Monitoring and Modeling of Outdoor Air Pollution
Summary and Keywords
Air pollution has been a major threat to human health, ecosystems, and agricultural crops ever since the onset of widespread use of fossil fuel combustion and emissions of harmful substances into ambient air. As a basis for the development, implementation, and compliance assessment of air pollution control policies, monitoring networks for priority air pollutants were established, primarily for regulatory purposes. With increasing understanding of emission sources and the release and environmental fate of chemicals and toxic substances into ambient air, as well as atmospheric transport and chemical conversion processes, increasingly complex air pollution models have entered the scene. Today, highly accurate equipment is available to measure trace gases and aerosols in the atmosphere. In addition, sophisticated atmospheric chemistry transport models—which are routinely compared to and validated and assessed against measurements—are used to model dispersion and chemical processes affecting the composition of the atmosphere, and the resulting ambient concentrations of harmful pollutants. The models also provide methods to quantify the deposition of pollutants, such as acidifying and eutrophying substances, in vegetation, soils, and freshwater ecosystems. This article provides a general overview of the underlying concepts and key features of monitoring and modeling systems for outdoor air pollution.
Air pollution was already recognized as a serious threat to human health and well-being in the early 1950s, when the great London Smog of 1952 caused thousands of deaths and contributed to health problems in a large urban population (Bell & Davis, 2001). But even before the mid-20th century, concerns about deteriorating urban air quality had been reported in fast-growing cities in the United States, such as San Francisco and New York City (Lioy & Georgopoulos, 2011). Today, extreme levels of air pollution affect many major urban areas in the world, in particular the rapidly expanding megacities in China, India, South East Asia, and South America. In contrast, urban and rural concentrations of key air pollutants have declined since major efforts to reduce emissions have been introduced in European countries, the United States, and Japan.
In a comprehensive study to assess the global burden of disease (GBD), Lim et al. (2012) identified air pollution—both indoor air pollution from solid fuels, and ambient particulate matter pollution—as one of the top ten risk factors for human health on a global scale.
Here, outdoor air pollution by key air pollutants serves as a focus for discussion of the development of air pollution challenges, of approaches to monitoring and modeling pollution, and of emerging and future challenges posed by air pollution for human and environmental health.
Development of Air Pollution Challenges
Brief History of Air Pollution
In the 1970s, first evidence of forest dieback (Waldsterben; see Krahl-Urban et al., 1988) was reported, and in the 1980s, substantive research efforts revealed a strong relationship between the emission of sulfur dioxide from fossil fuel combustion in coal-fired power plants and the levels of acidity in precipitation. Sulfur dioxide (SO2) became the focus of efforts to reduce emissions, and as early model analyses revealed the long-range transport of emissions across country boundaries, the transboundary nature of the challenge required an international approach. The 1979 Geneva Convention on Long-range Transboundary Air Pollution (CLRTAP), under the United Nations Economic Commission for Europe (UNECE), established its first protocol on the reduction of sulfur emissions in 1987, with the objective of reducing emissions or their transboundary fluxes by at least 30%. Subsequent research identified further threats to human health and ecosystems stemming from high concentrations of tropospheric (or ground-level) ozone and its precursors (mainly nitrogen oxides and volatile organic compounds), leading to further CLRTAP protocols on their emissions (1991 and 1997, respectively). In 1998, with recognition that the original sulfur reduction targets would still leave major areas of European ecosystems at risk of exceeding critical loads for acidification, a second protocol was ratified to further reduce sulfur emissions.
The latest CLRTAP protocol recognized the close relationships between the key air pollutants sulfur dioxide, nitrogen oxides, volatile organic compounds, and ammonia, and the potential for trade-offs and unintended consequences of emission-control strategies focusing on a single pollutant. This protocol, widely known as the Gothenburg Protocol, was signed in 1999, entering into force in 2005. In 2012, it was revised and amended to include emissions of fine particulate matter (PM2.5), taking account of emerging evidence of the adverse health effects of exposure to particulate matter and the widespread exceedance of guideline values set by the World Health Organisation (WHO).
While the UNECE incorporates 56 member states in Europe, Asia, and North America, the European Union, currently with 28 members, has developed stringent emission controls and air quality limit values. In contrast to the UNECE protocols, which—upon ratification by the parties to the CLRTAP convention—have to be translated into national legislation, an EC directive is a legal act of the European Union, which requires member states to achieve a particular result without dictating the means of achieving that result. It can be distinguished from regulations that are self-executing and do not require any implementing measures. The Air Quality Framework Directive of 1996 and its new version of 2008 (EC Directives 96/62/EC and 2008/50/EC) provided the basis for so-called daughter directives setting ambient concentration standards, averaging periods, and target years for a wide range of air pollutants, including priority heavy metals and persistent organic pollutants (POPs). As a parallel development between the UNECE CLRTAP and the EU, the Protocol to Abate Acidification, Eutrophication and Ground-level Ozone (Gothenburg Protocol) and the EC National Emission Ceilings Directive (NECD, Directive 2001/81/EC) both focus on the simultaneous reduction of key air pollutants contributing to acidification, eutrophication, and human health effects. While the NECD ceilings vary slightly from the emission limits adopted in the Gothenburg Protocol, the overarching rationale is to address transboundary pollution effects by reducing emissions at the sources. Other directives directly address specific emission sources, for instance the EC Large Combustion Plants (LCP) Directive (Directive 2001/80/EC) or the EC Medium Combustion Plants Directive (Directive 2015/2193).
In parallel to the developments in Europe, the U.S. Clean Air Act, first established in 1955, was subsequently amended to detail regulatory requirements to attain air-quality limit values on a national scale. The amendments in 1990, for instance, included acid rain, the release of ozone-depleting substances, and toxic air pollution. In Japan, the Basic Environment Law was established in 1993 and sets out the basic policies for environmental conservation, including regulations for environmental pollution control and environmental quality standards. Similar regulations have been introduced in many countries and regions, and a good overview is provided by UNEP in a global report. Finally, activities like the Global Atmospheric Pollution Forum (GAP Forum) support the development of solutions to air-pollution-related problems by promoting effective cooperation at regional, hemispheric, and global levels.
Key Air Pollutants
Many chemical substances, particles, and aerosols can be classed as air pollutants and have adverse effects on human health and ecosystem functioning (Figure 1). Here, the focus is on five pollutants that are the primary concern for the exceedance of critical loads and levels for the acidification and eutrophication of soils and freshwater ecosystems, and that are known to have substantial public health impacts.
Sulfur Dioxide (SO2)
SO2 is emitted mainly from the combustion of solid fossil fuels, such as hard coal or lignite, and, to a lesser extent, from the combustion of heavy and light fuel oil (e.g., shipping and diesel vehicles). Identified as a main contributor to acid rain, emissions of SO2 have substantially declined since the 1970s, with flue gas desulfurization technologies for large combustion plants and fuel sulfur content reductions for liquid fuels having resulted in reductions of more than 95%. SO2 contributes to the formation of secondary inorganic aerosols (SIA), together with ammonia and nitrogen oxides.
Nitrogen Oxides (NOx)
Nitrogen oxides are the sum of nitric acid (NO) and nitrogen dioxide (NO2). Primary emissions of NO from combustion processes are typically converted to NO2 under ambient conditions, depending on the composition of the atmosphere and the presence of other trace gases. NOx are emitted from combustion processes as thermal, fuel, or prompt NOx, depending on the formation process and the origin of the nitrogen used in the formation. As is the case for SO2, large combustion plants have traditionally been major contributors to NOx emissions, but stringent emission controls and the implementation of catalytic and noncatalytic converters have reduced emissions substantially. Residential combustion sources (e.g., natural gas boilers) and other industrial combustion processes contribute to some extent, but the bulk of NOx emissions can be attributed to road transport. Technological advances and increasingly stringent emission-control legislation have resulted in overall reductions of NOx emissions from the road-transport sector, despite increasing vehicle fleets and mileage. NOx are involved in the formation of SIA and contribute to the formation of ground-level ozone (O3) concentrations, in the presence of sunlight and non-methane volatile organic compounds (NMVOCs).
The by far largest share of anthropogenic ammonia emissions stems from agricultural activities, especially animal husbandry, manure management and application, and mineral fertilizer use. NH3 contributes to local acidification and eutrophication effects, affecting ecosystem functioning and biodiversity. Another important role of ammonia is enabling the formation of SIA, both ammonium nitrates and ammonium sulfates, which can be transported through the atmosphere and lead to deposition of acidifying and eutrophying substances, as well as affect urban air quality as fine particulate matter far away from the emission sources. In contrast to the case with most other air pollutants, policies reducing ammonia emissions have so far had only moderate success, and, in Europe, NH3 emissions have remained relatively stable, with the exception of Eastern and Central European countries, where the decline in mineral fertilizer use and in animal numbers as a consequence of the economic restructuring process have led emissions to decline.
Non-Methane Volatile Organic Compounds (NMVOCs)
Volatile organic compounds are a group of substances, rather than a single pollutant. Typically, methane is excluded to account for those compounds that contribute to air pollution, whereas methane (CH4) itself is known to contribute to the formation of ground-level ozone and is a powerful greenhouse gas. NMVOCs are primarily emitted from the use of chemicals and solvents in paints, glues, and adhesives, in both industrial and domestic use. NMVOC emissions occur, to a lesser extent, in the combustion of solid and liquid fuels ( e.g., in road transport) and the evaporation of gasoline from fuel tanks. NMVOCs contribute to the formation of ground-level ozone and some of the compounds have toxic or carcinogenic effects, such as, for instance, benzene. The individual compounds differ in terms of their reactivity, volatility, ozone-forming potential, and contribution to the formation of organic aerosols (for background information the role of NMVOCs in atmospheric composition change, see Seinfeld & Pandis, 2016).
Particulate Matter (PM)
Similar to NMVOCs, particulate matter (PM) describes a group of pollutants; however, in contrast to NMVOCs, the distinguishing factors for air pollution purposes are typically size distribution and composition. The most used classes are coarse particulate matter, or PM10, which describes particles with an aerodynamic diameter of <10 µm, and fine particulate matter, PM2.5 (< 2.5 µm diameter). More recently, even finer particle sizes of < 1 µm, or ultrafine particles (UFP) with a diameter of < 0.1 µm, have been considered, due to their ability to penetrate deep into the human respiratory tract and even enter the bloodstream, with the potential to cause systemic inflammation and other harmful effects on human health. PM components of interest are organic carbon (OC) or elemental carbon (EC), which can be used to determine source fingerprints and support source attribution of measured particles. Finally, PM can be split into primary PM, which is directly emitted from a source (e.g., combustion particles from road transport vehicles), and secondary PM (e.g., secondary inorganic or organic aerosols, formed in the atmosphere from precursor emissions). Emissions of PM10 and PM2.5 have been regulated for both stationary and mobile sources and direct emissions of PM have declined, albeit less so than emissions of SO2 and NOx.
Current Approaches to Monitoring Air Pollution
Scientific Discovery–Driven Measurements
Measuring the composition of the atmosphere has a long tradition and a substantial amount of literature and many textbooks exist that detail the techniques and approaches established over the past decades (e.g., Emeis, 2010; Heard, 2006). Scientific measurements of atmospheric components, including trace gases acting as air pollutants and climate forcing gases, include long-term activities (e.g., CO2 measurements at the Mauna Loa Observatory in Hawaii, spanning several decades) as well short, intensive measurement campaigns, where multiple scientific teams utilize a multitude of instruments to gain an in-depth understanding of atmospheric processes at finer scale (e.g., the Clean Air for London, ClearFlo, experiment in 2010; Bohnenstengel et al., 2014). Such measurement campaigns can combine ground-based instruments with aircraft, balloon, or other airborne platforms, as well as utilizing satellite and other remote sensing approaches (e.g., Light Detection And Ranging—LIDAR). The strength of these intense measurement activities lies in the ability to collect and analyze highly resolved and detailed data, which can underpin experiments in laboratories and provide insight into how atmospheric chemistry and physical processes operate in real-world conditions.
Regulatory Monitoring—Ground-Based Networks
One specific application of air-pollution measurement approaches is the deployment and operation of regulatory monitoring networks. The primary objective for establishing such networks is to monitor ambient levels of air pollution to safeguard public health by ensuring that limit values set for the protection of human health are not exceeded. Regulatory networks are set up to monitor atmospheric concentrations of priority pollutants (for instance, carbon monoxide, nitrogen oxides, particulate matter, sulfur dioxide, ammonia, and volatile organic compounds), deposition of nitrogen or sulfur components (to gauge the input of acidifying substances or nutrients into terrestrial and freshwater ecosystems), or concentrations and deposition of heavy metals (e.g., arsenic, cadmium, chromium, lead, copper, mercury, vanadium, zinc, or mercury).
In general, regulatory monitoring networks contribute to several objectives:
• Checking if statutory air-quality limit values are complied with (e.g., compliance with air-quality directives of the European Commission)
• Providing information to the public about air quality
• Providing information for local air-quality review and assessments (e.g., to design local and national air-quality strategies)
• Identification of long-term trends in air-pollution concentrations
• Assessment of the effectiveness of policies and measures to reduce air-pollution effects.
The European Monitoring and Evaluation Programme (EMEP) is a pan-European activity that has emerged from the investigation of the causes of forest dieback and the identification of sulfur emissions from fossil fuel combustion as the main source, back in the 1970s. It is a scientifically based and policy-driven program under the UNECE CLRTAP for international cooperation to solve transboundary air-pollution problems. Within EMEP, a vast amount of monitoring data from national networks are compiled and made available, and the Chemical Co-ordination Centre (CCC) of EMEP fosters international collaboration and exchange (e.g., to promote quality assurance and control and the implementation of common standards and approaches for atmospheric measurements).
The EMEP program and the air-quality directives of the European Commission are the main regulatory drivers for national governments in Europe to implement detailed monitoring strategies and to operate networks.
To date, ground-based monitoring networks have contributed the lion’s share of routine data on atmospheric composition and air-pollution levels and trends. They have supported the assessment of effectiveness of policies to reduce air pollution by providing data over many years, showing the decline of some of the priority air pollutants (notably nitrogen oxides, sulfur dioxide, and heavy metals), while for other pollutants they reveal relatively little change (e.g., ammonia).
However, monitoring networks are expensive to operate and maintain. In the view of budget cuts and economic pressures on national governments and local authorities, the number of monitoring sites and the components measured are in decline. And while national governments strive to optimize their monitoring costs, this reduction in monitoring activities has potentially substantial consequences for scientific research, which relies on the routinely collected data for the validation of models and the analysis of atmospheric processes. In other regions, in particular in developing and emerging economies, observation data from routine monitoring activities are even scarcer, if available at all.
An additional, albeit different, challenge is the use of regulatory monitoring data for the quantification of public health effects, i.e., to determine associations between respiratory or cardiovascular health issues and air pollution. Even in countries with a high number of monitoring sites, there are typically only a few sites in urban areas and the monitoring sites in rural areas are even more sparsely distributed. Consequently, the spatial and temporal variability of air pollutant levels, and hence the exposure of the population, can be quantified based on only a small number of data points that are not representative. Recent research (e.g., by Willocks et al., 2012) has illustrated that identifying associations between health effects of air pollution and monitoring data may fail due to the lack of spatial coverage and detail in existing monitoring networks.
Some of these challenges may be overcome by utilizing other monitoring techniques, e.g., remote sensing or earth observations (see section “Remote Sensing and Earth Observation”), or by crowdsourcing air-pollution measurement data from emerging low-cost, ubiquitous sensor technologies (see section “Emerging New Technologies”). In addition, modeling, particularly integrated approaches using sensor and model data, shows promising results (Reis et al., 2015).
Remote Sensing and Earth Observation
Satellites and airborne platforms, as well as ground-based instruments capable of conducting measurements of atmospheric composition across a wider area, have become widely used both to complement traditional site-based observations and to provide spatial and temporal coverage not currently available. While space-based platforms have substantially increased in number and data quality, they require additional information to calculate atmospheric concentrations because they observe a quantity representing the concentration of a trace gas or aerosol across the whole atmospheric column to the earth’s surface. In addition, surface albedo, cloud cover, and interferences between different gases and aerosols need to be taken into account. A major advantage of satellite measurements is that they can cover a large area, and by conducting repeated measurements at the same location, can provide vital insights into temporal emission patterns.
The EUMETSAT Satellite Application Facility on Ozone and Atmospheric Chemistry Monitoring (O3M SAF), for instance, provides data products on ozone and atmospheric chemistry products, while the Global Ozone Monitoring Experiment–2 (GOME-2) provides a detailed picture of the total atmospheric content of ozone and the vertical ozone profile in the atmosphere. The recent development of the Sentinel missions will generate an unprecedented amount and quality of environmental data and contribute to the already substantial range of earth observation services operated by the European Space Agency (ESA) or NASA. Other space-borne platforms launched by Japan, China, and India will further contribute to the capacity for earth observation. The Japanese Aerospace Exploration Agency (JAXA) operates, for instance, several satellite platforms, including GOSAT for the observation of greenhouse gas emissions.
Aerosols are measured from space by quantifying the aerosol optical depth (AOD) as an indicator for the abundance of particles in the atmospheric column. Figure 2 displays measurements of AOD made with a multi-angle imaging spectroradiometer (MISR) on board the NASA TERRA satellite as an example.
Emerging New Technologies—Low-Cost Sensors and Citizen Science
In addition to traditional ground-based monitoring technologies and remote sensing, the decrease in size and cost of gas and particle sensors and their increased sensitivity have led to the development of low-cost sensor devices. At price ranges from a few tens to hundreds of €, these sensors enable, for the first time, citizens and civil society groups to take their own measurements of air quality and conduct analyses of the data generated. The potential for crowdsourcing of air-quality data based on these developments is substantial, and activities like the Open Air Laboratory (OPAL), a U.K.-based citizen science network, demonstrate the power of public engagement and citizen scientists contributing to a data-rich picture of the environment.
To date, however, many of the low-cost sensors being marketed have not undergone rigorous testing and validation, making it difficult to directly compare data generated by these sensors with official monitoring results. While this is creating challenges for the dialogue between citizen scientists and regulatory agencies, the emerging scientific activities underpinning the assessment, and developing quality assurance and control approaches for low-cost sensors, will contribute to overcoming the challenges. In the context of budgetary constraints and the reduction in the number and frequency of regulatory measurements over recent years, complementing existing monitoring networks with medium- or low-cost sensors is a promising way forward.
Purpose and Scope of Emission Inventories
Emission inventories are compiled for a variety of purposes; most often, they are based on legal requirements to document and quantify emissions of air pollutants or greenhouse gases into the environment and/or provide relevant input data for atmospheric dispersion modeling. While these two aspects are not inherently different, requirements for the compilation of inventories can vary with their purpose. The Intergovernmental Panel for Climate Change (IPCC) highlights five key requirements for the generation of emission inventories—transparency, consistency, completeness, comparability, and accuracy—and has published comprehensive materials as guidelines for the compilation of greenhouse gas inventories and quality assurance/quality control (QA/QC). The criteria and the guidance are equally valid for the generation of any emission inventory dataset and are not limited to greenhouse gases.
Emission Inventories for Compliance Monitoring
A number of national and international agreements and laws exist that limit the release of potentially harmful substances into the environment. Air pollutants have been subject to substantial regulation over time and emission inventories play a vital role in documenting emissions and emission reductions. Based on an agreed methodology and calculation rules, inventories for a base year are developed, and using the same concepts, for forthcoming (target) years, supporting compliance monitoring as well as allowing to measure progress towards targets in the future.
This process requires detailed documentation of the emission sources included or excluded in the agreement, the basis for calculations both for the base year and for the target years, and mechanisms for potential recalculations if errors are observed or new scientific knowledge emerges.
Emission Inventories for Modeling
Emission inventories for modeling purposes are often governed by different key requirements, first and foremost completeness and transparency. While regulations at times exclude some emission sources from reporting (e.g., international shipping or aircraft emissions, emissions from natural sources), inventories for modeling aim to be comprehensive. Only by including all sources and all relevant substances can atmospheric dispersion models achieve results that can be validated against measurements and hence provide a robust basis for achieving a better understanding of atmospheric transport processes or chemistry processes.
The spatial resolution and geographical representation of emission inventories mainly depend on their geographical coverage and the complexity of the emission sources included in the inventory. The resolution may range from 1° × 1° on a global or hemispheric scale, to 50 km × 50 km or 10 km × 10 km on a regional scale (e.g., for Europe), and finally to 5 km × 5 km or 1 km × 1 km on a national scale. In general, the spatial resolution of an emission inventory is a function of the level of detail, in which data for the spatial disaggregation are available, and the amount of resources/time that can be dedicated to generate and process the increasingly large datasets. While, in general, emission inventories are most often displayed mapped in a grid, the underlying data distinguish between area, line, and point sources (see Figure 3 and the discussion of types of sources below).
Most emission inventories, in particular those addressing compliance monitoring, provide static figures for a reporting period, often one calendar year. For modeling purposes, however, the actual temporal distribution of emissions occurring at different times of the day, during different seasons, and with varying time patterns, need to be resolved. Temporal resolution is relevant because not only the exact location of the emission source, but also the point in time when the substance is released into the air, are required to model the fate and transport of the substance from its origin.
Methods to achieve a sufficient temporal resolution of emissions can range from simple time-fractions of emissions (derived, for instance, from operating-time patterns of a combustion plant) to complex and sophisticated functions, including dependency on meteorological parameters like ambient temperature (e.g., for emissions of ammonia in agriculture or biogenic/natural emissions). For the latter, the typical approach is to use annual total values as input from the inventory and process these based on functions built into the atmospheric dispersion model, as all relevant parameters are often available within the model (temperature, precipitation, wind direction, and speed, etc.). Other temporal disaggregations that are mainly based on activity patterns (e.g., road traffic peak times, working times, holidays) can be generated offline, based on statistical input data, such as road traffic counts or power generated.
Current emission inventories are often very detailed regarding their sectoral disaggregation, in most cases reflecting the level on which inventory compilation takes place. However, the rationale behind the sectoral allocation and split reflects as well the main purpose for which an inventory has been designed. This means that an inventory primarily developed to account for air-pollutant emissions will be split according to technologies, will include combustion as well as noncombustion emissions, and likely will have a high level of detail regarding the compound speciation (see section “Compound Speciation”). On the other hand, inventories containing emissions of greenhouse gases will be dominated by sectors with fossil fuel combustion, at times combining sectoral activities that are related only by the fuel type used, or showing a sectoral allocation that is less driven by technologies than by economic sectors.
In most cases, inventories contain information on individual substances, such as sulfur dioxide, carbon monoxide, or ammonia, which can be directly used in atmospheric dispersion models because their physical and chemical characteristics are known. However, some pollutants and compounds comprise a set of different substances, and information on the speciation within the set is vital for models to process them accordingly. This is the case for:
• Nitrogen oxides (NOx): The split between NO and NO2 is relevant, and NOx is most often shown in inventories as the sum of both and is reported with the molecular weight of NO2.
• Non-methane volatile organic compounds (NMVOCs): For the purpose of EU legislation, NMVOCs are defined as “. . . any organic compound having an initial boiling point less than or equal to 250 °C measured at a standard atmospheric pressure of 101.3 kPa and [that] can do damage to visual or audible senses.” For reporting in inventories, NMVOCs are combined in one group; however, the various substances in NMVOCs have quite different chemical and physical characteristics. For example, one aspect is their varying potential to contribute to the formation of tropospheric ozone.
• Particulate matter: Information is needed on the size distribution (PM10, PM2.5, PM1, PM0.1) as well as on particle mass, particle numbers, and—particularly relevant for health impact assessment—the chemical composition (BC, OC, EC, heavy metals, organics).
Typically, emission inventories contain aggregated information on speciation by sector or subsector. For example, getting the speciation right is the main prerequisite for modeling photochemical smog (especially tropospheric ozone), as the ozone creation potential of different NMVOCs can vary significantly. Similarly, the speciation of PM determines the physical characteristics of the particulates modeled with regard to (re)suspension and deposition.
Types of Sources
Emission sources can be classified in different ways: according to their industrial or economic sector, their operation, their characteristics, and so on. On a basic level, stationary and mobile sources are often distinguished, but a more detailed differentiation into area, line, and point sources is more useful, in particular with regard to the use of emission data in atmospheric dispersion models.
Point Sources, Line Sources, and Area Sources
Three main types of emission sources can be distinguished based on the physical characteristics of the source:
• Area sources
• Line sources
• Point sources
Area sources comprise the sum of all diffuse, small sources, which are best displayed aggregated into grid cells, mainly marked by a low density of emissions per area and low stack heights. This category typically covers residential and commercial combustion sources, household application of paints and solvents, small off-road engines, agriculture, etc.
Line sources are sources with a vector, containing mobile emission sources and characterized by a defined throughput of activity rates (e.g., vehicles per hour) at a certain node. A network of line sources, such as the network of major roads and motorways, of a country can provide information on the applicability and effect of specific measures to reduce road-transport emissions. Major line sources are typically clearly visible even when intersected with a gridded emission inventory of area sources due to the relatively high emission density along the lines, whereas the emission height is regarded low.
Point sources (often also called large point sources, LPS) are made up of large individual emission sources (e.g., large combustion plants, industrial facilities, or animal housing in the pig and poultry sector), often with high stacks injecting the emissions directly into the troposphere at heights between 100 and 300 m above ground. Due to the significant source strength and the generally higher temperature of exhaust gases, plumes for point sources can often display very specific characteristics and are highly relevant for atmospheric dispersion modeling. Point source emissions present the highest emission density levels and when plotted as individual points on a gridded map, typically exceed the emission levels of the surrounding area by a large margin.
Stationary Sources: Public Power Plants, Industrial Combustion, Residential and Commercial Combustion, Industrial Production Processes
Stationary combustion of fossil fuels—mainly hard coal, lignite, and natural gas—has contributed a significant share of emissions of carbon monoxide, nitrogen oxides, and sulfur dioxide. However, due to the implementation of stringent emission limit values and control measures, these large sources have been subject to substantial reductions of their unit emissions (for some processes, > 90% reduction) since the 1970s. In current years, the majority of emissions of NOx have been emitted by road transport and stationary combustion, while the bulk of SO2 emissions stem from public power generation. Emissions from residential and commercial combustion contributed only a small share of NOx and SO2 emissions; however, both CO and PM emissions from small combustion units can be substantial regionally. Due to the combustion of large amounts of fossil fuels, power generation is one of the main sources of CO2 emissions. A trend toward replacing coal- and lignite-fired power plants with natural gas-fuelled plants has been observed for some time, which could lead to even lower emissions per unit of power generated. However, this heavily depends on relative fuel prices and the type of power generated. Where plants burning hard coal and lignite are mainly generating base-load power, natural gas plants typically provide mid-load and peak-load generating capacity, as they can be powered up comparatively faster than coal-fired plants.
Industrial production processes contribute to emissions of a variety of pollutants and greenhouse gases (GHGs), depending on the processes applied and on production cycles and input materials.
Mobile Sources: Road Transport, Off-road Transport and Machinery, Railways, Shipping and Aircraft Emissions
Mobile emission sources are typically split into two major groups: road transport and other mobile sources. However, in the view of the diversity of these two sectors, other classifications would equally apply, for instance, distinguishing between private and public transport, passenger and freight transport, and so on.
As one of the major contributors to emissions of NOx and NMVOCs, road-transport emissions have been subject to stringent regulations from the early 1990s, namely under the EC legislation commonly known as the EURO standards. Within this sector, passenger cars, light-duty vehicles, heavy-duty vehicles, and mopeds and motorcycles contribute different shares of emissions of mainly ozone precursors and particulate matter. Differences exist, for instance, between passenger cars (NOx and NMVOCs emissions) and heavy-duty vehicles (mainly NOx emissions), as well as between gasoline-fuelled and diesel-fuelled vehicles (diesel engines being more efficient, but emitting relatively more NOx and PM than gasoline engines, and less NMVOCs). Particulate matter emissions—both exhaust and noncombustion—of motor vehicles have emerged as a major concern in recent years, mainly due to epidemiologic studies linking respiratory health effects and cancer to exposure to fine particulate matter (PM10, PM2.5 and below). Diesel engines in particular have been under pressure due to their soot emissions, and more stringent regulations have been established in the EURO 5 and EURO 6 standards for diesel engines in particular.
Among other mobile sources, national and international shipping (including recreational boats), railways, and off-road vehicles (tractors and machinery used in agriculture and forestry, military vehicles) are included. In addition to these, equipment used for power generation and construction engines are covered, as well as handheld equipment (lawn mowers, chainsaws, and so on). And while road-transport sources have been subject to intensive measurements to establish both test-cycle and in-use emission factors, many of the off-road and equipment emissions are not well understood, and quantifying emissions from these sources is typically regarded as subject to major uncertainties.
Finally, shipping emissions and, to a lesser degree, emissions from aircraft, are projected to increase significantly due to the global demand for freight and passenger transport in the future. Emissions from these sources have not been subject to substantial regulation in the past and only recently activities to reduce SO2 emissions from international shipping have been launched. While the current share of these emissions relative to road transport is comparatively low, their steady increase and the decline of road-transport emissions at the same time due to implemented emission-control regulations have put these sources in the spotlight. In some coastal areas, deposition of acidifying substances originating from ship-related SO2 emissions already contributes to a large extent to ecosystem damage and biodiversity losses.
Agricultural activities are a main source of ammonia (NH3) emissions in Europe. In contrast to all other sources of anthropogenic emissions, agriculture is to some extent a borderline case, as emission patterns are on the one hand determined by human activities (e.g., manure handling and management or fertilizer application on agricultural fields), but on the other hand influenced by meteorological and climatic conditions. However, other emissions may be subject to weather influences as well; for example, cold-start emissions from road vehicles and residential combustion emissions are higher in winter, while evaporation losses from gasoline or solvent use may be higher during the summer. The weather may, for instance, affect the times cattle stay in housing or when thawing soils make manure application feasible. In addition to that, temperature plays a major role in many of the biogeochemical processes that lead to emissions of NH3, as well as of NOx and N2O, from agricultural soils.
Natural and Biogenic Sources
So far, only anthropogenic emission sources have been discussed, but natural and biogenic sources emit at times significant amounts of trace substances that can make substantial contributions to air pollution. Biogenic VOCs, for instance, can be quite reactive and lead to the formation of ground-level ozone with anthropogenic NOx emissions (e.g., isoprenes and monoterpenes), and pollen and other natural debris (fungal spores, etc.) can contribute to high concentrations of PM. In almost all cases, emissions from natural/biogenic sources are not influenced by human activities (with the exception of forest fires and biomass burning, where many fires are initiated by people), but instead show distinct seasonal cycles, with high emissions typically during the summer and the growing season.
(See Figure 4)
The basic ingredients for the calculation of emissions are activity rates (e.g., annual mileage of a car, amount of fuel burned in a power plant, or amount of paint applied) and emission factors. Activity rates are often based on statistics or direct observations and can be determined quite accurately, depending on the effort spent. Emission factors are often a function of several process-related parameters. The formation of nitrogen oxides, as an example, in a vehicle engine is determined by the load factor, the fuel-air ratio, the nitrogen content of the fuel, the temperature at which the fuel is burned, and so on.
In an optimal case, emission factors would be based on detailed measurements of all relevant parameters, and the resulting emission functions would then be applied to determine (close to) real-world emissions from all sources. As this is impractical, different approaches are taken to determine emission factors, based on representative measurements or standardized emission profiles.
Large Point Sources
For many large point sources (LPS), legislation requires continuous measurements of exhaust gas streams and detailed reporting of emissions to regulatory bodies. This is the case under the EPER/E-PRTR, where installations subject to regulations by the IPPC directive are obliged to report emissions of a number of substances into air and water to a central database. While this is first and foremost an issue of compliance monitoring with regard to compliance with emission limit values, the information gathered on emissions can be used to derive emission factors. In addition to emission data, information on installed capacities and technologies is typically available and can be used to derive an EF based on, for example, electricity produced or fuel consumed.
A specific aspect of LPS is often the existence of high stacks to disperse emissions widely. The stacks inject emissions at a height of 100 to 300 m above ground, typically with high temperatures, and the resulting plume can rise significantly above the outlet level of the stack. For many large power plants, parameters like stack height and diameter are available in databases used for regulatory reporting of LPS emissions. However, where such data are not readily available, it can be assumed that high-stack emissions are injected at a height that will effectively carry pollutants above any low-inversion layer.
If sufficient data on stack parameters and meteorological conditions are available, the calculation of plume rise is typically based on the work published by Briggs (1965, 1968; see Figure 5). Briggs divided air pollution plumes into four general categories:
• Cold jet plumes in calm ambient air conditions
• Cold jet plumes in windy ambient air conditions
• Hot, buoyant plumes in calm ambient air conditions
• Hot, buoyant plumes in windy ambient air conditions
In contrast to the case with large point sources, in-situ measurements are not a viable option to monitor emissions from mobile sources or machinery. For passenger cars and light-duty vehicles, standardized driving cycles, such as the New European Driving Cycle (NEDC), are used to determine emission factors by vehicle type to determine comparable emission data for type approval. The NEDC consists of four repeated ECE-15 driving cycles and an extra-urban driving cycle, or EUDC. The NEDC is supposed to represent the typical usage of a car in Europe by running a car engine through predefined stages with preset speeds and load factors on a chassis dynamometer. However, with recently emerging evidence that test cycles may only poorly represent real-world driving conditions, on-road emission measurements and measurements on vehicles have been conducted. These reveal, as did previous studies using tunnel measurements or stationary measurements near roads, in some cases substantially higher emissions of nitrogen oxides, in particular from diesel passenger cars. In the EU, a new driving cycle is under development, aiming to be more representative of real-world driving conditions, and it is expected to be established as a standard soon (Figure 6).
Calculation of Emission Inventories
The calculation of emissions can be based on a variety of methods, mainly determined by the processes causing the emissions and the substance emitted. Carbon dioxide emissions, for instance, originate from the carbon content of fossil fuels in most cases and hence are calculated based on the type and amount of fuel consumed (see section “Fuel Consumption”). Processes where the chemical transformation plays a role, as in the emission of nitrogen oxides from internal combustion engines or the evaporation of organic solvents from paints and varnishes, are often determined by a fraction of a substance emitted relative to a certain activity. Here, the combination of activity rate and emission factor is used to calculate the emission (see section “Emission Factors”). In addition, mass balance approaches can be applied, where the content of a specific substance in a system may determine emissions, as is the case with the nitrogen content in animal manure and the resulting potential emissions of NH3 from manure application. While different application techniques may result in different emissions, the overall determining factor is the nitrogen content available in the manure (see section “Inverse Modeling”). Last, but not least, for some emissions, the source strength cannot be measured, and some emission sources are even unknown or are not accounted for in inventories. In these cases, inverse modeling techniques can help to identify the location and amount of emissions, as is described in section “Inverse Modeling”.
The use of fuel consumption to calculate emissions is best suited for substances where emissions have a linear relationship with fuel input and no significant process-influence on the formation of primary or secondary pollutants. The formation of carbon dioxide (CO2) from fuel carbon content, for instance, is one example. In this case, emissions are expressed relative to the mass of fuel input as:
where eCO2 is specific emissions of CO2 per unit of fuel mass, CCfuel is fuel carbon content, and Remitted is rate of carbon emitted.
For most processes, especially the combustion of fuels and industrial production, the amount of a substance released from the process under typical conditions has been well established either by measurements or by the theoretical description of the process and its chemical characteristics. Most emission factors provided in scientific literature and guidelines for the calculation of emissions are based on measurements, but it has to be noted that in most cases, only a comparatively small number of relevant measurements are available and the resulting average emission factors should be regarded as empirical values for certain types of installations, rather than ultimately correct and accurate representations of the emissions of a specific facility or process.
In most cases, emission factors (EFs) are distinguished into ”uncontrolled” and ”controlled,” where uncontrolled refers to a process without the addition of any primary or secondary emission control measures, whereas controlled EFs result from a process with control measures incorporated.
where Epol is emission of a pollutant (pol) resulting from an activity (A) with a certain EF, EFpol is the specific emission factor of a pollutant for this process, and A is the activity rate (e.g., km driven, fuel consumed, electricity produced, fertilizer applied).
This bottom-up approach of emission calculation requires data on both EFs and activity levels, the latter typically being available from energy and production statistics. Emission factors for a variety of processes and pollutants have been thoroughly described in the Emep/Corinair Emission Inventory Guidebook, which is available online and contains a wealth of information for inventory compilation.
A mass balance approach to determining emissions is often used to calculate the release of volatile organic compounds in use of organic solvents (e.g., coating or painting activities). This approach can also be used to quantify emissions of trace substances from combustion processes, such as sulfur dioxide from coal combustion.
With a mass balance approach, the content of a specific substance or trace on the input side and its content on the output side of a process are compared, with the difference between input and output amounts marking the release into the environment. As an example, the solvent content of a coating agent applied in a closed process could be quantified easily, with the amount of solvent recovered through the treatment of the waste air stream providing the output value. The difference between both values would be the emission of the solvent through ventilation or incomplete sealing of the process chamber.
The mass balance approach can generate reliable and cost-effective estimates of average emissions for certain pollutants, such as SO2, NMVOCs, or CO2, and it is an effective technique for estimating emissions from evaporation sources and sources where the measurement of low-level, intermittent emission would be difficult, costly, or uncertain (e.g., for trace elements).
A technique that is often applied to identify unknown emission sources or when information on emissions is difficult to obtain is the inverse application of atmospheric transport models to infer the emission source strength from the resulting concentrations measured. For this purpose, the following information has to be available: measured concentrations of the substance to be modeled (with sufficient temporal resolution) at a point, information about the meteorological conditions for the time to be modeled, and an atmospheric dispersion model capable of calculating trajectories of pollutant transport backward from the point of measurement to its (likely) point of origin.
Inverse modeling is applied in cases where measured concentrations and (forward) atmospheric modeling results cannot be reconciled and a “gap” remains between observed and modeled concentrations, with the aim of determining the location and strength of sources potentially missing in the inventories.
Urban Emission Inventories
The principles described above for the general compilation of emission inventories also apply for urban inventories. Yet, a number of aspects need to be considered for the compilation and use of urban emission inventories:
• Urban areas are marked by a high population density, and a significant amount of national total emissions stem from urban sources
• The proximity of the sources to a large number of people requires high accuracy in the determination of emissions, due to the potential immediate exposure effects on human health
• The aforementioned proximity to emission sources leads to marked temporal patterns of emissions and resulting concentrations, making it necessary to determine emissions not only with a high spatial resolution, but with detailed temporal profiles as well.
This is particularly true for emissions from road sources and the ambient concentrations emerging from road sources in street canyons and curbside. Detailed street-level emissions are most often derived from either continuous automatic measurements of ambient concentrations or sophisticated traffic-flow models, which are based on automatic traffic-counting networks and combining this information with average fleet emission factors. In addition to this, stationary emission sources (small combustion sources, construction and machinery, paint application, small businesses, etc.) contribute to the overall ambient concentration levels in a street. However, in most cases, these are calculated as area sources, and for some individual contributors, potentially as point sources (see below).
While street pollution is markedly determined by vehicles (streets as line sources), urban emissions both contribute to, and are influenced by, background concentrations of pollutants. A city area can be seen as a point source itself, exporting ambient concentrations to the surrounding countryside, while at the same time, ambient concentrations of pollutants in the urban area are a mixture of local sources and background concentrations imported from the countryside. Depending on the location of an urban area, the influence of background relative to the contribution of urban emissions to ambient concentrations can be significant (e.g., from shipping emissions in coastal areas or power plant emissions, agricultural emissions, and other land-based sources). In contrast, urban emissions of nitrogen oxides and non-methane volatile organic compounds often contribute to peak concentrations of ground-level ozone in rural areas some distance downwind from the urban area, due to the mixture of ozone precursors and the time taken for photochemical reactions.
Inventories to Assess Global Air Pollution and Long-Range Transport
A range of emission inventories are compiled, often with quite different motivations and for a variety of purposes. In general, inventories compiled with the aim of driving atmospheric dispersion models can be distinguished from inventories set up for regulatory purposes. For example, the EDGAR global inventory (see section “EDGAR” and Figure 7) has been developed for modeling, and its spatial resolution and level of detail have been gradually improved over time. On the other hand, the EMEP/CORINAIR inventory for air pollutants and the IPCC/UNFCCC greenhouse gas emission inventory are primarily compiled based on official national data submissions and serve to monitor compliance with international protocols regulating emission ceilings and reduction targets for specific substances (see sections “EMEP/CORINAIR and “IPCC/UNFCCC”). In recent years, intensive harmonization efforts have been made to adapt the structure of the EMEP/CORINAIR inventory for air pollutants to be similar to that of the IPCC/UNFCCC.
The EMEP/CORINAIR inventory originates from the international protocols signed by countries in the frame of the UNECE CLRTAP. Within CLRTAP, specific protocols for individual substances (sulfur, nitrogen oxides, heavy metals, POPs) as well as more integrated agreements, such as the Gothenburg Protocol (designed as a multipollutant, multieffect protocol), have been developed and ratified by individual countries in the UNECE region. In order to monitor compliance with the emission reductions (or stabilization) relative to a target year, countries that have ratified a protocol are bound by international law to submit annual emission datasets to document either progress to a target or compliance with targets in a target year.
Emissions are reported to the Centre on Emission Inventories and Projections (CEIP), and after quality checks and analyses have been conducted, can be accessed by the general public at the CEIP website. The focus of the inventory is on anthropogenic emissions of air pollutants.
In addition to the emission inventory work, the CLRTAP is supported by a comprehensive monitoring program and modeling activities, which provide the scientific underpinning and modeling capacities to assess the environmental and health impacts of changes in emissions and the resulting changes in air quality and atmospheric depositions.
Similar to the EMEP/CORINAIR inventory, within the United Nations Framework Convention on Climate Change (UNFCCC), data on greenhouse gas emissions (GHGs) are collected based on submissions of countries that have ratified the convention. Again, this inventory is primarily designed for monitoring compliance, with a focus on anthropogenic sources of GHG emissions (dominated by fossil fuel combustion). In addition to that, land use and land-use change and forestry (in short, LULUCF) feature strongly in the UNFCCC inventories due to the objective of accounting for net emissions of GHGs, taking into account carbon sources as well as sinks. In this context, emissions from agricultural and seminatural sources (e.g., agricultural soils and livestock, grasslands, and wetlands, etc.) are calculated and reported, based on the scientific methodologies documented in the IPCC methodology reports.
Greenhouse gas datasets reported to UNFCCC are publicly accessible on the convention website as well, with substantial supplementary material on the scientific background of emissions, sources, and sinks available from the IPCC, especially the different assessment reports.
The EDGAR (Emissions Database for Global Atmospheric Research) inventory is—unlike the two inventories just discussed—an independent scientific endeavor that has a long track record of providing global datasets of atmospheric emissions for modeling. EDGAR is based upon bottom-up calculations of emissions for all relevant anthropogenic and biogenic/natural sources and its current release (version 4) covers direct GHGs, ozone precursor gases, acidifying gases, primary particles, primary aerosols, ozone-depleting substances, and mercury.
Apart from very detailed emission tables for a wide range of source sectors and countries, emission data can be downloaded and readily mapped on a 10 × 10 km grid from the EDGAR website after registration.
Other Inventories (GEIA, ACCENT, VULCAN)
In addition to the inventories discussed above, inventories are available for a variety of spatial scales and purposes. The Global Emission Inventory Activity (GEIA), for instance, is an integrating project of the IGAC/International Geosphere-Biosphere Programme and compiles links to emission inventory datasets on a global scale. GEIA has been working in close collaboration with the European network of excellence ACCENT, which has in turn contributed a substantial amount of underpinning science and networking to make emission data and related resources accessible to the scientific community.
Furthermore, individual projects or activities generate datasets that, in the best case, can be accessed and used by the general public for modeling studies and other research. One example of such work is the VULCAN project, a NASA/DOE-funded effort under the North American Carbon Program (NACP) to quantify North American fossil-fuel carbon dioxide (CO2) emissions at very high spatial and temporal resolution.
As most emission inventories have originated from the need to either control and/or monitor emissions of potentially harmful substances into the environment, different concepts for the aggregation of substances exist, reflecting the specific targets or objectives for their control.
Air Pollutants and Precursor Substances
Air pollutants are monitored primarily because of their harmful effects on human health and their effects on agricultural crops and seminatural and natural ecosystems. In this context, the following aggregation approaches can be distinguished:
▪ Ozone-forming potential (ozone precursors): NOx, NMVOC (CH4)
▪ Acidification potential (acidifying substances): SO2, NOx, NH3
▪ Eutrophication potential (eutrophying substances): NOx, NH3
▪ Precursors for secondary aerosol formation: SO2, NOx, NH3
▪ Heavy metals: As, Cd, Cr, Cu, Hg, Ni, Pb, Se, Zn
▪ Persistent organic pollutants (POPs): dioxins, PAH, PCB, PCP, PER, etc.
The ozone-forming potential (OFP) of emissions heavily depends on the relative abundance of NOx and NMVOCs in the presence of sunlight, as well as on the individual composition of the NMVOC emissions. Different types of NMVOCs can have significantly different OFPs (a different concept is termed photochemical ozone creation potential, POCP). Taking into account how ozone formation can be limited due to the availability of either NOx or NMVOCs, “NOx-limited” or “NMVOC-limited” regions can be determined. This knowledge is relevant for the design of specific regional control strategies to reduce concentrations of ground-level ozone.
The acidification potential is given in sulfur dioxide equivalents (SO2-Eq.). The acidification potential is described as the ability of a certain substance to form and release H+ ions. The reference substance is SO2. Similarly, the eutrophication potential is calculated in phosphate equivalents (PO4-Eq.). The calculation of these equivalents serves to make contributions of chemically different precursors comparable and to account for different chemical characteristics of, for example, reduced and oxidized forms of nitrogen being deposited.
In the case of greenhouse gas emissions, the different contributions to the change of global radiative forcing are used to derive factors to calculate equivalents for all greenhouse gases. The so-called global warming potential (GWP, see UNFCCC for an introduction to how it is measured and expressed) uses CO2 with a reference value 1, and the GWP of other GHGs is expressed as CO2 equivalents (CO2-Eq.). This is particularly important because, for the stabilization of GHG emissions, the efforts required to control CO2 vs. CH4 might be quite different, while a relatively modest reduction of CH4 emissions could have a significantly larger benefit to achieve climate targets than a large reduction of CO2 emissions. The use of the CO2-Eq. makes it possible to directly compare emissions of different trace gases contributing to global warming with varying GWP.
Apart from the concepts discussed above, other concepts exist with specific environmental objectives as driving forces. The ozone depletion potential (ODP) of a chemical compound, for instance, is the relative amount of degradation to the ozone layer it can cause. The substance trichlorofluoromethane (R-11, resp. CFC-11) has been set as a reference, at an ODP of 1.0.
Further to that, concepts of toxicity (both to humans or animals and to ecosystems) provide reference systems for the relative comparison of substances. Finally, there are substances where even a low-level exposure is regarded as potentially harmful (with no known existing thresholds), as is the case for endocrine disrupting chemicals/compounds (EDC) or cancerogenic substances (carcinogens).
Modeling Air Pollution
Modeling of air pollution using atmospheric models has emerged as a cost-effective means to complement existing monitoring approaches over time. With increases in computing power and improved understanding of atmospheric composition and processes, current air-pollution models can fairly accurately reproduce past air-pollutant concentrations (validated by comparisons with observations). A key advantage of these models is that models can be applied to project future levels of air pollutant concentrations. This requires that key input parameters such as emissions and meteorological driver can be robustly estimated, or approximated using representative current conditions. This is a vital capability for the ex-ante evaluation of air- pollution control measures and policies.
Different Modeling Concepts
Current atmospheric models simulate the physical and chemical processes in the atmosphere, including the chemical transformation and short- or long-range transport of air pollutants. They most often follow one of two key approaches:
• Lagrangian models follow an air parcel over time and calculate both the position and properties of the parcel according to the wind field. The path taken by the air parcel is referred to as its trajectory and is represented by the differential equation: ΔX/Δt = A [X(t)], where t = time, X = position vector, and A = wind speed vector. A detailed description of the development and application of Lagrangian models can be found in Lin et al. (2013).
• Eulerian models, in contrast, define reference points in a gridded system to represent atmospheric properties (temperature, pressure, chemical concentrations of tracer substances, etc.) over time. The simulated world is thus divided into sets of fixed grid cells, both horizontally and vertically. In the horizontal, the cells vary in size from several meters for urban models to hundreds or thousands of kilometers for regional or global models. On the vertical, different models have varying resolution of the troposphere up to the planetary boundary layer, and are typically more highly resolved near the surface, representing surface conditions to account for surface roughness, turbulence, and wet and dry deposition of pollutants. One example of a Eulerian model that is widely applied in Europe is the EMEP Unified Model; a similar model developed by the EPA is the Community Air Quality Model (CMAQ).
Both modeling concepts have advantages and disadvantages with regard to computational requirements and the level of detail they represent, but neither modeling concept is inherently superior. In recent years, a trend toward more complex Eulerian models has been noticeable, as computational limitations were eroded by the increase in computer power and memory size. This allowed for the development and application of advanced modeling systems, incorporating the state of the art of understanding of chemical reactions and atmospheric processes. Most recently, air quality and climate change processes are beginning to be investigated jointly with coupled models, as in the United Kingdom Chemistry and Aerosols (UKCA) model. This integrated modeling approach will in the future allow for more detailed investigations of interactions and effects of climate change on atmospheric composition, and vice versa, while previously, atmospheric models were typically run with precomputed meteorological input data.
A specific application of air-pollution models is the use of computational fluid dynamics (CFD) concepts to generate very high resolution models of urban atmospheric processes. Such CFD-based models are typically used to analyze air-pollution hotspots, such as street canyons, where the complex terrain in built-up urban areas, including the wake effects of passing vehicles as well as urban vegetation and other factors, are all taken into account. For example, models like the widely applied MISKAM model are often used for planning purposes and to investigate the effectiveness of local development or traffic-control measures on air pollution hotspots.
Applications of Air-Pollution Models
Providing a comprehensive overview of all application cases for air-pollution models is not possible here. However, overall air-pollution modeling can be categorized by:
• spatial resolution—global, regional, local/urban scale
• model concept—Lagrangian, Eulerian, or CFD
• model type—simple dispersion models or complex chemistry-transport models
• application—ex-post assessment or projection/forecasting of pollution events
Air-pollution models are widely used by national and local authorities to assess compliance with air-quality limit values (ex-post), as well as to quantify the impact of possible future developments (e.g., for regulatory purposes for ex-ante environmental impact assessments). In this context, models are applied to complement existing measurements or to quantify concentrations or depositions in areas where no monitoring sites exist (in this case, more complex models can provide more accurate representations of complex concentration or deposition fields than interpolation techniques).
Figures 8 and 9 provide examples of the analysis of air pollution using atmospheric chemistry transport models and serve to illustrate the abundance of applications of air-pollution models at different scales and for different purposes.
Integrated Assessment Models
While not specifically designed for application to air pollution, integrated assessment models have been instrumental in designing policy strategies to tackle major air-pollution challenges. As an example, the Regional Air Pollution Information and Simulation (RAINS) model, developed by the International Institute for Applied Systems Analysis (IIASA), has been applied in the past to address the transboundary nature of air pollution from emissions of sulfur dioxide. This integrated assessment model took into account emissions, control measures, and parameterized output from atmospheric models to identify optimal emission-control strategies to reduce the adverse effects of acidification of European forests and freshwater ecosystems. The model has since been further developed to incorporate greenhouse gases, with the new version termed the Greenhouse Gas and Air Pollution Interactions and Synergies (GAINS) model.
Such integrated assessment models (IAMs in short) play a vital role in providing robust and science-based policy support, for instance in the development and, more recently, revision of the Multi-Effect Protocol under UNECE CLRTAP and the European Commission’s National Emission Ceilings Directive (NECD).
Data Quality and Big Data Issues
As briefly discussed above under the section “Emerging New Technologies,” data quality—in particular with regard to emerging medium- and low-cost sensors applied in citizen science contexts—is a challenge for air-quality data in general. The current paradigms for big data appear to be based on a situation where an abundance of data sources enables advanced data analysis techniques to identify and discard outliers, with the power of large datasets being in their size. Even with low-cost sensors, there is currently no such abundance of air-quality measurement data available.
In addition, as air quality can be a highly political topic, releasing air-quality measurement data into the public domain is not always straightforward, and local authorities and regulators are investing a lot of effort into ensuring the high quality of data before they are released.
Model-Data Fusion and Sensor-Model Integration
A further challenge, identified by Reis et al. (2015), lies in the methodological and conceptual complexities of comparing and integrating datasets from inherently different sources. Where monitoring sites typically provide point measurements at a defined location and height and with specific parameters, air-pollution models may deliver a grid average. Like-for-like comparisons of these values reveal that local conditions and spatiotemporal variability of the pollutant concentration field will introduce uncertainties that need to be addressed. Including satellite remote sensing data, where the concentration reflects the whole atmospheric column, presents challenges for direct comparisons with ground-based or aircraft measurements and model results with distinct vertical profiles. Nevertheless, the strength in combining data from ground-, aircraft-, and space-based monitoring platforms with numerical models at different spatial and temporal resolutions lies in their ability to identify and—to an extent—correct for the gaps and shortcomings of each method on its own.
Integrated Systems for Air Pollution Management
Ideally, future air-pollution monitoring and modeling systems will be fully integrated, making use of emerging big data methods and cloud-based computing to allow for (near) real-time processing and integration of data from an array of different data sources. The added value of such integrated systems for providing robust scientific evidence to policy decision makers will likely result in more salient, strategic air pollution management decisions. The role of such integrated systems in supporting communication at the science–policy interface and in contributing to addressing complex synergies and unintended consequences of air-pollution control strategies is described in Reis et al. (2012).
Baumbach, G. (1996). Air quality control. Berlin: Springer Publishers.Find this resource:
Bohnenstengel, S. I., Belcher, S. E., Aiken, A., Allan, J. D., Allen, G., Bacak, A., … Zotter, P. (2014). Meteorology, air quality, and health in London: The ClearfLo Project. Bulletin of the American Meteorological Society, 96, 779–804.Find this resource:
Bell, M. L., & Davis, D. L. (2001). Reassessment of the lethal London fog of 1952: Novel indicators of acute and chronic consequences of acute exposure to air pollution. Environmental Health Perspectives, 109(Suppl. 3), 389–394.Find this resource:
Briggs, G. A. (1965). A plume rise model compared with observations. JAPCA, 15, 433–438.Find this resource:
Briggs, G. A. (1968). CONCAWE meeting: Discussion of the comparative consequences of different plume rise formulas. Atmospheric Environment, 2, 228–232.Find this resource:
Butler T. M., Lawrence, M. G., Gurjar, B. R., van Aardenne, J., Schultz, M., & Lelieveld, J. (2008). The representation of emissions from megacities in global emission inventories. Atmospheric Environment, 42(4), 703–719.Find this resource:
Emeis, S. (2010). Measurement methods in atmospheric sciences—In situ and remote. Stuttgart: Schweizerbart Publishers.Find this resource:
Finlayson-Pitts, B. J., & Pitts, J. N. (1999). Chemistry of the upper and lower atmosphere: Theory, experiments, and applications. Amsterdam, The Netherlands: Elsevier Publishers, Academic Press.Find this resource:
Foerstner, U. (1993). Umweltschutztechnik (4th ed., in German). Berlin: Springer Publishers.Find this resource:
Friedrich, R., & Reis, S. (Eds.) (2004). Emissions of air pollutants—Measurements, calculation, uncertainties—Results from the EUROTRAC Subproject GENEMIS. Berlin: Springer Publishers.Find this resource:
Heard, D. (Ed.) (2006). Analytical techniques for atmospheric measurement. New Jersey: Wiley-Blackwell.Find this resource:
Hoor, P., Borken-Kleefeld, J., Caro, D., Dessens, O., Endresen, O., Gauss, M., … Sausen, R. (2009). The impact of traffic emissions on atmospheric ozone and OH: Results from QUANTIFY. Atmospheric Chemistry and Physics, 9, 3113–3136.Find this resource:
Krahl-Urban, B., Papke, H. E., & Peters, K. (1988). Forest decline: Cause-effect research in the United States of North America and Federal Republic of Germany. Germany: Assessment Group for Biology, Ecology and Energy of the Julich Nuclear Research Center.Find this resource:
Lim, S. S., Vos, T., Flaxman, A. D., Danaei, G., Shibuya, K., Adair-Rohani, H., . . . Ezzati, M. (2012). A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: A systematic analysis for the Global Burden of Disease Study 2010. The Lancet, 380(9859), 2224–2260.Find this resource:
Lin, J., Brunner, D., Gerbig, C., Stohl, A., Luhar, A., & Webley, P. (Eds.) (2013). Lagrangian modeling of the atmosphere. New Jersey: Wiley Publishers.Find this resource:
Lioy, P. J., & Georgopoulos, P. G. (2011). New Jersey: A case study of the reduction in urban and suburban air pollution from the 1950s to 2010. Environmental Health Perspectives 119(10), 1351–1355.Find this resource:
Olivier, J. G. J., & Berdowski, J. J. M. (2001). Global emissions sources and sinks. In J. Berdowski, R. Guicherit, & B. J. Heij (Eds.), The climate system (pp. 33–78). Lisse, The Netherlands: A.A. Balkema Publishers/Swets & Zeitlinger Publishers.Find this resource:
Olivier, J. G. J., Berdowski, J. J. M., Peters, J.A.H.W., Bakker, J., en Visschedijk, A. J. H., & Bloos, J. P. J. (2001). Applications of EDGAR. Including a description of EDGAR 3.0: Reference database with trend data for 1970–1995. RIVM Report no. 773301 001/NOP Report no. 410200 051. Bilthoven: RIVM.Find this resource:
Olivier, J. G. J., Bouwman, A. F., van der Maas, C. W. M., Berdowski, J. J. M., Veldt, C., Bloos, J. P. J., … Haverlag, J. L. (1996). Description of EDGAR Version 2.0: A set of global emission inventories of greenhouse gases and ozone-depleting substances for all anthropogenic and most natural sources on a per country basis and on 1 degree x 1 degree grid. RIVM Rapport 771060002. Bilthoven: RIVM.Find this resource:
Olivier, J. G. J., Pulles, T., & Van Aardenne, J. A. (2008). Part III: Greenhouse gas emissions: 1. Shares and trends in greenhouse gas emissions; 2. Sources and methods; Total greenhouse gas emissions. In: CO2 emissions from fuel combustion 1971–2006 (pp. III.1–III.47). Paris: International Energy Agency (IEA). (Also available on CD ROM.)Find this resource:
Olivier, J. G. J., Van Aardenne, J. A., Dentener, F., Pagliari, V., Ganzeveld, L. N., & Peters, J.A.H.W. (2005). Recent trends in global greenhouse gas emissions: Regional trends 1970–2000 and spatial distribution of key sources in 2000. Environmental Sciences, 2(2–3), 81–99.Find this resource:
Power, H., & Baldasano, J. M. (1998). Air pollution emission inventory. Southampton: WIT Press.Find this resource:
Reis, S., Grennfelt, P., Klimont, Z., Amann, M., ApSimon, H., Hettelingh, J-P., … Williams, M. (2012). From acid rain to climate change. Science 338(6111), 1153–1154.Find this resource:
Reis, S., Seto, E., Northcross, A., Quinn, N. W., Convertino, M., Jones, R. L., …Wimberly, M. C. (2015). Integrating modelling and smart sensors for environmental and human health. Environmental Modelling and Software, 74.Find this resource:
Seinfeld, J. H., & Pandis, S. N. (2006). Atmospheric chemistry and physics: From air pollution to climate change (2nd ed.). New Jersey: Wiley-Blackwell Publishers.Find this resource:
Seinfeld, J. H., & Pandis, S. N. (2016). Atmospheric chemistry and physics: From air pollution to climate change (2nd ed.). New Jersey: Wiley-Blackwell Publishers.Find this resource:
Sutton, M. A., Reis, S., Baker, S. M. H. (Eds.) (2009). Atmospheric ammonia—Detecting emission changes and environmental impacts—Results of an expert workshop under the Convention on Long-range Transboundary Air Pollution. Berlin: Springer Publishers.Find this resource:
Sutton, M. A., Reis, S., Riddick, S. N., Dragosits, U., Nemitz, N., Theobald, M. R., … de Vries, W. (2013). Toward a climate-dependent paradigm of ammonia emission & deposition. Philosophical Transactions of the Royal Society (Series B), 368(1621)Find this resource:
Willocks, L., Bhaskar, A., Ramsay, C., Lee, D., Brewster, D., Fischbacher, C., … Scott, E. M. (2012). Cardiovascular disease and air pollution in Scotland: No association or insufficient data and study design? BMC Public Health, 12, 227.Find this resource:
Van Aardenne, J. A., Dentener, F. J., Olivier, J. G. J., Klein Goldewijk, C. G. M., & Lelieveld, J. (2001). A high resolution dataset of historical anthropogenic trace gas emissions for the period 1890–1990. Global Biogeochemical Cycles, 15(4), 909–928.Find this resource: