TABLE OF CONTENTS
1. Summary
2. Program Rationale and Background
2.1 Introduction
2.2 Background
2.3 Previous
Programs
2.4 General
Meteorological Factors
2.5 Trace
Constituent Observations Relevant to TOPSE
2.5.1 Ozone
2.5.2 Volatile Organic Carbon/CO/Peroxides
2.5.3 HOx and other radicals
2.5.4 Reactive nitrogen
3. Model Simulations and Meteorological Perspectives
3.1 Chemical
Box Models
3.2 Global
3-D Model
3.2.1 Distributions
3.2.2 Idealized tracer studies
3.3 Additional
Meteorological Considerations
5. Measurement and Modeling Priorities
5.1 Airborne
in-situ measurements during TOPSE
5.2 Airborne
LIDAR Measurements during TOPSE
6. Experiment Design and Observational Requirements
6.1 Overview
6.2 Aircraft
Instrumentation
6.3 Airborne Lidar
7. Relationship of TOPSE to other programs
7.1 Satellite
Observations
7.2 Other
Studies
8. References
One of the major research emphases in the Atmospheric Chemistry Division at NCAR has been the study of the photochemical and dynamic processes that determine the rates of formation and loss of oxidants throughout the atmosphere. Oxidation processes shape the chemical composition of the atmosphere through transformations which act on the large variety of trace gases emitted from natural and anthropogenic sources. Since these processes affect the distribution and trends of radiatively important trace gases in the atmosphere, there is a potential feedback between chemistry and climate which has implications for future global change.
The proposal described here is a continuation of measurements and modeling research in NCAR/ACD which has its roots in the Mauna Loa Observatory Photochemistry Experiments (MLOPEX 1 and 2). The MLOPEX studies were designed to probe photochemical processes during four seasons in the remote marine troposphere over the Pacific Ocean [Ridley and Robinson, 1992; Atlas and Ridley, 1996]. Because of the remote location of the MLOPEX site, measurements and modeling from the MLOPEX studies focused on "aged" air masses and "middle-aged" photochemical processes. Furthermore, we found that valuable information was gained by making measurements that spanned several seasons. Here we propose to extend the MLOPEX approach in order to examine the photochemical processes and relationships in the northern mid-latitudes over continental regions during the winter/spring transition. This places the proposed experiment at a time and location so that we can evaluate the rapid onset and early stages of active photochemistry in the northern hemisphere.
The experiment described here is called Tropospheric Ozone Production about the Spring Equinox (TOPSE). The overall goal of the TOPSE experiment is:
To address the overall goal of TOPSE, a range of specific questions has been formulated. Some of the specific questions that will be examined during TOPSE are:
1) What are the rates of in-situ ozone production and loss as a function of latitude and season in the mid-troposphere?
The rates of ozone photochemical production and loss are a function of photolysis rates and radical reactions. Appropriate measurements of radical species, nitrogen oxides, and radiation will allow calculation of net ozone production throughout the experimental period. Emphasis of the in-situ measurements will be on the mid-troposphere near 6 km.
2) How are radical species and reservoirs distributed over the northern hemisphere mid-troposphere as a function of latitude and how do these reservoirs evolve over the winter spring transition period?
This question is clearly linked to question 1 dealing with photochemical oxidant evolution during TOPSE. Measurements of radicals and reservoir species spanning the period of the spring equinox also provide a stringent test of photochemical models of the troposphere.
3) What are the major sources and reservoirs of nitrogen oxides in the northern mid latitude free troposphere and how do these sources and reservoirs change from winter to spring?
Understanding the sources, interactions, and transformations of reactive odd-nitrogen is critical to unraveling the rates of ozone production and loss. Related to this question are the following:
a) Is PAN (and
other organic nitrates) the major source of NOx
in the spring?
b) Does the
chemistry of NO3 radical have
a significant impact on the chemical composition of the wintertime free
troposphere?
c) What is
the impact of heterogeneous chemistry on the partitioning of
NOy and
on subsequent photochemical production?
4) What is the composition and seasonal evolution of volatile organic carbon (VOC) species in the northern mid-latitudes, and what is their significance to a springtime ozone maximum?
The composition of volatile organic trace gases can provide significant clues to the origins of airmasses in the TOPSE region. Changes in the composition of VOC also can be used to diagnose oxidation mechanisms and rates over the winter /spring transition. Furthermore, adequate characterization of VOC is necessary to evaluate radical sources and contributions to ozone production.
5) What is the potential role of stratospheric influx of ozone and nitrogen oxides on regional tropospheric chemistry, and how does this role change as a function of latitude and season?
This question is to be addressed by a combination of in-situ and remote sensing measurements that will be analyzed in the context of regional and global scale chemical transport models. Understanding the actual distribution and variation of mid-tropospheric ozone is, of course, a critical element in addressing this question.
To adequately answer these questions will require a large suite of sophisticated measurement and modeling capabilities which are currently available, or are under active development, in the atmospheric chemistry community. Furthermore, it is clear that collaboration of investigators from within and from outside of NCAR will be necessary for successful completion of a TOPSE research program. We intend in this document to describe an outline of the TOPSE project in order to: 1) solicit input from the atmospheric chemistry community regarding formulation of the TOPSE plan, 2) identify those interested in measurements or models related to TOPSE, and 3) promote collaboration with related on-going and planned research programs, both nationally and internationally.
The document to follow outlines
the rationale behind the formulation of the TOPSE experiment and reviews
some of the previous measurements relevant to TOPSE. Science objectives
are outlined (as above), and an experimental plan is proposed to meet these
objectives. As already noted, input from potential collaborators is desired
at this stage to optimize the proposed research and implementation plan.
2. Program Rationale and Background
Tropospheric ozone plays a central role in the oxidative chemistry of the troposphere, has an important impact on the radiative balance of the atmosphere, and is known to have detrimental effects on human health and agricultural crop production. For these reasons, understanding the processes that control the origin, trends, distribution, and effects of tropospheric ozone have been high priority themes to atmospheric chemistry research over the last few decades. Remarkable progress toward understanding tropospheric ozone has been made through a combination of remote sensing, in-situ studies, and chemical and transport modeling studies. Remote sensing studies have outlined broad scale and seasonal patterns in tropospheric ozone [e.g. Fishman and Brackett, 1997], while in-situ measurements of photochemically related radicals and ozone precursors focus on testing and constraining models of the chemical and dynamical processes in the troposphere which control ozone production and loss (e.g., MLOPEX, NARE, NASA PEM West, PEM Tropics, TRACE, etc.). Current development of sophisticated global-scale, three-dimensional chemistry transport models are now being used to synthesize, simulate, and predict ozone behavior throughout the global troposphere (e.g. GFDL, MOGUNTIA, MOZART, NASA/GISS, etc.).
As will be discussed in this proposal, the processes that control the concentrations and distribution of tropospheric ozone are a complex interplay of chemistry and dynamics. The sources, transformations, and interactions of radical and reservoir species in the HOx, NOx, and organic carbon families, as well as ozone itself, play a role in determining local rates of ozone production and loss. In addition to strictly chemical and photochemical processes, ozone distributions are affected at least as significantly by atmospheric dynamics. Ozone distributions are affected by direct advective transport of ozone and reactive precursors through the troposphere, by stratosphere/troposphere exchange process, and through convective processes which rapidly transport ozone and precursors to different environments of radiation, temperature, humidity, aerosol loading, etc.
A feature of tropospheric ozone that is thought to have significant contributions from both chemical and dynamic origins is the widely observed springtime maximum of ozone in the Northern Hemisphere mid latitudes. The origin of this maximum has been attributed to increased input of ozone from the stratosphere to the troposphere, but it has also been suggested that tropospheric photochemical processes are the major factor in producing the maximum. Model studies now predict that the timing and relative contributions of stratospheric input and photochemical production to the springtime ozone maximum in the Northern Hemisphere will depend on the specific location (e.g., Wang et al, 1998). Though a great deal of progress has been made in improving our understanding of tropospheric ozone, it is fair to say that a range of scientific issues relating to the seasonal maximum in tropospheric ozone are not completely understood. Resolution of the problem will require major advances in understanding stratosphere-troposphere exchange as well as an improved understanding of the rates, processes, and interactions of reservoirs and radicals involved in tropospheric oxidant balance.
The investigation proposed here is designed to combine model studies and simulations with a set of chemical and photochemical measurements taken over the critical winter-spring transition in the northern mid-latitude troposphere. The anticipated outcome will be a better understanding of the primary photochemical and dynamic processes that control the budgets of radicals and reservoirs in the free troposphere. Through appropriate comparisons of measurements and model predictions, the TOPSE experiment will provide important tests and constraints of photochemical models and will advance the development of global simulations of tropospheric ozone.
This document will first review some of the relevant observations and conclusions obtained from past measurements. Next, results from preliminary model studies done at NCAR, using both a 0-D chemical box model and a 3-D global CTM, will be presented that will provide a theoretical perspective on chemical and transport aspects of the proposed study. We then discuss the major scientific objectives of the TOPSE experiment and present an experimental plan to meet these objectives. Relationship of TOPSE to other observational programs (both from aircraft and satellites) will also be discussed. (Section 7).
The basic idea for the TOPSE study emerged from recalling a simple experiment conducted over 10 years ago at Harwell, UK, where PAN was measured at the surface and over the seasons [Penkett and Brice, 1986]. By filtering the PAN data to "background" conditions based on simultaneous halocarbon measurements, they obtained the result shown in Figure 2.2.1. They concluded that the spring maximum in "background" PAN was the result of springtime photochemistry and that part of the spring maximum in O3 in the NH was also the result of net in situ production and not just from stratospheric injection.
Figure 2.2.1
PAN concentrations between January
1980 and April 1991. CFCl3 < 198 pptv. CH3CCl3
<180 pptv [Penkett and Brice, 1986]
Their conclusion was also reinforced at the time by a few high latitude measurements of hydrocarbons that showed a build-up in the winter over northern Norway [Hov et al., 1984] (see below). Indeed, later measurements from a variety of programs confirmed higher concentrations of mostly anthropogenic hydrocarbons in the high latitude troposphere in winter. An example from aircraft measurements made mostly just north of Ireland is shown in Figure 2.2.2 [Penkett et al., 1993]. The maximum in ozone lags the maximum in NMHC and is thus qualitatively consistent with enhanced photochemistry contributing to the spring maximum in O3. However, the lag indicated in this figure is based on relatively few measurements. The behavior involving enhanced hydrocarbon levels has since been called a "springtime bloom" in photochemical activity. Of course, at least for ozone and PAN production, this spring bloom in activity assumes that active nitrogen and radical species were present in sufficient quantities for efficient O3 production; however, no simultaneous measurements of NOX and radicals were made during their study. It is possible, also, to have a spring bloom in oxidation without significant ozone production. That the spring enhancement in PAN at Harwell could have resulted from southward transport of elevated levels of PAN from high latitudes in winter was not discussed (see Section 2.5 and 3.2).
Figure 2.2.2
a. Seasonal variation in
the availability of reactive carbon in non-methane hydrocarbons over the
North Atlantic Ocean [Penkett et al., 1993].
b. Seasonal variation of free
tropospheric ozone over the North Atlantic Ocean [Penkett et al., 1993].
There have been a number of aircraft
and ground based programs conducted at higher latitudes over most seasons.
These programs, with the exception of standard ozone-sonde stations and
some individual studies (discussed later), are shown in Table 2.3.1. Results
from some of these studies are reviewed in Section 2.5. To be noted here
is the significant lack of measurements in the middle troposphere during
the winter/spring transition.
Table 2.3.1. Atmospheric
chemistry research programs with focus on Arctic or North-mid latitudes.
PROGRAM
LOCATION
SEASON/YEAR
Aircraft platforms
The mean synoptic meteorology and its seasonal change is a driving force behind the accumulation of some trace species at higher latitudes during winter and early spring. [Barrie, 1986; Sturges, 1991]. A more or less slow but steady input of species into a region of slow to insignificant removal processes causes the accumulation of species in latitudes north of approximately 60°N. The subsequent processing of these precursors may enhance ozone production rates in spring as insolation increases.
In late winter, conditions favor long range transport from industrial/populated regions of the NH to the Arctic (Figures 2.4.1 and 2.4.2). The early AGASP programs showed convincingly that the major sources of the accumulation over the Arctic were transport from Eurasia and to a lesser extent North America. Additionally in winter, the lower troposphere over the region is stable, cold (and therefore of low absolute humidity), and wet and dry deposition to snow and ice are much reduced for some species. Near stagnant conditions exist over much of the Arctic and parts of the Subarctic. Barrie [1986] described the January situation as a "dome 7-8 km deep with shallow tongues of air 4-5 km deep spilling southward over the landmasses" AGASP showed that much of the Arctic winter haze phenomenon was below 5-6 km, was very layered and horizontally inhomogeneous, and covered much of the Arctic basin.
In summer, the Arctic air mass is confined to higher latitudes beyond most major industrial/anthropogenic sources. Precipitation is more frequent with the advection of warmer air over the ice and the formation of low level stratus clouds. Thus removal processes are more important, insolation is low but near continuous, and major sources of pollution are well south of the Arctic front. Thus the Arctic region is much less stagnant and "cleaner", at least near the surface. Aerosol loadings in summer are 20-40 times less than in winter [Barrie, 1986].
During the winter/spring Polar Sunrise Experiment, synoptic meteorology and long-range transport events were significant influences on the trace gas chemistry [Hopper and Hart, 1994]. Many trace gases and aerosol components (including methane, CO2 and black carbon) were found to be highly correlated during the dark period of winter as a result of long range transport events from common sources, mainly combustion (populated) sources. With "sunrise" the chemical variability was much less, and this marked the weakening of the long-range transport events and the recession of the polar front to higher latitudes. Hopper et al. [1994] also noted that the quieter spring period (March) coincided with the dissipation of the polar vortex in the upper atmosphere.
Figure 2.4.1 Schematic representations
of near-surface air flow during air pollution transport events [Raatz,
1991]. Left: to Norwegian arctic a) from European source, b) from European/Soviet
source; Center: to Canadian Arctic a) from Eurasia, b) from N. America;
Right: to Alaskan Arctic a) from Eurasia, b) from Europe.
Figure 2.4.2 Figure taken from Barrie and Hoff [1984] showing the mean position of the position of the arctic front. Note that the front is elongated and extends south to nearly 40o N over Asia and North America.
Figure 2.4.3 Examples of trajectories during Summer ABLE 3A, B missions [Shipham et al., 1992, 1994].
In the Subarctic the meteorological
situation is governed by the usual 3-7 day synoptic scale changes. For
example, during the ABLE 3 missions, the mid-lower troposphere (M-LT) region
was sometimes influenced by air masses having trajectory origins over the
remote North Pacific Ocean (cf. Figure 2.4.3). Stratospheric intrusions
were also frequently encountered during the summer ABLE missions. The tundra
and boreal forest regions become a strong sink of ozone (and other species
such as HNO3) via deposition.
Biomass burning plumes were a significant perturbation to background photochemistry
especially during ABLE 3B/NOWES.
2.5 Trace Constituent Observations Relevant to TOPSE
Because of its central role in atmospheric chemistry and climate, a relatively large number of ozone observations have been made from aircraft, balloon-borne sensors, and satellite platforms. Satellite observations of the tropospheric ozone column demonstrate the general seasonal behavior of ozone [Plate 1 from Fishman and Brackett, 1997]. At northern mid latitudes, the onset of spring coincides with an increase in tropospheric ozone column of 10 - 15 DU. To examine this seasonal effect on different altitude levels, Figure 2.5.1 shows a compilation of the average seasonal variation of ozone at various pressure altitudes from ozone-sonde observations made over stations located mostly within central North America [Hauglustaine et al., private communication]. (More discussion of the ozone seasonal trend is provided in Section 3.2). Several points arise from Figure 2.5.1:
2. The seasonal variation at the
surface is not always the same as for the lower to middle troposphere (L-MT,
altitudes less than about 6-7 km or less than ~400 mb, or altitudes that
the C130 aircraft can access). The distinction between the surface
versus the L-MT variations is important.
Differences between the surface and the L-MT will also be seen to hold
for some other trace species.
shown above: Alert, (82 N, 52 W) Pressure levels: (Clockwise from top left): 900, 800, 600, 400, 200, 300, 500, 700 mbar.
shown above: Resolute, (75 N, 95 W) Pressure levels: (Clockwise from top left): 900, 800, 600, 400, 200, 300, 500, 700 mbar.
shown above: Goose Bay, (53 N, 80
W) Pressure levels: (Clockwise from top left): 900, 800, 600, 400, 200,
300, 500, 700 mbar.
shown above: Boulder, (40 N, 105 W) Pressure levels: (Clockwise from top left): 900, 800, 600, 400, 200, 300, 500, 700 mbar.
Figure 2.5.1 Compilation of the average seasonal variation of ozone at various pressure altitudes from ozone-sonde observations made over stations located mostly within central North America. These plots were provided by D. Hauglustaine using J. Logan’s (Harvard) archive.
Some of the surface and airborne measurement programs related to seasonal ozone budgets are briefly discussed next.
EUROTRAC/TOR. The Tropospheric Ozone Research (TOR) program was established as part of a collaborative European effort to improve the understanding of ozone over the European continent. Solberg et al. [1997] used data of several years of surface ozone measurements from monitoring sites extending from 58o N to 79o N. These data were sorted by integrated NOx emissions into "background conditions" and "anthropogenically influenced conditions". The average seasonal cycle of ozone under "background conditions" showed a maximum in March/April at all sites. These same sites also show a maximum in the average seasonal cycle of C2-C5 alkanes, alkenes, and alkynes during late January/early February [Solberg et al., 1996]. At Mace Head, Ireland (53o N), measurements performed 1990-1994 reveal a spring maximum in both O3 and CO under unpolluted conditions [Simmonds et al., 1997]. At this location, the average spring maximum in ozone is 7-8 ppbv above the 34.8 ppbv mean value for the 4-year measurement period. Bazhanov and Rodhe [1997] also observed the spring maximum at another TOR station located at Areskutan, Sweden (63.4o N). Using back trajectory calculations and 7Be tracer measurements, these authors found that (at this location) the stratospheric contribution to the ozone budget is at its maximum in springtime and approximately equals the net photochemical production of ozone production in the air mass. The results still leave ambiguity as to whether the source of the spring maximum is mostly due to the effects of stratospheric/tropospheric exchange or photochemical production from precursors that accumulate during the winter.
ABLE 3B. Some studies of the ozone budget in the TOPSE geographical region were carried out as part of the NASA GTE/ABLE missions, though these focused on the summer season. The purpose of ABLE 3B was to characterize the chemistry and dynamics of the troposphere in Subarctic regions of eastern Canada during the summer of 1990. In addition to examining the photochemical conditions influencing production and loss of tropospheric ozone, objectives included characterizing the influence of regional biosphere-atmosphere exchange, determining the importance of regional biomass fires and of long-range transport of anthropogenic pollution as a source of trace gases to remote eastern Canada. ABLE 3B characterized five factors which influence regional ozone production, loss and concentrations: (1) stratosphere-troposphere exchange, (2) emissions from local and distant biomass burning, (3) transport of industrial pollutants, (4) biospheric emissions of ozone precursors, and (5) long-range transport of tropical air to northern high latitudes [Harriss et al., 1994].
Stratosphere to troposphere exchange was identified as an important influence on ozone observed in the ABLE 3B region. Remote sensing of ozone using a differential absorption LIDAR (DIAL) indicated that, along the aircraft flight tracks, over one-third of the troposphere had ozone enhancements due to stratospheric intrusions [Browell et al., 1994]. Meteorological analyses of the region indicated that stratospheric air was present at mid-tropospheric altitudes during 40% of the ABLE-3B period [Bachmeier et al., 1994]. In situ measurements found an increase of ozone mixing ratios with altitude, which suggested stratospheric influence [Anderson et al., 1994]. However, frequency of detection of stratospherically influenced air does not provide a quantitative measure of the relative importance of irreversible transport from the stratosphere in the tropospheric ozone budget. Mauzerall et al. [1996] went beyond these analyses and concluded that stratospheric influx of O3 is of secondary importance relative to in-situ production of O3.
The data obtained during the ABLE programs do not directly address the issue of the sources of ozone during the winter/spring transition. Certainly, at high latitudes in winter in the M-LT (no solar insolation), O3 must be maintained by import to the region from the stratosphere and/or lower latitudes. O3 disappears only episodically at high latitudes at or near the surface (< about 500 m) with the reappearance of sunlight in spring and with production of halogen radicals (e.g., Br/BrO, Cl/ClO) from precursors generated in the marine environment. No aircraft or other program to date has been able to quantitatively determine the ratio of stratospheric input to in situ production over a larger scale in the troposphere, and the direct measurements that relate to the problem are notably sparse. Large-scale models will likely provide better answers as their spatial resolution is increased.
PEM-WEST B. Results from the GTE PEM-WEST B study (February-March 1994) establish that net ozone production can occur at extratropical latitudes at low altitudes even under wintertime conditions [Crawford et al., 1997b]. The NASA DC-8 covered a large latitudinal range along the western Pacific rim with the objective to study chemical processes and long range transport of atmospheric trace species over the north-west Pacific Ocean and to estimate the magnitude of human impact over this same region [Hoell et al., 1997]. Ozone concentrations were sampled at latitudes of 10o S to 50o N and altitudes of 0-12 km. The extratropical component of the data set (20o N to 50o N) revealed two distinctly different photochemical environments north and south of 30°N. A marked decrease in the altitude of the tropopause near 30 o N and a simultaneous increase in total ozone column density defined this distinction. This difference in overhead ozone resulted in highly elevated values of j(O(1D)) for 20-30o N relative to 30-50oN. Diurnally averaged O3 photochemical production and destruction rates for 20-30o N exceeded those of 30-50o N by factors of 3-13 and 4-14 respectively. Despite these large differences, it was found that both regimes exhibit net photochemical production of ozone at all altitudes. For altitudes < 2 km, net ozone production in these latitude regions was estimated to be approximately equal. The low altitude outflow of continental emissions (as was apparent from observed elevated levels of NO, CO, and numerous NMHCs [Talbot et al., 1997]) was a critical factor in creating conditions for net ozone production in both regions. NO levels at altitudes < 2 km were measured to be 20-30 pptv in both regions. For the 30-50o N region however, these levels of NO are of 20 pptv in excess of the calculated value of NOcritical (the value of NO where ozone production and destruction balance) whereas these levels are only 4 pptv in excess in the 20-30o N region. While most of the low altitude (< 2km) NOx observed is thought to be anthropogenic in origin, Crawford et al. [1997b] show that recycling of NOx from PAN could contribute as much as 50%.
2.5.2 Volatile Organic Carbon/CO/Peroxides
Non-Methane Hydrocarbons/CO.
It is well established from ground and aircraft programs that many hydrocarbons and CO have a seasonal maximum in winter/early spring in the Arctic and in the Subarctic (AGASP, AASE, TROPOZ, PEM, ABLE, NOWES, TOR, [Beine et al.,1996]) throughout the vertical extent of the troposphere. This accumulation is in addition to the "normal" latitude gradient found on average in the Northern Hemisphere. In winter, the accumulated hydrocarbons in the Arctic are mostly of anthropogenic origin. From ratios of certain hydrocarbons, the Polar Sunrise Experiment group has inferred a more important role for Br compared to Cl atoms in catalytically removing ozone from the surface layer (e.g., Jobson et al., 1994b). Penkett et al. [1993] have shown a possible role for NO3 oxidation of some hydrocarbons from unexpected hydrocarbon ratios. This oxidation process could be a nighttime/wintertime source of HOx/ROx and subsequent oxidation products. Further examples of the seasonal variation of some hydrocarbons and CO are given in Figures 2.5.2 - 2.5.4. Figure 2.5.2 gives data for some particular hydrocarbons from the same data set shown in Figure 2.2.2. Surface measurements in northern Ontario and northern Saskatchewan show the same trends (Figure 2.5.3). Results from the winter AASE II DC-8 flights (Figure 2.5.4) show also that the trend is preserved at altitudes above 5 km.
Figure 2.5.2 Seasonal variation of some hydrocarbons collected over the North Atlantic [Penkett et al., 1993].
Figure 2.5.3 TOP Seasonal variation of some hydrocarbons collected
in northern Saskatchewan [Jobson et al., 1994]. BOTTOM. CO
mixing ratios measured at Barrow, AK [Novelli et al., 1994].
Figure 2.5.4 Results of various hydrocarbon measurements obtained
at altitude > 5 km during AASE II [Anderson et al., 1993].
Peroxides, Carbonyls, other VOC.
Peroxides and carbonyls are considered together here because both are chemical intermediates and radical reservoir species that can influence rates of ozone production. Comparison of predicted to modeled concentrations of these species also can provide significant diagnostic information regarding proposed chemical mechanisms. One interesting example of this was shown for surface ozone depletion events during Polar Sunrise. At the surface at Alert during the 1992 Polar Sunrise Experiment, de Serves [1994] measured CH2O and total peroxides. In the winter (dark) period of the study peroxides were low (as expected) at 10 - 40 pptv. In the sunlit period (about April 1 to end of study April 15) mixing ratios of peroxides increased to the 100-400 pptv range. Peroxides were near zero in ozone depletion events. Sigg et al. [1992] measured peroxides in Greenland in summer at the surface and found values near 1500 pptv, which is consistent with the increase measured by deServes [1994] at Alert in early spring.
However, the behavior of formaldehyde was observed to be different from earlier PSE studies. In the 1988 study, Bottenheim et al. [1990] reported that CH2O was less than 40 pptv usually, and acetaldehyde was near 65 pptv. The early studies observed that CH2O (and acetone) was low during surface ozone depletion events; low formaldehyde was initially predicted as a necessary consequence of Br atom reactions involved in the ozone depletion. However, de Serves [1994] observed the opposite effect: formaldehyde was actually enhanced in ozone depletion events. This observation suggested that heterogeneous processes are a necessary part of the chemical destruction mechanism for ozone if Br atoms remain involved. In the dark period of the study, CH2O ranged from 100-700 pptv and correlated well with methane and other species, as might be expected for the episodes of transport to the region in winter. de Serves [1994] estimated that ozone/alkene reactions can account for CH2O production since in the dark the lifetime of formaldehyde is long (months). Rather large values of ethene and propene were measured in the winter/dark period. In the sunlit portion of the experiment CH2O was in the range of 30-600 pptv (not very different from the dark period). Correlations with other species like methane and CO2 were not present.
In the free troposphere, formaldehyde and peroxide were measured during the TROPOZ II flights that covered high latitudes in Jan. 1991 [Arlander et al., 1995; Perros, 1993]. The aldehyde results are from a cartridge technique that agreed reasonably well with on-board TDL measurements. In the L-MT region near 50°N HOOH was 1-3 ppbv and near 1 ppbv most of the time. Formaldehyde ranged between 100 and 200 pptv.
Other carbonyl and oxygenated hydrocarbon measurements are less frequent at high latitudes. Singh et al. [1995] from the PEM-West DC-8 missions in February and March over the North Pacific show an increase of acetone and several alcohols with increasing latitude up to about 40°N (Figure 2.5.5). Whether this trend persists to higher latitudes is likely for acetone but is debatable for the alcohols. At the surface, acetone was measured at Alert during one of the Polar Sunrise studies, and C1 - C3 carbonyls were measured at several surface sites in Europe [Solberg et al., 1996]. Acetone measured at Alert was comparable to levels found at northern latitudes over Norway (@400 pptv in winter). During the winter/spring transition over Northern Europe, carbonyls begin to show a dramatic increase with a maximum occurring in early to mid-summer. At Birkenes, Norway, total C1 - C3 carbonyls range from about 1 - 2 ppbC in winter to about 5 - 7 ppbC in May to June. These measurements indicate that C1 - C3 carbonyls represent from 20 - 50% of the total VOC load during winter and spring. Carbonyls as a source of HOx/ROx at altitudes that can be sampled by the C-130 may be less significant than in the upper troposphere. However, at high latitudes in early spring the low insolation and low absolute water will minimize OH production from O3 photolysis compared to low latitude, summer conditions, and other radical sources may be important.
Figure 2.5.5 Latitudinal acetone and methanol distributions obtained
in the 5-10 km altitude range during PEM-West B [Singh et al., 1995].
Measurements of HOx
from aircraft in the region have been obtained in the upper troposphere
and lower stratosphere as part of the recent POLARIS and SONEX campaigns.
Neither of these programs focused on the winter/spring seasons or on the
M-LT region. Hausmann and Platt [1994] measured BrO at the surface during
the 1992 Polar Sunrise Experiment at Alert. Levels as large as 17 pptv
were measured during depletion events. The lack of relevant radical measurements
in the TOPSE region leads to a large uncertainty in evaluating rates of
in-situ ozone production and loss.
NOX, NOY, HNO3, NO3, PAN, Other Organic Nitrates.
Along with HOx species, the sources, transformations, and mixing ratios of reactive odd-nitrogen are among the most critical factors in contributing to ozone production. In winter and spring most of the data are restricted to surface measurements (Polar Sunrise Experiment {PSE}, at Alert or locations in Alaska [Honrath and Jaffe, 1990, 1992; Jaffe et al., 1991]. Reactive nitrogen observations from year to year are variable but there are some general trends relevant to the TOPSE experiment. Airborne measurements in winter and spring over the continent have also been made, but with an emphasis on the upper troposphere and lower stratosphere (AASE 1 & 2, TROPOZ/STRATOZ, TOTE/VOTE, SONEX). L-MT sampling is sparse in winter/spring except for the early measurements of Dickerson [1985]. In contrast, the ABLE 3 and ABLE/NOWES programs have provided a wealth of measurements over the continent, but in the summer season. Because of the central role of nitrogen oxides to ozone chemistry, we describe in some detail earlier observations relevant to TOPSE.
a) Surface measurements at Alert, NWT (82.5°N) in winter/spring. Measurements from several campaigns at Alert in winter and spring were focused on understanding episodes of surface (< 500m) depletion of O3. These measurements have provided an interesting picture of odd nitrogen partitioning that is in some respects quite different from normal mid-latitude spring/summer/autumn chemistry. Specifically, PAN and other organic nitrates appear to dominate the NOy budget. Figure 2.5.6 shows the strong winter (~200 pptv) versus summer (<15 pptv) contrast in PAN; in addition to the winter build-up, a stronger maximum occurs in April/May as solar insolation returns and/or synoptic transport changes. PAN is well correlated with sulfate or the occurrence of Arctic haze episodes (Figure 2.5.7). In the latest PSE (1992) much larger PAN mixing ratios of 400-600 pptv were observed but the overall conclusion that organic nitrates were dominant did not change. The major role for PAN is not that different from the lower free troposphere at mid-latitudes. However, the sum of simple alkyl nitrate abundances was roughly 30 - 50% of mean PAN abundances, very different from mid-latitude studies in summer.
2.5.6
2.5.7
Figure 2.5.6 Plot showing the strong contrast between winter and summer PAN levels obtained at Alert during 1988 [Barrie and Bottenheim 1991].
Figure 2.5.7 Plot showing the correlation between PAN and total sulphate at Alert during April 1986 [Barrie and Bottenheim, 1991].
Measurements of NO2 via the luminol technique were generally less than 100 pptv or about half of the PAN abundance in the earlier PSE experiments. Larger abundances were observed during haze episodes and lower values during surface O3 depletion events (it is necessary for NO2 to be low if Br/Cl are responsible for the ozone depletion). Since abundances of NO2 from luminol instruments are usually larger than chemiluminescence /photolysis measurements, the role of NOx (mostly NO2) might even be smaller than implied. For example, the latest PSE showed upper limit mixing ratios of NOx ~50 pptv in low altitude aircraft flights. There are a few filter measurements of nitric acid or of total inorganic nitrate (NO3- + HNO3) from the Alert studies: values were usually less than 45 pptv and also suggest a minor role for HNO3 relative to organic nitrates and NO2. This result is also substantially different from model predictions for lower latitudes and perhaps for high latitudes.
Both the PSE studies and those by [Honrath and Jaffe, 1990, 1992; Jaffe et al., 1991] show a summer minimum in PAN at the surface that does not likely reflect the L-MT region. The ABLE 3 missions and the ABLE/NOWES mission at lower latitudes showed that PAN was usually the dominant species (relative to NOx) in the colder atmosphere above the surface (see later discussion and Figure 2.5.14) and that the L-MT region would not be considered all that "clean" in summer.
The longest record of PAN measurements
(>5 yr ) in North America was obtained by AES Canada [Bottenheim et al.,
1994] in Nova Scotia (~46°N), which would be near the lower latitude
of the planned TOPSE flights. While this location is subject to outflow
from the northeastern U.S., the surface site does show a March maximum
in PAN and O3 (Figure 2.5.8),
but ozone declines much less rapidly than PAN toward late summer or early
autumn. Both O3 and PAN have
highest average mixing ratios and least variability about the average in
late winter/early spring compared to summer.
Figure 2.5.8 Plot of observed PAN and ozone in Nova Scotia (~46o N). Note both ozone and PAN exhibit a maximum in March, yet ozone declines much less rapidly toward late summer or early autumn [Bottenheim et al., 1994].
b) Surface Measurements in Alaska. D. Jaffe and co-workers have made surface measurements at the Barrow CMDL station (71°N), Poker Flats (65°N) and also at Svalbard, Norway (79°N). At Barrow, they see a strong modulation in NOy with maxima of 600-700 pptv in March (similar to Dickerson’s results over northern Norway) decreasing to ~100 pptv in summer. Daytime NO in March was usually around 10 pptv or less but there were pulses to higher values that they attribute to PAN decomposition. At Poker Flats, they observed much larger NOx in spring (median = 90 pptv) that they calculate is large enough for net ozone production of 1-4 ppbv/day; median PAN was 80 pptv. In northern Norway, median NOx was a factor of 4 lower (24 pptv), and PAN was much higher (median = 237 pptv) than in Alaska. They suggest the main difference in PAN was due to lower median temperatures in Norway (-8°C) versus Poker Flats, (6°C). They present an analysis of their data that argues that PAN at northern latitudes in spring can be a significant source of NOx. This is a common argument and one also concluded from the ABLE 3 and ABLE/NOWES aircraft programs. In Beine et al. [1996], median PAN for March to May was 132 pptv and alkyl nitrates were 34 pptv. The alkyl nitrate role was not as strong as measured in the PSE programs. The average alkyl nitrate/PAN ratio was about a factor of two lower than measured in the PSE experiments. PAN also did not decrease as much as did hydrocarbons and alkyl nitrates over the March 15 - May 14, 1993 period.
c) Measurements above the surface in winter/spring. Observations made at the surface or within the surface layer do not usually reflect chemical conditions in the L-MT.
As mentioned earlier, most recent
measurements in winter/spring at high latitudes have been
at the altitude of the upper troposphere and lower stratosphere. One exception
is Dickerson [1985], who first measured NOy
at high latitudes over northern Norway, though it should be noted that
the sampling techniques he used are not necessarily comparable to current
methods for NOy measurement.
Some NOy profiles in March 1983
at high latitudes are shown in Figure 2.5.9. A typical NOy
mixing ratio obtained from the figures presented in Dickerson [1985] for
the lower troposphere (< 2 km) at high latitude is in the range of 700-900
pptv; in the 2-6 km region NOy
is about 400-500 pptv. These levels of NOy
would not be considered a large build-up of reactive nitrogen species compared
to lower latitudes.
Figure 2.5.9 NOy profiles obtained by Dickerson et al. [1985] over Northern Norway in March 1983.
d) PEM-WEST B Experiment. Results from the GTE PEM-WEST B study (February-March 1994) are applicable to the discussion of springtime northern hemispheric conditions [Hoell et al., 1997]. NOx values obtained from flights in the -10°S to 60°N region of the western-Pacific basin are summarized in Figure 2.5.10. The data are grouped into high and low NOx regimes depending on the origin of the air mass. Ten-day back trajectory calculations showed that the high NOx data were from air masses with a continental origin while the low NOx data were from marine air masses. Chemical characterization confirmed the results of these back trajectory calculations [Talbot et al., 1997]. Differences in PAN were observed with median levels above 4 km ranging from 5-10 pptv in the low NOx regime, but reaching 50-65 pptv for the high NOx conditions. At these altitudes, Crawford et al. [1997a] calculate that the observed 148 pptv of HNO3 and 55 pptv of PAN, can only explain 20-30% of the observed NOx. While not ruling out the possibility of an additional pathway for (re)generating NOx, Crawford et al. [1997a] submit that the elevated NOx was the product of lightning and that the NOx had not fully converted to HNO3. For lower altitudes (< 4 km), anthropogenic emissions are thought to be the dominate source of NOx; however model simulations [Crawford et al., 1997b] suggest that recycling of NOx from PAN could have contributed only as much as 20-50% to the NOx at altitudes below 1 km and 20% between 1 and 2 km.
Figure 2.5.10. High and low NOx regime NO altitude distributions. Median values were derived from data binned into 1-km increments from 0-2 lm and 2 km increments from 2-10 km. [Crawford at al., 1996b].
e) AASE I and II, STRATOZ/TROPOZ Experiments. For winter conditions, the TROPOZ II program (conducted January, 1991) is most appropriate to this discussion. The Caravelle jet aircraft covered a large latitude range in both hemispheres mostly along coastlines on each side of the Atlantic. Data was collected in the 50-60°N region on both southbound and northbound flight legs. Results are summarized in Table 2.5.1 and Figure 2.5.11. The NO data are from Rohrer et al. [1997], the formaldehyde data from Arlander et al. [1995], and the peroxide and PAN data are from Perros [1993, 1994]. NO2 measurements were not made during this campaign. CO and other hydrocarbon data are also available, but are very limited or absent above about 45°N. The northbound NO data is fairly consistent with the NOx AASE data but that for the southbound category has NO mixing ratios alone much larger than AASE data below 6 km. Differences in the south and northbound legs were not seen in formaldehyde data but were seen in the peroxide and PAN data. Whether a peculiar synoptic scale transport event occurred is not known. It is also noted that the PAN mixing ratios are lower than would be implied from the PSE results.
Results from AASE II (Norway, Alaska, etc., Feb.-Mar., 1992) are quite comparable to AASE I and Dickerson [1985] studies (Figure 2.5.12, Weinheimer et al. [1994]). Based on data in Figure 2.5.12, in the L-MT region (1.5-3.5 km below the tropopause in the figure) NOx was 30-90 pptv, and NOy was 300-900 pptv.
In summary, one would not conclude from these data that there is much of a build-up of NOy constituents in winter and therefore that any "spring bloom" in ozone production might be very much NOx-limited.
Figure 2.5.11 Plot of NO mixing ratio obtained from northbound and
southbound legs during TROPOZ II [Roher et al.1997].
Table 2.5.11. Results from TROPOZ II (January, 1991) for the 2-6 km altitude region at latitudes north of or ~50°N
| Northbound | Southbound | |
| NO (not NOx) | 20-60 pptv | 200-400 pptv |
| CH2O | 100-200 pptv | 100-200 pptv |
| HOOH | 1 ppbv | 1-3 ppbv |
| PAN | 100 pptv | 40-60 pptv. |
Figure 2.5.12 Results of NOx measurements obtained during AASE II [Weinheimer et al., 1994].
f) Measurements above the surface in summer: ABLE Programs. There is a wealth of information for the summer period covering latitudes from central Canada to Alaska, Newfoundland, and Greenland as a result of the ABLE 3A and ABLE 3B/NOWES aircraft and ground-based programs. In very general terms the results from ABLE 3A and B were similar at least regarding nitrogen speciation. ABLE 3B/NOWES was conducted in summer at lower latitudes over Canada than ABLE 3A and experienced a larger variety of source inputs and a very inhomogeneous (altitude and geographic) atmosphere: biogenic isoprene, "clean" Arctic air, "clean" Pacific air, long range transport from industrial regions to the south, and especially natural biomass fire emissions from Alaska and Canada. (Note that the TOPSE study period is expected to be prior to most natural fires.)
Figure 2.5.13 and 2.5.14 show selected observations from the ABLE 3A summer Alaska experiment. Fire plumes were encountered but less frequently. In contrast to the observations of PAN by Beine et al. [1996] that showed a distinct and low summer minimum (mostly below the detection limit of 15 pptv), it is clear that PAN remains the dominant NOy constituent above the boundary layer. Certainly, the L-MT region would not be considered clean, although NOx levels were quite low much of the time: median NO and NO2 were 8.5 and 25 pptv and roughly independent of altitude [Sandholm et al., 1992]. Bakwin et al. [1992] also showed very low mixing ratios of NOx (10-15 pptv) and NOy (130-200 pptv) at the tundra surface during the experiment and that deposition of HNO3 was the dominant loss of NOy. These NOx values are low compared to some of the observations of Jaffe et al. in summer discussed above, but such inhomogeneity in NOx is not unexpected. NOx from the surface and aircraft during ABLE 3A was sufficiently small for net ozone destruction to occur much of the time.
During ABLE 3-B, biomass fire plumes were found to be an important source of ozone precursors to the region. Biomass fires were estimated to account for approximately 70% of the input to the Subarctic for most hydrocarbons and acetone and more than 50% for CO [Wofsy et al, 1994]. NOx mixing ratios in air impacted by biomass burning were usually < 50 pptv, however HNO3 and PAN were typically 100-300 pptv representing a two to three fold increase over the regional background [Talbot et al., 1994]. As pointed out by Fan et al.[1994] and Singh et al. [1994], dispersal of PAN formed in biomass-burning plumes may serve to sustain background NOx levels in the region and thus contribute significantly to in situ production of ozone. Mean NO mixing ratios within the ABLE-3B region were found to be approximately 10 pptv [Sandholm et al., 1994]. As shown by Mauzerall et al. [1996], due to high ozone production efficiency, this NO level was found to be sufficient to sustain in situ photochemical production of ozone as the largest source of ozone to the region.
Figure 2.5.13 Vertical soundings of various trace species obtained during ABLE 3A [Blake et al., 1992; Harris et al., 1992; Sandholm et al., 1992; ABLE 3A special issue].
Figure 2.5.14 Results of PAN, O3, and NOx/NOy measurements obtained during ABLE 3A [Singh et al., 1992; Talbot et al., 1992; Sandholm et al., 1992, ABLE 3A special issue].
HO2NO2, HONO, NO3, N2O5.
Information on these nitrogen species is generally scarce, but they can be significant in the chemistry of the TOPSE region. Hausmann and Platt [1994] obtained upper limits to HONO from their DOAS measurements at Alert. The measurements ranged from 50-200 pptv. Li [1994] inferred HONO mixing ratios in the dark period of the experiment of up to 70 pptv from nitrite measurements. The studies for PSE note that HONO can compete with O3 and formaldehyde photolysis as the source of OH in the early Polar Sunrise period.
Calculations at the surface suggest typical NO3 of a few pptv and N2O5 of perhaps 50 pptv if one assumes equilibrium conditions. Estimates obtained in the model simulations described in Section 3 suggest different levels of HONO, NO3 and N2O5. Such estimates are quite sensitive to temperature, insolation, and having a more accurate knowledge of the NO2 mixing ratios. No studies have examined in any detail the chemistry associated with "dark" or wintertime oxidation processes.
Much of the earlier work in the Arctic focused on studies of the aerosol sources and loadings (the Arctic Haze phenomenon). Because of this emphasis, the bulk composition, source identification (primarily anthropogenic), size distributions, areal distribution, etc. are reasonably well known for Arctic haze. Formation studies have also been made. On a mass basis the sulfate/sulfuric acid aerosols were dominant (~50%) followed by organic aerosols (~25%), black carbon (10%), and the remainder other substances. Particles in the 0.005-0.2 micron diameter range are most numerous. Few particles above 1-2 microns are haze related. Particles in spring are often more numerous, of larger size than in winter, and there is evidence that the composition changes from mostly sulfuric acid in winter to a higher fraction of ammonium sulfate in spring [Barrie, 1986; Sturges, 1991].
Heterogeneous reactions on aerosols have been suggested as a critical part of the surface ozone depletion in the Arctic, but the role of aerosols in relation to oxidant chemistry throughout the winter/spring transition and in the M-LT region is largely unexplored.
3. Model simulations and meteorological perspectives.
Model simulations provide valuable insight into potential areas of investigation for TOPSE, and they are critical to subsequent data analysis. A variety of modeling approaches and model scales can be applied to examine different aspects of the atmospheric chemical system. For planning the TOPSE experiment, simulations of atmospheric chemistry have been run with both a detailed chemical box model and with a global 3-D chemical transport model. Additional and more detailed discussion of meteorological factors affecting transport and chemistry also is presented below.
Model Description and Initial Conditions.
A box model employing the NCAR Master Mechanism was run for 20 days for various sets of initial conditions that were selected to span ranges of latitude, season, and initial NOx level that are relevant to the TOPSE experiment. The model includes detailed hydrocarbon chemistry with ~1000 species and ~3000 reactions for the present simulations. All simulations were performed for an altitude of about 5 km. Cases were run for 12 combinations of latitude (40°N and 70°N), time of the year (December, March, June), and initial NOx level (50 pptv and 250 pptv). For all runs HNO3 was initialized at 200 pptv and PAN at 500 pptv. Thus the total NOy was 750 pptv for the six low-NOx cases and 950 pptv for the six high-NOx cases; there was no continuous source of NOx in the model, so NOy is conserved. Hydrocarbons were initialized to levels characteristic of high northern latitudes for the given month (the same for both latitudes), based largely on Penkett et al. [1993]. Nonmethane hydrocarbon species include the following with, for illustration, the March initial conditions given in parentheses: ethane (2500 pptv), ethylene (100 pptv), acetylene (790 pptv), propane (1170 pptv), propylene (50 pptv), n-butane (490 pptv), isobutane (266 pptv), 2-butene (10 pptv), n-pentane (166 pptv), isopentane (100 pptv), benzene (140 pptv), and toluene (75 pptv). For all cases, O3 was initialized to 50 ppbv. Temperature was selected appropriate to the altitude and month of a given simulation; it ranged 240-263 K.
Limitations of the box-model approach should be noted at the outset. The lack of a simulation of transport is always a limitation, but especially so in some of the cases modeled here where the characteristic chemical times are so long. For a real air parcel at 70°N, the chemistry may well be dominated by processing that occurs during excursions to more southern latitudes rather than by the processing at 70°N that is reflected in the results to be described here. In other words, some of the chemistry represented here may never be actually be realized, because it is so slow compared to the time scales for transport to latitudes where the chemistry is faster. Nonetheless the time scales are representative of the chemical rates experienced while at 70°N, and so are representative in this sense.
NOy Partitioning.
The NOx lifetimes were mostly in the range of 1-3 days, with conversion to PANs, other organic nitrates, and (often to a lesser extent) HNO3 occurring on this time scale. Nitrophenols were at times a significant sink for the NOx. From about day 2 through day 20 there were slower interconversions among PANs, other organic nitrates, HNO3, and, nitrophenols. For example, for the December, 40°N simulation, the NOx lifetime was about 1 day, and for the low-NOx case, the initial 50 pptv had decreased to 8 pptv after 2 days, leading to the net formation of 16 pptv of PANs, 10 pptv of other organic nitrates, 9 pptv of HNO3, and 6 pptv of nitrophenols. During day 2 through day 20, NOx settled to a level of 3 pptv, while the PANs steadily declined at -2 pptv dy-1 and HNO3 at -0.6 pptv dy-1 and while other organic nitrates increased at +1 pptv dy-1 and nitrophenols at +0.7 pptv dy-1. For March at 40°N, the results are similar, again with an initial 1-day loss rate for NOx, but with a shift in the products at 2 days, favoring more organic nitrates at the expense of HNO3: +16 pptv of non – PAN organic nitrates, +14 pptv of PANs, +6 pptv of HNO3, +5 pptv of nitrophenols. As for December, over days 2-20, there was a decrease in PANs (-0.8 pptv dy-1) and HNO3 (-2.4 pptv dy-1), while there was an increase in organic nitrates (+1.9 pptv dy-1) and nitrophenols (+0.9 pptv dy-1). For March at 70°N, the NOx lifetime is a bit longer, at 2.5 days, but the principal initial products are again primarily PANs and other organic nitrates. In this case, however, for days 5-20, the PANs increased (+0.5 pptv dy-1) rather than decreased, while again there was a decrease in HNO3 (-0.7 pptv dy-1) and an increase in the other organic nitrates (+0.3 pptv dy-1).
After 20 days of simulation, the NOy was predominantly in the form of PANs (40-70% of NOy; 370-600pptv), and HNO3 (20-50%, 160-450 pptv) for all of the runs (Table3.1.1). Other organic nitrates were also generally significant (2-8%, 16-74 pptv), except for the December 70°N simulations in which there was insufficient chemical processing to generate precursors to the formation of organic nitrates. The set of species which interconvert over a diurnal cycle with NOx (NOx + HO2NO2 + 2N2O5 + HONO + NO3) comprised 0.1-5% of NOy (2-42 pptv) at the end of 20 days, with NOx and HO2NO2 being the most significant. The other NOx+ species were small fractions of NOy2N2O5 was 5x10-6 to 2.4%, (< .01 – 23 pptv), NO3 was 10-6 to 2x10-3 (< 01-2 pptv) , and HONO was 10-7 to 2x10-4 (< .01 – 0.2 pptv). The NO3 reached the higher values, 2x10-3 (2 pptv), only in the two June simulations at 40°N. Excluding those cases, the highest NO3/NOy fractions were 8x10-4 (0.8 pptv) for the high-NOx case and 3x10-4 (0.2 pptv) for the low-NOx case in December at 70°N. Next highest NO3/NOy fractions were in March at 40°. Nitrophenols were present at levels of 0.2 to 3% (2-24 pptv). These partitionings did not differ greatly in going from the low-NOx to high-NOx runs. We note the following caveats regarding NOy partitioning calculated by the box model: (i) the partitioning for HNO3 and PAN is heavily influenced by the choice of initial conditions, while the partitioning for the NOx + species is not, (ii) the model lacks significant sinks for non-PAN organic nitrates and nitrophenols, so these may be overestimated, and (iii) there is no heterogeneous chemistry in the model.
To summarize the NOy partitioning, freshly emitted NOx is converted into PANs, other organic nitrates, HNO3, and nitrophenols in 1-3 days. After that initial loss of NOx there are slow interconversions among these latter four groups of species such that, after 20 days, the predominant NOy species are the PANs and HNO3, and there are significant amounts (usually 1-20%) of other organic nitrates, NOx, HO2NO2, and nitrophenols.
Table 3.1.1 For the reactive nitrogen
species, or groups of species, this is a tabulation of the diurnal maximum
values at day 20 for the 12 model runs (‘Lo’ and ‘Hi’ denote low- and high-NOx
cases). Approximate diel steady state is reached (more or less) in about
5-10 days for most of the cases (70°N is the main exception), so these
values are fairly representative for days 5-20. Species amounts are expressed
as a fraction of total NOy, as a fraction of total NOx
(NOx + HO2NO2 + 2N2O5
+ NO3 + HONO), and as absolute mixing ratios (pptv). For NOx
and HO2NO2 there is a significant diurnal cycle resulting
from interchange between the two. As a result, some of the diurnal maxima,
expressed as a fraction NOx+, add to signficantly greater than
100%, with the excess over 100% being comparable to the amplitude of the
diurnal cycle in each of NOx and HO2NO2,
as these are often the dominant NOx+ species.
|
|
|
|||||||||||||||
|
|
|
|
|
|
|
|||||||||||
|
|
|
|
|
|
|
|
|
|
|
|
|
|||||
| NOx+/NOy | 0.40% | 0.38% | 0.66% | 0.66% | 4.6% | 4.4% | 0.7% | 3.0% | 0.20% | 0.22% | 2.8% | 3.4% | ||||
| PANs/NOy | 65% | 61% | 67% | 63% | 49% | 43% | 67% | 53% | 70% | 63% | 69% | 60% | ||||
| HNO3/NOy | 26% | 28% | 22% | 24% | 40% | 47% | 32% | 44% | 26% | 27% | 22% | 30% | ||||
| OrgNitr/NOy | 4.0% | 5.2% | 6.6% | 7.8% | 3.5% | 3.3% | 6E-4 | 0.12% | 2.1% | 4.2% | 2.5% | 2.8% | ||||
| Nitroph/NOy | 1.7% | 2.5% | 1.5% | 1.8% | 1.1% | 0.9% | 0.27% | 0.22% | 0.52% | 1.2% | 0.2% | 0.24% | ||||
| NOx/NOy | 0.40% | 0.35% | 0.61% | 0.61% | 4.30% | 4.20% | 0.24% | 0.58% | 0.12% | 0.13% | 2.1% | 2.6% | ||||
| HO2NO2/NOy | 0.14% | 0.12% | 0.19% | 0.20% | 1.0% | 1.0% | 3E-4 | 8E-4 | 0.11% | 0.12% | 1.3% | 1.5% | ||||
| 2N2O5/NOy | 3E-5 | 4E-5 | 1E-4 | 2E-4 | 0.65% | 0.79% | 0.35% | 2.4% | 1E-5 | 2E-5 | 5E-6 | 1E-5 | ||||
| NO3/NOy | 2E-5 | 2E-5 | 1E-4 | 2E-4 | 0.23% | 0.23% | 3E-4 | 8E-4 | 1E-5 | 1E-5 | 1E-6 | 2E-6 | ||||
| HONO/NOy | 3E-6 | 3E-6 | 9E-6 | 1E-5 | 2E-4 | 2E-4 | 5E-8 | 9E-8 | 6E-7 | 8E-7 | 4E-5 | 5E-5 | ||||
| NOx/NOx+ | 98% | 98% | 96% | 95% | 94% | 93% | 37% | 18% | 73% | 72% | 78% | 77% | ||||
| HO2NO2/NOx+ | 33% | 32% | 30% | 31% | 23% | 24% | 5% | 2.4% | 58% | 57% | 42% | 44% | ||||
| 2N2O5/NOx+ | 0.8% | 1.1% | 2.2% | 3.7% | 15% | 18% | 57% | 79% | 0.65% | 1.2% | 2E-4 | 2E-4 | ||||
| NO3/NOx+ | 0.6% | 0.7% | 1.8% | 2.6% | 5.1% | 5.3% | 5.0% | 2.4% | 0.6% | 0.7% | 3E-5 | 4E-5 | ||||
| HONO/NOx+ | 8E-4 | 8E-4 | 0.14% | 0.15% | 0.40% | 0.40% | 7E-6 | 3E-6 | 4E-4 | 4E-4 | 0.13% | 0.14% | ||||
| NOx+ (pptv) | 3 | 4 | 5 | 6 | 35 | 42 | 5 | 29 | 2 | 2 | 21 | 32 | ||||
| PANs (pptv) | 488 | 580 | 503 | 599 | 368 | 409 | 503 | 504 | 525 | 599 | 518 | 570 | ||||
| HNO3 (pptv) | 195 | 266 | 165 | 228 | 300 | 447 | 240 | 418 | 195 | 257 | 165 | 285 | ||||
| OrgNitr (pptv) | 30 | 49 | 50 | 74 | 26 | 31 | 0 | 1 | 16 | 40 | 19 | 27 | ||||
| Nitroph (pptv) | 13 | 24 | 11 | 17 | 8 | 9 | 2 | 2 | 4 | 11 | 2 | 2 | ||||
| NOx (pptv) | 3 | 3 | 5 | 6 | 32 | 40 | 2 | 6 | 1 | 1 | 16 | 25 | ||||
| HO2NO2 (pptv) | 1 | 1 | 1 | 2 | 8 | 10 | 0.2 | 1 | 0.8 | 1 | 10 | 14 | ||||
| N2O5 (pptv) | 0.01 | 0.02 | 0.05 | 0.11 | 2 | 4 | 1 | 11 | 0.005 | 0.01 | 0.00 | 0.01 | ||||
| NO3 (pptv) | 0.02 | 0.02 | 0.09 | 0.2 | 2 | 2 | 0.2 | 0.8 | 0.008 | 0.01 | 8E-4 | 0.001 | ||||
| HONO (pptv) | 0.002 | 0.003 | 0.007 | 0.01 | 0.1 | 0.2 | 4E-5 | 9E-5 | 5E-4 | 0.001 | 0.03 | 0.05 | ||||
O3 Production
The box model exhibits near-neutral or positive net O3 production rates only for the first few days of the simulations, before the initial NOx has had a chance to decay. After that the O3 tendency is negative. The transition to net O3 destruction usually occurs a couple of days later for the high-NOx runs than for the low-NOx runs. Thus the runs show little in the way of a "spring bloom" in O3 production. This is of course partly a result of the fact that the box model has no continuous source of NOx; there is only the initial infusion of NOx. This indicates that without injections of fresh NOx every few days or so there is not enough NOx released from the reservoir species to sustain net production of O3, at least in the box model. Transport could result in temperature increases which may favor the release of NOx from PAN, but given the fact that PAN is only twice the amount of NOx in the high- NOx simulations, such a source could not sustain net O3 production for too long, certainly not on a seasonal time scale.
Hydrocarbon Lifetimes
Most hydrocarbon lifetimes are governed by reaction with OH. Box-model estimates of mean OH are shown in Figure 3.1.1 for the 12 cases modeled. The expected trends are seen, with more OH in summer than winter. For 40°N, the difference is about a factor of 10, while for 70°N, the difference is more than a factor of 100. Also as expected, 40°N always has more OH than 70°N, with the difference being nearly a factor of 100 in December, but only about a factor of 2 in June. The differences in OH between low- and high-NOx cases are not large, usually no more than 50%. The hydrocarbon lifetimes, estimated from the box model, are given in Table 3.1.2. They show trends with latitude and with season that for the most part correlate with the trends in OH (a notable exception being 2-butene, which is lost via reaction with O3 and so shows no variation). For example, for toluene, which is lost via reaction with OH, at 40°N the lifetimes in the low-NOx runs are 8.0, 2.5, and 1.1 days for December, March, and June. For 70°N, the low-NOx lifetimes are 351, 10, and 1.6 days for December, March, and June. With respect to the toluene lifetime, March at 70°N is comparable to December at 40°N, while in June there is little difference between the low- and high-NOx cases. It appears the duration of the day in June at 70°N offsets the lower sun. As a second example, the lifetimes of propylene, which is lost via reaction with O3 as well as with OH, are 1.2, 0.4, and 0.2 days in December, March, and June at 40°N for the low-NOx cases, while at 70°N, the lifetimes are 7.8, 1.3, and 0.7 days for December, March, and June.
Table 3.1.2. Hydrocarbon
lifetimes (exponential) for the 12 box-model cases, expressed in days,
except when in years as noted.
|
|
||||||
|
|
|
|
||||
|
|
|
|
|
|
|
|
| toluene |
|
|
|
|
|
|
| benzene |
|
|
|
|
|
|
| ethylene |
|
|
|
|
|
|
| propylene |
|
|
|
|
|
|
| 2-butene |
|
|
|
|
|
|
| ethane |
|
|
|
|
|
|
| propane |
|
|
|
|
|
|
| n-butane |
|
|
|
|
|
|
| n-pentane |
|
|
|
|
|
|
|
|
||||||
|
|
|
|
||||
|
|
|
|
|
|
|
|
| toluene |
|
|
|
|
|
|
| benzene |
|
|
|
|
|
|
| ethylene |
|
|
|
|
|
|
| propylene |
|
|
|
|
|
|
| 2-butene |
|
|
|
|
|
|
| ethane |
|
|
|
|
|
|
| propane |
|
|
|
|
|
|
| n-butane |
|
|
|
|
|
|
| n-pentane |
|
|
|
|
|
|
Figure 3.1.1 Latitudinal
and seasonal estimates of mean daily OH for high NOx
and low NOx
simulations.
Heterogeneous Chemistry.
As suggested by the results of
the chemical box model simulations, in the region under study either net
photochemical production or destruction may occur. The overall efficiency
of ozone production will thus be very sensitive to local inputs of NOx,
but also to reactions that tend to favor the repartitioning of NOy
towards NOx. While the photolysis
of HNO3 and the thermal decomposition
of PAN represent gas phase processes that accomplish this repartitioning
there may exist other, heterogeneous, processes that affect the partitioning.
Furthermore, any processes that remove free radicals or reservoir species
that can photolyze to give free radicals will reduce the potential for
ozone formation when sunlight returns. Since the airmasses to be sampled
may have spent long periods in low or zero insolation, the possibility
clearly exists for the slow conversion or removal of reservoir species
on aerosol particles. The extent of this processing is currently not known.
It appears that nitrogen reservoirs such as PAN or HO2NO2
are removed only slowly on cold sulfuric acid particles, while N2O5
is hydrolyzed relatively rapidly, and the mechanism for the formation of
HONO on atmospheric particles has long been the subject of speculation.
Furthermore, the exact nature of the aerosol particles that will be encountered
on the mission is not certain, but they will probably consist of sulfuric
acid on a solid core of anthropogenic origin. Thus, the possibility exists
for a substantial effect on the formation of ozone due to heterogeneous
reactions. Previous airborne experiments carried out at midlatitudes have
not been able to explain the NOx/NOy
ratio very well, and it is clear that in the late winter/early spring the
ozone formation should be very sensitive to this ratio. Contemporaneous
measurements of NOx and
NOy as well as aerosols made
during TOPSE will enable new insight into this problem.
3.2 Global 3-D Simulations Relevant to TOPSE
3.2.1 Distribution of Species in the Northern Mid-to-High Latitudes Simulated by MOZART
MOZART (Model for OZone And Related chemical Tracers) is a global chemical transport model developed in the framework of the National Center for Atmospheric Research (NCAR) Community Climate Model (CCM) and includes a detailed representation of tropospheric chemistry. The model provides the distribution of 56 chemical species at a spatial resolution of 2.8 degrees in both latitude and longitude, with 25 levels in the vertical (from the surface to the upper stratosphere) and a time step of 20 min. In the present version of the model, the meteorological information is supplied from a 2-year run of the NCAR CCM. The version of CCM used in the present study (CCM2, W0.5 library) is intermediate between CCM2 [Hack et al., 1993] and CCM3 [Kiehl et al., 1996]. The MATCH model described by Rasch et al. [1995, 1997] forms the meteorological component of MOZART. The model is described in detail by Brasseur et al. [1998] and evaluated by comparison with observations by Hauglustaine et al. [1998].
In this section, we present results of model simulation and comparison to observations for selected chemical species relevant to the proposed TOPSE experimental regime. Figure 3.2.1a shows the calculated and observed seasonal variation of CO surface mixing ratios for 4 selected mid- and high-latitude stations from the CMDL network [Novelli et al., 1992]. Carbon monoxide exhibits maximum values during late winter (Feb - March). This is particularly visible at high latitudes (Point Barrow, Alaska) where CO ranges from about 100 ppbv in summer to 200 ppbv during winter. Similarly, Figure 3.2.1b shows the ethane and propane seasonal cycle calculated by MOZART and observed at Birkenes (Norway) and Waldhof (Germany) [Solberg et al., 1996]. These hydrocarbons exhibit a high temporal variability and generally a strong seasonal cycle. A build-up of the mixing ratio during winter and a minimum during summer are clearly visible. The model reproduces the observed seasonal cycles for CO, C2H6, and C3H8. The calculated mixing ratios are also consistent with the C2H6 and C3H8 measurements at Poker Flat, Alaska by Herring et al. [1997], and with those discussed earlier in this proposal.
Figure 3.2.1 a. MOZART calculated and observed seasonal variation
of the CO surface mixing ratio for 4 mid - high latitude stations from
the CMDL monitoring network [Novelli et al., 1992].
b. MOZART calculated and observed seasonal variations of ethane
and propane [Solberg et al., 1996].
Figures 3.2.2a and 3.2.2b illustrate the calculated CO distribution in the northern hemisphere at 500 mb in January and July. The CO maximum observed at the surface during winter and shown in Figure 3.2.1 is also visible in the free troposphere. A region of CO mixing ratio larger than 120 ppbv (and reaching 180 ppbv) appears to be confined around the North Pole with a sharp meridional gradient near 60°N. This maximum appears to have a rather uniform structure in longitude. In summer, much lower mixing ratios (generally less than about 100 ppbv) are calculated at high latitudes. This feature (maximum located around the North Pole during winter and spring) is also predicted by the model for CH4, and other hydrocarbons. Results from tracer simulations performed with MOZART in order to provide additional information on the formation of this high latitude maximum in spring are presented in the next section.
Figure 3.2.2 a. Calculated CO distribution in the northern
hemisphere at 500 mb in January.
Figure 3.2.2 b. Calculated CO distribution in the northern hemisphere at 500 mb in July.
Though isoprene is not expected to play a significant role in ozone chemistry during TOPSE, this reactive hydrocarbon may contribute to extending ozone production through the summer months. The calculated surface and 24-hour isoprene mixing ratio is shown in Figure 3.2.3 for summer conditions. Over the continents, the mixing ratio is generally in the range 100 - 500 pptv in the latitude band 40°N-70°N. North of 70°N, the mixing ratio is lower than 50 pptv. In January and March, the model calculated mixing ratios do not exceed 100 pptv north of 50°N. The values provided by the model over Alaska are in the range 1 - 5 pptv in March and 100 - 300 pptv in July. These values are on the low side of the measurements reported by Herring et al. [1997].
Figure 3.2.3 Calculated surface and 24 hour isoprene mixing ratios for summer conditions.
The NOx distribution calculated by MOZART in the mid-troposphere (500 mb) does not exhibit strong seasonal variations at high latitudes. North of about 60°N, the NOx mixing ratio at 500 mb is generally in the range 5 - 20 pptv in winter and spring and somewhat higher (10 - 30 pptv) during summer. At high latitudes, NO3 is lower than 1 pptv and N2O5 lower than about 2 pptv during all seasons. HO2NO2 simulated by the model is in the range 10 - 20 pptv during spring and lower than 10 pptv in July. The major NOy reservoir species (according to the model) are HNO3 and PAN (other nitrates contribute less than 20 pptv during all seasons), and this is quite different from the observations at the surface (Section 2.5.5). In the model, HNO3 and PAN do exhibit a strong seasonal cycle. PAN shows a maximum during March reaching 200 - 300 pptv, and a minimum during summertime of about 100 pptv. Figure 3.2.4 shows that the PAN maximum is broadly confined at high latitudes during March at 500 mb. North of 60°N, the PAN mixing ratio is in the range 150 - 250 pptv at 500 mb. The vertical distribution of PAN is illustrated in Figure 3.2.5 for January and June conditions and compared to the STRATOZ II [Rudolph et al., 1987] and TROPOZ II [Perros, 1994] measured distributions. Model results have been sampled along the flight track of the campaigns (roughly east coast of Canada, western North Atlantic, and west coast of Southern America). In the northern hemisphere, the highest concentrations (100 - 200 pptv) of this compound are found near the surface in January and at higher altitudes in June. This seasonal difference results from the different transport regimes prevailing in winter and summer. During winter, high concentrations are found in the lower troposphere at mid- and high-latitudes where PAN is more stable. During summer, PAN is transported to the mid- and upper-troposphere by diabatic motions where the lifetime against thermal decomposition is longer (about 10 - 15 days at this altitude) than near the surface. During summer, the geographical structure of the free tropospheric maximum in PAN is not confined at high latitudes as it is during spring, but the maximum appears located over the continental polluted regions where deep convection takes place. HNO3 also shows a strong seasonal cycle in the model. Maximum values north of 40-50°N appear during spring, but do not show a stable polar and zonal structure as PAN. During spring mixing ratios reach 150 - 250 pptv at 500 mb over Northern America and 200 - 300 pptv over Scandinavia. During summer, values in the range 50 - 150 pptv are calculated north of 60°N. The model results suggest an additional 100 - 200 pptv of NOy (present as PAN and HNO3) in the free troposphere at high latitudes in March in comparison to July, and a rather constant 20 - 30 pptv of NOx.
Figure 3.2.4 Calculated northern hemisphere PAN mixing ratios for March.
Figure 3.2.5 Comparison of MOZART calculated and measured PAN mixing ratios obtained during June (STRATOZ III [Rudolph et al, 1987]) and January (TROPOZ II [Perros, 1994]).
The ozone mixing ratio seasonal cycle amplitudes at 600 mb and 500 mb from climatological soundings and calculated by MOZART are compared in Figure 3.2.6. The model results have been sampled at the location of the stations. The results are expressed as deviations from the annual mean at a given latitude. Observations clearly show a broad maximum in ozone during spring-early summer (April-June in the northern hemisphere) for latitudes north of about 20 - 30°N. North of 60°N, the maximum appears in late spring, and around 40°N a second maximum appears during summer with ozone mixing ratios 10 - 15 ppbv higher than the annual mean. The model reproduces the seasonal variation of ozone as well as the amplitude of the seasonal cycle. Maximum ozone levels in the model appear in April-May at high latitudes in the northern hemisphere and during summer at mid-latitudes. The amplitude of the calculated maximum is 8 - 12 ppbv, lower than observations by about 5 ppbv.
In order to investigate further the origin of this maximum, a tracer for stratospheric ozone has been considered in MOZART. This tracer (O3s) is similar to O3 above the tropopause, but is not modified by photochemical production in the troposphere. The seasonal cycle of O3s is shown in Figure 3.2.7 and compared to the ozone soundings. The calculated O3s seasonal cycle shows a marked maximum during winter and spring at high latitudes when transport from the stratosphere is more intense. Comparison with the calculated ozone seasonal cycle indicates that photochemistry decreases ozone during wintertime at high latitudes and is responsible for the maximum simulated at mid-latitudes in late spring-early summer. At 40°N, in July, photochemistry contributes about 15 ppbv to the ozone maximum (-5 ppbv for O3s in comparison to 10 ppbv for O3). At higher latitudes in spring (April-May), stratospheric input produces the maximum in O3 and O3s of 10 ppbv simulated by MOZART in April-May. Stratospheric input appears to be responsible for the spring (April-May) maximum simulated by MOZART at high latitudes. Photochemistry contributes to the seasonal maximum in May-July and is responsible for the higher values calculated in summer at mid-latitudes. Available observations, though, suggest a mix of sources early in the spring (see Section 2).
Figure 3.2.6 Comparison of ozone mixing ratio seasonal cycle amplitude
at 600 mb and 500 mb obtained from climatological soundings with MOZART
calculated values.
Figure 3.2.7 Comparison of ozone with stratospheric origin calculated
by MOZART with ozone soundings.
3.2.2 Idealized Tracer Experiments
Idealized tracers have been included in MOZART to provide additional information on the CO maximum simulated in the free troposphere during spring and confined at high latitudes (see Fig. 3.2.2). Figure 3.2.8a shows the distribution of a CO-like tracer with emissions identical to CO south of 60°N and set to zero north of this latitude. The tracer is uniformly initialized to zero on November 1 and has the same rate constant for reaction with OH CO. After 4 months, the tracer shows a polar maximum similar to the real CO. The difference between the North Pole and 40°N is about 40 - 50 ppbv for the tracer and for CO. Figure 3.2.8b shows the distribution of a similar tracer not affected by reaction with OH (photochemically inert). In this case, a polar maximum with a structure identical to the real CO (fig. 3.2.2a) is simulated with a difference of 40 - 50 ppbv in the mixing ratio between 40°N and the North Pole. From these results (and in agreement with observations), it appears that the CO (and NMHCs) polar maximum in winter and spring is mainly of dynamic origin and associated with advection from mid-latitude source regions to the Arctic and accumulation north of the polar front. Figure 3.2.9 shows the zonal mean cross-sections of CO, of the inert tracer with emissions south of 60°N, and of the inert tracer with emissions north of 60°N solely for March conditions. The CO polar (north of about 50°N-60°N) maximum extends from 3 km to about 10 km altitude (fig. 3.2.9a). The inert tracer with mid-latitude emissions shows a similar pattern. Mid-latitude emissions are advected along isentropic trajectories to the Arctic mid-troposphere. Emissions above 60°N do not contribute significantly to the observed mid-troposphere maximum (Fig. 3.2.9c) as they remain confined in the lower cold troposphere. Figures 3.2.10 compares the different regimes of pollutant advection to the Arctic in winter and spring. Inert-tracers with mid-latitude emissions are shown one month after initialization. The tracers are initialized on January 1 and March 1 in Figure 3.2.10a and 3.2.10b respectively. During winter, pollution is efficiently advected into the Arctic and remains confined in the lower troposphere. In spring, as the continental masses warm, adiabatic motions transport pollutants to higher latitudes. Lower mixing ratios are calculated in the lower troposphere in the Arctic in comparison to winter, but higher mixing ratios are predicted above about 5 km. When oxidation by OH is considered, lower mixing ratios are calculated in March at all altitudes in polar regions.
Figure 3.2.8 a. Distribution of a CO-like tracer with emissions
identical to CO south of 60°N and zero north of this latitude.
b. Distribution of a similar tracer not affected by
reaction with OH.
Figure 3.2.9 a. Zonal mean cross-sections of CO.
b. Zonal mean cross-sections of inert tracer with emissions
south of 60°N.
c. Zonal mean cross-sections of inert tracer with emissions
north of 60°N solely for March conditions.
Figure 3.2.10 a. Comparison between the different regimes
of pollution advection to the Arctic in the winter. The tracers were
initialized on January 1 and July 1 in a and b, respectively.
b. Initialized in July.
3.3 Additional Meteorological Considerations
There is now rather convincing evidence for the accumulation of hydrocarbons at high latitudes in the Northern Hemisphere. This accumulation is most likely due to a combination of high latitude hydrocarbon emissions, reduced photochemical reactivity at high latitudes and slow meridional transport.
Though the exact role of transport and meteorology in the accumulation of hydrocarbons during the winter months is not known. Penkett et al. [1993] suggests that the reproducibility of the seasonal cycle of hydrocarbons indicates their seasonal accumulation is not a local phenomena, but it is characteristic of the troposphere north of the polar front. They further postulate this region of the troposphere is rather homogeneous in nature. South of the front during winter, concentrations of hydrocarbons are more typical of the summer months, pointing to a large north-south concentration gradient of hydrocarbons across the front. It is clear that relatively slow meridional transport is required to realize an accumulation of species during the winter months, and to maintain the observed meridional gradient of methane. The concentration gradient at the polar front suggests, in particular, that transport across the front is slow.
Although the troposphere undergoes dramatic changes between winter and summer, it is not clear if the seasonal cycle in hydrocarbons can be related to these changes in transport. Differences between the winter and summer circulation can be related to differences in the heating of the atmosphere, and therefore in how parcels are transported across isentropic surfaces. Without heating, parcels remain trapped on isentropic surfaces, which form a dome over the pole, with the isentropes increasing equatorward and with height. Therefore, without heating, a parcel is trapped on a surface that has its equatorward boundaries on the Earth's surface and is elevated over the surface as one moves poleward. This configuration of isentropes implies that moving a parcel to higher isentropic surfaces through heating also allows it to move equatorward; cooling a parcel restricts the parcels equatorward transport and is equivalent to an effective poleward transport. Heating can occur through the latent heat of condensation associated with convection, through the latent heating associated with precipitation within synoptic storms, in the boundary layer in association with surface heat fluxes, and in diffusive heating as air parcels are mixed together. These heating processes usually occur in sporadic events. By contrast, in the free troposphere parcels cool at approximately 1 degree/day, providing a slow sinking motion into the dome of isentropically cooler air at lower altitudes.
During the summer months the land and atmosphere warm and convective heating becomes more important, providing a mechanism to vigorously stir the atmosphere. During winter the landmass is cold, resulting in much less convection. The moisture available for latent heating is reduced during winter due to the dryer and colder atmosphere. On the other hand, during the winter months the baroclinicity of the atmosphere is larger (the meridional temperature gradient is stronger) resulting in more vigorous storms and stronger horizontal stirring.
Without any process for heating during the winter months, the dome of cold polar air would expand as parcels cool and sink, with very little transport out of the cold polar air. However, the northern hemisphere is not zonally symmetric and very strong land-sea contrasts exist. As discussed in more detail below this results in the formation of strong cyclones over the oceans, and rather vigorous stirring of the arctic region during the winter months. Vigorous heating occurs when the cold polar airmass is advected over the warmer oceans by these cyclones, allowing strong equatorward transport. The rapid potential vorticity destruction that occurs as polar air moves over the ocean prevents a cold polar vortex from forming. We believe it is this heating which allows parcels to escape from the cold dome of air which characterizes much of the polar regions.
The overall effect of the transport changes between winter and summer is not clear. The concentration gradient across the polar front apparently decreases during the summer months [Penkett et al.,1993]. The accumulation of hydrocarbons north of the polar front (especially during the wintertime) could be due primarily to differences in meridional transport between the winter and summer months, due to differences in vertical transport, due to differences in the location of sources with respect to the polar front between the winter and summer months, or due to seasonal differences in the photo-degradation of hydrocarbons between the two air-masses demarcated by the polar front. We examine each of these hypotheses below.
If north-south transport across the polar front were generally slower during the winter months than the summer months, one might expect an accumulation of hydrocarbons during wintertime. This hypothesis would suggest that during the springtime transition the barrier to north-south transport breaks down, the concentration of hydrocarbons north of the polar front decreases, and the hemisphere as a whole is flooded by the wintertime accumulation of hydrocarbons. To our knowledge, there is no concrete evidence for slower meridional transport during the winter months. Certainly the maximum in the PAN concentration over England during clean periods in May is consistent with the local degradation of accumulated hydrocarbons, and not necessarily with a rapid change in tropospheric transport characteristics.
The greater importance of convection during the summer months might result in a greater dilution of the hydrocarbon emissions during summer than during winter. During winter, the strong surface inversion may act to keep the hydrocarbons trapped near the surface. This hypothesis would suggest the decrease in hydrocarbons during the spring months occurs due to increased dilution by vertical stirring. The Penkett and Brice [1986] observation of a May peak in PAN at the surface over England, while it does not contradict this hypothesis, does not lend it support. Measurements and model results [Hauglustaine, personal communication] indicate that species also accumulate at higher altitudes during winter.
The polar front is located further south during the winter months than during the summer months. Therefore, the net flux of hydrocarbons is larger north of the front during winter than during summer. With weak transport across the front this might result in an accumulation of hydrocarbons during the winter months. However, the area into which these emissions are input is also larger in winter, resulting in a greater dilution of the hydrocarbons. The net effect is not clear. Moreover, a distinct seasonal cycle is observed in methane in polar and mid-latitude locations of the Southern Hemisphere [Steele et. al., 1987], with the methane maximum in the early fall months. This is approximately a month or two later than the corresponding maximum in the Northern Hemisphere In the Southern Hemisphere, the high latitude sources of methane are small, suggesting the location of emissions does not explain the seasonal cycle, at least in the Southern Hemisphere.
Seasonal and spatial differences in the photo-degradation of hydrocarbons, coupled with slow transport across the polar front, may also result in a seasonal accumulation of hydrocarbons north of the polar front, and a strong gradient of hydrocarbons across the front. The correlation between the winter to summer ratio of hydrocarbons and their reaction rate with the hydroxyl radical supports the importance of chemistry in causing the seasonal cycle of hydrocarbons at high latitudes. The methane maximum in the Southern Hemisphere is consistent with this explanation, as is the large peak in PAN over England during May.
The Arctic climatology consists of westerlies in the middle troposphere (500 mb). Wind speeds are higher during the winter months than during summer, with an elongated flow from northeastern Canada to southeastern Asia. During the summer months, the 500 mb flow is more nearly zonal, but with regions of relatively low heights over Baffin Island and Kamchatka. At lower levels, the pressure field is dominated by land-sea contrasts. During winter the warm ocean results in the formation of an Icelandic low and an Aleutian low. A Siberian high is the dominant feature over land with a weaker high situated over northwestern Canada. It is believed that these lower atmospheric structures are primarily responsible for advecting pollution into the Arctic in winter. In spring there is a transition as the landmasses warm and an anticyclone migrates over the cold pole. In the summer months, the flow over the pole is much weaker than during the winter months.
Worthy et al. [1994] show that the wintertime variability at Alert (82° 28’N) is high for PAN, black carbon, CN, CO2 and CH4 due to frequent transport events on approximately a synoptic timescale. The concentrations of these species tend to be anticorrelated with ozone. Although these events contribute to the observed winter time concentration maximum, the minimum concentration values are also higher in the winter months than the summer months, and show a tendency to gradually increase during the winter months. This increase is particularly notable for excess (non-sea salt) sulfate (Barrie et al., 1981). This suggests that the Arctic winter time maximum in pollutants is not solely due to seasonal pollution episodes. The background pollutant concentration also exhibits a slow increase. The increase in the methane background concentration is rather similar between the two hemispheres (the methane concentration difference at north polar latitudes between the minimum wintertime concentration and the summertime concentration is only slightly larger than what Steele et al. finds in the Southern Hemisphere). During March, the variability in the concentration of pollutants appears to rapidly decrease over the Arctic, as the absolute concentrations decrease. This is due to a springtime transition in the meteorology as an anticyclone migrates over the cold polar latitudes, decreasing the frequency of long-range transport events.
There is good evidence that the very strong seasonal cycle of hydrocarbons in the polar regions of the Northern Hemisphere is largely caused by the meteorology. During the winter months there is strong transport across the pole to the Canadian and Alaskan arctic from northern Eurasia. These transport episodes have been documented by [Raatz,1984]: strong transport from source regions to the arctic is often characterized by stagnation over the source region, followed by the approach of a cyclone and rapid poleward transport. Primary input pathways to the polar region occur in association with the Aleutian low, the Icelandic low and with the Siberian high [Barrie, 1986]. Of these pathways, the pathway over Eurasia is the most important, partly due to the fact that parcels transported over Eurasia encounter much less precipitation. Maximum concentrations of aerosols occur above the surface, but usually appear to be confined to the lowest 1-3 km of the atmosphere. Theoretical arguments based on the potential temperature structure of the atmosphere that the pollutants responsible for these aerosol layers originate poleward of the polar front [Carlson, 1981].
Inflow to the polar regions occurs primarily on the eastern side of the Aleutian and Icelandic lows, outflow occurs primarily on their westerly sides. Honrath et al. [1996] examines Arctic outflow events into the North Atlantic in association with NARE, using isentropic trajectories. Due to characteristic slope of isentropic surfaces parcels subside as the travel southward from the pole. Honrath et al. [1996] show that the integrated southward flux decreases with latitude. Seasonally the flux is largest at 55°N in January, and at 45°N in May. The isentropic outflow of air from 70°N to 50°N is slow, with a 175-day exchange time for the arctic region (air north of 70°N and below 6 km). It is expected that diabatic processes will be extremely important as the colder polar air flows over the Atlantic, preventing the return of the arctic outflow to cold polar dome of air. Precipitation processes, as the air moves south in the vicinity of the Icelandic low, may also be important in cleansing the air. Diabatic processes may enhance the equatorward transport in relation to that found by Honrath et al. In addition, considerable outflow is expected to occur in conjunction with the Aleutian Low, although this has not been documented. In a study of two wintertime periods (approximately 5 and 3 weeks, respectively) Barrie found only 50% of the 10 day 850 mb trajectories left the Arctic, also suggesting a rather slow mixing time.
During winter the polar front stretches from south Florida to Ireland, and from south of Japan to near the Washington coast. The mean front is not readily discernable over the western U.S. The Arctic front stretches from Alaska to the Northeastern U.S. The latter front is a shallow feature separating a true Arctic airmass from a more modified airmass, often of polar maritime air. It is usually restricted to lower atmosphere with weak temperature gradients aloft. During winter, the prevalence of strong low level surface inversions in the Arctic tends to obscure the front. To our knowledge, concentration gradients of various species have not been documented across the Arctic front, although at some locations (e.g. Barrow) northerly winds (from Europe) have the highest aerosol concentrations. Often the true Arctic air consists of a shallow airmass stretching along the east side of the Canadian Rockies, with the west side of the Rockies influenced by cool polar oceanic air. This maritime air can, of course, make further intrusions into Canada. Warmer Pacific air from south of the polar front, or air from the Gulf of Mexico, can also make intrusions into the Western U.S. and Canada.
During the summer months the polar
front stretches across Alaska towards Hudson Bay, and from the northeast
U.S. to north of Ireland. Most of the U.S. is influenced by marine air
from the Atlantic, or Pacific, originating from south of the polar front.
The arctic airmass is confined to the highest northern latitudes. The spring
months represent a transition between these regimes, with the polar front
often over the central U.S.
4. Scientific Objectives of TOPSE
The information reviewed in the previous sections indicates a fascinating complexity to the factors that can impact atmospheric photochemistry in the Northern Hemisphere winter and spring. Many measurements indicate a build-up of mostly anthropogenic fuels in late winter to early spring in the LT, MT and even in the UT. This buildup appears to ready the atmospheric chemical system for a burst of photochemical oxidant production when insolation, temperatures, and circulation patterns change in early spring. At the same time, stratosphere-troposphere exchange processes can have a profound influence on tropospheric chemical distributions. Though some of the factors which affect the chemistry of northern hemisphere mid-troposphere have been included in models, there is only limited information on the distribution of most trace constituents with latitude and season for L-MT altitudes in this geographic region and over the winter spring transition. Most of the data is either at high altitude or near the surface or during other seasons. No research programs to date have focused on this region of the atmosphere during the period most relevant to photochemical processing, ozone formation, and seasonal chemical evolution. Recognizing this data gap and also recognizing the availability of improved measurement and modeling capabilities to address the problem has provided the impetus for the current proposal.
The experiment described here is Tropospheric Ozone Production about the Spring Equinox (TOPSE), which describes the broad area of investigation. The overall goal of the experiment is:
Reasons for the latitude component (from the C-130 base at Jeffco to 75°N) are largely related to the access to changes in insolation. The latitude component provides (with the right flight planning) contrast in radical (and other species) abundances from dark or low sun to higher sun on each mission. It is also a basic question as to whether transport or chemistry controls the dispersion of species that have collected after sunrise at higher latitudes. Is the high latitude region a reactor with sunrise or does sunrise cause synoptic changes that "disperse" the collection? Thus the latitude component provides contrast with different synoptic conditions between mid-latitudes and high latitudes (north of the polar front). Similarly for the contrast between seasons, there is a transition from low or absent insolation to higher insolation, which corresponds to a change from a regime which is photochemically weak (or absent) to a regime of greater photochemical activity (at least in terms of OH being the primary oxidant).
To address the overall goal of TOPSE, a range of specific questions has been formulated. Some of the specific questions will be examined during TOPSE are:
1) What are the rates of in-situ ozone production and loss as a function of latitude and season in the mid-troposphere?
The rates of ozone photochemical production and loss are a function of photolysis rates and radical reactions. Appropriate measurements of radical species, nitrogen oxides, and radiation will allow calculation of net ozone production throughout the experimental period.
2) How are radical species and reservoirs distributed over the Northern Hemisphere mid-troposphere as a function of latitude and how do these reservoirs evolve over the winter spring transition?
This question is clearly linked to question 1 dealing with photochemical oxidant evolution during TOPSE. Measurements of radicals and reservoir species spanning the period of spring equinox also provide a stringent test of photochemical models of the troposphere.
3) What are the major sources and reservoirs of nitrogen oxides in the northern mid latitude free troposphere and how do these sources and reservoirs change from winter to spring?
Understanding the sources, interactions, and transformations of reactive odd-nitrogen (along with odd-hydrogen) is critical to unraveling the rates of ozone production and loss. There is an interesting and ill-defined interplay among the reservoirs of reactive nitrogen. Evidence is that NOy constituents do not build up to the same extent, and NOy mixing ratios are not large compared to more remote regions. Why not? What happens to the different constituents that make up NOy? Related to this question are the following;
a) Is the spring bloom in ozone production limited by the availability and supply of NOx? Is there a spring bloom in oxidation rates, but not in ozone production?
b) Is PAN (and other organic nitrates) the major source of NOx in the spring? How significant is HONO as a radical source in early spring?
c) Does the chemistry of NO3 radical have a significant impact on the chemical composition of the wintertime free troposphere? If NO3 cannot be directly measured, is there an inferred role of NO3 oxidation apparent from the chemical measurements?
4) What is the composition and seasonal evolution of volatile organic carbon (VOC) species in the northern mid-latitudes, and what is their significance to a springtime ozone maximum?
The composition of volatile organic trace gases can provide significant
clues to the origins of air masses in the TOPSE region. Changes in the
composition of VOC also can be used to diagnose oxidation mechanisms and
rates over the winter /spring transition. Furthermore, adequate characterization
of VOC is necessary to evaluate radical sources and contributions to ozone
production.
5) What is the potential role of stratospheric influx of ozone and nitrogen oxides on regional tropospheric chemistry, and how does this role change as a function of latitude and season?
This question is to be addressed by a combination of in-situ and remote sensing measurements that will be analyzed in the context of regional and global scale chemical transport models. Understanding the actual distribution and variation of mid-tropospheric ozone is, of course, a critical element in addressing this question.
5. Measurements and Modeling Priorities during TOPSE
The examination of the questions outlined above requires an appropriate suite of measurements, models, and meteorological analyses. We have already described some of the modeling studies done for the TOPSE planning (Section 3), and it is clear that continued application of both chemical process models and 3-D chemistry transport models will be an important part of the TOPSE effort. Further, a highly integrated and focused suite of in-situ and remote sensing measurements is necessary to achieve the TOPSE objectives. This section will briefly review the measurement needs currently identified for the TOPSE experiment and how these measurements relate to the specific TOPSE objectives. (These measurements are in addition to the normal aircraft sensors described in Section 6.)
5.1 Airborne in-situ measurements during TOPSE
The in-situ measurements are needed to address issues related to fast photochemistry, tracer distributions, and seasonal changes. One of the primary questions addresses the in-situ rates of ozone production and loss. As has been noted during MLOPEX and other experiments with a focus on local ozone production, most benefit is obtained from as complete a suite as possible of photochemically related measurements, which can be used to test our current understanding of processes affecting oxidant chemistry. Also, it is necessary to place the measurements of fast photochemistry in the context of radical sources and reservoirs, and also in relation to air mass sources and chemical compositions. Thus, the full suite of measurements desired for TOPSE include a wide range of radiation, radical, and tracer measurements.
Measurements currently identified as high priority for TOPSE are given in Table 5.1.1. It is expected that this list of measurements will evolve as the TOPSE plan develops with input from colleagues outside of NCAR/ACD. It is recognized, too, that currently there may not be suitable techniques for all species identified in Table 5.1, and space may not be available on the aircraft for all experiments. Airborne LIDAR measurements are discussed separately in the next section.
Table 5.1.1 In-situ measurement
needs identified for TOPSE.
Ozone (in-situ)
H2O
CO/CO2
Photolysis
Rates
NO2, O3, CH2O, CH3OOH, etc.
Radicals:
OH
HO2 (HO2+RO2)
Peroxides:
H2O2, CH3OOH
other speciated peroxides
Volatile
Organic Carbon
C2 - C8 NMHC
Halocarbons
Formaldehyde
Other carbonyl
acetone acetaldehyde, other
Organic acids
Stratospheric
Tracer:
N2O
Be7 (or other isotope?)
Reactive
Nitrogen
NO
NO2
HNO3
PAN, other peroxyacyl nitrates
Alkyl nitrates
NOy
Other nitrogen:
NO3, HONO, HO2NO2, N2O5
Aerosol:
size distribution
chemical composition
5.2 Airborne LIDAR Measurements during TOPSE
Measurement of vertical sections of ozone (and aerosols) is a critical element to achieving the scientific objectives of TOPSE. Airborne LIDAR provides time series of vertical profiles of ozone and aerosol scattering. The horizontal resolution is of the order of a few minutes, which translates as a few tens of kilometers [Browell et al., 1996], with a precision better than 5%. The vertical resolution is a few hundred meters. The resolution and precision is even higher for the aerosol measurements. By combining zenith and nadir measurements, continuous cross-sections from the ground to the lower stratosphere can be obtained along the flight track. LIDAR data provides a two dimensional view of the atmospheric stratification through the remote sensing of ozone and aerosols. In this way, LIDAR data give a qualitative look at the two dimensional atmospheric structure. Furthermore, to better understand stratospheric ozone sources, one can use the well-known fact that, in the stratosphere, ozone and potential vorticity (PV) are well correlated [Danielsen, 1968], due to their common source in the stratosphere and quasi-conservative behavior.
To date, a large fraction of airborne LIDAR observations has been used for the description of structure in the ozone field such as tropopause folds. For example, Wu et al. [1997] use LIDAR measurements of ozone in conjunction with potential vorticity from meteorological analysis to describe the large scale circulation features over the western Pacific ocean. Similarly, the CNRS/Service d’Aéronomie has conducted a number of missions utilizing the airborne LIDAR proposed as part of the TOPSE experiment.
The Service d'Aéronomie (CNRS/SA) acted as a coordinator of a campaign, the objective of which was to study the development of a cutoff low over Europe during the spring/summer season. An interesting event occurred on June 19, 1996, where a cutoff low formed over the Atlantic ocean and was present south west of Ireland on this day. On the following days (June 20th to 22nd) the cutoff low slowly moved eastwards before being absorbed by a large cyclonic system over Scandinavia. This process has led to a re-activation of the cutoff low over southern France on June 23rd-24th. This system remained over France during the two following days before decaying and moving over Eastern Europe.
The F27 aircraft flew on June 20, 21, 22, 25, 26 equipped with the upward looking ozone lidar providing ozone profile between 5 and 13 km. Aircraft in-situ measurement at 4.5 km included: ozone, humidity, aerosol extinction, IR radiometers, and winds. An example of the atmospheric vertical cross section measured by the airborne lidar on June 22 is shown in Fig. 5.2.1. Observations have shown that:
Figure 5.2.1 Ozone vertical cross section measured by the airborne lidar in a COL (a). The COL position is observed on the PV distribution at 315 K produced by MM5 (b). The modeled PV and relative humidity vertical cross sections are in good agreement with the ozone observations (c, d).
The work of a Ph.D. student, François Ravetta, who has used mesoscale simulations performed at the University of Cologne with MM5, focussed on the study of the cut-off low (COL) event observed during the June 1996 campaign. He has shown that the diabatic effect related to the COL decay above Spain can be calculated using the model simulations and this will be compared to the ozone evolution to derive any irreversible ozone transfer due to this mechanism. Many ozone layers are found in the frontal zone surrounding the COL with no associated potential vorticity signatures even using the mesoscale model. However, a good correlation is often obtained between the observed ozone layers and the simulated dry layers in the front. The challenge is now to separate the effect of a true PV decay implying an ozone transfer from the effect due to a model uncertainty resulting in the lack of PV layers in the frontal zone of a COL.
A second campaign was organized in February, 1997, during the FASTEX experiment, and it has produced a large data set in well-defined meteorological conditions. The winter time period was well suited for identifying thin ozone layers associated to transport from the stratosphere since very little photochemical production exists at this period of the year. Aircraft data have been obtained during 7 flights corresponding to 2 observation periods:
During TOPSE, one option is to have ozone cross-sections in the polar front and the Arctic fronts from 3 to 11 km obtained by ALTO mounted on the Mystere 20 flying at 11 km. For this option, the aircraft will fly to Boulder, CO, from Europe and will be flown across North America during a 10-day period in March 2000. An alternate option is to mount the LIDAR on the C-130 aircraft. The flight plans will try to accommodate both the need for studies of ozone transport across the tropopause in the polar front and the need for a bi-dimensional description of the atmosphere needed for interpreting the in-situ measurements.
LIDAR data is necessary to put the airborne in-situ measurements into the larger atmospheric context. For example, on a north-south transect, the altitude and frequency of polluted atmospheric layers is likely to change with latitude and with airmass. However, in the Arctic the available evidence indicates that the polluted layers are concentrated in the lower troposphere. Without LIDAR, and sampling at a constant height, the aircraft could produce a biased picture of the atmospheric structure on a north-south transect. Flying at an altitude of five kilometers in the 40 – 75°N latitudes, it is possible that the aircraft would sample a greater abundance of stratospheric air than polluted air. LIDAR allows one to account for this type of sampling bias. While this does not provide a quantitative picture of the chemistry of these remote layers, sampling of a similar layer (albeit at a different location) can provide a picture of the chemistry by analogy. Real time LIDAR data might help identify atmospheric layers of interest to sample.
Figure 5.2.2 Airborne lidar ozone vertical cross section between Biarritz and Paris on February 5th, 1997 showing large ozone concentration along the jet axis developing above France on this day (top panel). The MM5 modeled potential vorticity cross section shows the penetration of a stratospheric layer in the frontal zone, but PV remains larger than 2 PVU above 400 hPa (6km) at the end of the flight (bottom panel).
LIDAR will also be beneficial to atmospheric modeling in a highly stratified atmosphere. Experience in the episodic modeling of atmospheric chemistry in the vicinity of Hawaii indicates that the model can often produce a similar chemical structure to the observed atmospheric structure, if one allows for a vertical offset in the modeled and observed chemical stratification. Point measurements (e.g., those at Mauna Loa Observatory) are much more difficult to compare with observations. Therefore, if one can relate the aircraft measurements to a particular atmospheric layer, and that layer is also simulated in the model, albeit at a slightly different location, comparisons between model and data will be greatly facilitated. A simulation that reproduces the approximate location of the LIDAR observed ozone layers will have much more credibility and can be used to calculate the stratospheric input of ozone.
The LIDAR data can also be used as initial conditions for backward trajectory analysis to recreate the dynamic evolution of the ozone field prior to the measurements. Data assimilation of LIDAR data, although possible, seems to provide very little constraint to the model evolution [B. Khattatov, personal communication]. At best, the model results would be constrained for a small region around the flight track. Due to the location and planned timing of the TOPSE experiment, a lot of activity at the tropopause level (mainly tropopause folds) can be expected. In order to resolve the scales of motion seen in the observations, the regional models (i.e., MM5 and RCTM) seem the most appropriate tools to take advantage of the measurements.
It is clear that LIDAR data can
be very useful in the identification of air masses characterized by significant
departure from the background values. In particular, stratospheric and
polluted air masses are usually easily identified in LIDAR vertical sections.
Also, aerosol measurements can identify cloudy areas (including cirrus)
and polluted areas.
6. Experiment Design and Observational Requirements
6.1
Overview
The NSF C-130 aircraft is requested
for a minimum of 24 flights (216 flight hours) during the transition from
winter through spring (February through May) of the year 2000. The flights
will originate from Jefferson County Airport, Broomfield, Colorado and
make use of an acceptable northern base (Yellowknife, Northwest Territories).
Since takeoff from Jeffco with full fuel is not possible, a low-level flight
to Cheyenne, Wyoming for fueling is necessary, followed by transit to Yellowknife.
From Yellowknife, flights will continue as far north as possible (about
75°N) and return to Yellowknife. A typical flight sequence is summarized
in Table 6.1.1. Plans are for transit legs at 23 kft, with one or more
spirals to one kft above the surface. We can also measure vertical profiles
during the takeoff and landing sequences. The locations of the spirals
will be selected by analyzing the meteorological conditions, the ozone
and aerosol LIDAR results, and consultation with the flight crew. The goal
of these spiral patterns is to determine the degree of layering and the
representativeness of the transit altitude. It has been suggested that
if sufficient flight hours are allocated, more than one flight out
of Yellowknife could be performed (in other words, a flight sequence would
comprise the transit flights plus two or more flights north out of Yellowknife
as opposed to just the one listed). This additional flight would allow
some more detailed vertical profiling as well as an opportunity to examine
a broader longitude range. The latter objective may be especially important
in maximizing scientific interactions with other research programs. Furthermore,
we are considering flights extending south of Jeffco to obtain a more complete
latitude survey at selected times during the mission. However, the listed
3-flight sequence represents a minimum required to achieve the science
goals of the experiment.
The 3-flight sequences will be repeated from the winter, through the transition, and into spring at regular intervals as the solar elevation and the meteorology changes. Proposed timing to cover the period of interest is summarized in Table 6.1.2.
Table 6.1.1. Summary of
typical flight sequence consisting of transit flight Jeffco to Yellowknife
(via Cheyenne), a flight north and return to Yellowknife and the Yellowknife-Jeffco
transit.
| Flight Leg | Action | Time (hours) | Fuel (lbs) | Distance (nmi) |
| Jeffco-Cheyenne | Transit | 0.75 | 3,600 | 170 |
| Cheyenne-Yellowknife | Takeoff | 0.5 | 2,500 | 100 |
| Transit | 5.6 | 24,600 | 1530 | |
| 2 Spirals | 2.0 | 10,000 | 0 | |
| Total | 8.85 | 40,700 | 1800 | |
| Yellowknife-north & return | Takeoff | 0.5 | 2,500 | 100 |
| North leg | 3.65 | 16,100 | 1000 | |
| 1 Spiral | 1.0 | 5,000 | 0 | |
| South leg | 4.0 | 17,600 | 1100 | |
| Total | 9.15 | 41,200 | 2200 | |
| Yellowknife-Jeffco | Takeoff | 0.5 | 2,500 | 100 |
| Transit | 6.25 | 27,500 | 1700 | |
| 2 Spirals | 2.0 | 10,000 | 0 | |
| Total | 8.75 | 40,000 | 1800 | |
| Total for all legs | 26.65 | 121,200 | 5,800 |
Table 6.1.2. Suggested timing
for 3-flight sequences throughout winter-spring period of the year 2000,
leading to a request of 216 flight hours. The exact timing of these flight
sequences is negotiable, as long as the pre- and post-equinox time periods
are covered thoroughly.
| Sequence | Dates | Hours |
| A | 1st week February | 27 |
| B | 3rd week February | 27 |
| C | 1st week March | 27 |
| D | 3rd week March | 27 |
| E | 1st week April | 27 |
| F | 3rd week April | 27 |
| G | 1st week May | 27 |
| H | 3rd week May | 27 |
| Total | 216 |
The requested RAF instrumentation and the user-supplied instrumentation proposed to be involved in the TOPSE campaign are summarized in Table 6.2.1. Full details of the user-supplied instruments are found in the "Request for Aviation Support" document.
Table 6.2.1. Characteristics
of RAF-requested and user-supplied instrumentation. Some
details of size, weight and inlet requirements remain preliminary or unknown.
Many of these instruments and PIs have experience making measurements aboard
aircraft in the troposphere including some that have installed their instruments
aboard the NCAR/NSF C-130. We anticipate that 10-12 hours of flight testing
(in addition to the research request) before the TOPSE deployment will
be adequate. Individual PIs may arrange for more extensive testing well
in advance of the campaign time frame.
| Standard RAF Sensors | |||
| Measurement | Range | Accuracy | Resolution |
| a. A/C latitude, longitude | ±90°, ±180° | ?1.0 nmi hr-1 | 0.0014° |
| b. A/C ground speed | 0-200 m/s | ?1.0 nmi hr-1 | 0.04 m/s |
| c. A/C vertical velocity | ±200 m/s | ±0.10 m/s | 0.012 m/s |
| d. A/C true heading | 0 to 360° | ±0.05° | 0.00275° |
| e. A/C pitch angle | ±45° | ±0.05° | 0.00275° |
| f. A/C roll angle | ±45° | ±0.05° | 0.00275° |
| g. static pressure | 200 to 1035 mbar | ±1 mbar | 0.07 mbar |
| h. indicated airspeed | 0 to 125 mbar | 0.7 mbar | 0.006 mbar |
| i. total air temperature | -60 to 40°C | ±0.5°C | ±0.006°C |
| j. total air temperature (de-iced) | -60 to 40°C | ±0.5°C | ±0.006°C |
| k. dew-point temperature | -60 to 40°C | ±0.5°C>0°C
±1.0°C<0°C |
±0.006°C |
| l. angle of attack | ±10° | ±0.134° | ±0.002° |
| m. angle of side slip | ±5° | ±0.096° | ±0.002° |
| n. ice detector | 0 to 0.5mm inc | - | 0.0005 mm |
| o. geometric altitude | 0 to 0.8 km | - | 0.1 m |
| Other RAF sensors | |||
| Measurement | Range | Accuracy | Resolution |
| a. geometric (radar) alt | 0 to 12 km | 9.7 m | 0.1 m |
| b. abs. Humidity (fast) | 0.1 to 25 g m3 | ±5% | 0.0002 |
| c. Ophir III temperature | -60 to 40°C | - | 0.006°C |
| d. radiom. Sfc./sky temperature | selectable | ±1.0°C | 0.005°C |
| e. infrared radiation | 0 to 600 w/m2 | ||
| f. visible radiation | 0 to 1500 w/m2 | - | 0.12 w/m2 |
| g. ultraviolet radiation | 0 to 200 w/m2 | - | 0.12 w/m2 |
| h. aerosol concentration | 0 to 104 cm-3 | ±6% | selectable |
| i. cloud liquid water | 0 to 5 g/m3 | - | 0.005g/m3 |
| j. aerosol spectra | 0.12 - 3.12 mm | - | 0.25 - 0.375 prog. wtd |
| k. cloud drop spectrum | 0.5 - 47 mm | - | selectable |
| l. cloud drop spectrum | 40 to 620 mm | - | 10 mm |
| m. cloud drop spectrum | 25 to 600 mm | - | 25 mm |
| n. hydrometer spectrum | 200 - 6400 mm | - | 200 mm |
| Other sensors from RAF Bulletin Nos. 3 or 4 | |||
| Measurement | Range | Accuracy | Resolution |
| a. Aircraft position | 3-D position | 25/35/0.1 | 0.5/0.5/0.5 |
| b. Static pressure | 250-1035 mb | ±1 mb | 0.07 mb |
| c. Cabin pressure | 600-1035 mb | ±1 mb | 0.07 mb |
| d. Absolute humidity | 0.1-25 g/m3 | ±5% | 0.2% |
| e. Photography (date, time and GPS stamped, color) | 3 video cameras | ||
| f. Cloud droplet spectrum (FSSP-300) | 0.3-20 mm | - | 0.05-2.0 |
| g. Aerosol conc. (TSI) | 0-104 ct cm3 | ±6% | 1 ct cm-3 |
| h. SABL aerosol LIDAR (wing pod) | |||
| i. UV H2O Sensor | |||
| j. other H2O sensors?[CRYO?] | |||
| k. CO2 | 0-500 | ±5% | 0.5 ppm |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Figure 6.2.1 NCAR C-130 instrumentation layouts. Top figure shows layout without installation of LIDAR. Bottom figure shows instrumentation layout including LIDAR. Potential instrumentation is identified in Table 6.2.1
An airborne Nd-YAG-based ultraviolet (UV) differential absorption lidar (DIAL) has been developed to describe the atmospheric ozone distribution in the troposphere and the lower stratosphere. The NASA ozone airborne lidar has been used during the last 15 years for global and regional scale analysis of the ozone distribution both in the troposphere and the lower stratosphere [Browell, 1983]. A detailed description of this instrument and comparisons with other ozone data were reported in a recent paper accepted for publication in Applied Optics [Ancellet, in press].
In ALTO, a Nd-YAG laser is used in combination with a single Raman cell filled with D2 generate 3 wavelengths (266/289/316 nm) simultaneously transmitted to the atmosphere at a 20-Hz repetition rate. A 40-cm telescope is used to collect the backscattered light either pointing upward or downward.. The backscattered signals are recorded respectively by a photo-counting unit and a waveform recorder. The sampling resolution is respectively 15 m and 37.5 m for the analog and photocounting detection mode. Hardware averaging of the waveform recorder and counter outputs has been implemented to have enough time for further data processing on the computer. A typical 600-shot average corresponds to a 30-s sampling for an ozone profile. A real time analysis of the signals is performed to produce a 2-D plot of the ozone concentrations along the flight track. To compute the ozone concentration at a given range, one calculates the logarithmic derivative of the lidar signal, implying a degradation of the range resolution from 300 m at 1 km to 1000 m at 6 km [Ancellet, 1989]. The analog and photo-counting signal derivatives are combined to achieve a measurement range up to 7 km. The ozone concentrations are calculated for two wavelength pairs: 266/289 nm and 289/316 nm. The 266/289 nm pair is less sensitive to systematic errors (beam alignment, aerosol interference), but it can only be used for ranges less than 1.8 km as the 266 nm signal is rapidly absorbed in the atmosphere [Papayannis, 1990]. Above 1.8 km, the pair 289/316 nm is used, except at ranges where large aerosol interferences are observed, since large ozone error is generally associated to these interferences for this wavelength pair. The ozone mixing ratio is calculated from the ozone molecular density using an atmospheric molecular density model. In addition to ozone, the aerosol backscatter coefficient is calculated at 316 nm where ozone absorption is negligible, using a backward integration scheme described by Browell and co-workers [Browell, 1985]. This is very necessary to check for any interference by the aerosol differential backscatter in the ozone calculation.
Table 6.3.1 ALTO lidar characteristics
for an ozone measurement at a given range
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
The performances of the lidar are summarized in Table 6.3.1 for each
wavelength pair separately. We should stress that the decrease of the performances
with range may be changed because ozone accuracy and vertical and horizontal
resolution are often not desired at the same time. For example doubling
the horizontal resolution will permit a 1.4 and 1.7 improvement respectively
for the ozone accuracy and the vertical resolution.
Each flight sequence is composed of three types of flights that have been described earlier. They consist mostly of extended flight periods near the aircraft ceiling (18 - 23 kft) with spirals (or stair-step patterns) down and up to examine the vertical structure of the lower troposphere in selected regions. These flights are summarized in Figure 6.4.1. While the exact location of the spirals is not finalized, their structure will entail a near-vertical, slow descent (over 30 minutes) followed by a slow ascent (also 30 minutes or so). The ascent portions may have to be replaced with stair-step patterns to accommodate time and fuel requirements. The slow spirals will allow the measurements made at the slowest time scales to take several measurements during each spiral. Information about the vertical structure will also be obtained during takeoff and landing sequences, by flying at a low-level until out of the immediate area of the city, followed by ascent to the transit altitude. This will allow measurement of six profiles on each Jeffco-Yellowknife leg and four profiles on the leg north out of Yellowknife.
Figure 6.4.1 Sample flight profiles for the different
flight legs of TOPSE.
7. Relationship of TOPSE to other programs.
The TOPSE experiment will yield a larger scientific return if it can be related to or be coordinated with other observational programs planned for the year 2000. At this time, we can identify several planned or ongoing studies that are related to TOPSE. These studies include satellite, airborne, and surface platforms.
7.1 Relationship to satellite observations.
Over the last decade or so, it has become clear that satellite observations can provide significant information relating to tropospheric chemistry. A variety of analyses can be performed with satellite retrievals. For example, we have analyzed the SAGE II ozone data for the tropospheric ozone seasonal cycle. Figure 7.1.1 shows the seasonal cycle of mean tropospheric mixing ratio and the standard deviation, derived from the SAGE II ozone profile data (Pan, unpublished). In this analysis, the tropopause pressure was determined first, using the NMC temperature profile and the thermal definition of the tropopause. The profile points with pressure higher than the computed thermal tropopause and between 400 and 700 mb, and in latitude range between 30°-80°N were used to compute the mean and standard deviation. The seasonal cycle shown in the figure has a clear signal of springtime ozone maximum. There is a seasonal variation of latitude range in SAGE II sampling and the high value of ozone for August is at least partially due to the sampling of higher latitude range in August.
Figure 7.1.1 Seasonal cycle for tropospheric ozone derived from SAGE II ozone profile data [L. Pan, unpublished].
Compared to available ozone-sonde measurements, the values shown in Figure 7.1.1 are higher on average. It is known that SAGE II measurements for the tropospheric ozone have larger uncertainties compared with the stratospheric measurements. Although the averaging could reduce the random error, possible systematic errors for this altitude range are not well characterized.
There is a notable difference between these results and the tropospheric ozone residual derived from TOMS total ozone and SAGE II profiles [Fishman et al; 1990, 1997], where the derived ozone maximum for Northern Hemisphere is in the summer season.
Use of satellite data analysis to support the aircraft measurements and the model activity More tropospheric trace gas data from spaceborne instruments are expected to become available before and during the Spring of 2000. Analyses of these data can be a helpful component in parallel to the aircraft and model activity for achieving the scientific objectives. The satellite data analyses can help further refine the scientific objective of the experiments in the preparation stage. During the experiment, it can provide background and boundary conditions for the aircraft measurements and also help validate the model results.
Data products potentially useful for the TOPSE project in the preparation and during the experiments include:
We are aware of several other planned and continuing activities related to TOPSE. These are briefly described below. At this time, no specific arrangements or collaborations have been made with these activities.
Maximum oxidation rates in the free troposphere (MAXOX). The MAXOX program is a collaborative research effort involving scientists from the United Kingdom, Norway, Germany, and France. The main objective of the MAXOX project is to investigate how the annual cycle of ozone in the troposphere is influenced by atmospheric pollution on a continental and hemispheric scale from the surface to the upper troposphere. Significantly, one of the aspects of the program will be to study the chemistry and origin of the spring/summertime ozone maximum over continental Europe and the north Atlantic. The UK Met Research C-130 will contain a payload that contains many of the same measurements proposed for TOPSE. Coordination of TOPSE with MAXOX would be a valuable link to extend the range of chemical conditions which can be observed and which can impact photochemical processes.
Atmospheric Environment Service (AES) Canada. AES Canada has a long history of scientific research in atmospheric chemistry, especially related to ozone research. The series of Polar Sunrise studies has yielded intriguing data and insight related to tropospheric ozone depletion processes and halogen chemistry. Also, long term monitoring studies have been conducted which relate seasonal cycles, long-term trends, and atmospheric chemistry. Clear benefits can be seen from coordinating the TOPSE flights with ongoing research in AES.
NASA Program: SOLVE. Preliminary discussions have begun at NASA regarding a study of chemical processes in the upper troposphere and stratosphere of the Arctic region (SAGE Ozone Loss Validation Experiment, SOLVE). This study may involve the NASA DC-8 and ER-2 aircraft. Measurements of in-situ photochemistry of the UT/LS region aboard the NASA aircraft around the same season as the TOPSE experiment would be a valuable opportunity to examine processes throughout the atmospheric column.
Other Programs? We encourage others planning or considering research programs relevant to the TOPSE effort to contact NCAR/ACD about potential collaborations.
Anderson, B. E., G.L. Gregory, J.D. Bradshaw, J.E. Collins Jr., G.W. Sachse, M.C. Shipham, C.H. Hudgins, Summertime tropospheric ozone distributions over central and eastern Canada, J. Geophys. Res., 99, 1781-1792, 1994.
Arlander, D. W., D. Brüning, U. Schmidt, and D. H. Ehhalt, The troposphere distribution of formaldehyde during TROPOZ II, J. Atmos. Chem., 22, 251-269, 1995.
Atlas, E.L., and B.A. Ridley, The Mauna Loa Observatory photochemistry experiment: Introduction, J. Geophys. Res., 101, 14,531-14,541, 1996.
Bachmeier, A.S., M.C. Shipham, E.V. Browell, W.B. Grant, and J.M. Klassa, Stratospheric/tropospheric exchange affecting the northern wetlands regions of Canada during summer 1990, J. Geophys. Res., 99, 1793-1804, 1994.
Bakwin, P.S., S.C. Wofsy, S.M. Fan, and D.R. Fitzjarrald, Measurements of NOx and NOy concentrations and fluxes over Arctic tundra, J. Geophys. Res., 97, 16545-16557, 1992.
Barrie, L., Arctic Air Pollution: An Overview of Current Knowledge, Atmos. Environ., 20, 643, 1986.
Barrie L.A., R.M. Hoff, and S.M. Daggupaty, The influence of mid-latitudinal pollution sources on haze in the Canadian arctic, Atmos. Environ., 15, 1407-1419, 1981.
Barrie, L.A., and J.W. Bottenheim, Pollution of the Arctic Atmosphere, W. T. Sturges, Ed., Environmental Management Series, Elsevier Applied Science, pp. 155-183, New York, 1991.
Barrie, L.A., and R.M. Hoff, The oxidation rate and residence time of sulphur dioxide in the Arctic atmosphere, Atmos. Environ., 18, 2711-2722, 1984.
Bazhanov, V., and H. Rodhe, Tropospheric ozone at the Swedish mountain site Areskutan: budget and trends, J. Atmos. Chem., 28, 61-76, 1997.
Beine, H. J., D. A. Jaffe, D. R. Blake, E. Atlas, and J. Harris, Measurements of PAN, alkyl nitrate, ozone, and hydrocarbons during spring in interior Alaska, J. Geophys. Res., 101, 12,613-12,619, 1996.
Blake, D.R., D.F. Hurst, T.W. Smith Jr., W.J. Whipple, T.Y. Chen, N.J. Blake, and F.S. Rowland, Summertime measurements of selected nonmethane hydrocarbons in the Arctic and subarctic during the 1988 Arctic Boundary Layer Experiment, J. Geophys. Res., 97, 16559-16588, 1992.
Bottenheim, J.W., A.G. Gallant, and K.A. Brice, Measurements of NOy species and O3 at 82o N latitude, Geophys. Res. Lett., 13, 113-116,1986.
Bottenheim, J.W., L.A. Barrie, E. Atlas, L.E. Heidt, H. Niki, R.A. Rasmussen, and P.B. Shepson, Depletion of lower tropospheric ozone during Arctic Spring: the polar sunrise experiment, J. Geophys. Res., 95, 18,555-18,568, 1990.
Bottenheim, J. W., A. Sirois, K. Brice, and A. J. Gallant, Five years of continuous observations of PAN and ozone at a rural location in eastern Canada, J. Geophys. Res., 99, 5333-5352, 1994.
Brasseur, G. P., D. A. Hauglustaine, S. Walters, P. J. Rasch, J.-F. Muller, C. Granier, and X. X. Tie, MOZART: a global chemical transport model for ozone and related chemical tracers, Part 1. Model description, J. Geophys. Res., Submitted,1998.
Browell, E.V., A.F. Carter, S.T. Shipley, R.J. Allen, C.F. Butler, M.N. Mayo, J.H. Siviter, Jr., and W.M. Hall, NASA multipurpose airborne DIAL system and measurments of ozone and aerosol profiles, Appl. Opt., 22, 522, 1983.
Browell, E.V., I. Syed, and S.T. Shipley, Ultraviolet DIAL measurments of O3 profiles of spatially inhomogenous aerosols, Appl. Opt., 24, 2827-2836, 1985.
Browell, E.V., M.A. Fenn, C.F. Butler, W.B. Grant, R.C. Harriss, and M.C. Shipham, Ozone and aerosol distributions in the summertime troposphere over Canada, J. Geophys. Res., 99, 1739-1755, 1994.
Browell, E.V., M.A. Fenn, C.F. Butler, W.B. Grant, M.B. Clayton, J. Fishman, A.S. Bachmeier, B.E. Anderson, G.L. Gregory, H.E. Fuelberg, J.D. Bradshaw, S.T. Sandholm, D.R. Blake, B.G. Heikes, G.W. Sachse, H.B. Singh, and R.W. Talbot, Ozone and aerosol distributions and air mass characteristics over the South Atlantic Basin during the burning season, J. Geophys. Res., 101, 24043-24068, 1996.
Carlson, T.B., Speculations on the movement of polluted air to the arctic, Atmos. Environ., 15, 1473-1477, 1981.
Crawford, J.H., D.D. Davis, G. Chen, J. Bradshaw, S. Sandholm, Y. Kondo, J. Merrill, S. Liu, E. Browell, G. Gregory, B. Anderson, G. Saches, J. Barrick, D. Blake, R. Talbot, and R. Pueschel, Implications of large scale shifts in the tropospheric NOx levels in the remote tropical Pacific, J. Geophys. Res., 102, 28447-28468, 1997a.
Crawford, J.H., D. Davis, G. Chen, J. Bradshaw, S. Sandholm, Y. Kondo, S. Liu, E. Browell, G. Gregory, B. Anderson, G. Sachse, J. Collins, J. Barrick, D. Blake, R. Talbot, and H. Singh, An assessment of ozone photochemistry in the extratropical western North Pacific: impact of continental outflow during the late winter/early spring, J. Geophys. Res., 102, 28469-28487, 1997b.
Danielson, E.F., Stratospheric-Tropospheric exchange based on radioactivity, ozone, and potential vorticity, J. Atmos. Sci., 25, 502-518, 1968.
de Serves, C., Gas phase formaldehyde and peroxide measurements in the Arctic atmosphere, J. Geophys. Res., 99, 25391-25398, 1994.
Dickerson, R. R., Reactive nitrogen compounds in the Arctic, J. Geophys. Res., 90, 10.739-10,743, 1985.
Fan, S.M., D.J. Jacob, D.L. Mauzerall,
J.D. Bradshaw, S.T. Sandholm, D.R. Blake, H.B. Singh, R.W. Talbot, G.L.
Gregory, and G.W. Sachse, Origin of tropospheric NOx over sub-arctic
Canada in summer, J. Geophys.Res., 99, 16867-16877, 1994.
Fishman, J., C.E. Watson, J.C.
Larsen, and J.A. Logan, Distribution of tropospheric ozone determined from
satellite data, J. Geophys. Res., 95, 3599-3617, 1990.
Fishman, J. and V. G. Brackett,
The climatological distribution of tropospheric ozone derived from satellite
measurements using version 7 total ozone mapping spectrometer and stratospheric
aerosol and gas experiment data set, J. Geophys. Res., 102,
19275-19278, 1997.
Hack, J. J., B. A. Boville, B. P. Briegleb, J. T. Kiehl, P. J. Rasch, and D. L. Williamson, Description of the NCAR Community Climate Model (CCM2), NCAR Tech. Note NCAR/TN-382, Nat. Cent. for Atmos. Res., Boulder, Colo., 1993.
Harriss, R.C., G.W. Sachse, G.F. Hill, L. Wade, K.B. Bartlett, J.E. Collins, L.P. Steele, and P.C. Novelli, Carbon monoxide and methane in the North American Arctic and subarctic troposphere: July-August 1988, J. Geophys. Res., 97, 16589-16599, 1992.
Harriss, R.C., S.C. Wofsy, J.M. Jr. Hoell, R.J. Bendura, J.W. Drewy, R.J. McNeal, D. Pierce, V.Rabine, R.L. Snell, The Arctic Boundary Layer Expedition (ABLE-3B): July-August 1990, J. Geophys. Res., 99, 1635-1643, 1994.
Hauglustaine, D. A., G. P. Brasseur, S. Walters, P. J. Rasch, J.-F. Muller, L. K. Emmons, and M. A. Carroll, MOZART: a global chemical transport model for ozone and related chemical tracers, Part 2. Model results and evaluation, J. Geophys. Res., Submitted,1998.
Hausmann, M., and U. Platt, Spectroscopic measurement of bromine oxide and ozone in the high Arctic during the polar sunrise experiment, J. Geophys. Res., 99, 25399-25413, 1994.
Herring, J.A., D.A. Jaffe, H.J. Beine, S. Madronich, and D.R. Blake, High-latitude springtime photochemistry. Part II: Sensitivity studies of ozone production, J. Atmos. Chem., 27, 155-178, 1997.
Hoell, J.M., D.D. Davis, S.C. Liu, R.E. Newell, H. Akimoto, R.J. McNeal, and R.J. Bendura, The Pacific exploratory mission-west phase B: February-March 1994, J. Geophys. Res., 102, 28223-28239, 1997.
Honrath, R. E., and D. A. Jaffe, Measurements of nitrogen oxides in the Arctic, Geophys. Res. Lett., 17, 611-614, 1990.
Honrath, R. E., and D. Jaffe, The seasonal cycle of nitrogen oxides in the Arctic troposphere at Barrow, Alaska, J. Geophys. Res., 97, 20,615-20,630, 1992.
Hopper, J.F., B. Peters, Y. Yokouchi, H. Niki, B.T. Jobson, P.B. Shepson, and K. Muthuramu, Chemical and meteorological observations at ice camp SWAN during Polar Sunrise Experiment 1992, J. Geophys. Res.,99, 25489-25498, 1994.
Hopper, J. F., and W. Hart, Meteorological aspects of the 1992 Polar Sunrise Experiment, J. Geophys. Res., 99, 25,315-25,328, 1994.
Hov, Ø., S. A. Penkett, I. S. A. Isaksen, and A. Semb, Organic gases in the Norwegian Arctic, Geophys. Res. Lett., 11, 425-428, 1984.
Jaffe, D. A., R. E. Honrath, J. Herring, S.-M. Li, and J. D. Kahl, Measurements of nitrogen oxides at Barrow, Alaska during spring: Evidence for regional and northern hemispheric sources of pollution, J. Geophys. Res., 96, 7395-7405, 1991.
Jobson, B.T., Z. Wu, H. Niki, and L.A. Barrie, Seasonal trends of isoprene, C2-C5 alkanes, and acetylene at a remote boreal site in Canada, J. Geophys. Res., 99, 1589-1599, 1994a.
Jobson, B.T., H. Niki, Y. Yokouchi, J. Bottenheim, F. Hopper, and R. Leaitch, Measurements of C2-C6 hydrocarbons during the Polar Sunrise 1992 Experiment: evidence for Cl atom and Br atom chemistry, J. Geophys. Res., 99, 25,355-25,368, 1994b.
Kiehl, J. T., J. J. Hack, G. B. Bonan, B. A. Boville, B. P. Briegleb, D. L. Williamson, P. J. Rasch, Description of the NCAR Community Climate Model (CCM3), NCAR Tech. Note NCAR/TN-420, Nat. Cent. for Atmos. Res., Boulder, Colo., 1996.
Li, S.M., Equilibrium of particle nitrate with gas phase HONO: tropospheric measurements in the high Arctic during polar sunrise, J. Geophys. Res., 99, 25469-25478, 1994.
Logan, J., Trends in the vertical distribution of ozone: an analysis of ozonsonde data, J. Geophys. Res., 99, 25,553-25,585, 1994.
Mauzerall, D. L., D. J. Jacob, S. -M. Fan, J. D. Bradshaw, G. L. Gregory, G. W. Sachse, and D. R. Blake, Origin of tropospheric ozone at remote high northern latitudes in summer, J. Geophys. Res., 101, 4175-4188, 1996.
Novelli, P. C., L. P. Steele, and P. Tans, Mixing ratios of carbon monoxide in the troposphere, J. Geophys. Res., 97, 20,731, 1992.
Papayannis, A., G. Ancellet, J. Pelon, Multiwavelength lidar for ozone measurements in the troposphere and lower stratosphere, Appl. Opt., 29, 467, 1990.
Penkett, S. A., and K. A. Brice, The spring maximum in photo-oxidants in the northern Hemisphere troposphere, Nature, 319,655-657, 1986.
Penkett, S. A., N. J. Blake, P. Lightman, A. R. W. Marsh, and P. Anwyl, The seasonal variation of nonmethane hydrocarbons in the free troposphere over the North Atlantic Ocean: Possible evidence for extensive reaction of hydrocarbons with the nitrate radical, J. Geophys. Res., 98, 2865-2885, 1993.
Perros, P. E., Large -scale distribution of hydrogen peroxide from aircraft measurements during the TROPOZ II experiment, Atmos. Environ., 27A, 1695-1708, 1993.
Perros, P. E., Large -scale distribution of peroxyacetylnitrate from aircraft measurements during the TROPOZ II experiment, J. Geophys. Res., 99, 8269-8279, 1994.
Raatz, W.E., Tropospheric circulation patterns during Arctic Gas and Aerosol Sampling Program (AGASP), March/April 1983, Geophys. Res. Lett., 11, 449, 1984.
Raatz, W.E., Pollution of the Arctic Atmosphere, W. T. Sturges, Ed., Environmental Management Series, Elsevier Applied Science, pp. 13-42, New York, 1991.
Rasch, P. J., N. M. Mahowald, and B. E. Eaton, Representation of transport, convection, and the hydrologic cycle in chemical transport models: implications for the modeling of short lived and soluble species, J. Geophys. Res., 23, 28127-28138, 1997.
Rasch, P. J., B. A. Boville, and G. P. Brasseur, A three-dimensional general circulation model with coupled chemistry for the middle atmosphere, J. Geophys. Res., 100, 9041-9071, 1995.
Ridley, B.A., and E. Robinson, The Mauna Loa Observatory Photochemistry Experiment, J. Geophys. Res., 97, 10285, 1992.
Rohrer, F., D. Brüning, and D. H. Ehhalt, Tropospheric mixing ratios of NO obtained during TROPOZ II in the latitude region 67oN-56oS, J. Geophys. Res., 102, 25429-25449, 1997.
Rudolph, J. B.Vierkorn-Rudolph, and F.X. Meixner, Large-scale distribution of peroxyacetylnitrate results from the STRATOZ III flights, J. Geophys. Res., 92, 1987.
Sandholm, S., J. Olson, J. Bradshaw, R. Talbot, H. Singh, G. Gregory, D. Blake, B. Anderson, G. Sachse, J. Barrik, J. Collins, K. Klemm, B. Lefer, O. Klemm, K. Gorzelska, D. Herlth, and D.O. O’Hara, Summertime partitioning and budget of NOy compounds in the troposphere over Alaska and Canada: ABLE 3B, J. Geophys. Res., 99,1837-1861, 1994.
Sandholm, S.T., J.D. Bradshaw, G. Chen, H.B. Singh, R.W. Talbot, G.L. Gregory, D.R. Blake, G.W. Sachse, E.V. Browell, J.D.W. Barrik, M.A. Shipham, A.S. Bachmeier, and D. Owen, , Summertime tropospheric observations related to NxOy distribution and partitioning over Alaska: Arctic Boundary Layer Expedition 3A J. Geophys. Res., 97, 16481-16509, 1992.
Shipham M.C., A.S. Bachmeier, D.R. Cahoon, and E.V. Browell, Meteorological overview of the Arctic Boundary Layer Expedition (ABLE 3A) flight series, J. Geophys. Res., 97, 16395-16419, 1992.
Shipham, M.C., A.S. Bachmeier, D.R. Cahoon Jr., G.L. Gregory, B.E. Anderson, and E.V. Browell, A meteorological interpretation of the Arctic Boundary Layer Expedition (ABLE) 3B flight series, J. Geophys. Res., 99, 1645-1657, 1994.
Sigg, A., T. Staffelbach, and A. Neftel, Gas phase measurements of H2O2 in Greenland and their meaning for the interpretation of H2O2 records in ice cores, J. Atmos. Chem., 14, 223-232, 1992.
Simmonds, P.G., S. Seuring, G. Nickless, and R.G. Derwent, Segregation and interpolation of ozone and carbon monoxide measurements by air mass origin at the TOR station Mace Head, Ireland from 1987-1995, J. Atmos. Chem., 28, 45-59, 1997.
Singh, H. B., M. Kanakidou, P.J. Crutzen, and D.J. Jacob, High concentrations and photochemical fate of oxygenated hydrocarbons in the global troposphere, Nature, 378, 50, 1995.
Singh, H.B., D. O’Hare, D. Herlth, J.D. Bradshaw, S.T. Sandholm, G.L. Gregory, G.W. Sachse, D.R. Blake, P.J. Crutzen, and M.A. Kanakidou, Atmospheric measurements of peroxyacetyl nitrate and other organic nitrates at high latitudes: possible sources and sinks, J. Geophys. Res., 97, 16511-16522, 1992a.
Singh, H.B., D. Herlth, D. O’Hara, K. Zahnle, J.D. Bradshaw, S.T. Sandholm, R. Talbot, P.J. Crutzen, and M. Kanakidou, Relationship of peroxyacetyl nitrate and total nitrogen at northern high latitudes: influence of reservoir species, J. Geophys. Res., 97, 16523-16530, 1992b.
Singh, H.B., D. Herlth, D. O’Hara, K. Zahnle, J.D. Bradshaw, S.T. Sandholm, R. Talbot, G.L. Gregory, G.W. Sachse, D.R. Blake, and S.C Wofsy, Summertime distribution of PAN and other reactive nitrogen species in the northern high-latitude atmosphere of eastern Canada, J. Geophys. Res., 99, 1821-1835, 1994.
Solberg, S., C. Dye, N. Schmidbauer, A. Herzog, and R. Gehrig, Carbonyls and nonmethane hydrocarbons at rural European sites from the Mediterranean to the Arctic, J. Atmos. Chem., 25, 33-66, 1996.
Solberg, S., F. Stordal, and O. Høv, Tropospheric ozone at high latitudes in clean and polluted air masses, a climatological study, J. Atmos. Chem., 28, 111-123, 1997.
Steele, L.P., P.J. Fraser, R.A. Rasmussen, M.A.K. Khalil, T.J. Conway, A.J. Crawford, R.H. Gammon, K.A. Masarie, and K.W. Thoning, The global distribution of methane in the troposphere, J. Atmos. Chem., 5, 125-171, 1987.
Sturges, W. T., ed., Pollution of the Arctic Atmosphere, Environmental Management Series, Elsevier Applied Science, New York, 1991.
Sturges, W.T., R.C. Schnell, S. Landsberger, S.J. Oltmans, J.M. Harris, and S.-M. Li, Chemical and meteorological influences on surface ozone destruction at Barrow, Alaska, during Spring 1989, Atmos. Environ., 27A, 2851-2863, 1993.
Talbot, R.W., J.E. Dibb, B.L. Lefer, J.D. Bradshaw, S.T. Sandholm, D.R. Blake, N.J. Blake, G.W. Sachse, J.E. Collins, B.G. Heikes, J.T. Merrill, G.L. Gregory, B.E. Anderson, H.B. Singh, D.C. Thornton, A.R. Bandy, and R.F. Pueschel, Chemical characteristics of continental outflow from Asia to the troposphere over the western Pacific Ocean during February - March 1994: Results from PEM-West B, J. Geophys. Res., 102, 28255-28274, 1997.
Talbot, R. W., et al., Summertime distribution and relations of reactive odd nitrogen species and NOy in the troposphere over Canada, J. Geophys. Res., 99, 1863-1885, 1994.
Talbot, R. W., A. S. Vijgen, and R. C. Harriss, Soluble species in the summer arctic troposphere: Acidic gases, aerosols, and precipitation. J. Geophys. Res., 97, 16,531-16,544, 1992.
Wang, Y., J. A. Logan, and D. J. Jacob, Global simulation of tropospheric O3-NOx-hydrocarbon chemistry, 2, Model evaluation and global ozone budget, J. Geophys. Res., 103, 10,727-10,756, 1998.
Weinheimer, A. J., J.G. Walega, and J.E. Collins, Meridional distributions of NOx, NOy, and other species in the lower stratosphere and upper troposphere during AASE II, Geophys. Res. Lett., 21, 2583-2586, 1994.
Wofsy, S. C., S.-M. Fan, D.R. Blake, J.D. Bradshaw, S.T. Sandholm, H.B. Singh, G.W. Sachse, R.C. Harriss, Factors influencing atmospheric composition over subarctic North America during summer, J. Geophys. Res., 99, 1887-1897, 1994.
Worthy, D.E.J., N.B.A. Trivett, J. F. Hopper, and J. W. Bottenheim, Analysis of long-range transport events at Alert, Northewest Territories, during the Polar Sunrise Experiment, J. Geophys. Res., 99, 25,329- 25,344, 1994.
Wu, Z., R.E. Newell, Y. Zhu, B.E.
Anderson, E.V. Browell, G.L. Gregory, G.W. Sachse, and J.E. Collins, Atmospheric
layers measured from the NASA DC-8 during PEM-West B and comparisons with
PEM-West A, J. Geophys. Res., 102, 28353-28365, 1997.
Field Studies
MLOPEX - Mauna Loa Observatory Photochemistry EXperiment
NARE - North Atlantic Regional Experiment
NASA AASE - NASA Airborne Arctic Stratospheric Experiment
NASA ABLE - NASA Arctic Boundary Layer Expedition
NASA PEM - NASA Pacific Exploratory Mission
NASA TRACE A - NASA TRansport and Atmospheric Chemistry near the Equator - Atlantic
NOWES - NOrthern WEtlands Study
PSE - Polar Sunrise Experiment
Models
CCM Community Climate Model
CTM Chemistry Transport Model
GFDL Geophysical Fluid Dynamics Laboratory
MATCH Model for Atmospheric Transport and Chemistry
MOGUNITA Model of the Global Universal Tracer Transport in the Atmosphere
MOZART Model for Ozone and Related Chemical Tracers
Satellite Programs
GOME Global Ozone Monitoring Experiment
MOPITT Measurement of Pollution in the Troposphere
TOMS Total Ozone Mapping Spectrometer
SAGE Stratospheric Aerosol and Gas Experiment