CO2, NOx, and Particle Emissions from Aircraft and Support Activities

Article pubs.acs.org/est

CO2, NOx, and Particle Emissions from Aircraft and Support Activities at a Regional Airport Michael E. Klapmeyer and Linsey C. Marr* Department of Civil and Environmental Engineering, Virginia Tech, 418 Durham Hall, Blacksburg, Virginia, 24061, United States S Supporting Information *

ABSTRACT: The goal of this research was to quantify emissions of carbon dioxide (CO2), nitrogen oxides (NOx), particle number, and black carbon (BC) from in-use aircraft and related activity at a regional airport. Pollutant concentrations were measured adjacent to the airfield and passenger terminal at the Roanoke Regional Airport in Virginia. Observed NOx emission indices (EIs) for jet-powered, commuter aircraft were generally lower than those contained in the International Civil Aviation Organization databank for both taxi (same as idle) and takeoff engine settings. NOx EIs ranged from 1.9 to 3.7 g (kg fuel)−1 across five types of aircraft during taxiing, whereas EIs were consistently higher, 8.8−20.6 g (kg fuel)−1, during takeoff. Particle number EIs ranged from 1.4 × 1016 to 7.1 × 1016 (kg fuel)−1 and were slightly higher in taxi mode than in takeoff mode for four of the five types of aircraft. Diurnal patterns in CO2 and NOx concentrations were influenced mainly by atmospheric conditions, while patterns in particle number concentrations were attributable mainly to patterns in aircraft activity. CO2 and NOx fluxes measured by eddy covariance were higher at the terminal than at the airfield and were lower than found in urban areas.

■

INTRODUCTION The Federal Aviation Administration predicts that air travel will more than double in the next 20 years.1 This growth fuels continuing concerns about the health and environmental effects of airport emissions, which include carbon dioxide (CO2), nitrogen oxides (NOx), particulate matter (PM), and other trace compounds, many of which are responsible for both direct and indirect effects on radiative forcing.2 Emissions from in-use aircraft at major airports have been measured in previous studies (e.g., Popp et al.3 and Herndon et al.4) and were found to be consistent with those in the International Civil Aviation Organization (ICAO) Aircraft Engine Emissions Databank, which compiles information from engine manufacturers.5 The ICAO’s emission indices are widely used for modeling airport emissions. On the contrary, a study by Schürmann et al.6 revealed differences at Zurich Airport of up to a factor of 2 when comparing measurements to the ICAO emission indices, with higher realworld emissions for some engines and slightly lower ones for others. Any deviation from the ICAO indices is typically thought to be the result of engine-to-engine variability, the effects of pollutant aging as the plume travels to the receptor, and engine thrust settings inconsistent with ICAO standards. Prior studies have not addressed emissions from commuter aircraft, which are the focus of this research. Additional studies indicate that aircraft and related activities by ground support equipment can impair air quality near major © 2012 American Chemical Society

and small, general aviation airports, contributing to elevated concentrations of NOx, ozone, PM, and black carbon (BC).7−10 While these studies have steadily advanced the collective understanding of aircraft operations and their effects on local air quality, there is little information about midsized regional airports. Regional airports are of concern because they often have runway lengths capable of accommodating larger aircraft akin to major airports without the associated zoning restrictions intended to keep residential development at a distance. Consequently, the potential for exposure of nearby residents to emissions associated with airport operations may be larger at regional airports compared to major ones. The objective of this research was to quantify emissions of CO2, NOx, and particles that result from in-use aircraft and associated activity at a regional airport. Observed NOx and particle concentrations were translated into emission indices, which were compared with the ICAO databank. Additionally, surface−atmosphere exchange fluxes of CO2 and NOx were measured by eddy covariance, the first known application of this technique at an airport, for characterization of the spatial and temporal variability of airport-related emissions. Received: Revised: Accepted: Published: 10974

June 11, 2012 September 8, 2012 September 10, 2012 September 10, 2012 dx.doi.org/10.1021/es302346x | Environ. Sci. Technol. 2012, 46, 10974−10981

Environmental Science & Technology

■

Article

EXPERIMENTAL METHODS Instrumentation. This study employed a mobile eddy covariance laboratory, a modified television news van dubbed the Flux Laboratory for the Atmospheric Measurement of Emissions (FLAME), with a mast extending nearly 15 m above ground level.11,12 The FLAME is described in greater detail in the Supporting Information (SI). The following parameters were measured during the field study: CO2 (LiCor 7000), NOx (EcoPhysics CLD 88 Y), particle number (TSI Ultrafine Condensation Particle Counter 3025A), BC (Magee Scientific microAeth AE51), and wind velocity and temperature (Applied Technologies K-Probe). All analyzers were calibrated according to the manufacturers′ recommendations. Measurement Sites and Sampling Times. Measurements took place at the Roanoke Regional Airport in western Virginia. It is utilized primarily by airlines′ regional affiliates operating commuter aircraft such as the Canadair Regional Jet (CRJ2), which has dual turbofan jet engines, and the twinengine turboprop De Havilland Canada Dash 8 (DHC8). The airport supports more than 50 scheduled airline flights per day. The airport has two intersecting runways, with its longest extending ∼2000 m, making it well equipped to accommodate larger airliners, airfreight carriers, and a variety of military aircraft. The mobile laboratory was deployed at either of two sampling locations at the airport, shown in Figure 1. The first

conducted at the terminal site exclusively on days when winds were northwesterly, thus transporting emissions mainly from the taxiways and gate areas to the mobile laboratory. In order to comply with requirements of the eddy covariance method, sites were also selected to allow the van’s mast to extend to at least twice the height of upwind buildings and trees.13 In order to characterize the seasonal variability of pollutant concentrations and fluxes, measurements were carried out during summer, autumn, and winter between July 2011 and February 2012. Spring was excluded because of its similarity to autumn in terms of temperature and wind. During each season, measurements were conducted for four days at each location, for a total of 24 sampling days. Sampling took place only on days when wind direction met the criteria stipulated above, average wind speed was at least 2 m s−1, and temperature fell within the desired bounds. Summer, autumn, and winter sampling were defined by daily high temperatures above 27 °C, between 10 and 21 °C, and below 10 °C, respectively. SI Table S1 summarizes the sampling dates and temperatures encountered during each season. Sampling took place between 8:00 and 18:00 local time during periods of daylight savings and between 7:00 and 17:00 when daylight savings was not observed. For practical considerations and due to challenges in applying eddy covariance during stable atmospheric conditions, measurements were not conducted at night. Emission Indices. Emission indices (EI) of NOx and particles in terms of grams of NOx or number of particles emitted per kilogram of fuel burned were calculated on the basis of concentration ratios measured in the plumes of individual aircraft during the landing−takeoff cycle (e.g., SI Figure S1).4,14 For these calculations, 1 s averages of the raw 10 Hz data were used so that all species would have a time response similar to that of the slowest instrument. Each plume was identified through visual inspection, and plume durations ranged from 2 to 168 s, with a median of 17 s. The NOx EI (g NOx as NO2 per kg jet fuel) was calculated according to eq 1: EI NOx = EICO2 ×

ΔNOx 46 × ΔCO2 44

(1)

The ratio ΔNOx/ΔCO2 is the slope of NOx concentrations plotted against CO2 concentrations within the plume. The fit between NOx and CO2 in the plumes was excellent; R2 averaged 0.99 across all plumes. The value of EICO2 is 3160 g CO2 per kg jet fuel, assuming that the carbon content in jet fuel is constant and that, upon combustion, all carbon in the fuel is converted completely to CO2.4,14 The quantity 46/44 converts from a molar ratio of NO2-to-CO2 to a mass ratio. Equation 1 is valid anywhere in the exhaust plume if equal dilution and negligible chemical losses are assumed on a time scale of ∼100 s.15−17 Particle number EIs were calculated similarly with the inverse of the molar volume of air, adjusted for ambient temperature, in place of the molecular mass of NO2 to account for the different units of particle number concentration.14,18 The fit between particle number and CO2 in the plumes was good; R2 averaged 0.83 across all plumes, not as high as for NOx probably because particle number varies with degree of dilution due to the dependence of coagulation and condensation on concentration. Eddy Covariance. Flux (F) was calculated as the covariance between the vertical wind velocity (w) and gas concentration (c), as shown in eq 2. The primes denote the fluctuating component of each scalar, calculated as the instantaneous

Figure 1. Sampling sites at the Roanoke Regional Airport in Virginia (inset).

site was adjacent to the air traffic control tower in the center of the airfield, ∼250 m from the centerline of the primary departure runway and ∼120 m from the associated parallel taxiway. Under selected wind conditions, this site isolated emissions from takeoff and climbout events. The second site was located in an overflow parking lot, that was closed to other vehicles, ∼225 m from the airport terminal. Under selected wind conditions, this site isolated emissions from terminal activities. Measurements were collected at the airfield site exclusively on days when winds originated primarily from the southwest to northwest. Planes took off toward the southwest direction on the runway adjacent to the site. Thus, plumes from individual aircraft were sampled relatively soon after exiting engines (i.e., ∼1 min) and were very clearly resolved against background levels (Figure S1 in the SI). Measurements were 10975

dx.doi.org/10.1021/es302346x | Environ. Sci. Technol. 2012, 46, 10974−10981


Article

Figure 2. NOx and particle number EIs for aircraft in both taxi and takeoff modes compared to values from the ICAO where available.

Table 1. NOx and Particle Number EIs by Aircraft Type and Engine Mode aircraft

engine

CRJ2

GE CF34−3B1 turbofan

DHC8

P&W PW123 turboprop

ERJ145

AE 3007A1 RR turbofan

MD83

P&W JT8D−219 turbofan

C560

P&W JTD15−5 turbofan

mode taxi/idle takeoff taxi/idle takeoff taxi/idle takeoff taxi/idle takeoff taxi/idle takeoff

NOx EI (g kg−1)

sample size

ICAO NOx EI (g kg−1)

± ± ± ± ± ± ± ± ± ±

49 28 28 16 19 14 9 6 6 4

3.72 11.28 N/A N/A 3.66 19.01 3.88 22.86 1.66 11.13

3.0 8.8 3.7 10.5 2.9 12.3 2.9 20.6 1.9 9.1

0.7 1.8 0.7 4.3 0.5 2.3 0.3 2.4 0.4 0.8

sample size

± ± ± ± ± ± ± ± ± ±

46 27 25 16 18 14 8 5 6 4

6.0 5.0 7.1 6.8 5.4 6.0 3.9 1.4 5.6 3.7

3.4 3.1 4.9 6.3 4.6 4.2 1.9 1.0 3.3 4.6

be lossless, was calculated up to the attenuation frequency. The value of the normalized integral at the attenuation frequency represented the fraction of flux that was actually captured by the slower analyzers (SI Figure S3). An attenuation frequency of 0.4 s−1 for both CO2 and NOx was determined by visual inspection of spectra and cumulative integrals and agreed with the analyzers′ nominal response times. Correction factors applied to the measured CO2 and NOx fluxes ranged from 1.07 to 1.18, depending on the season. Although it is not feasible to accurately isolate and quantify uncertainties from eddy covariance data, on the basis of previous work dedicated to the estimation of such uncertainties, the systematic error in fluxes is likely ≤25% and random error is ≤20%.21,29−34

departure from the respective least-squares regression lines for vertical wind velocity and concentration (represented by the angle brackets) over a 30 min time period, a conventional interval for eddy covariance measurements. The overbars indicate the time-average mean. F = w′c′ = [w(t ) − ⟨w⟩][c(t ) − ⟨c⟩]

particle EI ( × 1016 kg−1)

(2)

Measurements were subjected to standard postprocessing and quality assurance and control procedures, including despiking,19−21 rotation by the planar fit method,22 lag correction,23,24 stationarity testing,25,26 and spectral analysis,27 described in greater detail in the SI. The spectra of temperature and cospectra of temperature with vertical wind velocity had slopes of −2/3 and −4/3, obeying their respective power decay laws,27 but those of CO2 and NOx deviated from these slopes at higher frequencies due to the analyzers′ slower response times (SI Figure S2). The CO2 analyzer was capable of 20 Hz response, but an incorrect setting in the analyzer’s data filtering, discovered after the campaign, led to attenuation at higher frequencies. Losses in fluxes at high frequencies were corrected by assuming spectral similarity with temperature using a previously published method.28 Briefly, the integral of the raw cospectrum of temperature (i.e., the flux), which is assumed to

■

RESULTS Emission Indices. Over the course of 12 days of sampling at the airfield, 221 unambiguous taxi and takeoff events were captured for calculation of EIs. Plume identification at the terminal was not feasible because of the frequent presence of other sources. Of the 221 events, over 80% were associated with five types of airframes: Canadair Regional Jet (CRJ2), De Havilland Canada Dash 8 (DHC8), Embraer Jet 145 (ERJ145), McDonnell Douglas 83 (MD83), and Cessna 560 Citation 10976



Article

Figure 3. Average seasonal and overall fluxes (whiskers show standard deviations) for CO2 and NOx at the airfield and terminal during daytime hours.

thrust was unknown. The resulting BC EIs for this limited data set were 0.2 g BC (kg fuel)−1 and 0.5 g BC (kg fuel)−1. Although the sample size is small, the results are a valuable addition to the limited database of in-use BC emissions from turboprop aircraft. Concentrations. CO2, NOx, particle number, and BC concentrations were examined in order to determine whether they were higher at the site influenced by activities around the terminal v. the one dominated by aircraft taxiing and taking off and whether any differences had a seasonal dependence. SI Table S2 summarizes the average seasonal concentrations of CO2, NOx, particle number, and BC during sampling at both the airfield and terminal locations. SI Figures S4−S7 show concentration box plots at 30 min intervals. CO2 concentrations and seasonal variations were consistent with global trends. A diurnal pattern was evident at both sampling locations, although less pronounced during the winter, consistent with observations from other studies in urban areas.12,13,26 Average CO2 concentrations during each season differed little between the two sampling locations, except during autumn, when the average concentration at the terminal was 9 ppm higher than at the airfield. This difference was likely an artifact of the sampling schedule, which led to measurements at the airfield in October, when vegetation was still taking up CO2, and at the terminal in November and December, when vegetation had become dormant. Diurnal patterns in NOx concentrations were similar to those of CO2. The highest 30 min average concentration of 72 ppb was measured at the airfield during especially stable conditions, whereas minimum concentrations were consistently below 1 ppb during all seasons at both locations. Unlike CO2, no seasonal variations were evident. Average particle number concentrations ranged from 0.15 × 104 cm−3 to 17.4 × 104 cm−3. Unlike the gaseous species, particle number concentrations were highly variable in time and appeared to be more sensitive to upwind activity. Concentrations were higher at the terminal during both summer and autumn, likely because emissions at the terminal were more persistent than those encountered at the airfield and included contributions from ground support equipment and auxiliary

(C560). Excluding the DHC8, which was powered by twin turboprop engines, all aircraft were powered by twin turbofan jet engines. The C560 is a small business jet while the others are commuter aircraft. Figure 2 shows NOx and particle number EIs by type of aircraft for both taxi (i.e., idle) and takeoff modes, and Table 1 lists the corresponding engine models and values of the EIs. In the figure, EIs from the ICAO emissions databank are indicated with a red triangle, where available. The databank does not contain EIs for particle number, nor does it contain EIs for turboprop engines. The average NOx EI for jet-powered aircraft was consistently lower than ICAO values for both taxi and takeoff engine modes, with the exception of the C560 during taxi. Average EIs of the CRJ2, ERJ145, and MD83 in taxi mode were similar, 2.9−3.0 g NOx (kg fuel)−1, and all were ∼20% lower than ICAO values. For the C560, the average EI was 1.9 g NOx (kg fuel)−1, 14% higher than the ICAO value. During takeoffs, average NOx EIs were 3 to 7 times higher than their respective values during taxiing, depending on the type of aircraft, and ranged from 8.8 g NOx (kg fuel)−1 for the CRJ2 to 20.6 g NOx (kg fuel)−1 for the MD83. Observed NOx EIs in takeoff mode were 10−35% lower than ICAO values. Average particle number EIs during taxi mode showed little variation among the different aircraft, ranging from 3.9 × 1016 particles (kg fuel)−1 for the MD83 to 7.1 × 1016 particles (kg fuel)−1 for the DHC8. Unlike NOx EIs where values in takeoff mode were consistently higher than in taxi mode, relative differences for particle numbers were mixed. Average takeoff EIs ranged from 1.4 × 1016 particles (kg fuel)−1 for the MD83 to 6.8 × 1016 particles (kg fuel)−1 for both the DHC8. Particle number EIs were lower during takeoff relative to taxi mode for four out of the five aircraft types, although the difference was not statistically significant. In most cases, BC concentrations in aircraft plumes were not sufficiently elevated above background to allow for calculation of EIs. However, clear spikes in BC were observed on two occasions when DHC8 aircraft, parked ∼60 m away outside a maintenance hangar, provided several minutes of uninterrupted emissions. The two events had NOx EIs indicative of an engine in idle mode (i.e., 7% thrust setting), although the exact engine 10977



■

power units. In contrast, average winter concentrations were comparable between the airfield and terminal probably because some sampling occurred on weekend days when airport activity was lower and/or because of construction upwind of the airfield sampling location. BC concentrations ranged from 0.01 μg m−3 during a period of inactivity at the terminal in winter to 3.6 μg m−3 during very stable atmospheric conditions at the airfield in autumn. Patterns in BC concentrations were more similar to those of CO2 and NOx than to those of particle number, especially at the airfield where most aircraft activity had little appreciable impact on BC levels. Clear spikes in BC at the airfield were observed on two occasions when DHC8 turboprop aircraft, parked adjacent to a maintenance hangar approximately 60 m away, started their engines and advanced the throttles for several minutes before departing the area, creating instantaneous peak BC concentrations in excess of 30 μg m−3. No seasonal pattern was apparent in BC. Fluxes. Figure 3 presents average seasonal fluxes of CO2 and NOx at the airfield and terminal. In all cases, average fluxes were positive, indicating that the airfield and terminal served as sources of emissions during daylight hours. Fluxes did not appear to vary by season, at least within the uncertainties of the eddy covariance method and the limited duration of this field campaign. Time series of fluxes are shown in SI Figure S8. Diurnal patterns were not evident except that fluxes were higher around 11:00 and in the late afternoon at the terminal. CO2 fluxes ranged from −0.3 mg m−2 s−1 to 1.4 mg m−2 s−1 (30 min averages) throughout the campaign, with the minimum value occurring on the airfield during the summer and the maximum occurring at the terminal during the autumn. Unlike CO2 concentrations, fluxes did not exhibit a diurnal pattern throughout the day at the two sampling locations. Rather, peak values were associated with periods of greater aircraft activity. The average CO2 flux was lower at the airfield than at the terminal likely due to less persistent aircraft activity as well as the uptake of CO2 by grass on the airfield. Average CO2 fluxes on the airfield were lowest during autumn, when grass was actively growing and not stressed by the warm, dry conditions encountered during the summer. NOx fluxes ranged from −1.1 μg m−2 s−1 to 2.6 μg m−2 s−1 (30 min averages) throughout the study, with the minimum and maximum values both occurring on the airfield during the autumn. As with CO2, NOx fluxes were maximal during periods of higher-than-normal aircraft activity. Extreme negative values for NOx flux at both locations were also correlated with periods of increased aircraft activity. At the airfield, the largest negative flux resulted from touch-and-go maneuvers by a U.S. Air Force C-40 transport aircraft, a military version of the Boeing 737. During this 30 min period, the C-40 periodically landed on the runway and subsequently took off without coming to a complete stop. The engine mode for these maneuvers was indicative of a takeoff, with instantaneous NOx concentrations exceeding 200 ppb on several occasions. The magnitude of the NOx emissions coupled with the frequency at which the C-40 was airborne during its maneuvers (i.e., elevated source of emissions) resulted in a period characterized by NO x deposition. A downward NOx flux at the terminal was also observed when aircraft were exclusively utilizing the far side of the terminal, parking in excess of 300 m from the mobile laboratory’s position. Effects of the building wake may have contributed to a downward flux.

Article

DISCUSSION

Comparison of Emission Indices to Previous Studies. Observed NOx EIs were slightly lower than those in the ICAO databank, but trends in emissions with increased engine throttle were consistent. Average EIs of the five most frequently encountered aircraft at the Roanoke airport ranged from 1.9 to 3.7 g NOx (kg fuel)−1 for taxi events and from 8.8 to 20.6 g NOx (kg fuel)−1 for takeoffs. These observations agree with results from previous airport and aircraft studies investigating NOx emissions.3,4,14,35 SI Table S3 shows a comparison of results from different studies. Carslaw et al.36 studied NOx emission rates at London’s Heathrow Airport and also found reasonable agreement between observations and ICAO’s EIs for NOx, and differences of up to 41% were attributed to aircraft operational factors such as takeoff weight and engine thrust settings. While measured NOx emissions in the present study were at most 25% lower than ICAO values for taxi events with idle thrust settings, Schäfer et al.37 measured NOx emissions for idling aircraft at three major European airports and found them to be about 50% lower than ICAO values. Particle number EIs at the Roanoke airport fell within the range of values reported in the literature for studies in Santa Monica, Atlanta, New York, and Boston,8,18,35 and were most similar to those measured in Santa Monica, which is also a regional airport whose traffic is dominated by smaller jets. SI Table S4 shows a comparison of results from different studies. The minimum particle size detected by the particle counter employed in each study is also shown, as this factor is likely to affect the results; it does not appear to be the controlling factor, however. Ambient temperature may also affect results, as particle number in aircraft exhaust plumes has been shown to be sensitive to postemission condensation.38 In all studies, particle emissions were roughly two times lower during takeoff relative to taxi mode for most aircraft. Herndon et al.18 concluded that particle number emissions, when compared to NOx emissions, did not produce as consistent a trend in EIs and were not controlled by the same engine conditions. The limited number of BC EIs observed in this study, 0.2 and 0.5 g BC (kg fuel)−1 from the DHC8 turboprop engines, were comparable to the upper end of those measured for idling jet engines during the comprehensive Aircraft Particle Emissions eXperiment (APEX).39 Results from APEX indicated that, in general, BC EIs were highest at takeoff and climbout and lowest during idle and approach modes, a finding that validates previous observations.40 Of the seven different jet engines studied in APEX, BC EIs at idle thrust settings ranged from 0.02 to 0.38 g BC (kg fuel)−1. APEX also found that larger engines did not always produce the most BC and that older technologies and colder engine temperature contributed to higher EIs. Emission factors at the lower end of the range found in APEX would not have been detectable in the present study because the rise in BC concentrations in the diluted plumes above background levels would have been at the limit of the aethalometer’s resolution. Temporal and Spatial Variability of Concentrations and Fluxes. Diurnal patterns in CO2 and NOx concentrations appeared to be influenced mainly by atmospheric conditions, with elevated concentrations occurring in the early morning hours and then decreasing through day, coincident with a growing mixing depth. In contrast, the daily behavior of particle number concentrations and of CO2 and NOx fluxes was attributable mainly to patterns in aircraft activity. From a 10978



Article

seasonal standpoint, NOx concentrations as well as fluxes of all pollutants did not exhibit significant variability. CO2 seasonal variations were consistent with trends in the northern hemisphere, and particle number concentrations were generally lower during warmer months. Fluxes of CO2 and NOx were higher at the terminal v. the airfield, indicating that that ground-support equipment and motor vehicle traffic around the terminal were important contributors to overall emissions at this airport. When averaged over all seasons, CO2 flux was eight times higher at the terminal compared to the airfield, while NOx fluxes were only two times higher at the terminal (SI Figure S8). One possible explanation for the larger disparity between CO2 fluxes is the contribution of uptake by grass on the airfield during summer and autumn. Although average fluxes at the terminal were consistently higher than at the airfield, pollutant concentrations did not vary significantly between the two locations studied. This apparent discrepancy is possible because concentrations reflect a combination of regional and local influences, while fluxes reflect local influences only. Flux footprints extended up to a few hundred meters in the upwind direction, so they would have incorporated activities at the terminal. The result indicates that compared to regional background sources, emissions associated with the Roanoke airport have a relatively small impact on local air quality. Comparison of Fluxes to Previous Studies. Because this is the first known study to target airport emissions using eddy covariance, comparisons with previous work involving eddy covariance are limited to urban areas, whose major sources of emissions are typically motor vehicles. Net fluxes of CO2 and NOx at the Roanoke airport were predominantly positive (i.e., upward) yet generally lower in magnitude than flux measurements in other urban areas. For CO2, average daytime seasonal fluxes ranged from 0.0 mg m−2 s−1 at the airfield in autumn to 0.3 mg m−2 s−1 at the terminal in winter; the average across both sampling locations was 0.1 mg m−2 s−1 over the entire study. SI Table S5 shows fluxes measured in other cities for comparison. CO2 fluxes measured in Münster, Germany were slightly higher, with weekday summertime values ranging from 0.2 to 0.5 mg m−2 s−1.41 A study in Helsinki, Finland reported average seasonal CO2 emissions to be fairly consistent over an urban district at 0.2 mg m−2 s−1, while fluxes measured over vegetative cover ranged from 0.2 mg m−2 s−1 in the winter to −0.1 mg m−2 s−1 in the summer.42 Velasco et al.26 found higher average CO2 fluxes in Mexico City, ranging from ∼0.4 to 0.8 mg m−2 s−1 during the 10 h sampling period observed in this study. Finally, a study in the border cities of Tijuana, Mexico and San Diego, California measured average daytime fluxes ranging from 0.2 to 0.7 mg m−2 s−1, with a combined average of 0.4 mg m−2 s−1 over the entire study.43 CO2 fluxes at the Roanoke airport were lower than those measured in the aforementioned urban areas and similar to those found over an area with mainly vegetative cover. These results are consistent with the observation that flux footprints at the airport often included undeveloped land. NOx fluxes have been measured in far fewer locations. Average daytime seasonal NOx fluxes at the airport ranged from 0.1 μg m−2 s−1 at the airfield in both autumn and winter to 0.6 μg m−2 s−1 at the terminal in winter, with a combined average of 0.3 μg m−2 s−1 for the entire study. These values were lower than observations from Norfolk, Virginia, where daytime fluxes measured across 13 sampling locations in a 144 km2 urban area ranged from 1.5 to 5.7 μg m−2 s−1 during weekday sampling.11

The Roanoke airport values were more similar to, albeit still lower than, observations in Tijuana−San Diego, where average daytime NOx fluxes ranged from 0.8 to 2.2 μg m−2 s−1, with a combined average of 1.8 μg m−2 s−1 for the entire study.43 By contrast, a study of a forest located in the northeastern U.S., with minimal anthropogenic influences, reported average NOx fluxes as downward, with values as large as −0.3 μg m−2 s−1.44 Comparison to the National Emissions Inventory. Measured NOx fluxes were compared to the 2008 National Emissions Inventory (NEI) developed by the U.S. Environmental Protection Agency (EPA). The Roanoke Regional Airport, like all airports in the inventory, is designated as a point source, and its emissions estimates encompass aircraft, ground support equipment, and auxiliary power units. In-flight emissions that occur above 915 m (3000 ft) (i.e., outside the landing−takeoff cycle) are designated as nonpoint emissions and are detailed in the inventory on a per-state basis.45 Although the database of flux observations was small12 days at each of two sitesthe comparison illustrates the potential for the use of eddy covariance measurements for evaluation of emission inventories. According to the 2008 NEI, the Roanoke airport emits 53.6 tons of NOx per year. This annual estimate was spatially averaged over the 3.66 km2 (904 acre) area occupied by the Roanoke airport, resulting in an area-averaged flux of 0.4 μg m−2 s−1. This value would be expected to be lower in 2011 because total landings and takeoffs at the airport decreased by 22% between 2008 and 2011.46 The NOx flux measured by eddy covariance, averaged over all seasons and sampling locations (Figure 3), was 0.3 μg m−2 s−1, 25% lower than the 2008 inventory’s area-averaged flux. Thus, the agreement between the inventory and eddy covariance flux measurements is good, but a larger set of flux measurements at more sites, to achieve sufficient spatial representativeness of the entire airport area, and over more days would be needed to conclude whether observations offer validation of the inventory. The relative impact of NOx emissions from the Roanoke airport can be approximated from additional data from the 2008 inventory. It estimates total NOx emissions within the City of Roanoke to be 2316 tons yr−1. NOx emissions from onroad mobile sources are estimated at 1358 tons yr−1, while point, area, and off-road mobile sources account for ∼1000 tons yr−1. Thus, airport activities within the City of Roanoke are estimated to account for only 2% of total NOx emissions and to produce 25 times less NOx emissions than all gasoline and diesel powered on-road vehicles.

■

ASSOCIATED CONTENT

S Supporting Information *

Gas and particle analyzers used, sampling dates and temperatures, QA/QC procedures, and concentration and flux time series and statistics are available. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*Phone: (540) 231-6071; fax: (540) 231-7916; e-mail: lmarr@ vt.edu. Notes

The authors declare no competing financial interest. 10979



■

Article

Knighton, W. B. Gas turbine engine emissions-part I: Volatile organic compounds and nitrogen oxides. Trans. ASME: J. Eng. Gas Turbines Power 2010, 132 (6), 061504 DOI: 10.1115/1.4000131. (15) Danilin, M. Y.; Kruger, B. C.; Ebel, A. Short-term atmospheric effects of high-altitude aircraft emissions. Ann. Geophys. 1992, 10 (11− 12), 904−911. (16) Schulte, P.; Schlager, H. In-flight measurements of cruise altitude nitric oxide emission indices of commercial jet aircraft. Geophys. Res. Lett. 1996, 23 (2), 165−168, DOI: 10.1029/95gl03691. (17) Weibring, G.; Zellner, R. Chemistry in plumes of high-flying aircraft with H2 combustion engines: A modelling study. Ann. Geophys. 1994, 12 (5), 403−408, DOI: 10.1007/s00585-994-0403-y. (18) Herndon, S. C.; Onasch, T. B.; Frank, B. P.; Marr, L. C.; Jayne, J. T.; Canagaratna, M. R.; Grygas, J.; Lanni, T.; Anderson, B. E.; Worsnop, D.; Miake-Lye, R. C. Particulate emissions from in-use commercial aircraft. Aerosol Sci. Technol. 2005, 39 (8), 799−809, DOI: 10.1080/02786820500247363. (19) Hojstrup, J. A statistical data screening procedure. Meas. Sci. Technol. 1993, 4 (2), 153−157, DOI: 10.1088/0957-0233/4/2/003. (20) Schmid, H. P.; Grimmond, C. S. B.; Cropley, F.; Offerle, B.; Su, H.-B. Measurements of CO2 and energy fluxes over a mixed hardwood forest in the mid-western United States. Agric. For. Meteorol. 2000, 103 (4), 357−374, DOI: 10.1016/S0168-1923(00)00140-4. (21) Vickers, D.; Mahrt, L. Quality control and flux sampling problems for tower and aircraft data. J. Atmos. Oceanic Technol. 1997, 14 (3), 512−526, DOI: 10.1175/1520-0426(1997) 0142.0.CO;2. (22) Wilczak, J. M.; Oncley, S. P.; Stage, S. A. Sonic anemometer tilt correction algorithms. Boundary Layer Meteorol. 2001, 99 (1), 127− 150, DOI: 10.1023/A:1018966204465. (23) Leuning, R. A. Y.; Judd, M. J. The relative merits of open- and closed-path analysers for measurement of eddy fluxes. Global Change Biol. 1996, 2 (3), 241−253, DOI: 10.1111/j.13652486.1996.tb00076.x. (24) Nemitz, E.; Jimenez, J. L.; Huffman, J. A.; Ulbrich, I. M.; Canagaratna, M. R.; Worsnop, D. R.; Guenther, A. B. An eddycovariance system for the measurement of surface/atmosphere exchange fluxes of submicron aerosol chemical speciesFirst application above an urban area. Aerosol Sci. Technol. 2008, 42 (8), 636−657, DOI: 10.1080/02786820802227352. (25) Cullen, N. J.; Steffen, K.; Blanken, P. D. Nonstationarity of turbulent heat fluxes at Summit, Greenland. Boundary Layer Meteorol. 2007, 122 (2), 439−455, DOI: 10.1007/s10546-006-9112-2. (26) Velasco, E.; Pressley, S.; Allwine, E.; Westberg, H.; Lamb, B. Measurements of CO2 fluxes from the Mexico City urban landscape. Atmos. Environ. 2005, 39 (38), 7433−7446, DOI: 10.1016/j.atmosenv.2005.08.038. (27) Roth, M.; Oke, T. R.; Steyn, D. G. Velocity and temperature spectra and cospectra in an unstable suburban atmosphere. Boundary Layer Meteorol. 1989, 47 (1), 309−320, DOI: 10.1007/BF00122336. (28) Damay, P. E.; Maro, D.; Coppalle, A.; Lamaud, E.; Connan, O.; Hébert, D.; Talbaut, M.; Irvine, M. Size-resolved eddy covariance measurements of fine particle vertical fluxes. J. Aerosol Sci. 2009, 40 (12), 1050−1058, DOI: 10.1016/j.jaerosci.2009.09.010. (29) Foken, T.; Wichura, B. Tools for quality assessment of surfacebased flux measurements. Agric. For. Meteorol. 1996, 78 (1−2), 83− 105, DOI: 10.1016/0168-1923(95)02248-1. (30) Hollinger, D. Y.; Richardson, A. D. Uncertainty in eddy covariance measurements and its application to physiological models. Tree Physiol. 2005, 25 (7), 873−885, DOI: 10.1093/treephys/ 25.7.873. (31) Kruijt, B.; Elbers, J. A.; vonRandow, C.; Araujo, A. C.; Oliveira, P. J.; Culf, A. D.; Manzi, A. O.; Nobre, A. D.; Kabat, P.; Moors, E. J.; Process, H. The robustness of eddy correlation fluxes for Amazon rain forest conditions. Ecol. Appl. 2004, 14 (4), S101−S113, DOI: 10.1093/treephys/25.7.873. (32) Lenschow, D. H.; Mann, J.; Kristensen, L. How long is long enough when measuring fluxes and other turbulence statistics. J. Atmos.

ACKNOWLEDGMENTS This research was supported by the National Science Foundation (CBET-0547107 and CBET-0715162). We gratefully acknowledge the assistance of H. Rakha, J. Bryson, and the Virginia Tech Transportation Institute for use and maintenance of the van, as well as A. Plaza, E. Gonzalez, and the Roanoke Regional Airport Commission for their exceptional support and cooperation. We also thank the Department of the Air Force for their sponsorship of Lieutenant Colonel Klapmeyer; however, the views expressed in this article are those of the authors and do not reflect the official policy or position of the Air Force, the Department of Defense, or the U.S. Government.

■

REFERENCES

(1) Federal Aviation Administration FAA Aerospace Forecast (Fiscal Years 2011−2031). http://www.faa.gov/about/office_org/ headquarters_offices/apl/aviation_forecasts/ (accessed 21 December 2011). (2) Wuebbles, D.; Gupta, M.; Ko, M. Evaluating the impacts of aviation on climate change. EOS Trans., Am. Geophys. Union 2007, 88 (14), 157, DOI: 10.1029/2007eo140001. (3) Popp, P. J.; Bishop, G. A.; Stedman, D. H. Method for commercial aircraft nitric oxide emission measurements. Environ. Sci. Technol. 1999, 33 (9), 1542−1544, DOI: 10. 1021/es981194+. (4) Herndon, S. C.; Shorter, J. H.; Zahniser, M. S.; Nelson, D. D.; Jayne, J.; Brown, R. C.; Miake-Lye, R. C.; Waitz, I.; Silva, P.; Lanni, T.; Demerjian, K.; Kolb, C. E. NO and NO2 emission ratios measured from in-use commercial aircraft during taxi and takeoff. Environ. Sci. Technol. 2004, 38 (22), 6078−6084, DOI: 10.1021/es049701c. (5) Annex 16 to the Convention on International Civil Aviation Environmental Protection, Volume II − Aircraft Engine Emissions; International Civil Aviation Organization, 2008. (6) Schürmann, G.; Schäfer, K.; Jahn, C.; Hoffmann, H.; Bauerfeind, M.; Fleuti, E.; Rappenglück, B. The impact of NOx, CO and VOC emissions on the air quality of Zurich airport. Atmos. Environ. 2007, 41 (1), 103−118, DOI: 10. 1016/j.atmosenv.2006.07.030. (7) Carslaw, D. C.; Beevers, S. D.; Ropkins, K.; Bell, M. C. Detecting and quantifying aircraft and other on-airport contributions to ambient nitrogen oxides in the vicinity of a large international airport. Atmos. Environ. 2006, 40 (28), 5424−5434. , DOI: 10.1016/j.atmosenv.2006.04.062. (8) Hu, S.; Fruin, S.; Kozawa, K.; Mara, S.; Winer, A. M.; Paulson, S. E. Aircraft emission impacts in a neighborhood adjacent to a general aviation airport in Southern California. Environ. Sci. Technol. 2009, 43 (21), 8039−8045, DOI: 10.1021/es900975f. (9) Unal, A.; Hu, Y.; Chang, M. E.; Talat Odman, M.; Russell, A. G. Airport related emissions and impacts on air quality: Application to the Atlanta International Airport. Atmos. Environ. 2005, 39 (32), 5787− 5798, DOI: 10.1016/j.atmosenv.2005.05.051. (10) Westerdahl, D.; Fruin, S. A.; Fine, P. L.; Sioutas, C. The Los Angeles International Airport as a source of ultrafine particles and other pollutants to nearby communities. Atmos. Environ. 2008, 42 (13), 3143−3155, DOI: 10.1016/j.atmosenv.2007.09.006. (11) Moore, T. O. Application of a mobile flux lab for the atmospheric measurement of emissions (FLAME). Ph.D. Dissertation, Virginia Tech, Blacksburg, VA, 2009. (12) Moore, T. O.; Doughty, D. C.; Marr, L. C. Demonstration of a mobile flux laboratory for the atmospheric measurement of emissions (FLAME) to assess emissions inventories. J. Environ. Monit. 2009, 11 (2), 259−268, DOI: 10.1039/b810798j. (13) Grimmond, C. S. B.; King, T. S.; Cropley, F. D.; Nowak, D. J.; Souch, C. Local-scale fluxes of carbon dioxide in urban environments: Methodological challenges and results from Chicago. Environ. Pollut. 2002, 116 (Supplement 1), S243−S254, DOI: 10.1016/S02697491(01)00256-1. (14) Timko, M. T.; Herndon, S. C.; Wood, E. C.; Onasch, T. B.; Northway, M. J.; Jayne, J. T.; Canagaratna, M. R.; Miake-Lye, R. C.; 10980



Article

Oceanic Technol. 1994, 11 (3), 661−673, DOI: 10.1175/15200426(1994)0112.0.CO;2. (33) Loescher, H. W.; Law, B. E.; Mahrt, L.; Hollinger, D. Y.; Campbell, J.; Wofsy, S. C. Uncertainties in, and interpretation of, carbon flux estimates using the eddy covariance technique. J. Geophys. Res. 2006, 111 (D21), D21S90 DOI: 10.1029/2005jd006932. (34) Massman, W. J.; Lee, X. Eddy covariance flux corrections and uncertainties in long-term studies of carbon and energy exchanges. Agric. For. Meteorol. 2002, 113 (1−4), 121−144, DOI: 10.1016/S01681923(02)00105-3. (35) Herndon, S. C.; Jayne, J. T.; Lobo, P.; Onasch, T. B.; Fleming, G.; Hagen, D. E.; Whitefield, P. D.; Miake-Lye, R. C. Commercial aircraft engine emissions characterization of in-use aircraft at Hartsfield-Jackson Atlanta International Airport. Environ. Sci. Technol. 2008, 42 (6), 1877−1883, DOI: 10.1021/es072029+. (36) Carslaw, D. C.; Ropkins, K.; Laxen, D.; Moorcroft, S.; Marner, B.; Williams, M. L. Near-field commercial aircraft contribution to nitrogen oxides by engine, aircraft type, and airline by individual plume sampling. Environ. Sci. Technol. 2008, 42 (6), 1871−1876, DOI: 10.1021/es071926a. (37) Schäfer, K.; Jahn, C.; Sturm, P.; Lechner, B.; Bacher, M. Aircraft emission measurements by remote sensing methodologies at airports. Atmos. Environ. 2003, 37 (37), 5261−5271, DOI: 10.1016/j.atmosenv.2003.09.002. (38) Wey, C. C.; Anderson, B. E. Aircraft Particle Emissions eXperiment (APEX), NASA/TM-2006−214382; ARL-TR-3903; NASA: Washington, DC, 2006. (39) Kinsey, J. S.; Hays, M. D. Chemical characterization of the fine particle emissions from commercial aircraft engines during the Aircraft Particle Emissions eXperiment (APEX) 1 to 3. Environ. Sci. Technol. 2011, 45 (8), 3415−3421. (40) Petzold, A.; Döpelheuer, A. Reexamination of black carbon mass emission indices of a jet engine. Aerosol Sci. Technol. 1998, 29 (4), 355−356, DOI: 10.1080/02786829808965575. (41) Dahlkotter, F.; Griessbaum, F.; Schmidt, A.; Klemm, O. Direct measurement of CO2 and particle emissions from an urban area. Meteorol. Z. 2010, 19 (6), 565−575, DOI: 10.1127/0941-2948/2010/ 0492. (42) Vesala, T.; Järvi, L.; Launiainen, S.; Sogachev, A.; Rannik, Ü .; Mammarella, I.; Siivola, E.; Keronen, P.; Rinne, J.; Riikonen, A. N. U.; Nikinmaa, E. Surface−atmosphere interactions over complex urban terrain in Helsinki, Finland. Tellus B 2008, 60 (2), 188−199, DOI: 10.1111/j.1600-0889.2007.00312.x. (43) Klapmeyer, M. E. Characterization of urban air pollutant emissions by eddy covariance using a mobile flux laboratory. Ph.D. Dissertation, Virginia Tech, Blacksburg, VA, 2012. (44) Horii, C. V.; William Munger, J.; Wofsy, S. C.; Zahniser, M.; Nelson, D.; Barry McManus, J. Atmospheric reactive nitrogen concentration and flux budgets at a Northeastern U.S. forest site. Agric. For. Meteorol. 2005, 133, 210−225, DOI: 10.1016/j.agrformet.2006.03.005. (45) 2008 Emissions Inventory System Implementation Plan: Reporting Instructions for Airports, Locomotives, and Commercial Marine Vessels; U.S. Environmental Protection Agency, 2009. (46) Roanoke Regional Airport Air Traffic Reports. www. roanokeairport.com/Home/AboutUs/ RoanokeRegionalAirportComm/AirTrafficReports (accessed March 2012).

10981


CO2, NOx, and Particle Emissions from Aircraft and Support Activities

Recommend Documents