Guidance for Quantifying the Contribution of Airport Emissions to Local Air Quality (2012)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

8Current State of Airport Air Quality Assessments and Considerations 2.1 Introduction This chapter serves as a starting point for better understanding the guidance and findings presented in later chapters of this report. Overviews of the various air quality regulations, guid- ance, and modeling and measurement capabilities, as well as a survey of recent airport air quality studies, are provided. This background material is intended to summarize the current state of knowledge and capabilities. Although presented in relatively simple terms, this section is not intended to serve as a primer because the reader needs to already have a basic understanding of air quality and airport concepts. Various USEPA and FAA guidance documents exist related to air quality in general or to the combination of air quality and airports. Although these documents provide a good under- standing on how to comply with air quality regulations (e.g., what assessments to conduct), they generally have gaps related to their application to airports and to comprehensively under- standing airport contributions to local air quality (e.g., comprehensive coverage of most or all pollutants). These documents are referenced throughout this report, rather than repeating their contents. The desire to answer the question of when to conduct modeling instead of measurements and vice versa serves as the impetus to better understand the associated capabilities. Although significant advances have been made in recent years to both emissions and atmospheric disper- sion modeling capabilities, there are still significant limitations such that measurements are necessary to more comprehensively understand airport air quality contributions. As stated by the USEPA, air quality measurements are generally conducted for one or more of the following purposes (USEPAj 2011): • To judge compliance with and/or progress made toward meeting ambient air quality standards; • To activate emergency control procedures that prevent or alleviate air pollution episodes; • To observe pollution trends throughout the region, including non-urban areas; and • To provide a database for research evaluation of effects: urban, land-use, and transportation planning; development and evaluation of abatement strategies; and development and valida- tion of diffusion models. It is generally agreed that air quality field measurements are more accurate than modeled results for most cases. As such, notwithstanding errors in measurement techniques including calibration mistakes, measurement data are usually considered the “gold standard.” However, current practices at airports in the United States revolve around modeling much more than measurements. The aforementioned FAA guidance documents mainly provide directions on conducting modeling work and virtually no information on conducting measurements. The modeling work is also focused on emissions modeling. C h a p t e r 2

Current State of airport air Quality assessments and Considerations 9 2.2 Existing Regulatory Framework The applicable regulations begin with the National Environmental Policy Act (NEPA) (Pub. L. 1982 Latest). This 4-page document and its amendments establish a broad national policy to protect the quality of the human environment and provide for the establishment of a Council on Environmental Quality (CEQ) (WH 2011). NEPA requires a detailed statement by all agen- cies of the Federal Government when conducting a major federal action. This statement is to include descriptions of • The environmental impacts of the proposed action; • Any unavoidable adverse environmental impacts; • Alternatives, including no action; • The relationship between short-term uses of the environment and maintenance of long-term ecological productivity • Irreversible and irretrievable commitments of resources; and • Secondary/cumulative effects of implementing the proposed action. In 1963, the original Clean Air Act (CAA) was passed but did not have “teeth.” The CAA was amended in 1967 and provided authority to establish air quality standards. Subsequent amendments have led to the Clean Air Act Amendments of 1990 (CAAA) which is now in effect (USC Title 42, Chapter 85). The established National Ambient Air Quality Standards (NAAQS) include specified criteria pollutants consisting of ozone, carbon monoxide, partic- ulates, sulfur dioxide, nitrogen dioxide, and lead. Table 2-1 shows the established standards. Some states, as well as local jurisdictions, have additional state requirements pertaining to local air quality concentrations. Although no ambient standards yet exist for HAPs, HAPs have increasingly become issues of public concern because of the potential for health impacts. As part of this concern, Pb from piston-powered GA aircraft have in recent years gained significant attention because Pb is known for neurological damage among other concerns. Modeling can account for the emissions of most HAPs, but dispersion modeling is limited to assumptions of non-reactivity (i.e., no formation of secondary pollutants). sdradnatS yradnoceS sdradnatS yramirP Pollutant Level Averaging Time Level Averaging Time Carbon Monoxide (CO) 9 ppm 8-hour None 35 ppm 1-hour Lead (Pb) 0.15 µg/m3 Rolling 3-Month Average Same as Primary Nitrogen Dioxide (NO2) enoN ruoh-1 bpp 001 53 ppb Annual (Arithmetic Average) Same as Primary Particulate Matter (PM10) 150 ppb 24-hour Same as Primary Particulate Matter (PM2.5) 15.0 µg/m 3 Annual (Arithmetic Average) Same as Primary 35 µg/m3 24-hour Same as Primary Ozone (O3) 0.075 ppm 8-hour Same as Primary Sulfur Dioxide (SO2) 75 ppb 1-hour 0.5 ppm 3-hour Source: USEPAa 2010. Table 2-1. National ambient air quality standards.

10 Guidance for Quantifying the Contribution of airport emissions to Local air Quality UFPs emitted from jet engines have also gained significant attention because their size (<0.1 µm) allows them to penetrate deep into lungs, causing various respiratory problems. Further studies are needed to continue to characterize these emissions and to understand the actual health impacts. In a larger context, measurements for a defined geographical area are used to determine if the area is in compliance with the NAAQS. If not, these areas are designated as nonattainment areas (NAAs) and the state must develop a State Implementation Plan (SIP) to ensure attainment of the standards in the future. Measurements are then used to determine if the area has come into compliance. Areas previously designated nonattainment pursuant to the Clean Air Act Amend- ments of 1990 (CAAA90) that come into compliance are subsequently re-designated to attain- ment and termed “maintenance” areas. A “maintenance” plan, or revision to the applicable SIP is required (USC Title 42, Chapter 85). Of direct importance to airports in NAAs is the General Conformity Rule. This is a strat- egy to achieve and maintain (conform) to the NAAQS required in Section 176(c)(1) of the CAAA. Conformity statements are used to ensure that local concentrations are not exacerbated by new government projects. Measurements may be required during major actions at airports to provide baselines and show improvements. Although airports fall under General Conformity, Transportation Conformity may also apply if funding includes any highway or transit project that receives funding assistance and approval through the Federal-Aid Highway program or the Federal mass transit program. These actions could trigger a need for measurements. These conformity determinations can be accomplished through emissions modeling (USC Title 42, Chapter 85, Section 176(c)(4)). During the preparation of environmental statements, DOT/FAA-Specific DOT Order 5610.1C (Procedures for Considering Environmental Impacts) must be followed. It provides instructions for implementing NEPA, with application to DOT programs, including FAA actions. The order also provides instructions for implementing other environmental laws and executive orders. Again, measurement guidance is not included in these documents. The amended Airport and Airway Improvement Act of 1982 provides that a grant application for an airport project will not be approved unless the State certifies that there is reasonable assur- ance that the project will be in compliance with applicable air quality standards (49 USC 2215). FAA Order 5050.4 (The Airport Environmental Handbook) provides guidance for FAA procedures for processing environmental assessments, findings of no significant impact, and environmental impact statements for airport development proposals and other airport actions as required by various laws and regulations. Unfortunately, measurement of local area concentrations is not included. FAA Order 1050.1, Policies and Procedures for Con- sidering Environmental Impacts, documents procedures in compliance with NEPA require- ments, DOT Order 5610.1C, and other environment-related statutes, although air quality is included in this document. Other documents are available for Department of Defense (DoD) airbases. DoD Directive 6050.1 (Environmental Effects in the United States of DoD Actions), U.S. Air Force Policy Directive (AFPD) 32-70 (Environmental Quality), U.S. Air Force Instruction (AFI) 32-7061 (Environmental Impact Analysis Process or EIAP), and the Environmental Impact Analysis Process (Desk Reference) are available and provide air quality guidance to comply with NEPA and other regulations. In addition to federal requirements, there often are state and/or local air quality require- ments applicable to airports. These requirements differ from location to location and must be addressed on a project-by-project basis.

Current State of airport air Quality assessments and Considerations 11 In summary, many of the environmental laws and regulations apply directly to local air quality and must be considered by airports. Establishment of background concentrations and the existing environment are required. As air quality models are limited, monitoring would provide accurate assessments of these concentrations at points and extrapolation could be used to determine local area concentrations. But, the guidance material is almost devoid of any guidance on local air quality monitoring. 2.3 State of Guidance from FAA and USEPA The FAA provides considerable guidance on meeting air quality requirements at airports (e.g., Draper 1997). Many airport studies only require an emissions inventory. However, when emis- sions are found to increase due to some project initiation, it may be necessary to determine local area concentrations either through modeling or measurements (e.g., when additional emissions exceed the General Conformity de minimis levels and there are no other alternatives). Although the guidance does provide some information dealing with air quality issues (e.g., background concentrations), there is very little in regard to actual dispersion modeling and no information regarding measurements. The USEPA provides a dedicated website (USEPAg 2011) that provides both regulatory and modeling information for aircraft. The modeling information is largely based on the use of the FAA’s EDMS and the USEPA’s NONROAD model. Some sources of emissions data and methods for developing and evaluating emissions inventories are also provided. Virtually no direct information on measurements or atmospheric dispersion modeling is provided. The USEPA’s Support Center for Regulatory Atmospheric Modeling (SCRAM) site (USEPAd 2011) provides a wealth of information on air quality models and dispersion modeling. In addition to downloadable models, the site provides a comprehensive collection of model user’s guides, model formulation documents, and model assessment (e.g., validation reports). Supporting guidance documents on the site include those for permit modeling and SIP attain- ment demonstrations. Referenced on the site is Appendix W to 40 Code of Federal Regulations (CFR) Part 51 which provides guidance on air quality models, including modeling techniques and data requirements (CFR 2005). Although these resources exist, any air quality model- ing work should conform to the protocols set forth by the appropriate agency (e.g., state air resources board). Although not specific for airports or the vicinity of airports, the USEPA also offers sig- nificant information related to ambient monitoring. A good beginning point for this infor- mation is the Ambient Monitoring Technology Information Center (AMTIC) site, which is a part of the Technology Transfer Network (TTN) (USEPAa,e,f,h 2011). AMTIC is perti- nent to the exchange of ambient monitoring-related information and contains all Federal Regulations pertaining to ambient monitoring, as well as ambient monitoring-QA/QC- related information. Information on ambient monitoring-related publications, ambient monitoring news, field and laboratory studies of interest, and available related training is also included. Although these monitoring and modeling resources are available, an ideal combination of airport guidance with measurements and air quality (concentration) modeling for airports is lacking. That is, although the EPA materials cover both measurements and modeling, they are not specific to airports. As a result, methods used for stationary sources have to be extrapolated to account for the diverse and mobile sources found at an airport.

12 Guidance for Quantifying the Contribution of airport emissions to Local air Quality 2.4 Measurement Capabilities and Limitations The USEPA promulgates various Federal Reference Methods (FRMs) that specify the meth- ods to monitor certain pollutants. Adherence to these standards is crucial for use of the mea- sured data in any regulatory-type study. The FRMs encompass both real-time and sampling methods. For example, continuous analyzers are specified for certain criteria gases (e.g., CO and NOx) while filter-based sampling methods have typically been specified for particulate matter (PM). These methods have been employed as part of the USEPA’s various monitoring networks, including the Air Quality System (AQS), Chemical Speciation Network (CSN), and the Inter- agency Monitoring of Protected Visual Environments (IMPROVE). Although such methods have been commonly used, new measurement techniques and equip- ment are being developed to provide continuous measurements of various volatile organic com- pound (VOC) species and PM, including Tapered Element Oscillating Microbalances (TEOMs), aethalometers, rotating drum impactors (RDIs), and the first Federal Equivalent Method (FEM) for PM2.5—Beta Attenuation Monitors (BAM). Continuous measurement equipment is already used in some of the USEPA’s monitoring networks and is considered necessary to help improve the assessment of health impacts (USEPA 2008). Even with these types of improvements, significant limitations still exist (e.g., the inability to practically measure a comprehensive set of individual VOC species in real time). Usually, dedi- cated monitors can only measure THCs and non-methane HCs or some individual HCs, such as formaldehyde (HCHO). Similarly, dedicated equipment needs to be employed to monitor the various PM components (e.g., black carbon and nitrates). Also, there are no methods to com- prehensively monitor individual metals in real time. These have to be measured from samples using techniques such as x-ray fluorescence (XRF). Although costs of such monitors have decreased, they are still relatively expensive and need infrastructure to support their use. They need reliable sources of power as well as shelter to pro- tect them from the weather (they usually need a climate-controlled shelter to properly operate). Although portable solutions are possible, they are usually employed in a stationary environment in part to help reduce the potential for damages to the equipment. Human resource require- ments to use and maintain the equipment also need to be considered, as well as supplemental materials (e.g., calibration gases). As a result, sampling is performed to overcome the limitations and costs of continuous moni- tors. VOC and PM samples are usually submitted to USEPA-certified laboratories. Specific pro- tocols for collecting, handling, and shipping of the samples should be followed for each sampling and analysis method specified by the USEPA (e.g., USEPAe 2011). Depending on what monitoring equipment and methods are employed, sometimes a trad- eoff is necessary between spatial and temporal coverage. The resources available (e.g., funding, equipment, and personnel) and the requirements of the project will dictate where to place moni- tors and how to define an appropriate measurement schedule. Given that there are no specific guidance materials for monitoring at airports, existing USEPA protocols for their ambient mon- itoring networks can be followed (USEPAf,h 2011). However, due to resources and depending on the potential use of the collected data, deviations from such protocols have been common with past airport air quality measurement efforts. 2.5 Modeling Capabilities and Limitations For both airport emissions and dispersion modeling, the FAA promotes the use of the Emissions and Dispersion Modeling System (EDMS) (FAA 2007). Currently at Version 5.1.3 and expected to be incorporated into the FAA’s Aviation Emissions Design Tool (AEDT) (FAAa 2011), EDMS is

Current State of airport air Quality assessments and Considerations 13 the FA-required model for all aviation sources and is listed among the USEPA’s “preferred” guide- line models. EDMS uses Eurocontrol’s Base of Aircraft Data (BADA) for aircraft performance modeling and the Boeing Fuel Flow Method 2 (BFFM2) for emissions modeling (Eurocontrol 2011 and DuBois 2006). Atmospheric dispersion modeling is conducted through the use of the USEPA’s American Meteorological Society/Environmental Protection Agency Regulatory Model (AERMOD) which has been integrated into EDMS (USEPAa,b 2004). EDMS development was started in the mid-1980s as a complex source microcomputer model by the FAA and U.S. Air Force as a simpler replacement to the main frame model called the Airport Vicinity Air Pollution (AVAP) model (FAA 2007). Both were designed to assess the air quality impacts of proposed airport development projects, but AVAP was difficult to use and required main frame computer access. EDMS operates in the Windows operating regime and on personal computers, which has led to much more widespread use in the United States as well as abroad. EDMS provides a graphical user interface (GUI) that allows the user to model various cases within a single study. The user may calculate an emission inventory based on defined emission index/factor databases (EI/EF) using screen input. The EI/EFs in EDMS include a large array of aircraft engines by operational mode, APUs, GSE, motor vehicles using MOBILE6.1 (reference), a partial set of factors, and stationary sources. The informational database for these EI/EFs may be reviewed. After sufficient information has been supplied to allow the successful completion of an emission inventory, the user may, upon entry of additional data such as meteorology, run a dispersion analysis. Earlier versions of EDMS ran the Point/Area/Line (PAL) (reference) model for most sources and CALINE3 (reference) for motor vehicles. Since Version 4.0, EDMS has used the EPA-preferred dispersion model, AERMOD (USEPAa,b 2004). Since 1998, the FAA has made it a policy to “require” the use of EDMS to perform air quality analyses for aviation sources instead of just being the “preferred” model. Given that EDMS is specifically designed for airports, it facilitates the modeling of airport sources. It provides various default operating times for most sources, as well as emission factors for virtually all airport sources. In addition, the representation of sources is already defined in EDMS so that the user does not need to worry about such details. Although this can be limiting to some expert users, it promotes consistency and simplifies the modeling of airports sources. For example, the rectangular shape of taxiways and runways is automatically determined by EDMS for dispersion modeling—the user only needs to specify the locations of the end points. EDMS’ GUI also simplifies the use of the USEPA models, MOBILE6.2 and AERMOD. Without the interface, each model would need to be run in the legacy command prompt (character- mode) interfaces such as the Disk Operating System (DOS). Although most studies (especially relatively simple studies) can be conducted in EDMS without any “special” work, many expert users have used EDMS components and other tools externally to improve the overall modeling work. For example, the use of MOBILE6.2 outside of EDMS has been frequently conducted since it allows more control over the development of GAV-related emissions inventories. Other possibilities include the use of AERMOD outside of EDMS for simi- lar purposes as well as various other emissions (e.g., NONROAD) and dispersion models (e.g., CAL3QHC). This robustness is possible because both emissions and atmospheric concentrations are mathematically additive—component pieces can be added and subtracted as necessary. In recent years, significant improvements have been implemented in EDMS, including both higher fidelity methods and data. Some of these improvements include, but are not limited to, the incorporation of HAPs speciation, BADA, BFFM2, the First Order Approximation (FOA) for particulate matter (PM) emissions (Wayson 2009), and continued updates to the EPA MOBILE-series [being replaced by the Motor Vehicle Emissions Simulator (MOVES)] and NONROAD models whose data are incorporated into EDMS (USEPA 2005 and 2010b). All

14 Guidance for Quantifying the Contribution of airport emissions to Local air Quality of these improvements have led to more accurate and more robust emissions modeling and emissions inventory development capabilities. Since Version 5.1, EDMS has been able to estimate emissions of individual organic gas (OG) species, including 44 known HAPs and 352 non-HAP species (FAA 2008). This is accomplished through the use of a speciation profile applied to the total OG (TOG) (FAA 2009). With the inclusion of BADA and BFFM2, EDMS’ emissions modeling capabilities were noticeably improved, in part because the dynamic aircraft performance modeling allowed more accurate determinations of fuel flow (essentially, power level) and emissions indices (EIs) used to predict emissions for each segment. As shown through the work conducted by the Society of Automotive Engineers (SAE) A21 group, BFFM2 provides less than 5% error for modeling emis- sions of nitrogen oxides (NOx) per mode (e.g., takeoff and climbout) as compared to 50 to 90% errors using the simpler International Civil Aviation Organization (ICAO) reference methods in older versions of EDMS (SAE 2009). Although the errors in NOx seem to have improved, emis- sions of carbon monoxide (CO) and total hydrocarbons (total HC or THC) need to be studied further. Errors for CO and THC, especially at the lower power settings (i.e., idle and taxi) can be greater than 200%. Until recently, there were no recognized viable methods to model PM from aircraft. Hence, EDMS did not provide any results, leaving a gap between regulatory requirements and model- ing capabilities. With the implementation of the FOA in EDMS, that gap was filled, but only as a temporary measure. Originally developed by Volpe and FAA, the FOA, which continues to be promulgated through ICAO’s Committee on Aviation Environmental Protection (CAEP), is based on the use of the SMOKE Number (SN) data from the ICAO emissions databank. Given that these SNs have large errors, their use is intended to be a temporary, stop-gap measure until mass-based PM EI data become available. The SAE E31 group is developing guidance to help standardize the measurement of aircraft engine exhaust emissions. In addition to these emissions modeling advancements, various atmospheric dispersion-related improvements have also been made. The basis for these improvements was the implementation of AERMOD in EDMS 4.0. This allowed EDMS to use the latest state-of-the-art dispersion model to assess local air quality impacts. In the last two decades, drastic improvements in the characteriza- tion of the atmosphere were made, which led to the improvements represented in the USEPA’s workhorse dispersion model, AERMOD. AERMOD represents a significant improvement over its predecessor, the Industrial Source Complex (ISC) model, which was largely intended for industrial sources such as powerplant stacks. Refinements to the use of AERMOD in EDMS involved the incorporation of more accurate jet engine exhaust release heights and more representative initial dispersion coefficients (i.e., initial dispersion sigma values). Even with all of these enhancements, the overall air quality modeling capabilities of EDMS, and local air quality modeling in general, still have room for improvement. Even though it has been approximately 45 years since air quality modeling was first applied to assess airport air quality, significant uncertainties are still associated with predicting atmospheric behavior and local concentrations. The combination of errors associated with emission factors, source activities, meteorological data, and terrain information can make predicting concentrations dif- ficult. Errors in any one of these input datasets can have significant impacts on the accuracy of predicted concentrations. A case in point is aircraft CO and THC emissions at lower power settings (e.g., idle and taxi). Errors in estimating fuel burn at low-power settings can cause significant errors in modeling emissions of these pollutants (Kim 2008). This is important because any errors in emissions will propagate to the modeled concentrations. That is, modeled concentrations will only be as accurate as the emissions data allow. Also, speciation of THC needs to be further researched. In

Current State of airport air Quality assessments and Considerations 15 addition to data from Spicer (2004), data from various other airport studies and recent measure- ment campaigns such as those through the TRB Airport Cooperative Research Program (ACRP 2011) and the Aircraft Particle Emissions eXperiments (APEX) (Whitefield 2007) are beginning to provide a better understanding of the types of species emitted and their emissions character- istics (USEPAb,c 2009), but more work is required in this area. Previous studies on errors associated with atmospheric dispersion modeling (at least those involving the use of Gaussian methods) have highlighted the difficulties of accurately predict- ing atmospheric concentrations (Ellis 1980, API 1980, Bowne 1983, Benarie 1987, Seibert 2000, CEC/CARB 2002, and Stiggins 2002). The inherent uncertainties associated with dispersion components (e.g., initial dispersion values, atmospheric rise, and atmospheric profiles) have resulted in a general understanding that concentrations can be predicted to an accuracy level of within a factor of 2. Unlike a stationary powerplant stack emitting pollutants at a reasonably constant rate over time, an airport as a whole presents a much more difficult source to model. Although the most visible source at an airport is aircraft, there are various others sources [e.g., ground support equipment (GSE) and ground access vehicles (GAV)], some of which may even exceed air- craft emissions during the Landing and Takeoff (LTO) cycle. Furthermore, much of the airport sources are mobile, meaning that, in addition to fuel-consumption activities, the location of the sources must also be predicted. With such varied sources at an airport, properly modeling their emissions and the associated atmospheric dispersion can be challenging. In addition to these difficulties, one of the biggest limitations of EDMS and AERMOD in general is the lack of any real capabilities for chemical transformations. The reason for this is that the Gaussian plume framework within AERMOD is not well-suited for modeling chemi- cal transformation. Generally, this type of modeling is handled through grid-based chemistry transport models. The mass conservation techniques and the detailed atmospheric chemical mechanism implementations used with grid models like the Community Multiscale Air Quality (CMAQ) model and the Comprehensive Air quality model with extensions (CAMx) provide a more natural environment suited for the complex reaction modeling necessary to predict con- centrations of air pollutants like NO2, SO2, O3, various hydrocarbon species, etc. (Byun 1999, 2006, and ENVIRON 2006). CMAQ is a state-of-the-art, comprehensive, one-atmosphere air quality modeling system that treats gas-phase chemistry, particulate matter (PM), and hazardous air pollutants (HAPs). CMAQ simulates the numerous physical and chemical processes involved in the formation, transport, and destruction of air pollutants. Inputs to the model include emissions estimates (from aircraft as well as all other anthropogenic and biogenic sources), meteorological fields, and initial condition and boundary condition data. CMAQ has been applied to study air pol- lution from local scales (a few kilometers) to regional (hundreds of kilometers) to hemispheric (several thousands of kilometers). The model has been continually evaluated against both sur- face observations (Eder 2006 and Boylan 2006) and more recently against remote sensing data from satellite-based observations. CMAQ is an open-source community model (http://www. cmaq-model.org), and, under base funding from the EPA, the Center for Community Model- ing and Analyses System (CMAS) (http://www.cmascenter.org) provides support and training of CMAQ to a global user community. For gas-phase chemistry, CMAQ has the latest version of the Carbon Bond chemical mecha- nism (Carbon Bond 2005) described by Yarwood (2005). Particulate matter is treated in CMAQ as the sum of three modes—Aitken (diameters up to approximately 0.1 microns for the mass distribution), accumulation (mass distribution in the range of 0.1 to 2.5 microns) and coarse (particles of size 2.5 to 10 microns). Additional details on the treatment of particulate matter

16 Guidance for Quantifying the Contribution of airport emissions to Local air Quality in CMAQ are described elsewhere (Binkowski 2003). CMAQ’s treatment of HAPs is described by Luecken (2006). The last version of CMAQ that was released during the performance of this project was 4.7.1. This version has several updates to the algorithms for the formation of the Secondary Organic Aerosol (SOA), a key component of PM2.5 in several areas of the country (Carlton 2010). Foley (2010) presents an evaluation of this CMAQ version focused on O3 and PM2.5. The treatment of coarse sea-salt particles is described by Kelly (2010), and a comprehen- sive multi-year evaluation of CMAQ’s wet deposition is described by Appel (2011). The USEPA has been preparing CMAQ v5.0 for release in early 2012, and it is expected to have additional updates to the treatment of particulate matter and several other new features. To be able to use a gridded model like CMAQ, one has to prepare emissions inputs from various anthropogenic as well as natural sources at the same gridded resolution and pre- pare meteorological inputs from a prognostic model, also at the same gridded resolution. The emissions processing is accomplished through the Sparse Matrix Operator Kernel Emis- sions (SMOKE) modeling system (Houyoux 2000), which is also an open-source community model, continuously developed and also supported by CMAS (http://www.smoke-model. org). SMOKE takes state- or county-level emissions inventories developed by USEPA or other agencies and performs chemical speciation, temporal allocation, and spatial allocation in a format that can be used by CMAQ. The meteorological inputs for CMAQ are prepared from either the Pennsylvania State University/NCAR mesoscale v3.7 model (MM5) (Grell 1994) or more recently from the Weather Research Forecast Model with its Advanced Research WRF core (WRF-ARW). The WRF-ARW is typically used in research settings such as this and based off the MM5 model. It uses a ter- rain following sigma vertical coordinate and the Arakawa-C grid in the horizontal. Addition- ally, it contains additional surface physics packages developed at the USEPA to help process meteorological parameters required by the CMAQ air quality model. More information about the WRF-ARW model can be found at http://www.mmm.ucar.edu/wrf/users/model.html. The MM5 or WRF outputs are processed through the Meteorology-Chemistry Interface Processors (MCIP) (Otte 2010) to prepare the meteorological inputs in a form that can be directly used by CMAQ. MCIP is also available via, and supported, by CMAS. 2.6 Survey of Recent Airport Air Quality Studies In order to better understand airport contributions to local air quality and atmospheric dis- persion modeling processes, there have been various recent studies. These have included, but are not limited to, projects and measurement programs conducted at (or being conducted at) the following airports: • Los Angeles International Airport (LAX) (LAWA 2011) – The ongoing LAX source apportionment study is considered the most comprehensive air quality measurement and modeling project conducted by the Los Angeles World Airports (LAWA) and possibly any airport authority nationwide. – The work involves the collection of criteria and hazardous air pollutants (HAPs), and the purpose is to associate (apportion) pollutants to their sources. • Teterboro Airport (TEB) (ENVIRON 2008) – The TEB study involved measurements of various pollutants, including volatile organic compounds (VOCs) and PM, to investigate health risks associated with airport operations. – The study generally found that the risks associated with the measurement locations were higher than at other New Jersey Department of Environmental Protection (NJDEP) moni- toring locations.

Current State of airport air Quality assessments and Considerations 17 • T. F. Green Airport (PVD) (KBE 2007) – The ongoing PVD monitoring work is conducted as part of a long-term plan to measure and better understand air quality near the airport. – Pollutants included in the measurement program include PM components and various VOCs. • Van Nuys Airport (VNY) and Santa Monica Municipal Airport (SMO) (SCAQMD 2010 and ICF 2010) – Under a grant from the USEPA, a community-scale air toxics monitoring study was conducted in the communities surrounding VNY and SMO. – Except for measurements conducted near runways, most concentrations of PM and VOCs were similar to those in other urban areas. The air toxics study also found that while relatively high concentrations of ultrafine particles (UFP) were noticed in the surrounding neighborhoods, further studies would be necessary to determine airport contributions. – Based on public concerns over lead (Pb) emissions from General Aviation (GA) aircraft, the USEPA commissioned a study to develop a method to evaluate lead concentrations near airports operating piston-powered aircraft. – The SMO Pb modeling work was conducted using EDMS with dispersion handled with AERMOD. The modeled results showed relatively good agreement with measured concentrations which generally tended to be higher than the regional background concentrations. • Boston Logan International Airport (BOS) (Massport 2011 and CDM 2010) – As the longest running airport air quality measurement program in the United States (begun in 1982), the Massachusetts Port Authority (Massport) NO2 program collects data to track air quality changes in and around the airport. – A 2-year Massport-commissioned set of ambient measurements was started in 2007 in the vicinity of the airport as part of the requirements for a Massachusetts Environmental Policy Act (MEPA) study related to the Logan Airside Improvements Project (LAIP). With a focus on air toxics, the first year’s baseline data has been established. • Chicago O’Hare International Airport (ORD) (IEPA 2002) – The purpose of the ORD study was to better understand airport air quality impacts on large urban areas by specifically focusing on HAPs or air toxics. – The study found that, in general, the concentrations measured near ORD were similar or lower than those found in other large U.S. cities. • Washington-Dulles International Airport (IAD) (Wayson 2003) – An intensive measurement campaign was conducted during the winter of 2002 to col- lect ambient CO data to help validate the dispersion modeling capabilities implemented into EDMS. – Real-time gas analyzers and Minivol bag samplers were used to collect CO samples from various locations in and around the airport. – Although the results from the project were not published, the study provided useful insights on typical concentration levels at the airport. • London-Heathrow International Airport (LHR) (UKDOT 2006) – A recent study was undertaken by the United Kingdom Department for Transport (UKDOT) to strengthen the understanding of air quality and assessment capabilities around the airport. – A combination of criteria pollutants and HAPs were measured using various types of equipment. – The resulting recommendations involved approvals to use the Atmospheric Dispersion Modeling System (ADMS) (CERC 2011) with limited use allowed for LASPORT. EDMS was not chosen as an acceptable model for the airport.

18 Guidance for Quantifying the Contribution of airport emissions to Local air Quality • Zurich International Airport (ZRH) (Duchene 2007) – A 2004 study involving the use of ALAQS showed that, based on improper modeling of thrust settings and times in mode (TIM), errors in NOx emissions can be up to about 25%. The results also indicated the important impacts that heat flux, plume dynamics, and terrain have on modeled concentrations. – Depending on the location of monitors, 8 of the 21 sites showed predominant NOx contributions by the airport. In addition, there have been various FAA efforts to study and further the science involved in assessing air quality models, including through the FAA’s Center of Excellence (CoE) Partner- ship for Air Transport Noise and Emissions Reduction (PARTNER) projects. For example, the ongoing work under FAA’s PARTNER Project 16 have shown that a 108 by 108 km domain centered on the airport captured most of the population exposure to primary components of PM2.5, but total public health risks were dominated by populations at greater distances from the airport—based on the impact of secondary PM (Arunachalam 2011). The purpose of all of these airport air quality projects/programs is to better understand local air quality contributions and health impacts posed by the airports. These studies have advanced and continue to advance the understanding of airport air quality impacts by contributing to the existing knowledge base. Given that measurements and modeling studies are resource-intensive, it is expected that, although the overall understanding will increase, it will be a long-term process to expand beyond the current state of infancy. The importance of continuing this process is clear when considering future impacts of airports on local air quality. Even with the advent of Next Generation Air Transportation (NextGen) technologies, the demand for air travel will ensure that airport emissions loadings will increase beyond present conditions (Woody 2011). As such, this ACRP project is intended to further advance the understanding of airport contributions to local air quality by providing guidance to better understanding and use of the combination of measurement with modeling techniques.