To assess the impact on the chemical composition of the atmosphere from a current and future fleet of aircraft, it is necessary to take the following steps:
Identify emissions associated with aircraft operation
Evaluate how each emission would change concentration of corresponding species in the atmosphere
Determine how those changes could alter concentrations of other species in the atmosphere.
For aircraft with conventional engines that use hydrocarbon fuels, emissions include H2O, carbon dioxide CO2), NOx, oxidation products from sulfur impurities, hydrocarbons, carbon moNOxide (CO), and soot. These products have the potential to perturb the atmosphere if their emissions are large enough to change background concentrations substantially. Other emissions, such as metallic particulates from engine wear and paints, have been considered and are thought to have minimal atmospheric effects (Stolarski and Wesoky, 1993). One environmental concern is how emissions could affect O3 in the atmosphere-both in their potential to deplete O3 in the lower stratosphere, leading to increases in UV-B radiation at the ground, and in their potential to increase O3 in the upper troposphere, leading to greenhouse warming. In addition, increases in water vapor in the LS could have a direct effect on radiative balance as well as chemistry.
To evaluate how emission of a species could change its background concentration, one could estimate the expected change in concentration from emission rate and residence time and compare that with the background concentration, or one could compare the emission rate directly with sources that sustain the background concentration. Consideration of either criterion points to different impacts from material emitted either to the UT or the LS. The residence time for material emitted in the UT is typically on the order of weeks, whereas residence time in the LS is on the order of months to a few years (Johnston, 1989). For NOx and H2O, background sources also differ in the UT and LS. In the UT, NOx is affected by surface sources as well as lightning sources in the whole troposphere. In contrast, NOx in the LS is sustained by downward transport of NOx and nitric acid (HNO3) from the middle stratosphere and transport of NOx produced by lightning in the tropical UT. In the case of H2O, the background concentration in the UT is orders of magnitude larger than that in the LS.
The character of the O3 budget is also very different in the UT and LS. In the UT, the transport and chemical time scales are on the order of weeks. The chemical transformation is dominated by reactions among oxides of hydrogen (HOx) and NOx radicals, which affect local production and removal rates of O3. Local concentrations of HOx species in the UT are controlled by concentrations of water, hydrocarbons, NOx, and CO, each of which is affected by how contributions from surface sources are redistributed in the UT by convection. In the LS, transport and chemical time scales are on the order of months. The O3 budget in the LS is maintained by a balance between transport and chemistry (chemical production balanced by transport out of the region in the tropical LS, and chemical removal balanced by transport into the region in the mid-latitude LS). Addition of NOx and H2O to the LS would modify the chemical production and destruction rates of O3. However, the efficiency of the added radicals in removing O3 is dependent on the amount of chlorine radicals in the background atmosphere and the extent of surface reactions that occur on sulfate particles and polar stratospheric clouds (PSCs). Previous modeling studies (Danilin et al., 1997; Weisenstein et al., 1998) have shown that sulfur emissions from supersonic aircraft can increase the surface area of the sulfate layer by about 50% in the Northern Hemisphere LS. Effects from the current subsonic fleet are less clear. Subsonic aircraft cruise in the troposphere or the very lowest part of the stratosphere (just above the tropopause); thus, the stratospheric impact from subsonic aircraft sulfur emissions would probably be less than that computed for the projected supersonic fleet. Whether any observed trend in the sulfate layer in the past decade can be ascribed to subsonic aircraft is currently under debate (see Hofmann, 1991, and Section 188.8.131.52). The amount of PSCs also would be increased as a result of H2O and NOx emissions from aircraft (see Section 3.3.6).
Numerical models of the atmosphere are used to calculate these changes. By solving a system of equations, these models simulate the transport and chemical interactions of trace gases to obtain their spatial and seasonal distributions. A typical model keeps track of the distributions of 50 species that interact via more than 100 reactions. Transport in the models is driven by winds and parameterization of mixing, which change with seasons. There are several ways to classify the models into different classes. One is by dimensionality: Two-dimensional (2-D) versus three-dimensional (3-D). 3-D models simulate the distributions of trace gases as functions of altitude, latitude, longitude, as well as season. 2-D models of the stratosphere simulate the zonal mean (averaged over longitudes) concentrations of species, taking advantage of the fact that many of the trace gases have uniform concentrations along latitude circles in the stratosphere. Another way to distinguish different types of models is to note whether the transport circulation is fixed or calculated in a consistent way with the model-generated trace gas concentrations.
General circulation models (GCMs) calculate temperature and transport circulation along with chemical composition. Alternatively, chemistry-transport models (CTMs) simulate the distribution of trace gases using temperature and transport circulation either from pre-calculated GCM results or derived from observations.
Because of intrinsic differences in chemistry and dynamics that control O3 and precursor species in the UT and LS, different models have been developed to examine the different regions. Models for the troposphere require better resolution immediately above the planetary boundary layer and a proper description of convection that carries material from the boundary layer to the free troposphere. The chemical scheme in these models places more emphasis on the role of non-methane hydrocarbons (NMHCs), acetone, and peroxyacetalnitrate (PAN). Models with emphasis on the stratosphere concentrate on large-scale transport from the equatorial LS to the mid-latitudes and the exchange of material between the stratosphere and the UT. The chemical scheme pays more attention to the coupling between the nitrogen, hydrogen, and halogen species and their sources in the stratosphere. The computation requirements are such that it has not been possible to develop a model that will treat both the UT and the LS in a satisfactory manner. Historically, two sets of models have been used to evaluate aircraft impact in the UT and LS. This approach is clearly unsatisfactory because a portion of the subsonic fleet operates in the LS. In this report, we essentially continue to use this approach.
A large number of model studies of the impact of NOx emissions from subsonic aircraft have been performed over the past 20 years (see Chapter 2 for an overview). During the past few years, these studies have been based on 3-D CTMs. Recent assessments of the atmospheric effects of aircraft emissions were completed by the National Aeronautics and Space Administration (NASA) (Friedl, 1997) and the European Community (Brasseur et al., 1998). For these reports, 3-D CTM studies of the ozone perturbation from the present-day aircraft fleet were performed with several models. The model studies used the NASA database in Friedl (1997) and the Deutsches Zentrum für Luft- und Raumfahrt (German Aerospace Center) (DLR)-2 (Schmitt and Brunner, 1997) database in Brasseur et al. (1998). Although there are clear differences in the calculated perturbations caused by aircraft emissions, all model calculations show significant increases in NOx concentration in the UT (up to 100% above those calculated without aircraft) in the latitude band where traffic is most frequent (30-60°N). Corresponding increases in O3 concentration in the UT are up to 10% above those calculated without aircraft. Comparisons revealed significant differences in calculated O3 perturbations among models, both in magnitude and in seasonal variation.
A projected fleet of high speed civil transport (HSCT) aircraft would fly at supersonic speeds in the LS at altitudes where stratospheric O3 concentrations are large and particularly vulnerable to emissions from these aircraft. The effects of NOx and H2O emissions from this projected fleet on stratospheric ozone were thought to be most important when NASA's Atmospheric Effects of Stratospheric Aircraft (AESA) program started in 1989 (Prather et al., 1992). Projected changes in the O3 column as a result of aircraft emission of NOx and H2O from six 2-D models were presented by Stolarski et al. (1995) for various scenarios of fleet size and EI(NOx) in g NO2 kg-1 fuel. The model predictions in Stolarski et al. (1995) showed generally that supersonic fleet sizes of 500-1,000 aircraft would result in some depletion of Northern Hemisphere averaged total column O3. More recently, the role of SO2 aircraft emissions has been studied carefully and found to have a potentially major influence on resultant O3 perturbation computed in models (Weisenstein et al., 1996, 1998).
The new model studies in this chapter focus on the impact on atmospheric chemical composition from a current and future fleet of subsonic aircraft flying in the UT and LS, and include a fleet of supersonic aircraft flying in the LS in one of the technology options. In the case of subsonic traffic, the estimated impact on atmospheric composition is based on 3-D CTMs; in the case of supersonic transports, the estimated impact is based on a combination of 2-D and 3-D CTMs.
Table 4-1: Description of models used in the subsonic assessment.
This chapter has many ties to other chapters in this document. The aircraft fleet emissions used in model computations for current (circa 1992) and future (roughly 2015 and 2050) aircraft are described in Chapter 9. A discussion of the validity of models used here to accurately represent the present atmosphere is contained in Chapter 2. Information about the interaction between aerosols and aircraft emissions is presented in Chapter 3. Model computations of the distribution of a passive tracer emitted according to aircraft fuel burn are also presented in Chapter 3, and are used here to generate soot distributions by scaling to EI(soot). This tracer experiment is also used to evaluate the different transport characteristics of the models with respect to stratosphere-troposphere exchange and upper tropospheric mixing. Results presented in this chapter are incorporated in Chapters 5 and 6.
An evaluation of the effects of subsonic and supersonic engine effluent on atmospheric trace constituents is presented in Sections 4.2 and 4.3, respectively. A discussion of uncertainties in model results is given in Section 4.4, and a discussion concerning the selection of model simulations used for some of the computations in Chapters 5 and 6 is presented in Section 4.5.
Other reports in this collection