Abstract

Extra-tropical stratospheric ozone levels have declined by more than 10% in the last 20 years. There have been catastrophic reductions in the amount of ozone at almost all levels in the stratosphere in the Antarctic spring. On some recent spring days at mid-latitudes, when ozone values are normally at their largest during the year, ozone columns have been lower than any previously recorded at any time of the year over the entire 30-year ozone record. Losses of ozone at mid-latitudes have been much greater than can be accounted for by current theories regarding ozone destruction. Current predictions for the future evolution of the ozone layer are based on these theories, and form the scientific basis for the terms of the Montreal Protocol of the Vienna Convention, which controls the global production and release of ozone-depleting substances.

Stratospheric ozone research over the next decade must focus on two areas: (1) determining the reason for the lack of agreement between current theory and observations, and (2) verifying that the Montreal Protocol is actually effective in returning stratospheric chlorine to its historical level and restoring the natural concentration of ozone in the upper atmosphere. This proposal to the Small Payload Program of the CSA suggests one particular avenue for investigating the first of these topics, while the resulting data will also be relevant to resolving the second.

This proposal describes a unique project, named MANTRA (Middle Atmosphere Nitrogen TRend Assessment), that will investigate changes in the odd-nitrogen budget of the stratosphere. It will involve two balloon flights which will use a complementary suite of balloon-borne instruments to measure a number of important atmospheric species. The balloon payload will include two instruments flown between 10 and 20 years ago and three instruments that employ state-of-the-art technology. This combination of instruments will allow a dual intercomparison, one between the measurements obtained by the same instruments after an interval of 20 years, and the other between the old and new measurement techniques. The resulting data will be used to perform a retrospective interpretation of the odd-nitrogen budget of the stratosphere, and to determine whether changes in odd nitrogen could be responsible for the unexpectedly large values of ozone depletion which have been observed at mid-latitudes during the last two decades. The balloon instrument complement will have the ability to measure the concentrations and vertical profiles of the significant components of the short-, medium- and long-term reservoirs of odd nitrogen, as well as the species which link the chlorine and nitrogen reaction cycles. In addition, the long- and medium-term reservoirs for chlorine radicals will be quantified.

As an additional component of the project, detailed background atmospheric studies will be conducted to characterize the local atmospheric conditions in the viscinity of the main balloon flights. These will include ozonesondes, radiosondes, and ground-based observations by three optical instruments. These data, together with Canadian Meteorological Centre (CMC) objective analyses and forecast output, will be important for an extensive modelling effort which will take place after the experimental work has been completed, in order to provide definitive conclusions regarding stratospheric processes based on the experimental products.

The large experimental effort needed for this project will be provided in a truly collaborative effort between two Canadian universities, a government agency, a provincial centre of excellence, and an industrial partner, and will also benefit from the participation of an international partner. The University of Toronto, York University, and the Atmospheric Environment Service of Environment Canada will all provide substantial personnel, material and financial support for the project. The Institute of Space and Terrestrial Science and NSERC will provide indirect support through the time of research personnel at the universities and through the use of equipment and materials provided under research grants. The University of Denver in Colorado will provide a high-resolution BOMEM interferometer and the staff to operate it and analyze the data. Extensive private sector support will be provided by Scientific Instrumentation Limited, using the funds from this project to construct the payload, provide launch and recovery services, and collect and distribute the data collected during the flights.

1.1 Scientific Objectives

1.1.1 Introduction

It is now generally accepted that ozone concentrations have declined significantly since about 1980 (WMO, 1991, 1995; Kerr, 1991) in response to the enhanced levels of chlorine in the stratosphere resulting from anthropogenic emissions of chlorofluorcarbons (CFCs). Catastrophic declines in ozone concentrations have been measured in the Antarctic in late winter and early spring, and decreases in Arctic ozone during the same seasonal period have been reported (e.g., Kerr et al., 1993). A general reduction of ozone throughout most of the year at mid-latitudes in both hemispheres has also been detected. Only tropical regions have not experienced a detectable decline in the ozone column.

In the early 1970s, Molina and Rowland (1974) predicted that the use of CFCs would lead to a source of free chlorine atoms in the stratosphere as a result of the photochemical breakdown of these molecules at high altitudes. Furthermore, they also suggested that the chlorine produced would participate in a sequence of "catalytic" reactions (reactions which are not net consumers of the activating chlorine atoms) to destroy ozone. Using gas-phase chemistry, it was expected that the greatest ozone loss would occur in the upper region of the stratosphere, at altitudes of 40 to 50 km.

Since the time of this prediction, the expected high-altitude ozone loss has been confirmed by measurements by the SBUV and SAGE satellite instruments and by ground-based Umkehr observations. However, the loss of ozone at high altitudes is not large enough to account completely for the reduction in total ozone amount averaged over the globe. In 1985, Farman et al. (1985) found that very large ozone losses were occurring in the Antarctic in the late winter and early spring. Further measurements indicated that these losses were actually occurring in the lower stratosphere, driven by heterogeneous chemistry which acts to tie up active NO_x (NO + NO₂) as nitric acid (HNO₃) while converting inactive chlorine to an active form.

Ozone is produced by the photolysis of O₂, mostly in equatorial regions. Most is destroyed locally by catalytic reactions involving NO_x, HO_x, ClO_x and BrO_x compounds. Below approximately 30 km, the excess ozone flows to higher latitudes where it is destroyed. NO_x plays a major role in the standard gas phase destruction of ozone in the 25 to 45 km region. Its role is modulated by the presence of chlorine, as chlorine nitrate (ClONO₂) can form and tie up both chlorine and NO_x in inactive forms. A similar situation occurs with BrONO₂. However, in the presence of sulphate aerosols, and at mid-latitude temperatures, the reaction

affects the partitioning of NO_x and BrO_x and also acts as a source of HO_x radicals, because HOBr photolyses rapidly to OH and Br. It also leads to the destruction of ozone (Danilin and McConnell, 1995; Lary et al., 1996). Similarly, the reaction

also modulates the NO_x partitioning.

The importance of odd nitrogen in controlling the ozone layer can be further illustrated by considering the role it plays in the polar vortex. Throughout the polar night, NO_x is converted into relatively inactive HNO₃ on the surfaces of ice particles and sulphate aerosols. In addition, during the long polar night, the reaction

or similar heterogeneous reactions can release chlorine from its reservoir species. When sunlight returns in the spring, the free chlorine from photolysis of Cl₂ rapidly attacks ozone.

Detailed dynamical and chemical modelling can now account for about half of the reduction in the ozone column observed at northern mid-latitudes. This discrepancy between the modelled and measured trends is a cause for some concern, since the projected behaviour of the ozone layer in response to the remedial measures taken under the Montreal Protocol is based on photochemical theory. If there is an unknown chemical mechanism in operation which is responsible for the discrepancy, then it is possible that the steps which have been taken under the Protocol may not arrest the decline of the ozone layer as intended.

To date, the possibility of long-term changes in the amounts of other trace constituents in the stratosphere has not been seriously considered as a potential source for this discrepancy in ozone values. Because NO_x concentrations control ozone destruction indirectly through their influence on the level of free chlorine the lower stratosphere, it follows that if the amount of total odd nitrogen decreases, or if the partitioning of the nitrogen compounds shifts from NO_x toward the longer-lived constituents, then the depletion of ozone by the chlorine destruction cycles would increase. Changes in the hydrogen/water/hydroxyl species may also be involved. Therefore, the discrepancy between the models and the observations might be explained in terms of a change in the total odd-nitrogen budget, defined as

There are several mechanisms which can change the concentration of NO_y. Decreases in the total ozone column can modify the NO_y source, N₂O. Also, the increasing levels of chlorine may lead to NO_y loss in the polar regions due to the formation of HNO₃ in and on polar stratospheric clouds (PSCs) and its subsequent sedimentation on the larger particles. In addition, the possibility of perturbation of stratospheric NO_y levels by the current fleet of subsonic aircraft is only now beginning to be addressed (Thompson et al., 1996). The injection of NO_x during the winter and springtime, when the tropopause is low, can also affect mid-latitude and polar NO_y budgets. Finally, the variable aerosol concentrations over the last 20 years have also modified the ozone chemistry and possibly perturbed NO_y amounts such that more NO_y may be lost via PSCs.

The research proposed herein is aimed at providing experimental results which can test the hypothesis that a change in the total odd-nitrogen budget may explain the anomalously large ozone depletions seen, on average, at northern mid-latitudes.

1.1.2 Historical Perspective

Long-term studies of odd nitrogen in the stratosphere are not readily available. The systematic monitoring of even the NO₂ column has only been done at a few locations and for only limited time intervals (e.g., Kerr, 1976; Kerr et al., 1982; McKenzie et al., 1991). Satellite-based observations provide sporadic coverage, because of the limitations of the technique (such as poor coverage of a local region using solar occultation, e.g., SAGE (McCormick et al., 1992)), the limited temporal interval coverage due to short instrument lifetime (e.g., LIMS and SME), and the problem of long-term calibration drift. In any event, the true state of the nitrogen budget cannot be assessed without making simultaneous estimates of several different nitrogen species so that the relative amounts of both short-lived species (e.g., NO₂) and long-lived species (e.g., HNO₃) can be determined.

In Canada, we are fortunate to have a data set collected in the 1970s and 1980s which includes basic measurements of NO₂ and HNO₃. While measurements have also been made by others over a range of latitudes and locations (e.g., MAP-GLOBUS in Europe; ATMOS; Murcray et al., 1987), the Canadian results predate the onset of ozone decline (generally thought to be approximately 1980; see Kerr, 1991; and Bojkov, 1990), connect to the Arctic through flights made using the same equipment in 1978 and 1979, and have been compared to satellite results from lower latitudes (e.g., LIMS (Gille and Russell, 1984)). These historical measurements could be usefully compared to current measurements to examine the possibility of detecting long-term changes in the stratospheric odd-nitrogen budget.

It has now been more than 10 years since the last full-scale Stratoprobe campaign was held in the lower mid-latitudes in Palestine, Texas (30 N) in 1985. The last flights held at higher latitudes in Canada (Yorkton, Saskatchewan) are now nearly 20 years in the past. Many questions about the stratosphere have been asked in the intervening years; some have been answered, but a number of basic issues are still not unambiguously understood. At the same time, a series of intense, international campaigns at mid and high latitudes are about to begin and a number of important satellite initiatives will be coming to fruition toward the end of this decade, such as EOS-AM and ENVISAT, as discussed in Section 1.3.

The balloon campaigns conducted at Churchill, Manitoba and at Yorkton, Saskatchewan in the 1970s made a substantial contribution to our understanding of the stratosphere. The early estimates of the odd-nitrogen budget which were published (Evans et al., 1976; 1977; 1981; 1982a, 1982b; 1985; Kerr and McElroy, 1976; Kerr et al., 1982; Ridley et al., 1984) at that time are frequently cited even in contemporary publications in the field (e.g., Webster et al., 1990). In situ resonance scattering spectra from the hydroxyl (OH) radical were also measured at Yorkton. Changes to the mid-latitude ozone layer which started in the 1980s (Stolarski et al., 1991; 1992; Kerr and McElroy, 1993; Hofmann et al., 1994) followed shortly after the end of the Stratoprobe flight series. It is now believed that chemistry which takes place on the surfaces of aerosol particles is of great importance in determining the balance of the chemistry of the stratosphere (Evans et al., 1985; Salawitch et al., 1994).

1.1.3 Scientific Goals

This proposal introduces a balloon mission to investigate changes in the odd-nitrogen budget of the stratosphere. The MANTRA (Middle Atmosphere Nitrogen TRend Assessment) project has as its primary focus a single balloon flight to take place in August, 1998 at mid-latitudes (Vanscoy, Saskatchewan, 52 N). However, some thought has also been given to a second flight to occur one year later, possibly at a higher latitude. This would enable an investigation of the latitude-dependence of the odd-nitrogen chemistry, and provide a back-up balloon for the primary mission. In addition, the MANTRA project could lead to highly desirable and extremely useful wintertime flights at high latitudes, possibly long-duration events using new technology, in conjunction with wintertime efforts of other nations (e.g., a possible NASA 1999 campaign out of Alaska to investigate NO_x, or the continuing European SESAME program to investigate Arctic and mid-latitude ozone loss).

It is proposed here that two of the instruments flown by AES during the 1970s and 1980s, a visible grating spectrophotometer for NO₂ measurements (Kerr and McElroy, 1976) and an interference-filter infrared emission instrument for HNO₃ measurements (Evans et al., 1977), be reflown to provide a data set which is directly comparable to the results of the earlier campaigns. These measurements will provide valuable diagnostic information about the odd-nitrogen budget because both NO₂ (representative of the odd-nitrogen content of the atmosphere) and HNO₃ (the most important nitrogen reservoir species) will be observed. In addition, these two important nitrogen compounds will be simultaneously measured using contemporary techniques, namely photodiode array spectrometry (McElroy, 1995; McElroy et al., 1994; 1995) in the case of NO₂, and solar absorption infrared interferometry for NO₂ and HNO₃ (Murcray, 1987; Murcray et al., 1987; Goldman et al., 1988). At the same time, a new spectrometric technique under development in a research initiative at the Institute of Space and Terrestrial Science, acousto-optic tunable filter spectroscopy, will be demonstrated in comparison with the old AES NO₂ spectrophotometer and the new photodiode array spectrometer. Supporting ground-based observations will also be performed by a Brewer spectrophotometer, a photodiode array instrument, and a zenith-sky grating spectrometer.

The primary scientific goals of this mission can be summarized as follows:

(1) To measure the total columns and vertical concentration profiles of those reactive species which control stratospheric ozone concentrations, specifically ozone, NO₂, HNO₃, NO, ClONO₂, HCl, and aerosols.

(2) To combine these measurements with those obtained from similar campaigns mounted at northern mid-latitudes over the past 20 years, in order to address the crucial question of why current theory and observations of ozone are in such poor agreement.

(3) In particular, to determine if there have been long-term changes in the amount of total odd-nitrogen or if the partitioning of the nitrogen compounds between NO_x and long-lived constituents has changed, and hence whether such changes could be responsible for the anomalously large ozone depletions seen, on average, at northern mid-latitudes.

A number of secondary scientific goals are also proposed:

(1) To use the measured concentrations of stratospheric chlorine species and ozone to investigate the issue of whether the Montreal Protocol is returning stratospheric chlorine to its historical level and restoring the natural concentration of ozone in the upper atmosphere.

(2) To undertake a careful study of the long-term trend in the vertical ozone profile at northern mid-latitudes. At the moment, this trend remains uncertain. It is clear that upper level ozone has declined in fairly good agreement with predictions based on gas-phase chlorine chemistry. However, it is not clear how important heterogeneous process have been in producing the large-scale changes in the total ozone amount which have been observed on the global scale.

(3) To compare measurements of the same species recorded by different instruments on the balloon, and perform an intercomparison and assessment of the old and new measurement techniques.

(4) To utilize the technological improvements in instruments to quantify more precisely the aerosol involvement both now and at the time of the original measurements.

(5) To take advantage of the opportunity offered by the combination of balloon-borne and ground-based instruments to investigate two techniques for the retrieval of NO₂ vertical profiles from ground-based zenith-sky spectra. This is discussed further in Section 1.5.1, below.

(6) To use the data obtained by the proposed experimental package for validation and ground-truthing of a number of new satellite instruments to be flown in the latter part of this century and the beginning of the next, including the OSIRIS instrument on the satellite Odin. It is important to note that it is still possible to refly instruments which were used to gather ground-truth data for the calibration and intercomparison of instruments flown in the 1970s and 1980s such as LIMS, SBUV, TOMS, and SME. Comparisons of this type will allow a much better retrospective examination of those data and will also improve our confidence concerning temporal changes which might be implied from the comparison of new instrument data to the extant data set.

1.1.4 Experimental Methods

As discussed above, the proposed mission will investigate long-term trends in the odd-nitrogen budget of the stratosphere and their possible role in explaining the anomalously large ozone depletions which have been observed at mid-latitudes during the last two decades. For this purpose, two balloon flights are proposed, the first in August, 1998 from Vanscoy, Saskatchewan, and the second in August, 1999. These flights are linked in two ways. Firstly, the initial flight of the instrument package, if successful, would prove the package for making measurements at higher latitudes in future years. The differences in the chemistry with latitude could be very effectively probed with this instrument. The exact flight plan would need to be planned with care as the selection of sites in northern Canada is restricted as are the logistics (assuming limited resources) and weather. Secondly, by purchasing both the balloons simultaneously, the second balloon is available as a back-up to the first. The second flight requires considerably fewer resources than the first, being about 65% of the cost of the first launch.

The balloon payload will include a suite of five instruments to measure the concentrations and vertical profiles of the significant components of the short-, medium- and long-term reservoirs of odd nitrogen, as well as those species which link the chlorine and nitrogen reaction cycles, specifically ozone, NO₂, HNO₃, NO, ClONO₂, HCl, and aerosols. Two of the instruments will be identical to those flown by AES in previous campaigns between 10 and 20 years ago, and three will employ more recent technology, thus allowing intercomparisons between the measurements obtained by the same instruments after an interval of 20 years, and between the old and new measurement techniques. Detailed supporting measurements will be made to characterize the local atmospheric conditions in the vicinity of the balloon flights. These will include ozonesondes, radiosondes, and ground-based observations by three optical instruments. These data, together with objective analyses and forecast output from CMC, will be used in an extensive modelling effort which will commence after the experimental work is complete, in order to interpret the data in terms of the underlying processes.

Most of the instruments described in the paragraphs below have been successfully flown in the past. The instruments will be refurbished based on the experience of past flights and are therefore not expected to provide any significant challenges. The launch facilities will be provided by Environment Canada, which has a fully equipped balloon launching station at Vanscoy. The equipment on hand includes a complete gondola system, a telemetry and command package, power and pointing systems, and all ground and launch support gear.

The primary components of the proposed experiment are described below, specifically, the five balloon-borne instruments, the measurement of J-values, the payload configuration, the ground-based component, and the subsequent modelling effort.

1.1.4.1 Experimental Methods: AES NO₂ Spectrophotometer

The AES NO₂ spectrophotometer is an antecedent of the Brewer Ozone Spectrophotometer now used to measure ozone in more than 30 countries. Five wavelengths (exit slits), carefully chosen to lie at high and low absorption points in the NO₂ spectrum and at stationary points in the solar spectrum, are measured in the 430 to 450 nm region of the spectrum. This provides a large differential absorption cross-section and a reduced sensitivity to small shifts in the precise position of the band centre wavelengths. The high signal-to-noise ratio of the system permits its use to measure the direct solar beam (Bloxam et al., 1975; McElroy, 1976; Kerr and McElroy, 1976) over a large range of atmospheric optical depths and also to make measurements of the light scattered by air molecules from the Earth's limb (McElroy, 1985, 1988; Roscoe et al., 1985, 1990).

This instrument and data analysis technique provide NO₂ concentrations of about 5% precision and 10% absolute accuracy for a 2-km-thick layer. The measurements made in 1976 should be comparable to new measurements at the 10% confidence level.

1.1.4.2 Experimental Methods: AES HNO₃ Radiometer

The AES HNO₃ radiometer is based on the balloon emission radiometer of Pick (1969). It uses a liquid-nitrogen-cooled detector and optics to make measurements of the atmospheric thermal emission in the 11.3 micron region. The emission of HNO₃ is isolated by using several different filters to provide signals on- and off-band. The instrument is pointed at about 19 elevation angle on the payload to increase the atmospheric slant path, and measurements of the atmospheric radiance profile are made during ascent or descent. The radiance profiles are differentiated with respect to height to assign a radiance change to each layer, from which the concentration of HNO₃ is calculated (Evans et al., 1976).

The HNO₃ concentrations obtained with this instrument are accurate to about 20% in a 5-km layer. The difference between the old and new observations should be accurate to about 30% for a 5-km-thick layer.

1.1.4.3 Experimental Methods: AES Photodiode Array Spectrometer

The AES photodiode array spectrometer is based on the AES SunPhotoSpectrometer which flew on STS-52 with Canadian Astronaut Steve MacLean and aboard the NASA ER-2 as part of the NASA Upper Atmospheric Research and High Speed Research Programs (McElroy, 1995). On the ER-2 the instrument is called the Composition and Photodissociative Flux Measurement (CPFM) Experiment, and is used as an absolute spectroradiometer and to provide ozone data using the technique of differential absorption spectrophotometry. The instrument has flown at 70,000 feet nearly 100 times with this program.

Spectra are recorded from 300 to 785 nm at a resolution varying from 1.2 to 4 nm (FWHM). Data will be collected by this instrument in both solar occultation mode and limb-scan mode, and will be analyzed using a spectral fitting code to determine concentration-height profiles of ozone, NO₂, O₂, O₄, and aerosol. Upper limits for the amount of OClO and BrO will also be estimated.

1.1.4.4 Experimental Methods: ISTS/York University Acousto-Optic Tunable Filter Spectrometer

The ISTS/York University acousto-optic tunable filter (AOTF) spectrometer uses an acousto-optic filter to scan the spectrum in the visible and near-UV regions (250-400 nm). It has a wavelength-dependent resolution varying from of 0.07 nm at 250 nm to 0.21 nm at 400 nm, and can scan very rapidly or step scan across a spectral region. It will be operated in the same mode (wavelength scanning) as the AES photodiode array spectrometer, so that the results from the two instruments can be carefully compared to establish the performance characteristics of the acousto-optic spectrometer.

The AOTF instrument will be used to make solar occultation and limb-scanning observations, in order to measure profiles of ozone and NO₂, and to set detection limits for OClO and BrO.

1.1.4.5 Experimental Methods: University of Denver High-Resolution Interferometer

The University of Denver high-resolution interferometer is a BOMEM Fourier transform spectrometer capable of up to 0.02 cm^-1 resolution (unapodized). It has been used in a large number of flights in the United States (e.g., Goldman et al., 1988; Murcray et al., 1987). This very high resolution permits the accurate measurement of a large number of atmospheric constituents using the solar occultation method. Of particular significance to this particular proposal are the measurements of HNO₃, NO, ClONO₂, and HCl.

It is anticipated that data will be obtained with an accuracy in a 5-km layer of the following (the constituents are in parentheses): 10% (HNO₃), 20% (NO), 15% (ClONO₂), and 10% (HCl). Data from this particular interferometer have been compared to a number of other instruments including ATMOS and CLAES, and it has been used in the laboratory to determine absorption cross-sections of a number of chemical species.

1.1.4.6 Experimental Methods: Measurement of J-Values

One concern in the analysis of the data to be collected is that the J-values as modelled may not be sufficiently precise to properly partition the odd-nitrogen family in the photochemical model simulations. For this reason, it is expected that a second AES photodiode array spectrometer may be flown on the balloon in its ER-2 configuration to make absolute radiometric measurements from which J-values for significant reactions such as the photolysis of ozone and the photodissociation of NO₂ can be determined. In this configuration, the instrument measures the direct solar beam, the brightness of a point on the horizon, and the brightness of the upwelling radiation from below the balloon. These three fields are appropriately integrated with the cross-sections of the relevant absorbers and are combined to produce the J-values.

1.1.4.7 Experimental Methods: An Additional AES High-Resolution Scattering Spectrometer

One additional instrument that is being considered for inclusion in the balloon payload after further study is the AES high-resolution scattering spectrometer for measurements of OH, which was flown on Stratoprobe payloads in July, 1975 from Yorkton and in September, 1976 from Palestine (McElroy and Wardle, 1978). The measurements were of polarized spectra of the sky at right angles to the sun and at about 10 above the horizon. The spectral range and resolution of the instrument are 305 to 311 nm, and 0.04 nm, respectively. The OH signal-to-noise ratio in the altitude range 25 to 40 km is high, but the lack of sufficiently detailed, very-high-resolution data on the solar spectrum precluded an absolute calibration. This problem can be solved today, and in any case, the calibration factor will not have changed between the earlier flights and today. Thus measurements made with this instrument of OH in 1998 would be directly comparable with those made in 1975 and 1976. The OH resonance scattering technique has been used since 1975 by others (Torr et al., 1987), but without the AES refinement of using polarization. The ability to measure OH during the MANTRA campaigns would further enhance the scientific value of the balloon payload.

1.1.4.8 Experimental Methods: Payload Configuration

The payload will be built up on a light-weight aluminum frame. The frame will carry two main structures - the University of Denver high-resolution interferometer and a solar-elevation pointing table. The entire gondola will be swung around by an azimuth-pointing joint so that the elevation pointing table will always face the sun. The elevation table will use an active-pointing system to track the sun in elevation. In this pointing mode, the elevation table will be used to make solar observations at positive solar elevation angles in order to obtain reference spectra, and at elevation angles below 0 in order to collect solar occultation data. The interferometer is equipped with its own pointing system and will track the sun independently of the rest of the gondola.

Limb-scan observations will be made by orienting the azimuth joint so that the elevation table points at 90 to the solar direction. The table will be scanned up and down on command through a range of angles corresponding to the tangent heights at which observations are required. The pointing direction will be stabilized in absolute spatial co-ordinates by feeding a vertical reference from a differential GPS system into the control system. This will ensure that the observations are made with the field-of-view pointing in a particular direction, and that all instruments observe the same region of the atmosphere.

The HNO₃ radiometer will be mounted to the gondola structure, but oriented at 90 to the solar direction so that the sun will not enter the instrument field-of-view. Observations will be made during the ascent and possibly during descent if the cryogens have not been consumed by that time.

1.1.4.9 Experimental Methods: Ground-Based Measurements

A detailed set of background atmospheric measurements will be conducted in order to characterize the local atmospheric conditions in the viscinity of the main balloon flights. These will be performed using ozonesondes, radiosondes, and ground-based observations by three optical instruments. These data, primarily ozone profiles and total column abundances, together with CMC objective analyses and forecast output, will be used in the post-flight modelling effort which will be undertaken in order to interpret the balloon-based measurements in terms of stratospheric processes. Further details regarding the ground-based component of the project are presented in Section 1.5.

In addition, data from the ground-based spectrometers and the balloon-borne instruments will be used to investigate two techniques for the retrieval of NO₂ vertical profiles from ground-based zenith-sky spectra. This is thus a secondary "bonus" objective of the project. The two techniques (McKenzie et al., 1991; Preston, 1995) will be compared, and the retrieved profiles will be validated using the height-resolved measurements made by the balloon-borne instruments. As noted above, NO₂ plays a highly altitude-dependent role in ozone chemistry, reducing ozone loss below 25 km by converting active chlorine and hydrogen species into reservoir species, and causing ozone loss between 25 and 40 km in a catalytic cycle that accounts for about 45% of the ozone removed by gas-phase processes (Crutzen, 1970; Webster et al., 1990). Clearly, vertical profiles of NO₂ are required in order to interpret the chemical processing that occurs at different altitudes. In addition, measurements of NO₂ at mid-latitudes can be used to determine the trajectories of air parcels, with NO₂-rich air generally associated with transport from lower latitudes and NO₂-poor air transported from the polar vortex (in winter) (Noxon, 1978; 1979; Noxon et al., 1979). Because air parcels at different altitudes may have followed very different trajectories, vertical profiles of NO₂ are needed in order to be able to use NO₂ as a photochemical tracer. Thus, the ability to retrieve vertical profiles of NO₂ using the ground-based zenith-sky spectrometer will offer a significant advantage in studying the role of this trace gas in stratospheric processes.

1.1.4.10 Experimental Methods: Modelling

As was the case for the interpretation of the original Stratoprobe data set, extensive modelling will be required to use the results of the balloon investigation effectively. Because of the great variability of stratospheric species, an extensive modelling effort will be required to ensure that the conclusions drawn from the data are actually related to the physics and chemistry of the stratosphere and not to any particularly anomalous events which happened to be sampled at the time of the balloon flight. Further details regarding the planned modelling activities are presented in Section 1.4.

1.2 The Science Team

Consistent with the requirements of the CSA Small Payloads Program, the proposed mission will be a collaborative effort between two Canadian universities, a government agency, a provincial centre of excellence, and an industrial partner, and will also benefit from the participation of an international partner. The specific organizations involved are the University of Toronto, York University, the Atmospheric Environment Service (AES), the Institute of Space and Terrestrial Science (ISTS), Scientific Instrumentation Limited (SIL), and Prof. F.J. Murcray of the University of Denver in Colorado. All groups will be seeking additional funding from other sources to support post-doctoral fellows, graduate students, and undergraduate students, as appropriate.

1.2.1 Principal Investigator Kimberly Strong

Prof. K. Strong (Assistant Professor of Physics, University of Toronto) will have overall responsibility for managing the MANTRA project. She will contribute her research experience with the detection of minor atmospheric species from ground-based and satellite instruments. She has participated in several projects involving ground-based UV-visible absorption spectroscopy, including the development of a portable zenith-sky spectrometer which was deployed in Cambridge, England in November, 1993 (Fish, Jones, and Strong, 1995), and a modelling study of a novel UV-visible ranging instrument which combined absorption spectroscopy with lidar profiling (Strong and Jones, 1995). She is a member of the Canadian Odin Aeronomy Science Team, and has developed optimal estimation algorithms for the retrieval of vertical concentration profiles from spectra of scattered sunlight that will be recorded by the Canadian OSIRIS (optical spectrograph and infrared imager) instrument on the Odin satellite. She has also been involved in the construction of an intracavity laser spectroscopy system to detect weakly absorbing gases, and in near-infrared laboratory spectroscopy of methane in support of the Galileo mission to Jupiter.

As Principal Investigator, Prof. Strong will oversee the MANTRA campaign, liasing between the various participants, and ensuring that the project achieves its scientific and technical objectives as described in this proposal. In addition to her general responsibilities to the project, she will have primary responsibility for several of the instruments. She will be building the ground-based zenith-sky grating spectrometer, developing algorithms to retrieve vertical column abundances from these spectra using the technique of differential optical absorption spectroscopy (DOAS), and investigating the retrieval of vertical profiles from these spectra. She will also operate both the ground-based and balloon-borne photodiode array spectrometers that will be provided by AES, and directly participate in the analysis of the data recorded with these instruments. Ms Rowan Dundas, who will be taking up a post-doctoral appointment early in 1997 to work on the grating zenith-sky spectrometer, will become involved in the balloon campaign.

1.2.2 Co-Investigator James R. Drummond

Prof. James R. Drummond (Professor of Physics, University of Toronto) is an experimentalist in the area of remote sounding instrumentation and specializes in measurements of atmospheric constituents using radiative techniques. He is currently the Principal Investigator on the Measurements Of Pollution In The Troposphere (MOPITT) space instrument and holds the COMDEV/BOMEM/AES /CSA/University of Toronto/NSERC Industrial Research Chair in Atmospheric Remote Sounding from Space at the University of Toronto. In addition, he has participated in balloon campaigns for middle atmospheric measurements from the mid 1970s (Drummond and Jarnot, 1978; Roscoe, Drummond, and Jarnot, 1981). During one balloon campaign in France in 1975, measurements of NO and NO₂ were made using correlation spectroscopy techniques (Drummond, 1977). He was also the Principal Investigator on the Canadian Toronto Balloon Radiometer (TORBAR) experiment (Drummond et al., 1989) and brings experience in all aspects of ballooning to the team.

Prof. Drummond's primary responsibility will be to supervise the interfacing of the University of Denver high-resolution interferometer, and this will involve liaising with the instrument provider and the launch contractor, and some in-house interface work at the University of Toronto. He will be involved in the flight preparations and the balloon flights, and will also participate in the analysis and interpretation of data recorded by the solar absorption spectrometer instrument as well as in achieving the overall science objectives.

1.2.3 Co-Investigator John C. McConnell

Prof. J.C. McConnell (Professor of Earth and Atmospheric Science, York University) will play an important role in the modelling activities of the MANTRA study. He has extensive chemical and radiative transfer modelling experience in both planetary and terrestrial atmospheres. Current modelling activity involves constructing global models of the Earth's troposphere and stratosphere. His group has successfully put an efficient but comprehensive chemical code into their chemical transport models, and also into the Canadian Middle Atmosphere General Circulation Model as part of the Middle Atmosphere Modelling (MAM) project. Radiative transfer work on both coherent and frequency redistribution scattering continues.

The group of Prof. McConnell will be involved with the modelling aspects of the proposed study, both chemical and radiative transfer, that will be critical to understanding the data obtained from the balloon campaign. Box models that have been built with current heterogeneous chemistry and photochemistry (Danilin and McConnell, 1995; Sandilands and McConnell, 1996; Chartrand and McConnell, 1997) will be used to reanalyze the historical balloon data as well as the projected flight data. This group has also put gas-phase chemistry into the Canadian spectral element finite forecast model, SEF, (Sandilands et al., 1996). Heterogeneous chemistry is currently being added and this should be completed by the end of January, 1997. This combined model will be used to expand the "region of influence" of the balloon and ground-based data.

In addition, the SEF+chemistry model can be used to do an intercomparison of measurements of odd nitrogen and ozone from satellites, such as GOME and Odin, that will be flying at the time of the balloon launch. The group is also working with the Canadian Middle Atmosphere Model (Shepherd, 1995; Beagley et al., 1996; de Grandpre et al., 1997). Those people who will be involved with some aspect of the study include Dr. J. Kaminski, Dr. D. Chartrand and Ms E. Templeton.

1.2.4 Co-Investigator C. Thomas McElroy

Dr. C.T. McElroy (Senior Research Scientist, Environment Canada) has extensive experience in remote sounding of the stratosphere from ground-based, balloon, aircraft, and Space Shuttle platforms. He was a co-inventor of the Brewer Ozone Spectrophotometer now in use in more than 30 countries to measure ozone, NO₂, SO₂, and UV-B, and designed a novel double spectrometer version of the Brewer instrument which is now in commercial production. He has played a key role in the measurement of NO₂, making the first measurements of NO₂ in the stratosphere using visible light spectroscopy from a balloon-based spectrophotometer, and developing an algorithm for retrieving vertical profiles from UV-visible measurements of zenith-sky light. He was Deputy Principal Investigator for the SunPhotometer Earth Atmosphere Measurement flown on the US Space Shuttle in October, 1992, and Principal Investigator for the Composition and Photochemical Flux Measurement (CPFM) experiment flown as part of the NASA SPADE (Stratospheric Photochemistry, Aerosol and Dynamics Expedition) project, and the Airborne Southern Hemisphere Ozone Experiment (ASHOE/MAESA) and Stratospheric Tracers of Atmospheric Transport (STRAT) projects of the NASA Upper Atmosphere Research Program.

Dr. McElroy will co-ordinate the analysis of the old NO₂ data, provide consultation and support for the balloon-borne and ground-based photodiode array spectrometers, and provide the analysis of the HNO₃ data (using the assistance of AES support staff and scientists, including Dr. Hans Fast). It is anticipated that AES will provide about one month of Dr. McElroy's time per year for two years, and about six person-months per year of other people's effort, in addition to the actual flight support efforts.

1.2.5 Co-Investigator Brian H. Solheim

Dr. B.H. Solheim (Senior Research Scientist and Adjunct Professor, York University) has extensive experience in optical aeronomy, with emphasis on deriving global wind, temperature, and emission rate distributions from the oxygen airglow. He helped to develop the data system, including algorithms and validation, for the WINDII instrument on the Upper Atmosphere Research Satellite, and has successfully managed that live data system. He was involved in the development of the TOI photometer and the retrieval of high altitude ozone from O₂(¹ ) rocket observations during the International Ozone Rocketsonde Intercomparison (1979). He is a member of the Odin Aeronomy Science Team, with responsibility for the ground segment, and is the Principal Investigator for the development of the acousto-optic tunable filter spectrometer.

Dr. Solheim's primary responsibility will be to operate and analyze the data from the AOTF spectrometer. He will be involved in the pre-flight calibration of the AOTF instrument and in the flight preparations. He will assist in formatting and archiving mission data so that it will be available on the internet after the flight.

1.2.6 Collaborating Partner Frank J. Murcray

Prof. F.J. Murcray (Professor of Physics, University of Denver) will be an international collaborating partner. He has designed and built ultraviolet, optical, and infrared instruments for operation on aircraft, balloons, and extreme environments on the ground. His primary research is in atmospheric remote sensing (Murcray, 1987; Murcray et al., 1987; Goldman et al., 1988). He is a science team member of the ILAS (Improved Limb Atmospheric Sounder, an instrument on the Japanese Advanced Earth Observation Satellite), and the TES (Tropospheric Emission Spectrometer, a NASA Earth Observing Satellite instrument). He is Principal Investigator on research programs for NASA, NSF, DoE, and DoD, and serves as a consultant to several aerospace companies.

Prof. Murcray will commit his BOMEM high-resolution interferometer to the MANTRA measurement campaign. This instrument is already built and has flown a number of times, making it proven flight technology. The University of Denver will supply the instrument and two flight support personnel for the project, while the CSA will cover the costs of instrument shipping to and from the launch site, supplies for the instrument while it is in Canada, and any costs associated with integration into the balloon payload. Analysis of data from the interferometer will be performed jointly at the University of Denver, the University of Toronto, and AES.

1.3 The Scientific Context

There are a number of important themes which need to be to be addressed regarding the application of space technologies in the context of the atmospheric sciences over the next few decades. These include climate change, ozone depletion, oxidants in the lower atmosphere, toxic chemicals in the atmosphere, and tropospheric pollution. Each of the world's space agencies has produced a plan which includes some or all of these topics as targets of technology development for the purpose of studying these issues on a global basis.

In relation to this project proposal, there are a number of initiatives which are closely related to the science to be addressed here. NASA, ESA and the Japanese Space Agency have all committed to investigate some aspects of the atmospheric system and have in some cases already put systems into operation. It is expected that by the time the first of the MANTRA flights takes place, there will be a number of satellites in place making measurements which will support and be served by the data to be collected under this initiative. Several total-ozone satellites are already in place at this time and should be still operational at the time the balloon flights take place.

This is a critical point in time for the study of the chemistry of the stratosphere. Ozone has been declining since about 1980. Measures have been adopted which should stop the increase of ozone-depleting substances around 2000 and lead to an eventual return to the pre-CFC period by the end of the next century. However, the terms of the Montreal Protocol, which has brought the required changes to the CFC production rates around the globe, are based on the currently accepted science of the ozone layer. Unfortunately the global depletion levels are twice as high as the modelled depletion rates, leading to speculation that there may be serious problems still to be faced in the next decade.

This balloon project is intended to make use of a unique Canadian resource to address one possible cause for the discrepancy between the modelled and observed ozone-depletion levels. While this is not the only possible explanation for the decrease, it is one which we are in a position to study and one which is certainly important in the overall chemistry of the stratosphere.

In addition to the direct benefit of addressing the important issue of the mechanism for ozone depletion as outlined in Section 1.1, this balloon program will provide a number of other valuable scientific returns. These include the development of new technology which has potential for future satellite missions (the AES photodiode array spectrometer and the AOTF spectrometer), the provision of an unusually good set of ground-truth data which will include multiple concentration profiles of ozone, NO₂, and HNO₃ that can be used to validate those from Odin, GOME, and other satellite instruments which are due to be launched between now and the time of the first MANTRA flight, as listed in Table 1. The project also provides an excellent opportunity to train the new scientists who will be essential to utilize the forthcoming data sets which will be provided at great cost by future satellite instruments.

Climate change, ozone depletion, and the verification of the Montreal Protocol are three of the most important priorities outlined for the CSA Space Science Program and are also highly rated by the Atmospheric Environment Service as well as the World Meteorological Organization. The number of satellites flying or scheduled to be launched in the next five years by the world's space agencies also underlines the significance of the issues to be addressed by this balloon project (see Table 1).

Table 1. Operational or planned satellite instruments that will be measuring atmospheric species related to the MANTRA proposal. C and P indicate that total columns and vertical profiles, respectively, will be measured. Generally, column amounts and profiles of OClO and BrO will only be measured under perturbed "ozone-hole" conditions. STROBE and CCOSM are competing Earth System Science Pathfinder satellite proposals for which Dr. C.T. McElroy is a Co-Investigator.

1.4 Related Modelling Studies

1.4.1 Photochemical Modelling

As noted in Section 1.1, there have been no major studies addressing the possibility that changes in stratospheric odd-nitrogen (NO_y) levels over the last 20 years may be linked to such stratospheric perturbations as varying ozone and aerosol concentrations. Likewise, little attention has been given to the changing partitioning of odd nitrogen in the stratosphere. (Changes in partitioning due to natural events such as volcanic episodes have been investigated). To some extent this is understandable, as the initial focus has been on decreasing ozone levels driven largely by increasing odd chlorine levels in the stratosphere.

From a measurement perspective, in order to determine total NO_y, it is necessary to measure as many of the NO_y species as possible. Below 35 km or so this is mainly NO, NO₂, and HNO₃. However, there are also non-trivial amounts of ClONO₂ and BrONO₂. During the last 20 years, chlorine levels have been changing, and given the interaction between nitrogen and chlorine chemistry, species such as ClONO₂ and BrONO₂ should be measured if a full analysis of the NO_y budget and partitioning is to be conducted. During the older Stratoprobe flights, ClONO₂ and BrONO₂ were not measured, however, ClONO₂ will be measured as part of this proposed campaign. In addition, HCl will be measured and this can act as a reasonable proxy for ClONO₂ if the stratospheric aerosol levels are known.

As part of the MANTRA study, the older balloon measurements will be reinterpreted using box and 3-D models developed and in use at York University (Danilin and McConnell, 1995; Chartrand and McConnell, 1997; de Grandpre et al., 1997) with current chemistry (Demore et al., 1994; and more recent laboratory measurements). Knowledge of contemporary chlorine and bromine levels will be taken into account, as will current stratospheric sulphate aerosol levels, which can play a major role in the modification of the nitrogen partitioning by converting NO_x to HNO₃ and producing copious amounts of HO_x (Danilin and McConnell, 1995; Lary et al., 1996; Chartrand and McConnell, 1997). Efforts will also be made to incorporate possible high altitude heterogeneous (ion or neutral) reactions in the polar regions that appear to convert N₂O₅ to HNO₃, similar to those which occur at lower altitudes (Kawa et al., 1995).

The models will also be used to compare the balloon and ground-based data from the MANTRA campaign, and also from other campaigns and satellites that may be proximal (but neither simultaneous nor co-located) in time and space. A 20-year run with a 3-D chemical transport model is planned. This will provide a backdrop for analyzing the historical Stratoprobe data and the contemporary MANTRA measurements, as well as much of other NO_y data that has been obtained over the years. Funding for this work, including student or PDF support, will be pursued from other sources, possibly AES or Transport Canada as part of other modelling efforts.

1.4.2 Data Assimilation of Ground-Based and Balloon Measurements

An important methodology for extending the "region of influence" of both balloon-based and ground-based measurements and enhancing the intercomparison of these measurements will be data assimilation. Data assimilation is a means whereby point information can be used to influence fields over larger areas. There are various methods for doing this. Most familiar is the data assimilation of standard meteorological data such as winds and temperatures with weather forecast models. However, species information can also be incorporated to adjust and build global fields using data assimilation.

If the requirement for simultaneity and co-location of measurements is relaxed, then much greater use can be made of the measurements obtained by the various instruments that will be used in the proposed mission. By means of data assimilation, 3-D modelling, trajectories, or some other suitable method, the supporting measurements can be used to fill out a 4-D field (time and 3-D space) that can be projected onto the measurement time and space domain. To some extent the procedure will depend on the quality of the stratospheric winds used. Currently, it is anticipated that wind fields obtained from the Canadian Meteorological Centre's forecast model, SEF, will be used. The operational top of the model is currently 10 mb or about 30 km. However, the top is being raised and it can also be run in a research mode. Furthermore, as part of the MAM (Middle Atmosphere Model) the CMC is developing a mechanistic model with a 95 km top than can be driven by the MAM or by SEF.

1.4.3 Current Modelling Work

The first chemical data experiment to explore the potential for applying the photochemical models and data assimilation to the data obtained from the MANTRA campaign will take place in the next few months, initially in support of the Odin satellite experiment. The weather forecast model will be run with stratospheric chemistry included (Sandilands et al., 1996) at T119 to provide a chemical data base for a period of about a month. Initially, the impact of the ozone data alone will be assessed. The ozone data will be convolved with the Odin orbit information to provide synthetic satellite data which will then be treated with current RPN data assimilation techniques to assess the impact of the Odin data on the ozone field. Ideally, the data assimilation procedures would use the radiance data as part of the assimilation methods. This experience gained from this data assimilation experiment will be invaluable in preparing for the analysis and assimilation of the data from the proposed balloon mission.

1.5 Supporting Ground-Based Measurements

Three ground-based instruments will be operated at the launch site, in conjunction with the balloon flights. Each of these instruments will be built using additional (non-CSA) funding: the Brewer spectrophotometer and photodiode array spectrometer will be provided by AES, and the University of Toronto zenith-sky grating spectrometer has been funded by NSERC. The primary role of these instruments will be to provide vertical column abundances of ozone and NO₂ in order to constrain and verify the vertically resolved measurements that will be made by the balloon-borne instruments. In addition, they will be compared against each other, used to set detection limits for the slant columns of OClO and BrO, and used to investigate the accuracy of two techniques for retrieving vertical profiles from NO₂ slant columns. In addition to these instruments, it is intended that ozonesondes and radiosondes also be launched from Vanscoy (funded by AES), again in conjunction with the balloon launch. These will be used to validate the ozone profiles that are obtained by the balloon-borne instruments.

1.5.1 University of Toronto Zenith-Sky Grating Spectrometer

A portable ground-based UV-visible zenith-sky spectrometer is being constructed at the University of Toronto, under the direction of Prof. K. Strong, with funds provided by NSERC. It will be capable of measuring ozone, NO₂, OClO, and BrO. This instrument will use a Czerny-Turner spectrometer with an adjustable entrance slit and a triple-grating mount to ensure flexibility in the available resolution and dispersion. A spectral resolution of 1.0, 0.25, and 0.125 nm will be obtained over wavelength ranges of 500, 125, and 62 nm, respectively. The detector will be a CCD array, back-illuminated for high quantum efficiency, thermoelectrically cooled, and having a UV-coating for response down to 300 nm.

During the MANTRA campaign, spectra will be recorded from the ground for the duration of the balloon flight, allowing the retrieval of column abundances using the differential optical absorption spectroscopy (DOAS) technique (Solomon et al., 1987). The feasibility of these column measurements is demonstrated in Table 2, which shows that measurements of the vertical columns of ozone and NO₂ will be feasible under both mid-latitude and perturbed polar conditions. Measurement of OClO and BrO at mid-latitudes is less likely, given that the detection limits are larger than or comparable to typical mid-latitude columns, but under perturbed polar conditions, the concentrations of OClO and BrO increase above the calculated detection limits. For the mid-latitude balloon campaign, it is anticipated that the zenith-sky DOAS instrument will be used to measure ozone and NO₂, and possibly to establish detection limits for OClO and BrO.

The precision of the measured vertical columns is likely to be primarily a function of the photon noise, as the CCD dark current is expected to be less than 0.2 electrons/pixel/second, and the readout noise to be about 4 electrons r.m.s.. The accuracy of the vertical columns will be determined by the systematic errors such as changes in wavelength calibration, uncertainties in the accuracy and temperature dependence of the laboratory absorption cross sections, the accuracy of the Ring spectrum, interference from tropospheric species, and the calculation of the air mass factors used to convert slant columns into vertical columns. Quoted values for measurements by similar instruments range from 6% precision and 10% accuracy for total ozone measured at a solar zenith angle of 90 with the SAOZ spectrometer (Goutail et al., 1994), to 15% accuracy for ozone in the UV, 30% accuracy for ozone in the visible, 25% accuracy for NO₂ (Sanders et al., 1989), and 15% to 35% for OClO and BrO (see references in Table 2). The Toronto spectrometer is expected to have better accuracy than SAOZ, given its improved design, including higher spectral resolution and spectral sampling.

The role of the ground-based zenith-sky DOAS instrument in the balloon campaign will be two-fold. Its primary purpose will be to measure vertical columns of ozone and NO₂ for comparison with the height-resolved measurements that will be made by the instruments on the balloon platform. However, the combination of ground-based and balloon-based measurements will provide a unique opportunity to validate the ability to retrieve vertical profiles of NO₂ from the ground. This was first attempted by comparing the ground-based observations of NO₂ by Brewer et al. (1973) to the results of a single-scattering model developed by Davis (1972), and verified by direct solar absorption measurements made during a balloon flight (Bloxam et al., 1975; McElroy, 1976). An algorithm for the analysis of twilight-sky observations, including the effects of the diurnal changes in NO₂ due to chemistry, was developed by McElroy in the late 1980s and published by McKenzie et al. (1991). More recently, Preston (1995) has successfully applied the optimal estimation approach of Rodgers (1976, 1990) to retrieve vertical profiles of NO₂ from the dependence of slant column on solar zenith angle. This technique was validated for altitudes from 10 to 35 km (at a vertical resolution of 5 to 7 km) by comparison with a single NO₂ profile measured using a balloon-borne spectrometer. The MANTRA campaign will investigate the application of both of these profiling techniques to spectra recorded by the ground-based instrument, using the height-resolved measurements made by the balloon-borne instruments for validation of the retrieved profiles.

Table 2. Determination of the feasibility of measuring stratospheric ozone, NO₂, OClO, and BrO with the zenith-sky DOAS instrument.

1.5.2 AES Photodiode Array Spectrometer

The ground-based photodiode array spectrometer will be the same as that flown on the balloon (described in Section 2.3.3), except that it will measure the sunlight scattered from the zenith sky. Spectra will be recorded from 300 to 785 nm at 0.5 nm steps, with a spectral resolution varying from 1.2 to 4 nm across this wavelength range. This instrument has a dynamic range of more than 10⁶, and a signal-to-noise ratio of better than 1000:1 if light levels are high. The field-of view is 2 with a sun diffuser, and 1.2 6 without it. Spectra will be analyzed using a spectral fitting code to determine vertical columns of ozone, NO₂, O₂, and aerosol. Upper limits for the amount of OClO and BrO will also be estimated.

1.5.3 AES Brewer Spectrophotometer

AES will contribute a Brewer spectrophotometer to the balloon campaign, to be operated at the launch site. This instrument is now used routinely world-wide, with more than 80 Brewers in operation. The Brewer is fully automated, and can make quasi-simultaneous observations of total column ozone, NO₂, SO₂, and UV-B radiation. Measurements are usually made using zenith-sky observations, but direct solar and lunar viewing can also be performed. In addition, vertical profiles of ozone can be derived using the Umkehr inversion technique, whereby the atmosphere is divided into a series of layers, and the ozone density in each layer will be determined, along with the total column, from a series of zenith-sky observations at twilight. The Brewer measurements of ozone columns and profiles, and NO₂ columns at the launch site will be used to validate and constrain the measurements made by the balloon-borne instruments.

1.5.4 AES Ozonesondes and Radiosondes

A total of twelve ozonesondes will be launched from Vanscoy in support of the primary balloon launch. These will be supplied by AES, who will also provide two support personnel for three weeks (six person-weeks total). These ozonesondes will be used to obtain accurate vertical profiles of ozone for comparison with the balloon-based measurements and for integration into the modelling studies. AES will also supply radiosondes to provide a record of winds, which will be needed as background information for interpretation of column and profile measurements.

1.6 Data Analysis

The following subsections describe the methods and techniques that will be used to analyze the data from each of the science instruments. In general terms, it is anticipated that there will five levels of data, as follows:

level 0 - raw telemetry data

level 1 - geolocated engineering units

level 2 - geophysical units (profiles, columns, etc.)

level 3 - gridded products

level 4 - integration with models.

Data will be archived and made available on the World Wide Web, as appropriate. Presentation of the results, both technical and scientific, arising from the MANTRA project will be done through the usual channels, at workshops, conferences, and in the refereed literature. Data analysis will be performed within a set time frame (12 months) after the flight, to ensure prompt dissemination of the results.

1.6.1 Balloon-Borne Instruments

1.6.1.1 Data Analysis: AES NO₂ Spectrophotometer

The AES NO₂ spectrophotometer measures radiance at five well-defined wavebands between 430 and 450 nm which lie at high and low absorption points in the NO₂ spectrum and at stationary points in the solar spectrum. This provides a large differential absorption cross-section and a reduced sensitivity to small shifts in the precise position of the band centre wavelengths. The analysis of data from the NO₂ spectrophotometer will be co-ordinated by Dr. McElroy, and will be performed primarily at AES.

Data collected in the solar occultation mode will be analyzed using standard codes developed for the analysis of the data collected during past flights. In this analysis, the data are reduced to slant column amounts using the (extraterrestrial) solar reference signal measured at high sun, and then the partial columns are converted to concentrations using an onion-peeling analysis. The onion-peeling algorithm accounts for the effects of refraction, while the differential absorption technique rejects the influences of other interferences (see McElroy, 1988).

Data collected in the limb-scanning mode will be analyzed using the single-scattering code developed for the analysis of the data published by McElroy (1985) for comparison purposes, and will be reanalyzed using a contemporary multiple-scattering code which includes the effects of polarization (McLinden et al., 1996). The instrument and data analysis technique provide NO₂ concentrations of about 5% precision and 10% absolute accuracy for a 2-km-thick layer. The 1976 data should be comparable to new measurements at the 10% confidence level.

1.6.1.2 Data Analysis: AES HNO₃ Radiometer

The AES HNO₃ radiometer measures the atmospheric thermal emission in the 11.3 micron region, using filters to isolate the emission of HNO₃ and provide signals on- and off-band. The atmospheric radiance profile is measured during ascent or descent, and the radiance profiles are differentiated with respect to height in order to determine the change in radiance for each layer, from which the concentration of HNO₃ can be determined (Evans et al., 1976). The HNO₃ concentrations obtained with this instrument are accurate to about 20% in a 5-km layer. The difference between the old and new observations should be accurate to about 30% for a 5-km thick layer. The analysis of data from the HNO₃ radiometer will be performed at AES by AES support staff, including Dr. Hans Fast, and will be under the direction of Dr. McElroy.

1.6.1.3 Data Analysis: AES Photodiode Array Spectrometer

The AES photodiode array spectrometer will record spectra in both solar occultation mode and limb-scan mode. These will be analyzed using a spectral fitting code across the 1024 wavebands to determine concentration-height profiles of ozone, NO₂, O₂, and aerosol, and upper limits for OClO and BrO columns. The second photodiode array instrument will make absolute radiometric measurements of the direct solar beam, the brightness of a point on the horizon, and the brightness of the upwelling radiation from below the balloon. The J-value contributions from each field are then combined to obtain total estimated J-values for reactions such as the photolysis of ozone and the photodissociation of NO₂ (McElroy et al., 1995). Data analysis for the photodiode array spectrometers will be done in collaboration between AES and the University of Toronto. Dr. McElroy will provide consultation and support, and Prof. Strong will participate in the reduction and analysis of spectra.

1.6.1.4 Data Analysis: ISTS/York University Acousto-Optic Tunable Filter Spectrometer

The AOTF data analysis will use essentially the same DOAS technique as the ground-based zenith-sky spectrometer. The data will first be corrected for dark current and then corrected for responsivity. The wavelength scale will be determined from the pre-flight calibration and the in-flight calibration, correcting for any temperature variations. At this point the spectra may be analyzed by the standard DOAS retrieval, including a shift and stretch of the wavelength based on the known spectral features. The solar occultation measurements require the same preliminary analysis (dark current subtraction, responsivity and wavelength corrections) before the "standard" occultation retrieval. ISTS and York University will perform for the data analysis, with Dr. Solheim having primary responsibility.

1.6.1.5 Data Analysis: University of Denver High-Resolution Interferometer

Data analysis for the high-resolution interferometer will be performed by a combination of three research groups: the University of Denver group, the Atmospheric Environment Service and the University of Toronto. Funding for this analysis will be sought from other sources and from internal funds as appropriate. Prof. James R. Drummond will have overall charge of the data analysis program.

The first-order conversion of the interferograms into concentration profiles is a well-established technique which has been encapsulated into several computer packages. This reduction should not present any serious difficulties, although it should be recognized that every balloon flight is a unique experience. The integration of the profiles obtained with this instrument with the data from the other instruments will be the combined responsibility of the members of the Science Team.

1.6.2 Ground-Based Instruments

1.6.2.1 Data Analysis: University of Toronto Zenith-Sky Grating Spectrometer

The detection of stratospheric constituents from the ground by measuring the absorption of sunlight scattered from the zenith sky is a well established technique, first applied to ozone (Dobson and Harrison, 1926; Dobson, 1957) and NO₂ (Brewer et al., 1973) using measurements of intensity at a few discrete wavelengths. This method was improved by Noxon (Noxon, 1975; Noxon et al., 1979), and the zenith-sky DOAS technique has since been adopted by various groups for the detection of a number of stratospheric species. In addition to ozone and NO₂, it has been successfully used to measure vertical columns of OClO and BrO (e.g., Solomon et al., 1989; Wahner and Schiller, 1992).

Data analysis of spectra recorded with the zenith-sky spectrometer will start with the necessary pre-processing procedures, including subtraction of dark current, flat fielding, and wavelength calibration. DOAS algorithms will then be used to ratio the spectra to a reference, calculate the differential optical depth, perform least squares fitting of the differential absorption cross sections, and calculate the path of scattered light through the atmosphere to convert the retrieved slant columns into vertical columns. This work will be performed at the University of Toronto, with funding from non-CSA sources, under the direction of Prof. Strong.

1.6.2.2 Data Analysis: AES Ground-Based Instruments

Analysis of data from the Brewer spectrophotometer, ozonesondes, and radiosondes will use standard proven techniques. AES will perform this analysis, with Dr. McElroy.