidation chemistry and in the formation of particles (Carpenter and Nightingale, 2015; Carpenter et al., 2001). Chlorine has also been postulated to be an important oxidant in the marine boundary layer, making a small contribution to atmospheric oxidation in some urban port cities (Riemer et al., 2008; Young et al., 2014), and according to some recent studies, may be an important oxidant over continents (Mielke et al., 2011; Thornton et al., 2010; Wild et al., 2016).
There is growing evidence that Criegee intermediates (highly reactive and unstable biradicals) may be important atmospheric oxidants. Criegee intermediates were
postulated almost 70 years ago (Criegee, 1948) and laboratory studies using indirect methods have inferred their existence and reaction kinetics (Calvert et al., 2015). The Criegee intermediate has recently been directly generated and measured in laboratory systems (Taatjes et al., 2013; Welz et al., 2012), and its existence in the atmosphere has been inferred from measurements of sulfuric acid, an oxidation product of the reaction between SO2 and Criegee intermediates (Mauldin et al., 2003, 2012). However, the role of the Criegee intermediate as an atmospheric oxidant has not been firmly established (Berndt et al., 2014; Huang et al., 2015; Nguyen et al., 2015b; Osborn and Taatjes, 2015; Taatjes et al., 2014).
Gas-Phase Oxidants in Globally Dispersed Forested Environments
Forests cover about one-third of the Earth’s land mass and their emissions of reduced gases strongly influence the global atmospheric oxidation potential. The chemistry in these environments is controlled by abundant biogenic VOCs and sparse nitrogen oxides so that the biogenic VOCs determine the destruction and the production of OH and O3. Measured OH has greatly exceeded modeled OH in studies of several forests (Carslaw et al., 2001; Hofzumahaus et al., 2009; Lelieveld et al., 2008; Ren et al., 2008; Tan et al., 2001; Taraborrelli et al., 2012; Whalley et al., 2011) but not all (McKeen et al., 1997). These higher-than-expected OH abundances in forests have led to the hypothesis that an unknown mechanism recycles HO2 to OH at a maximum rate without generating O3 (Rohrer et al., 2014). To explain these OH abundances, the photochemistry of isoprene has been explored in greater detail, and the mechanisms are now much better understood (Crounse et al., 2011, 2012, 2013; Paulot et al., 2009; Peeters and Muller, 2010; Peeters et al., 2009). Confounding this disagreement between measured and modeled OH are the recent discoveries of interferences in the most common detection method for HO2 (Fuchs et al., 2011; Whalley et al., 2013) and in some (Hens et al., 2014; Mao et al., 2012) but not all (Fuchs et al., 2016) laser-based instruments that measure OH. After these interferences were discovered, interference-free detection methods were devised and are being applied. For some recent field studies, when improved mechanisms are included in the model and interference-free OH measurement techniques are used, the observed OH levels agreed with the modeled OH levels to within the combined uncertainties (Feiner et al., 2016; Hens et al., 2014; Mao et al., 2012). While recent results are encouraging, more work is needed to develop a predictive understanding of the oxidation chemistry in forest atmospheres.
An important constraint on understanding atmospheric oxidation is OH reactivity (defined as the inverse of the OH lifetime [Kovacs and Brune, 2001]). In forests, unidentified OH reactivity is between 30 and 90 percent of the measured total (Di Carlo et al.,
2004; Nolscher et al., 2014; Sinha et al., 2008). However, in some cases, the estimated OH reactivity (constrained to measured primary chemical emissions) agrees within 20 percent of the measured values (Mao et al., 2012; Zannoni et al., 2015). Further work is needed to address the lack of consistent agreement between estimated and measured OH reactivity in order to develop a predictive understanding.
Gas-Phase Oxidants in Urban Regions
The atmospheric oxidation chemistry of urban areas and regions characterized by copious emissions of VOCs, NOx, and particles has been studied for decades and is thought to be understood. However, measured ozone generally exceeds modeled concentrations at high ozone abundances (Appel et al., 2007; Im et al., 2015). The agreement between measured and modeled OH and HO2 in cities is somewhat mixed, with good agreement in some studies and differences of a factor of two in others (Griffith et al., 2016; Stone et al., 2012). Some of this disagreement may be explained by the recently discovered interferences in OH and HO2 measurements, but persists even when interference-free methods are used (Ren et al., 2013). The measured and calculated OH reactivity only agree sometimes (Chatani et al., 2009; Griffith et al., 2016; Kovacs and Brune, 2001; Ren et al., 2003; Yoshino et al., 2006), suggesting the presence of unknown chemical species in some urban areas.
For ozone production, some studies indicate that the concentration of ozone-producing HO2 behaves as expected as NO increases (Mihelcic et al., 2003; Thornton et al., 2002); while for others, the measured-to-modeled HO2 ratio increases as a function of NO, resulting in more ozone production than is predicted by current models (Brune et al., 2016; Emmerson et al., 2005, 2007; Griffith et al., 2016; Kanaya et al., 2007, 2008; Martinez et al., 2003; Ren et al., 2003, 2006). Consistent with this result are direct measurements of ozone production that suggest the calculated rates are too low in the morning when NOx concentration is the greatest (Cazorla et al., 2012). These findings of greater-than-expected measured HO2 and measured ozone production versus NO appear to be inconsistent with the weekend/weekday effect, in which more ozone is produced on weekends when NOx levels are lower (Cleveland et al., 1974; Pollack et al., 2012). Further, many commonly used air quality models under-predict high ozone events and their decrease with NOx reductions. Possible causes for the discrepancies include the details of the chemical mechanism used, boundary conditions, meteorology, and long range transport (Appel et al., 2007; Gilliland et al., 2008; Im et al., 2015) Thus, the lack of consistent agreement between measured and modeled oxidants suggests an incomplete understanding of urban oxidation chemistry and the presence of other factors that control ozone amounts.
Elucidating the role of ROS in atmospheric chemistry and their impacts requires understanding their global abundance and spatial distribution as well as their transformations across a range of different environments. ROS is often used to describe particle components capable of catalytically generating oxidants in the condensed phase both in vivo and in vitro. ROS species include free radicals such as OH, HO2, superoxide anion, peroxynitrites, H2O2, organic peroxides (Pöschl and Shiraiwa, 2015), and transition metals (Charrier and Anastasio, 2011). These compounds are found both outdoors and indoors and there is significant recycling among the different oxidants and across the phases. As discussed in Chapter 4.2, ROS are thought to play a key role in inflammation, pulmonary oxidative stress, vascular dysfunction, atherosclerosis and lung cancer, so there are active research programs studying the health effects of ROS. In the atmospheric chemistry community, mechanistic studies of ROS in particles and on surfaces are in their infancy. A variety of analytical techniques have been developed to quantify PM-induced oxidative stress from different perspectives. Some systems mimic antioxidant loss (e.g., dithiothreitol assay; Cho et al., 2005), others measure cellular response such as macrophage ROS generation (Landreman et al., 2008), other stress expressions and cytokine activation (Wilson et al., 2010).
The most important oxidants in indoor environments are expected to be O3, OH, and NO3 (Gligorovski and Weschler, 2013; Waring and Wells, 2015). Of these, indoor O3 has received the most attention and is best understood at present (Britigan et al., 2006; Weschler, 2000). O3 is known to be lost by reaction on indoor surfaces, with VOCs and with human skin oils (Coleman et al., 2008; Singer et al., 2006; Wisthaler and Weschler, 2010). Some studies on the abundance, sources, and impact of OH indoors have been carried out (Gligorovski et al., 2014; Weschler and Shields, 1997), where OH abundances are expected to be one to two orders of magnitude smaller than outdoor levels due to low OH production, but much remains to be learned. Even less is known about the abundance of NO3 with only one published study of indoor measurements (Nøjgaard, 2010). It is not clear whether the Criegee intermediate or some other unidentified oxidants may be playing important roles indoors.
3.4 CHEMICAL PROCESSES INTERACT WITH ATMOSPHERIC DYNAMICS TO CONTROL THE DISTRIBUTION OF TRACE GASES AND PARTICLES
The distribution of atmospheric trace constituents reflects a balance between the effects of chemical sources and sinks, chemical transformations, and atmospheric transport. Dynamical processes in both the troposphere and stratosphere drive how quickly gases and particles are transported in these regions, as well as vertically between the two regions. Climate-related changes in these processes will drive changes in the distribution of the trace constituents, which can feed back on the dynamics, with implications for weather, air quality, and climate. In the troposphere, extratropical cyclones, which travel along the storm tracks, play an important role in the export of pollution from the continental planetary boundary layer to the global atmosphere. In the stratosphere, the dominant atmospheric transport process is the large-scale meridional overturning circulation, also known as the Brewer-Dobson circulation (Brewer, 1949; Dobson, 1956). Changes in this overturning circulation will influence the transport of pollution from the troposphere into the stratosphere, and the transport of ozone from the stratosphere to the troposphere. Furthermore, the dynamics of the troposphere and stratosphere are coupled, and changes in the stratospheric circulation can influence the storm tracks in the troposphere. Better understanding is needed for how climate-related changes in these transport processes affect the composition and chemistry of the atmosphere and how these chemical changes in turn feed back on atmospheric dynamics.
Stratospheric Overturning Circulation
Recent advances have been driven in part by the use of chemistry-climate models with increasing spatial resolution and more detailed tropospheric and stratospheric chemistry. These models have robustly shown, for example, that warming due to increased greenhouse gases results in intensification of the Brewer-Dobson circulation (SPARC, 2010), which has implications for the distribution of trace gases in the stratosphere and troposphere. Hegglin et al. (2014) reported a negative trend in water vapor in the lower stratosphere and a positive trend in the upper stratosphere, which they attributed to changes in the Brewer-Dobson circulation. Stratospheric water vapor, in addition to functioning as a greenhouse gas, also influences strato-
spheric chemistry by serving as a source of OH and contributing to the formation of polar stratospheric clouds that are key to important ozone-destroying heterogeneous chemistry (see Chapter 2). A positive trend in stratospheric water vapor that exceeded the increase expected from the oxidation of atmospheric methane has been suggested (Rosenlof et al., 2001; Rosenlof and Reid, 2008), but the more recent work by Hegglin et al. (2014) suggests that the trend inferred from balloon data may not be globally representative. This discrepancy highlights the need for better observational coverage—including satellite observations—to understand the changing composition of the atmosphere.
Changes in the Brewer-Dobson circulation also affect tropospheric ozone. Hegglin and Shepherd (2009) estimated that climate-induced strengthening of the Brewer-Dobson circulation could contribute to an increase in the flux of ozone into the troposphere of as much as 23 percent by 2095 (relative to 1965), although Neu et al. (2014) suggested a more modest increase of about 2 percent. While transport of ozone from the stratosphere represents only about 10 percent of the total budget of tropospheric ozone, it is an important source of ozone throughout the upper extratropical troposphere and may contribute significantly on an episodic basis to regional ozone budgets. For example, Lin et al. (2012) predicted that transport of ozone from the stratosphere accounted for 50–60 percent of surface ozone abundances in the western United States, two to three times greater than estimates from previous modeling studies.
Another robust feature in climate models of increased greenhouse gases is a poleward shift of the midlatitude westerly jets (between 30–60° latitude), associated with an expansion of the Hadley cells (low-latitude overturning circulations with rising air at the equator and sinking air at roughly 30° latitude) in the tropics (e.g., Lu et al., 2007; Yin, 2005). Comparisons of models with observations indicate that they generally underestimate the widening of the Hadley cells observed since 1979 (Johanson and Fu, 2009). This bias in the jet position in the models has implications for projections of changes in jet variability and thus midlatitude storm tracks. Barnes and Polvani (2013) suggested that changes in the variability of jets are linked to the mean latitudinal position of the jets. Consequently, models that have an equatorward bias in the jet position are biased in their projection of climate-related changes in jet variability and storm tracks. These models also incorrectly estimate climate-induced changes in summertime surface ozone variability since the maximum in surface ozone variability is linked to the position of the jets (Barnes and Fiore, 2013).
In the Southern Hemisphere, the shift in the jet is also linked to ozone depletion in the Antarctic stratosphere. Increases in greenhouse gases and stratospheric ozone loss have acted in concert to drive a positive trend in the southern annular mode, which is associated with the poleward shift of the westerly jet (Thompson and Solomon, 2002). The trend is robust in austral summer, reflecting the seasonal reduction in stratospheric ozone in austral spring and summer, and is accompanied by a summertime widening of the Hadley cell (Son et al., 2010). As the stratospheric ozone recovers, projected ozone-related changes to the circulation in the Southern Hemisphere will oppose the circulation response to increased greenhouse gases (Son et al., 2010); this change will complicate attempts to isolate the impact of increased greenhouse gases on storm tracks in the Southern Hemisphere.
Cyclones and Monsoons
Climate-related changes in the circulation of the troposphere have also influenced the frequency of extratropical cyclones, which play an important role in ventilating the planetary boundary layer. However, the impact on air quality of these changes in cyclone frequency is uncertain. Wu et al. (2008) suggested that a reduction in the frequency of these cyclones across the United States in the late 20th century may have reduced by 50 percent the air quality benefits expected from decreased anthropogenic emissions in the United States between 1980–2006. However, Turner et al. (2013) predict that there was little change in surface ozone exceedances, despite a decrease in the number of cyclones. These cyclones also influence the transport of continental pollution to the remote atmosphere. Much work has been done to examine intercontinental transport of pollution in the Northern Hemisphere (e.g., Auvray et al., 2007; Doherty et al., 2013; Fiore et al., 2009; Liang et al., 2004; Owen et al., 2006). Fewer studies (e.g., Staudt et al., 2001, 2002) have focused on the impact of pollution on the remote southern tropics and subtropics. For nutrients and pollutants of biogeochemical importance, long-range transport plays a key role in the movement of phosphorus from North Africa to the Amazon (Swap et al., 1992) or iron into the Southern Ocean (Gasso et al., 2010); how these will change under climate change is an important area of future study. It is unclear how changes in storm tracks in the Southern Hemisphere, combined with increasing emissions from countries in South America and Africa, for example, affect the composition and chemistry of the remote atmosphere in the Southern Hemisphere.
Aerosol particles can have an important impact on the radiation budget and dynamics of the atmosphere. Studies have linked changes in rainfall in the South Asian and East Asian monsoons to the radiative effects of aerosol particles (e.g., Bollasina et al.,
2013; Lee and Kim, 2010; Menon et al., 2002; Wang et al., 2015a). Surface cooling, driven largely by sulfate aerosol particles, tends to weaken the atmospheric monsoon circulation and results in reduced precipitation. In contrast, heating from black carbon enhances the circulation and leads to increased precipitation. Bollasina et al. (2014) argued that the influence of local aerosol particles on the dynamics of the atmosphere was the dominant driver for the late 20th century reduction in precipitation in South Asia. It has been suggested that anthropogenic aerosol particles in the Northern Hemisphere resulted in a southward shift in the Hadley circulation during the 20th century, producing a decrease in precipitation in the northern tropics and an increase in precipitation in the southern tropics (e.g., Hwang et al., 2013; Ming and Ramaswamy, 2011). Recent work by Allen (2015) suggested that future reductions in anthropogenic sulfate aerosol particles may drive a northward shift in precipitation in the tropics during the 21st century. However, Kirkby et al. (2016) showed the nucleation of aerosol particles with only highly oxygenated molecules, thus sulfate reduction may not reduce aerosol particle and cloud condensation nuclei (CCN) abundances as much as currently thought. Mechanisms for VOC oxidation and particle nucleation need to be included at the appropriate level of detail in climate models to better understand the influence of aerosol particles on the stability and large-scale circulation of the atmosphere.
3.5 PARTICLE CHEMICAL AND PHYSICAL PROPERTIES AFFECT CLOUD AND AEROSOL PARTICLE RADIATIVE PROPERTIES, CLOUD MICROPHYSICS, AND PRECIPITATION PROCESSES
Aerosol particles and clouds play a key role in the Earth’s energy budget by directly scattering and absorbing radiation. Aerosol particles can also directly affect atmospheric stability and moisture fluxes that drive convective processes. Aerosol particles are the nuclei upon which clouds form and variations thereof can modulate cloud properties, the radiation budget, and the hydrological cycle. Most CCN grow from particles formed by condensation of molecules produced by atmospheric chemistry, but that are initially too small to act as CCN. Clouds in turn can also strongly affect aerosol particles through chemical and microphysical processes. The transformations that occur in aerosol particles during their interactions with water and clouds can affect ecosystems by modulating nutrient fluxes, and changes in precipitation and radiation can alter ocean thermohaline circulation. The interactions between aerosol particles and clouds are multiscale (spanning from spatial scales of meters to thousands of kilome-
ters and timescales from minutes to months), multiphase (among water vapor, liquid water, ice, and all the components found in aerosol particles), and involve a multitude of dynamical and physical processes that couple them.
Inclusion of aerosol particle–cloud interactions in previous Intergovernmental Panel on Climate Change (IPCC) assessments completely changed the model responses to aerosol particles and revealed some fundamental problems with the treatment of cloud formation and cloud feedbacks in global models (Boucher et al., 2013). Combined with the large diversity of natural and anthropogenic sources of aerosol particles over space and time, aerosol–cloud–precipitation–radiation interactions constitute the largest source of uncertainty in assessments of climate sensitivity and anthropogenic climate change.
In terms of weather forecasts, there is a history of weather modification through cloud seeding with aerosol particles like silver iodide, but the scientific basis for its effectiveness has been difficult to ascertain (NRC, 2003). Nevertheless, with recent attention to the question of how particles affect precipitation, a combination of new measurements and chemically enhanced weather forecast models have opened up this field. For example, increases in aerosol particles have been linked to invigoration of convection and the occurrence of intense precipitation events (Koren et al., 2012). The use of interactive aerosol particles in weather forecast models is fairly recent, but is showing that interactive assimilated aerosol particles can increase the forecast accuracy (Kolusu et al., 2015; Sessions et al., 2015). For example, as part of the European Copernicus programme, atmospheric composition variables (greenhouse gases, aerosol particles, and chemical species) have been introduced in the European Centre for Medium-Range Weather Forecasts (ECMWF) model to improve the radiative heating and the Numerical Weather Prediction (NWP) system itself, and other forecast models are engaged in similar experiments (Baklanov et al., 2014; Eskes et al., 2015; Kong et al., 2015; Pleim et al., 2014).
Atmospheric aerosol particles are emitted from a wide variety of sources including soil and deserts, the ocean, volcanoes, biogenic activity, biomass burning, burning of fossil fuels, and numerous other anthropogenic activities. As discussed in Chapter 3.2, reactions of gaseous precursors to form low volatility products also comprise a major source of airborne particles. Aerosol particle composition and size distribution are determined by their sources and are further modified in the atmosphere by chemical reactions with gases, photochemistry, and through interactions with clouds (Boucher
et al., 2013). Cloud feedbacks and the resulting buffered response of clouds and cloud systems mitigate much of the small-scale responses of clouds and radiation to aerosol particles and challenge the understanding of aerosol particles in the Earth system (Stevens and Feingold, 2009). Nevertheless, considerable progress in understanding the links between aerosol particles, radiation, and clouds has been established; see Figure 3.4 for overview. The observational capacity for aerosols has advanced substantially in recent decades both via in situ instrumentation (e.g., Jayne et al., 2000) and satellite observations (e.g., Kaufman et al., 1997), thus improving our understanding of the distribution and composition of particles. Yet the heterogeneity and short lifetimes of these particles ensure that characterization of atmospheric aerosols remains an ongoing challenge.
Aerosol Direct Effect
Decades of research have established that aerosol particles directly scatter and absorb radiation and affect the planetary energy balance. Of all aerosol particle physical parameters that influence scattering efficiency, the uptake of water with increasing
humidity, which is controlled by the composition of the particles, is most important (e.g., Pilinis et al., 1995; Seinfeld and Pandis, 2006; Wagner et al., 2015). The highly variable relative humidity across the scales, and the nonlinear response of aerosol water uptake to it, introduces aerosol direct radiative forcing uncertainty. Recent model intercomparisons suggest that estimates of the global all-sky radiative effect of anthropogenic aerosol remains quite uncertain (−0.58 Wm−2 to −0.02 Wm−2) (Myhre et al., 2013). Furthermore, while estimates of aerosol direct radiative forcing are typically dominated by scattering aerosols, the degree of warming provided by absorbing aerosol has been an active area of research, particularly over the last decade.
The latest IPCC report ranked black carbon as the second most important climate warming agent, after CO2 (IPCC, 2013). However, uncertainties surrounding the sources, mixing state, and optical properties of black carbon lead to large uncertainty in its ultimate impact on the absorption of solar radiation (Bond et al., 2013). The mechanism behind and prevalence of coated black carbon and the associated absorption enhancement is also an active research area (e.g., Cappa et al., 2012). Network absorption measurements and airborne measurements of black carbon over remote regions suggest that models typically overestimate its lifetime, supporting a downward revision of the climate forcing attributed to black carbon (Wang et al., 2014a,b). Quantifying the sources, transport, and removal of black carbon is therefore key to understanding its radiative forcing.
Brown carbon (organic carbon aerosol that absorbs at ultraviolet and visible wavelengths) may also make an important contribution to solar absorption and climate forcing. Initial modeling studies suggest that brown carbon could globally contribute 20–30 percent of the total aerosol absorption at visible wavelengths (e.g., Feng et al., 2013; Wang et al., 2014b). However, the sources and composition of brown carbon, its location throughout the atmosphere (especially in the vertical) as well as its absorption characteristics and how these might evolve in the atmosphere are not well known (e.g., Forrister et al., 2015; Laskin et al., 2015; Saleh et al., 2014).
The AR5 assessment of aerosol-radiation interactions was more uncertain than previous IPCC assessments, reflecting a growing understanding of aerosol sources and properties and, correspondingly, an increasingly complex treatment in models. Much work is needed to reduce the uncertainty on this key climate-relevant metric.
Aerosol Indirect Effect
Clouds can respond to aerosol perturbations through a large suite of complex interactions that affect cloud extent, precipitation rate, and radiative properties. Here we
focus on those that are directly impacted by aerosol particle modulations, noting that the overall cloud responses to aerosol may be more complex than represented here owing to the nature of clouds themselves. Enhanced aerosol particle concentrations that act as CCN may affect the distribution and vertical extent of clouds with important feedbacks on the hydrological cycle (e.g., Andreae et al., 2004; Rosenfeld, 2006; Rosenfeld et al., 2008). Increased concentrations of large aerosol particles, termed giant CCN, may act as efficient collector drops and promote the formation of precipitation (e.g., Cheng et al., 2009; Levin and Cotton, 2009; Woodcock, 1950). The presence of large numbers of giant CCN (e.g., generated during storms) may also deplete water vapor availability in the early stages of cloud formation, strongly affecting the sensitivity of cloud droplets to aerosol particle variations (e.g., Ghan et al., 1998; Morales Betancourt and Nenes, 2014). Another effect of increased aerosol particles is enhancement of evaporation rates of droplets at cloud fringes, which may result in a negative buoyancy feedback from evaporative-entrainment and reduce cloudiness (e.g., Xue and Feingold, 2006). These complex dynamical interactions are neither well understood nor represented in large-scale models.
Particles that act as ice nuclei (IN) are rare in comparison with CCN; about one in a million aerosol particles acts as an IN. It is known that this ability is strongly a function of temperature, and controlled by the physical characteristics and chemical composition of the aerosol particles in a fundamentally different way from CCN. A quantitative understanding of what makes an effective ice nucleating particle and a generally accepted theory for ice nucleation is lacking. Part of this difficulty in establishing a theory is related to the surface chemical and morphological complexity of the IN. The other difficulty lies in the multiple ways that IN can catalyze the formation of ice (the so-called modes of freezing, e.g., deposition, immersion, deliquescence, contact). The small number of particles that do act as IN makes measurement challenging and also has profound implications for clouds that contain ice. For example, if one IN or cloud droplet freezes in an existing cloud, it begins growing at the expense of the existing water droplets because the vapor pressure of water over ice is lower than over liquid water. Hence, the introduction of ice nuclei to a mixed phase cloud modulates cloud droplet size and concentration, which in turn may affect its precipitation efficiency, lifetime, shortwave reflectivity, and longwave emissivity (e.g., Lohmann and Diehl, 2006; Lohmann and Feichter, 2005; Rosenfeld, 2006; Rosenfeld et al., 2008). Increases of IN can profoundly affect pure ice (cirrus) clouds as well, as IN form ice before the more populous supercooled haze droplets. The ice from IN strongly competes for water vapor with the supercooled drops, potentially reducing or completely inhibiting their freezing, strongly affecting crystal number and size, hence cloud lifetime and
An understanding of the behavior of and changes in ice and mixed-phase clouds requires identification of the number of IN in each location. A number of materials have been identified as ice nuclei, including mineral dust, soot, bacteria, fungal spores, pollen, crystalline soluble salts, glassy aqueous materials, and volcanic ash (Hoose and Möler, 2012; Moreno et al., 2013; Murray et al., 2012) with mineral dust being an especially important contributor to IN concentration (e.g., Atkinson et al., 2013; Baustian et al., 2012; Creamean et al., 2013; Cziczo et al., 2013; DeMott et al., 2003a,b; Kamphus et al., 2010; Pratt et al., 2009b). Highly variable contributions from bioparticles can regionally be very important for IN concentrations, especially in regions of agricultural activity (Tobo et al., 2013) and intermediate temperatures (Spracklen and Heald, 2014). Regionally and temporally, biomass burning can act as an IN source (McCluskey et al., 2014). Strong differences in the source and type of IN are expected to exist between land and ocean (Burrows et al., 2013), with potentially important impacts on cloud properties. Recently, biogenic organic material has been suggested to be an important source of IN in remote marine ocean environments (Wilson et al., 2015). While several modeling and empirical approaches describe heterogeneous ice nucleation (e.g., Barahona, 2012; Broadley et al., 2012; Connolly et al., 2009; DeMott et al., 2016; Herbert et al., 2014; Hoose et al., 2010; Khvorostyanov and Curry, 2005; Niedermeier et al., 2011; Vali, 1994), the lack of an accepted mechanistic description for ice nucleation processes severely limits predictive capability in this area.
Compared to pure liquid and pure ice clouds, the effects of aerosol particles on mixed-phase clouds are more complex, owing to thermodynamic and microphysical interactions between cloud particle types throughout the cloud (Cheng et al., 2010; Fan et al., 2013; Lance et al., 2011; Lebo and Morrison, 2014; Lebo and Seinfeld, 2011; Rosenfeld et al., 2008; Saleeby et al., 2011; Storer et al., 2010; Van den Heever et al., 2006). The effects of particles on precipitation are a function of the cloud microphysics and thermodynamics, where particles can suppress warm rain processes and can enhance or suppress cold rain processes (Cheng et al., 2010). The end result of particle modulations on precipitation therefore strongly depend on the specific conditions of cloud formation examined. Precipitation enhancement has been explained by the phase change and release of latent heat during the transport of liquid mass to freezing levels that then fuels the updraft velocity (Lebo and Seinfeld, 2011). Microphysical features of deep convective clouds are sensitive to the type and size of hygroscopic aerosol particles (Storer et al., 2010). In deep convective systems, smaller drops resulting from CCN addition can be lofted higher in the atmosphere given the same updraft velocity
(Rosenfeld et al., 2008). This process can invigorate convection, where the latent heat release from freezing enhances updrafts or secondary convection (Van den Heever et al., 2006). These changes in cloud microphysics affect precipitation of all hydrometeor types, driving thermodynamic and dynamic changes that can affect regional scale precipitation (Fan et al., 2013; Lebo and Morrison, 2014).
Connecting the effects of clouds on radiation and precipitation requires deep understanding of the relationship between clouds and aerosol particles. This relationship is directly controlled by atmospheric chemistry, and the potential responses that can prevail for each cloud type and state. The major mechanisms of interaction between aerosols and clouds have been evaluated on a microphysical level and estimated by cloud-resolving and large-scale models. Further research is needed to confirm aerosol particle influence on cloud-scale dynamics and to embed these influences in multiscale models that can ultimately evaluate changes in radiative balance and precipitation on regional and global scales.
Atmospheric chemistry plays a central role in the biogeochemical cycling of elements (e.g., carbon, nitrogen, sulfur) through the Earth system (Andreae and Crutzen, 1997; Duce et al., 2008; Nadelhoffer et al., 1999). Chemically reduced gases and particles are emitted from terrestrial and marine ecosystems, oxidized or otherwise transformed in the atmosphere, and deposited downwind. Biogeochemical cycles control the elements that are necessary for life on Earth. These cycles support the basic functioning and biodiversity of life (including humans), link biotic and abiotic systems, and ensure the continuous survival of ecosystems. Thus biogeochemical cycles also support the resource production associated with maintaining global food security (e.g., agriculture, fisheries) and certain energy sources (e.g., biofuels, wood). There is a critical need to understand the global distribution of fluxes of elements between the atmosphere and both the terrestrial and marine ecosystems, as well as how atmospheric chemical processes modulate the composition and bioavailability of these elements.
Furthermore, biogeochemical cycles are subject to the influence of ever-changing human activity and global climate and play an important role in the energy balance of the Earth system. The indirect climate forcing exerted by aerosol particles via biogeochemical cycles has recently been estimated to be as large as the direct impact
of aerosol particles on radiation (Mahowald, 2011). This estimate largely reflects the enhanced carbon uptake associated with nitrogen, phosphorus, and iron deposition to the biosphere.
New approaches and tools have largely enabled advances in connecting atmospheric chemistry processes with biogeochemical cycles. The advent of Earth system models, which dynamically couple reservoirs in the Earth system, has initiated new avenues of retrospective and predictive atmospheric chemistry modeling connected to biogeochemistry (e.g., investigation of the role of ozone on the land carbon sink [Sitch et al., 2007] and how climate change affects global dust sources [Mahowald, 2007]). Development and deployment of new analytical measurement approaches permit a more comprehensive characterization of atmospheric constituents relevant to biogeochemical cycling (e.g., new observations of atmospheric ammonia [Altieri et al., 2014; von Bobrutzki et al., 2010] and atmospheric trace metals such as divalent gaseous mercury [Landis et al., 2002; Park et al., 2013]). Global Earth observations from satellite platforms provide new constraints on the sources and transport of key nutrients over poorly observed regions (e.g., global ammonia emissions, constraints on NOx emissions from soils, and transport of dust over the Atlantic [Clarisse et al., 2009; Jaegle et al., 2004; Kaufman et al., 2005]). Additionally, long-term in situ concentration measurements allow the measurement of interannual fluctuations in important constituents, providing new information on the role of biogeochemical cycles at longer time scales (e.g., Prospero and Lamb, 2003).
The biosphere is reaching a state where human impacts are beginning to overwhelm natural ecosystems, which could lead to irreversible changes to the Earth system (Barnosky et al., 2012). Ecosystem changes can be abrupt and due to complex interactions and feedbacks and they are difficult to predict. Changes to the health of global ecosystems impact atmospheric composition by affecting biogenic emissions (see Section 3.1). For example, warming temperatures in the Arctic will result in reduced permafrost as well as increased emissions of methane and carbon dioxide from the land to the atmosphere (O’Connor et al., 2010). Land use and land cover change may play a role in controlling these surface-atmosphere exchange processes and the composition of the atmosphere (Heald and Spracklen, 2015; Unger, 2014). The ocean and atmosphere represent an important and strongly coupled biogeochemical system, where changes in atmospheric deposition of nutrients in the form of dust and other aerosols can induce phytoplankton blooms which change oceanic composition and emissions (Andreae and Crutzen, 1997; Boyd et al., 2007; Chien et al., 2016; Ito et al.,
Nutrient Exchange with the Biosphere: Nitrogen
Humans, through the development of industrial ammonia fertilizer, as well as fossil fuel emissions of nitrogen oxides, have dramatically enhanced the nitrogen supply of the atmosphere (Erisman et al., 2008). Atmospheric deposition of this nitrogen can act as a beneficial external nutrient supply in both terrestrial and marine ecosystems. More than half of the global oceans are depleted in key macronutrients such as nitrate required for phytoplankton growth, and atmospheric deposition provides a key input for nitrogen in these regions (Guieu et al., 2014). However in the terrestrial environment, nitrogen inputs may lead to oversaturation, threatening biodiversity and leading to eutrophication of aquatic ecosystems (Beem et al., 2010; Erisman et al., 2007). Recent work suggests nitrogen deposition will decline worldwide as increases in ammonia emissions will be offset by large decreases in nitrogen oxide emissions (Lamarque et al., 2013). Despite this potential emission shift, other studies show that sensitive ecosystems may continue to be threatened by excessive nitrogen deposition (Ellis et al., 2013).
Nutrient Exchange with the Biosphere: Iron and Phosphorus
Mineral dust represents a major source of iron and other micronutrients which are essential for marine primary productivity (Maher et al., 2010). Iron is the limiting nutrient for more than 25 percent of the surface ocean (Martin, 1990). There has been continuing interest in the relationship between atmospheric iron deposition and marine biological productivity given the proposed link between sea spray aerosols in biologically active regions and climate (Boyd et al., 2007). Changes in the atmospheric dust cycle could play a role in climate dynamics by altering the ocean-atmosphere carbon cycle and thus atmospheric carbon dioxide concentrations (Jickells et al., 2005; Mahowald et al., 2010; Parekh et al., 2006), although evidence for the importance of such a dust-ocean-climate feedback process is still limited (Schulz et al., 2012). Phosphorus, which is a limiting nutrient for many forest ecosystems including the Amazon, is supplied through the transport of smoke particles, bioparticles, and mineral dust (Mahowald et al., 2008; Ridley et al., 2012). Because dust is a vector for nutrient transport, better constraints on global natural and anthropogenic dust source regions using satellite observations (Ginoux et al., 2012), and examining contemporary trends in dust transport (Ridley et al., 2014) are areas that need to be further developed.
Bioavailability is largely governed by the chemical speciation of a nutrient and, in general, insoluble species are not bioavailable. For iron and phosphorus (and perhaps other nutrient trace metals), solubility increases during transport through the atmosphere (Baker et al., 2006; Longo et al., 2016). The causes of this increase are complex and uncertain, but recent work suggests that the interactions of aerosol particles with acids plays an important role in the process (e.g., Baker and Croot, 2010; Kumar et al., 2010; Longo et al., 2016; Meskhidze et al., 2005; Nenes et al., 2011). These interactions of metals with atmospheric acidity, whether from sulfate, nitrate, or organic acids, lie at the intersection of atmospheric biogeochemistry and more traditional atmospheric chemistry. Recent studies have also highlighted the importance of the mineralogy of the dust for the amount of easily solubilized iron (e.g., Journet et al., 2008; Schroth et al., 2009) as well as the role of combustion sources of soluble iron (Chuang et al., 2005; Guieu et al., 2005).
Ozone Exchange with the Biosphere
Tropospheric ozone can damage vegetation, leading to reductions in the global land carbon sink (Sitch et al., 2007) and declines in global agricultural productivity (Avnery et al., 2011; Tai et al., 2014). Although uptake of ozone by vegetation represents a harmful ecological impact, it is also one of the primary removal mechanisms for surface ozone. However, estimations of this impact are highly dependent on the description of land use and vegetation phenology in models (Park et al., 2014; Val Martin et al., 2014). Efforts have been made to routinely estimate ozone dry deposition velocities from near-surface concentration measurements (e.g., Clarke et al., 1997). While direct measurements of ozone deposition flux are available from an array of specific sites for short-term studies, long-term measurements are limited (e.g., Fares et al., 2010; Munger et al., 1996).
Organic Compound Exchange with the Biosphere
The deposition of trace gases and aerosol particles to terrestrial and marine ecosystems plays a vital role in controlling the lifetime of atmospheric chemical species, while at the same time, dictating the location and extent of impacts described above. Goldstein and Galbally (2007) highlighted that a poor understanding of the sinks of organic species affects the ability to constrain the global budget of VOCs, yet almost no measurement constraints are available to quantify this sink. Hallquist et al. (2009) estimated that deposition of oxidized vapors is the dominant sink of VOCs from the atmosphere, further emphasizing the need for new observations. New analytical ap-
proaches have been applied to estimate the surface fluxes of many gas-phase organic species, allowing measurement of the location of deposition, and thus its impact (Karl et al., 2010; Nguyen et al., 2015a; Park et al., 2013). Beem et al. (2010) and Kanakidou et al. (2012) showed that the wet deposition of organic nitrogen can make up an important fraction of total nitrogen deposition to terrestrial and marine ecosystems, respectively. However, while network measurements regularly report the wet deposition flux of inorganic acids throughout much of North America and Europe (e.g., the National Atmospheric Deposition Program in the United States), the wet removal of organic species (including organic nitrogen and organic acids) is not routinely monitored. In addition, much of the deposition occurs in the form of dry deposition, for which there are very limited observations. Ecosystem damage from acid deposition (i.e., acid rain) has been quantified and addressed in some regions (see Chapter 2); yet, the way in which this type of damage translates to reductions in global carbon uptake is not well known.
Relative to terrestrial sources, far fewer measurements have characterized VOC emissions from the ocean. Recent studies have quantified marine emissions of species such as isoprene and monoterpenes, and these limited measurements indicate that these emissions are extremely low (Carpenter et al., 2012). Several decades ago, scientists speculated the ocean could act as a “planetary thermostat” and regulate climate through biogenic organosulfur (dimethyl sulfide) emissions which undergo processing to ultimately produce sulfate aerosols that can seed marine clouds and lead to changes in cloud properties, the so-called CLAW hypothesis (Charlson et al., 1987). After decades of conflicting results from marine field studies, the evidence for the CLAW hypothesis remains elusive, however the links between the biological productivity of the ocean, sea spray and gas phase emissions, and clouds are still under debate (Quinn et al., 2015).
Mercury Exchange with the Biosphere
The atmosphere transports and transforms compounds that damage ecosystem and human health (e.g., mercury, ozone, polycyclic aromatic hydrocarbons). For example, the understanding of the global biogeochemical cycling of mercury has advanced considerably over the last decade, in large part due to new measurements and modeling of atmospheric mercury and its transformations (Ariya et al., 2015; Selin, 2009). Recent advances include better modeling of atmosphere-ocean coupling (Soerensen et al., 2010), a better understanding of polar mercury depletion events (Steffen et al., 2008), the identification of bromine as a globally important oxidant of elemental mercury (Ariya et al., 2015; Holmes et al., 2010; Obrist et al., 2011), especially in the high lat-
itudes, and a better understanding of mercury sources (Pirrone et al., 2010). However, the effect of the deposition of mercury and other toxic constituents on ecosystem productivity is poorly understood. Limited efforts have explored the biogeochemical cycles and environmental impacts of other contaminants, such as persistent organic pollutants (e.g., Lammel et al., 2009).
The topics described in this chapter span the discipline of atmospheric chemistry research. This chapter has discussed a few examples that have recently been recognized as being important for understanding and predicting atmospheric processes and impacts, but for which key areas of uncertainty remain. These issues not only contribute to advancing fundamental scientific knowledge in a number of fields, including chemistry, physics, and meteorology, but—as described in the next chapter—they also lie at the core of understanding the basis of, and possible solutions to, many issues central to human health and welfare.
This page intentionally left blank.