S T R A T E O L E

In addition to such unprecedented sampling of the wind field, a large number of drifting balloons can give valuable information on the mixing properties of air masses, taking into account the quasi-Lagrangian or semi-Lagrangian property of balloon motion. For example, the small scale turbulent energy and enstrophy can be evaluated, based on absolute and relative balloon dispersion (Babiano et al., 1990). Mixing properties can be quantified by the direct determination of directional diffusion coefficients (see below).

The observation of wave-turbulence fields is also of major importance for the under-standing of ozone destruction processes in the Antarctic region. Gravity waves - especially mountain waves - induce temperature oscillations which in some cases can produce temperatures cold enough to generate PSCs. Such events were indeed observed during the European Arctic Stratospheric Experiment (EASOE) (Ovarlez and Ovarlez, 1994). In addition, the transport of air parcels by waves induces the formation of laminae, i.e., small-scale vertical variations of ozone and other minor constituents. Such vertical fluctuations can reach large intensities when the wave happens to propagate near the vortex edge, as the gradient of minor constituent concentrations is large enough there to make concentrations sensitive to oblique transport. McIntyre (1989) has suggested that, even though the vortex edge may not be in general permeable to large-scale transport processes, it might become so in presence of small-scale waves. Generation of strong vertical fluctuations of ozone and water vapour by mountain waves has been indeed observed during EASOE (Teitelbaum et al., 1994; Teitelbaum et al., 1996). This phenomenon should significantly contribute to mixing and therefore to the transport of chemical tracers like chlorine. Conditions for generation of such waves could be favourable over Antarctica due to the existence of very strong winds and large orographic structures.

Of particular relevance to the vortex porosity problem is the low-frequency inertia-gravity wave field. In the middle-latitude stratosphere, cases have been observed in which horizontal parcel excursions due to these waves are of the order of tens or even hundreds of kilometres. Whether this is typical in the vortex edge is largely unknown; but low-frequency inertia-gravity waves at far smaller amplitudes must be easily detectable and measurable using precise positioning of the balloons at 15-minute time intervals. Other instruments should make it possible to gain information on gravity waves. The wind and density variations induced by waves result in changes in balloon drifting level, which could be recorded by pressure sensors.

II.2 Description of the dynamics of the vortex core

Vertical motions in the vortex core in winter and early spring are related to diabatic heating. This heating is mostly an infrared effect as the sun is near or below the horizon. The infrared radiative heating is determined by local absorption (or emissivity) which increases in the presence of particles, and the local temperature difference with the radiative surface below. Small differences (dT < 30 K) cause cooling; large differences (dT > 50 K) cause heating [Pollack and McKay, 1985]. In presence of large particles (PSCs type 2 and 3) infrared heating may be large enough to affect the dynamics and induce upward motion.

Recent results (Valero et al., 1993) have shown that there is a strong dependence of stratospheric heating rates on tropospheric brightness temperature. In this case a detailed and accurate knowledge of tropospheric radiative fluxes is needed to estimate dynamical processes in the lower stratosphere. Such measurements have not been obtained in Antarctic yet. Besides, diabatic heating is very sensitive to local stratospheric temperature. Its seems that large temperature changes are common from one side of the vortex to the other. But the effect of infrared radiation depends on the temperature differences between the troposphere and stratosphere: small (large) differences cause cooling (heating). Strong vortex radiative cooling associated with a warm stratosphere is a major mechanism also because it intensifies the diabatically forced descent in a warm sector of the vortex.

The seeding of the vortex core by long lived balloons is thus of primary importance to study radiation and their effects on the dynamics well inside the vortex. Previous experiments in the Arctic have shown large variations of the brightness temperature within the same region with a multimodal character of the temperature distributions (Valero et al., 1993). This means that the use of averaged - or even instantaneous - values could be misleading in representing tropospheric temperatures. These averaged values should thus be used with histogram and possibly modal averages. The use of long lived balloons is particularly attractive to make such measurements. In order to obtain statistically significant results, the life time of the balloons should be at least of 6 weeks.

Launches of balloons with life times exceeding 6 weeks would give a unique, and unprecedented, opportunity to answer, in addition, two key dynamical questions about the core part of the vortex. First, within this core it is unknown to what extent the motion is horizontally dispersive (layerwise-two-dimensionally turbulent) as compared with the opposite extreme possibility of little horizontal dispersion, as seen in many numerical vortex-dynamics experiments (e.g.: Juckes and McIntyre, 1987). Stratéole will provide the first information on this. Second, Stratéole will also provide the first direct view of the different transport regimes above and below the so-called transition isentrope, an isentropic surface probably located near potential temperature 400 K. Below the transition isentrope there should be relatively free dispersive exchange with middle latitudes, with possibly serious implications for midlatitude ozone chemistry.

The existence of this transition isentrope is expected from chemical tracer observations, from numerical modelling, and from theoretical arguments (McIntyre, 1995). It is expected furthermore that the lowest Stratéole level (70 hPa) should be very close to the transition. But, because of the hill-like topography of isentropic surfaces, the Stratéole levels will span the transition near the vortex centre only.

This is another unique first chance to probe an aspect of vortex dynamics especially critical to the ozone problem. Nevertheless, the scientific problems raised by vortex core dynamics are very different of those at the vortex edge. They can thus be studied separately.

II.3 Study of ozone chemistry

Beyond the unprecedented description of the dynamics of the polar vortex given by such an experiment, a lot of invaluable information on stratospheric chemistry could be obtained using long lived, properly instrumented, balloons. Indeed, unlike satellite or airborne measurements, the Lagrangian-like description of the Antarctic vortex offered by balloon measurements provide a unique opportunity to asses, over a long period, the causes and rate of ozone depletion at high spatial and temporal resolutions.

Considering the range of solar zenith angle in the vicinity of 60°S from September to November, measurements of integrated column content of O3 or NO2 or H2O by solar occultation techniques would have an horizontal resolution of the same order as the resolution of satellite measurements: 100 km for TOMS, 35 km for TOVS/HIRS2. Moreover, the precision of such balloon measurements should be higher, because they are performed looking upwards and thus avoid contamination by tropospheric effects. In addition, drifting balloon measurements can be repeated every 10 to 20 minutes, giving accurate sampling of the diurnal variations, an information not available from space borne instruments.

The combination of simultaneous measurements of ozone, aerosol (including ice particles), temperature and wind at a single location is particularly interesting for the understanding of the complex (homogeneous and heterogeneous) chemical processes involved in ozone depletion and their coupling with dynamics. In particular, uncertainties remain on ozone destruction rate, and on the role of heterogeneous chemistry, which could increase locally the concentration of chlorine radicals and participate in the catalytic ozone destruction cycles. These chemical processes are likely to be active also in winter midlatitudes, around the 45-60° band: the presence of PSCs in this latitude band has already been noticed, especially above mountains during episodes of strong generation of orographic waves.

Furthermore, the simultaneity of radiation measurements and in situ measurements of dynamical variables provides additional advantages over remote measurements. For example, the extinction at NO2 wavelengths is dominated by Rayleigh scattering and not by NO2 absorption. In order to determine NO2 column densities, a large correction for Rayleigh scattering (dependent on the in situ pressure) must be made. A similar situation applies, in varying degrees, to O3 and H2O column measurements.

By combining quasi-Lagrangian simultaneous measurements of temperature and water vapour at high resolution, balloons could give an unprecedented documentation on the threshold for PSC formation, their growth and their persistence as a function of small scale temperature variability. Adding simultaneous measurements of methane, one could obtain invaluable information on the factors controlling the subsidence rates of PSCs and on the process leading to the observed dryness of the southern polar vortex. Indeed, to a good approximation, the oxidation of methane produces two molecules of water in the middle and upper stratosphere. Thus the quantity 2*[CH4]+[H2O] should be quasi conservative until loss of water vapour by process such as sedimentation of ice crystal occurs. Measurement of methane over the period of ozone hole formation is particularly interesting, since it is render possible to distinguish between the chemical processes that destroy ozone from the dynamical processes, such as the diabatic descent in the vortex, which affect the vertical profiles of long-lived tracers.

NO2 column and aerosol extinction measurements from balloons should also document the role of sulphate/nitrate aerosols in the lower stratosphere, which are believed to offer sites for heterogeneous reactions. The nature and rates of these reactions that tend to remove nitrogen oxides are still uncertain, as they depend on the physical state of the aerosol (liquid or frozen), its concentration and composition, as well as the ambient temperature.

Erosion of the vortex by planetary wave activity transports air from the vortex edge region to lower latitudes. Although observational and modelling studies suggest a time scale of 3-4 months to replace a substantial fraction of the vortex air, the importance of this transport to in-situ chemical effects at mid-latitudes remains poorly known and is still under debate. A large number of drifting balloons should give a high-resolution picture of the vortex erosion processes occurring at the vortex edge. It seems therefore important to study in situ the degree of chemical perturbation of the air parcels peeled off from the vortex and to be able to clearly identify the export of chemically perturbed air towards mid-latitudes and its potential impact on the ozone contents in these regions.

III. THE STRATEOLE CONCEPT

In order to study the dynamics of the established winter polar vortex and its springtime breakdown, the Stratéole project has a goal to fly a large number of small constant level (superpressure, i.e. isopycnic) drifting balloons in the lower stratosphere instrumented, in addition to energy sources, with the following devices:

In addition to this basic instrumentation present on each balloon ñ hereafter referred as the dynamical payload ñ some gondola will host additional devices, 'the passengers'. The basic gondola is designed in such a way that any new sensor can be boarded easily on Stratéole gondolas as far as it fits the energy consumption, weight and data transmission rate imposed by the dynamical payload. For now, the envisaged sensors are:

Flux-gate magnetometers,

TDDR (Total Direct Diffuse Radiometer),

IRBBR (InfRared Broad Band Radiometer),

LAMA (Lightweight Atmospheric Methane Analyser).

The Stratéole aerostat and observing system are fully described in Appendix 1 and the envisaged passengers in Appendix 2. The reason which conducted to the choice of this observing system are given below.

The carriers selected for the Stratéole experiment are small superpressure (isopycnic) balloons, flying approximately at 50 and 70 hPa levels, able to carry a weight of around 10 to 20 kg, with an expected lifetime of 3 months.

An superpressure balloon has a closed "no-leak" envelope which keeps a constant volume once it has got fully inflated (and thus has reached its drifting level). >From then on, the trajectory is isopycnic due to Archimedes' principle. The choice of such carriers was based on the following considerations:

At present, the small isopycnic balloon stands as the only reliable technology ensuring the required average three-month lifetime around the 50-70 hPa levels. The reliability of such technology has been largely demonstrated by more than a thousand successful flights, during such large experiments as EOLE (Morel and Bandeen, 1973; Morel and Larchevêque, 1974; Desbois, 1975), TWERLE (TWERLE TEAM, 1977), TCLBS (Julian, 1981) and BALSAMINE (Cadet et al., 1981).

Contrary to other types of balloons like Infrared Montgolfières, the flying capacity of isopycnic balloons is independent of latitude and season. It is thus particularly well suited for an experiment like Stratéole, which must fly at high latitudes in late winter and Spring.

The Global Positioning System, based on an operational network of satellites, allows three-dimensional positioning of the platform with 100 m accuracy. Its utilisation for Stratéole will meet the balloon specifications. The localisation will be effected every 15 minutes. This frequency correspond to a spatial resolution (and scale representativeness) of the order of 20 to 100 km for the wind field, well suited to data assimilation in global stratospheric models. The wind velocity will be deduced from successive positions of the balloons with an accuracy better than 0.2 m-1.

Two drifting levels in the lower stratosphere are being considered, corresponding approximately to 50 and 70 hPa. The 50 hPa level was selected because it correspond to a level at which the vortex is well developed and where the ozone hole is the most important. The 70 hPa altitude corresponds more or less to a level, the "sub-vortex", where tropospheric perturbations are still able to penetrate weakening the vortex isolation and, consequently corresponds to a quite different dynamical regime for transport consideration (McIntyre, 1995). This is the region where mixing between polar and middle latitude air masses is supposed to be quite effective. Furthermore, flying balloons at this level allows to observe the ozone hole looking upward without contamination due to tropospheric effects faced by ground based or satellite measurements

IV. THE STRATEOLE LAUNCHING STRATEGY

The definition of the best observational strategy for the Stratéole rely on several problems, in particular the selection of launching bases and the number of balloons which should be launched. To answer these questions, several preliminary studies have been undertaken:

meteorological and logistical surveys of possible launching sites,

surveys of the Climatology of the vortex edge position,

numerical simulations of balloon trajectories.

IV.1 Surveys of Possible Launching Sites

There are important technical constraints in the choice of launching sites (see Appendix 1). First, due to the large number of balloons envisioned and to the duration of the experiment, the launching site should be of easy access and offer good logistic support. Second the balloon envelope is fragile and must not be shaken on the launch pad in order to avoid formation of leaks. Thus wind speed at the surface must be weaker than 2 m-1 during launch operation. Third, this choice is also constraint by the goal of the experiment and should result in the successful seeding of vortex edge with some balloons well inside the vortex core.

Thus a survey (Bidau, 1995) of the logistic and meteorology of the potential bases on the Antarctic continent or close to the continent was carried out and leaded to the selection of three potentials launching sites fulfilling the above requirements: Ushuaia (68°19 W, 54°48 S), Marambio (56°38 W, 64°14 S) and McMurdo (166°37 E, 77°51 S). Most of the stations on the continent were eliminated because of katabatic winds, rendering launch operations impossible, and because of very poor logistical structures. Also the numerical simulations of balloon trajectories and the climatological survey of the vortex edge location, which will be presented below, have shown that Ushuaia is at the northward acceptable limit for a launching site.

IV.2 Climatology of the Vortex Edge Position

As most of the balloons should be seeded close to the vortex edge, it is important to know the exact extension of the polar vortex and the location of its edge. Unfortunately the position and shape of the vortex change from to day and from year to year. For example, whereas the vortex is relatively symmetric around the pole during winter, it becomes more and more eccentric and elongated in spring because the amplitude of the stationary wave 1 is increa-sing more and more. Changes in phase of the QBO also produce an interannual variability.

The vortex core is characterised by large negative values of the Ertel potential vorticity (PV) whereas, at the mid-latitudes, the PV values are quite weak without large spatial variations. It is usual to define the vortex edge as being the line where the PV gradient is maximum. This definition is difficult to use in practice when working with observational data. This explain that alternative definitions were used in the studies summarised below.

For Stratéole, two studies of the Climatology of the vortex edge position were conducted. The study of Hart (1994) concerns the years 1985-1993 and uses ECMWF (interpolated on a grid 2.5°x2.5°) and Australian Bureau of Meteorology (T31 truncation) analysis on a daily base. Hart used the latitude of the maximum wind speed on the 50 hPa level to define the edge of the vortex. Investigations for sample months had show good correspondence between the location of the wind maxima at either the 50 or 70 hPa level, and the region of sharp PV gradient on the 450 and 500 K isentropic surfaces. At latitudes higher than 40°S for the end of winter and early spring these isentropic surfaces correspond to pressure levels 45-75 and 35-55 hPa respectively. The isolines -30 and -40 PV units were selected to define the vortex edge at these two levels. Edouard and Vautard (1994) used ECMWF analyses (T42 truncation) to study PV charts on the 475°K surface fro the period 1983-1992. The edge was defined has being the -50 PV units isoline. The survey made since these two studies largely confirms the conclusions reached at that time.

Figure 4, from Hart (1994), is an illustration of the variability of the position of vortex edge at 65°W (corresponding to South America and Palmer Peninsula) and 125° E (selected to represent several potential launch sites including McMurdo) for the period 15 August to 30 September. In this Figure, the circle indicates the mean, the box represents the standard deviation and the line indicates the range.

Figure 4 : Interannual variability of the vortex edge position at 65°W and 125°E (from Hart, 1994).

The conclusions of both studies are similar and the statistics of occurrence of the vortex above possible stations clearly show that the three stations experience very different situations:

for Edouard and Vautard: Ushuaia is at least 90% of time outside the vortex in August-October, McMurdo is at least 90% of time inside the vortex in August-September and at least 60% in October, Marambio is around 40 % of the time inside the vortex in August-October;

for Hart: the vortex edge is occasionally located northward of Ushuaia each year but this is exceptional for 3 of the 9 studied years, Marambio has good chance to be at the vortex edge, McMurdo is virtually always located inside the vortex except during spring when the vortex can be distorted.

IV.3 Numerical Simulations of Balloon Trajectories

Numerical simulations of balloon trajectories were also performed in order to respond to several questions (Trounday et al., 1995, Paparella et al., 1997; Wauben et al., 1997).

The first problem to solve was to determine if the vortex edge acts as a barrier for the Stratéole balloons and what difference exists between their movements (on an isopycnic surface) and those of air masses (on an isentropic surface with a good approximation). All the simulations undertaken indicate a greater tendency for balloons to cross the vortex edge. Nevertheless this barrier is very efficient for the balloons too. For example, it can be seen on Figure 5 (from Trounday et al. 1995) that balloons launched inside the vortex have a tendency to stay in the vortex whereas those launched inside tend to stay inside. This is also illustrated on Figure 6 (from Paparella et al., 1997), in which, the number of balloons (for the three stations) which drift inside, outside and at the edge of the vortex is displayed as a function of time (the day 0 is 15 August, 1993).

Figure 5 : Percentage of fluid parcels and balloons in 5°-wide latitude bands from 10°S to 85°S on September 1, 1997. From the simulations of Trounday et al. (1995).

Figure 6 : Number of balloons launched from Ushuaia (top left), Marambio (top right) and McMurdo (right) drifting inside outside and at the edge of the vortex as a function of time (from Paparella et al., 1997)

For balloons released near the edge of the vortex, there are about an equal number of crossings from inside to outside of the vortex as from outside to inside. But they tend to drift at the vortex edge as illustrated in Figure 7.

In fact the study of Paparella et al. (1997) shows that the balloon tends to stay in the region of kinetic energy in which they were seeded. In this study the definition of the vortex edge is based on quasi-geostrophic turbulence theory. For an axisymetric vortex, the edge is defined by the line where the kinetic energy is maximum. Indeed this line cannot be crossed by particles from outside (where the motions are chaotic) or inside (with regular motions). This heuristic definition has the advantage to be simple and gives very similar results to those obtained from the maximum of PV in most of the cases.

Another related study was carried out by Wauben et al. (1997). They were mainly interested in magnitude of transport out of the Antarctic vortex. They used the TKM transport model developed at KNMI and studied the years 1990-1993. Their main conclusions is that around 80 % of the outflow is by descending into the troposphere inside the vortex (pointing to the interest of studying this region during Stratéole) while the remaining 20 % occurs through quasi-horizontal mixing across the vortex edge as a result of planetary waves activity. The quasi-horizontal outflow also seems to appears in a more episodic way.

Figure 7 : Successive positions (one point every 30 minutes) of 30 balloons launched from Marambio. A balloon is launched every 2 days. The simulation is stopped 30 October. After Paparella et al. (1997).

A second goal of these numerical simulations is to explore the extent to which balloon trajectories can contribute to a better understanding of dynamical structure of the polar vortex. Dispersion properties of the flow have been explored by Trounday et al. (1995) by performing a quantitative analysis of the relative dispersion of balloon trajectories using the technique of Morel and Larchevêque (1974) in their study of Eole data. They considered "pair of balloons" defined as a sets of two balloons released from the same location two hours apart. "Chance pairs of balloons" defined as sets of two balloons are close to each other at a given time, regardless of their initial locations were also considered. Trounday et al. conclu-ded that, although individual trajectory may evidence chaotic behaviour, ensembles of trajectory clearly provide robust results on the properties of the flow. They were thus able to study the isotropy, homogeneity, stationarity and diffusivity of the flow showing, in par-ticular, that it exists very different regimes inside and outside the polar vortex. Separation rates of pair of balloons depend on the strength, size and deformation of the polar vortex.

Finally these numerical models were also used to verify that Stratéole goals could be reached taking account of realistic experimental conditions: possible launching sites, meteorological constraints at ground level, limits imposed by technical constraints on launching frequency... This served to define the launch strategy.

IV.4 The Launch Strategy

These three bases were finally selected because all of them have specific advantages for Stratéole:

Marambio (the main station) is the best location to seed the vortex edge (being generally situated at the edge of the vortex) but the potential launching frequency (around 30 balloons per month) is too weak when compared with the needed number of balloons,

Ushuaia allows to seed the vortex edge and exterior and to increase the number of balloon drifting simultaneously during the experiment,

McMurdo is the only base for which a successful seeding of the vortex core is guaranteed.

In order to study the vortex edge dynamics, the strategy is to have the largest number of balloons drifting in the October-November when the vortex begin to break. Thus launch should start early as August in Marambio and Ushuaia. There is no need for a specific launching frequency. Indeed large scale atmospheric motions will homogenise the balloon distribution after a certain amount of time. Nevertheless the possibility of launching balloons per pair should facilitate the study of atmospheric dispersion properties of the flow.

The study of the vortex core, based on the use of McMurdo, will begin later in September. This will allow to study it when it is well isolated up to its final breakdown in January.

The last point is to decide how many balloons should be launched from each station. This is not a simple question as more balloons are drifting, more information is obtained. Nevertheless based on numerical simulations (Trounday et al., 1995; Paparella et al., 1997) a compromise was reached which take the technical constraints into account. The Table 2 give the number of balloons which should be successfully launched from the different stations. This means that a larger number of balloons has to be launched (see Annexe 1).

Station	Marambio	Ushuaia	McMurdo
Number of balloons	70	35	15

Table 2

IV.5 VORCORE and VOREDGE

It has been decided that the Stratéole experiment will be divided in two separate phases:

The VORCORE experiment will concentrate on the vortex core dynamics and during which the launches will be made from McMurdo,

The VOREDGE experiment will concentrate on the dynamics of the vortex edge and on its role in the mixing of air masses of different origins. In this phase Ushuaia and Marambio will be used as launching bases.

Although both experiments could be carried out simultaneously this is not a necessity because the scientific questions to be solved in each phase are relatively uncorrelated. Also, VORCORE can be realised as early as 2000 because it is relatively easy to manage (small number of balloons and no logistical problems). VORDGE induces bigger problems of logistic to be solved. It could take place later than VORCORE.

V. CORRELATIVE MEASUREMENTS

Stratéole is primarily a dynamical experiment allowing an accurate description of the southern polar vortex and its breakdown. It must permit to understand the interaction which exists, in this region, between dynamics and chemistry. In order to obtain the highest benefits from Stratéole project, correlative experiments must be considered during which wind system and/or chemical species could be measured on different time/space scales.

V.1 Radiosondes, Ozonesondes and Ground Based Instruments

There exists, on and close to the Antarctic continent, several stations, a part is included in the WMO network, where radiosondes are launched on a regular basis. The interest of these soundings for Stratéole is obvious because:

they are made on a regular basis (two to four times per day),

they provide vertical profiles of temperature and pressure at some locations whereas Stratéole gives horizontal structures of these fields at only two levels. Nevertheless, it is necessary to obtain the high resolution profiles prior to reduction at standard levels.

In some of these stations, ozonesondes are also launched but on a less regular time basis. The complementarity of these measurements with Stratéole is very attractive as we can hope to follow the movements of ozone rich/poor air masses. A campaign where more frequent ozonesonde launches at the time of Stratéole should be developed in collaboration with scientists in charge of these ozonesondes. Furthermore there also exists lidars devoted to the measurements of ozone and aerosol profiles and other ground based instruments (providing total ozone) which should be included in such a correlative campaign.

V.2 Observations from Space

Another benefit should be obtained from the obvious complementary of Stratéole and satellite missions. Indeed Stratéole documents important small scale events, such as the export of chemically perturbed air from the vortex, which are out of reach for satellite instruments and their limited horizontal resolution. In addition, drifting balloon measurements can be repeated every 10 to 20 minutes, giving accurate sampling of short scale variability, an information not available from space borne instruments. On the other hand, the large number of chemical species which can monitored from space are a necessary complement of the partial chemical picture provided by Stratéole.

ENVISAT, an European Space Agency (ESA) Earth Observation mission, should be launched in 2000, boarding three instruments devoted to stratospheric chemistry. Stratéole experiment is planned to be performed during the ENVISAT mission and offers an unique opportunity for validation (in a region characterized by very sparse in-situ chemical data) and interpretation of these measurements.

V.3 Instrumented Aircrafts

The complementarity of Stratéole and possible instrumented aircraft flights during Stratéole experiment is clearly evident:

Stratéole provides a global horizontal description of the stratosphere, whereas aircraft makes measurements only along their trajectory;

Whereas Stratéole will provide measurements, mainly following the air masses, aircrafts are able to monitor the stratosphere perpendicularly to the air mass trajectory and/or at different altitudes. This must allow to document, among other things, the horizontal extension of possible filaments or isolated vortices observed by Stratéole balloons (see, for example Plumb et al., 1994);

Having no problems of power consumption and weight, aircrafts can host instruments able to measure some important minor species which cannot be monitored by Stratéole.

This evident complementarity shows that the use of instrumented aircrafts should be also envisaged during the Stratéole experiment.

VI. RELATED MODELLING STUDIES

Modelling activities are an important aspect of Stratéole experiment and will help to understand processes related to the vortex dynamics and to analyse the Stratéole experimental results.

The balloon trajectory simulations were a key element for the definition of a launching strategy and the choice of launching sites. They also allow to study mixing and diffusion properties related to the vortex dynamics, in association with tri-dimensional simulations of air masses trajectories (Trounday et al., 1995; Wauben et al., 1997; Paparella et al., 1997). On the other hand, contour advection models, with their excellent resolutions, will permit to study small scale phenomena documented by Stratéole balloons, such as filamentation, which govern the porosity of the vortex edge.

Stratéole will furnish an unprecedented set of wind observations at the high latitudes of the southern hemisphere - practically absent to date. In order to obtain the highest benefits from this observational set, these data will be assimilated in dynamical models and in particular in operational forecast models such as those used in ECMWF or BMRC. This will permit to obtain a better global view of the structure of the stratosphere. In turn, this assimilation should permit to study the systematic bias of these models. it should also allow to obtain better forecasts from these models during the Stratéole experiment.

The Stratéole concept provides a unique opportunity for Lagrangian chemical models to give new insights into the details of the chemical and physical processes occurring during the formation of the ozone hole. Because of their low computing cost, these models can use very short time steps (at sunrise and sunset for instance) and include complex mechanisms such as detailed microphysics. Using trajectory and temperature data constrained by gondola observations, the comparison between the computed and observed evolution of chemical species at the unprecedented temporal resolution of Stratéole will be a stringent test of our understanding of heterogeneous processes and ozone destruction kinetics.

The recent expansion in computer resources is also permitting the rapid development of quite detailed three-dimensional models of stratospheric chemistry. It is likely that these models will have achieved a high degree of complexity and high horizontal and vertical resolutions by the time Stratéole is organised. Because they resolve important processes which cannot be included in Lagrangian models (such as diabatic effects, erosion by planetary waves, mixing) the contribution of 3D chemical models to the interpretation of Stratéole data should be essential and serve as a link between the large scales of satellite observations (from which they are initialised) and the small scales of Stratéole.

APPENDIX 1

The Stratéole Technical System

The project team, including Balloon Divisions of Laboratoire de Météorologie Dynamique (LMD) and Centre National déEtudes Spatiales (CNES, The French Space Agency) have defined a complete system including balloons, gondolas, flight train, ground stations and operation procedures that are presented in this appendix.

The observing system

Pressurised balloon

The vehicle for the mission is a spherical superpressure polyester balloon, designed to drift 3 months at the desired constant stratospheric level. CNES already managed the manufacture and launch of more than 500 balloons for the Eole project (Morel and Bandeen, 1973) in 1971, but this know-how had to be recovered. This is now done at CNES and Zodiac Toulouse and two nominal Stratéole balloons have been launched successfully in December 1997 from Aire-sur-l'Adour (France). The principle of a superpressure balloon is that it keep gas-tight when the internal pressure exceeds the atmospheric pressure so that the internal mass of gas is constant. Due to the non tensible material, it can be assumed that, during the flight, the volume is constant as long as the internal pressure exceeds the external pressure. For more detail concerning Stratéole see, for example, Dubourg et al. (1997,a) or Dubourg et al. (1997, b).

The size of the superpressure balloon is limited by the strength of the material. For a given radius, a superpressure balloon is limited by the changes in internal temperature between a cold limit corresponding to a zero overpressure (for which the balloon descends) and a hot limit corresponding to the maximum allowed stress on the material (for which it fails).

Thanks to Marambio and Ushuaia balloon soundings performed in collaboration with teams in Argentina, a statistical model of Stratéole atmosphere was implemented (Mauroy, 1996a) showing temperature commonly equal to -90°C at flight level. Infrared fluxes, deduced from ERBE data, vary from less than 100 Wm-2 to 240 Wm-2. The nominal solution proposed by CNES for flying at 50 hPa is described in Table 1. For the 70 hPa level, the nominal choice is to use 8.50 m balloons, but attempts will be made to fly a 10 m balloon with an heavier payload (30 kg) for this mission.

Bilaminated polyester offers a good biaxial strength and a quite isotropic material. Its transparency gives a low efficient absorption coefficient minimising radiative effects on the balloon. The material is bilaminated to prevent micro-leak. Numerical models show that a 3 month duration can be obtained with this type of balloon if its equivalent leak (including diffusion and effusion) is less than a 0.15 mm wide pinhole (Mauroy, 1996b).

Flight level :	r = 92.5 gm-3
Spherical balloon, diameter :	10 m
Weight :	25.5 kg
Material :	2x23 mm bi-laminated polyester
Nominal weight at hook :	15 kg
Inflation gas :	helium
Free lift :	20 %

Table 1

In order to avoid any concentration of stress, the flight train is not suspended directly to the south pole of the balloon: a net embodies the envelop and the load is distributed on a large part of the bubble. Each balloon will be completely inflated and tested for tightness on all its surface before launch: the machine designed to perform these tests has been manufactured and is in process of validation.

The Gondola and the Flight Train

The mass allocation for the gondola for a nominal flight is 10 kg. Under definition at LMD, it is composed of a platform and a main payload dedicated to dynamical measurements. An allocation of extra mass and energy exists for a 'passenger payload' (see above).

Platform

The gondola is formed in two decoupled parts, the core which is very compact and hosts the electronics and the external structure which mainly serves as a thermal shelter. The external structure is made of polyestyrene with a cylindrical shape.

The onboard energy is given by a set of Cadmiun-Nickel batteries coupled with solar cell panels. With 2.34 kg, the power block will provide at least 55 Wh per day.

After 6 hours of insolation, used to reload the batteries and to heat the gondola, an instantaneous power of 30 W is available. The available energy depends, of course, on the insolation, i.e. on the drift period and balloon position. The Table 2 indicates the duration for which the instantaneous energy is available at different latitudes and time. Here the threshold is when solar energy at the gondola position is greater than 500 Wm-2.

One sees that, in all cases, instantaneous energy is available at least during 6 hours per day after 15 September; this period being even greater than 18 hours after 15 November for balloon close to the South Pole. So that a passenger needing 30 W can envision to function a large part of the day.

A control device based on the use of a resistant heater will complete thermal insulation provided by the structure. The goal is to maintain cold-fearing subsystems, like Argos oscillator, above -40°C.

	15 August	15 Sept.	15 Oct.	15 Nov.	15 Dec.
Ushuaia (-55°)	3h45	6h	8h15	10h30	11h30
Marambio (-65°)	2h30	6h	9h30	13h15	17h30
McMurdo (-77°)	0h	6h	14h	18h	18h

Table 2

The Argos satellite system has been chosen for the telemetry. An onboard Argos beacon (200 g, 4.32 W/day) will transmit the payload data. The transmission strategy is still under refinement. It is based on the repeated emission of data such that, after a certain amount of time, the probability of reception by an Argos satellite is 1. The data are then send to the ground segment. A second Argos address can be supplied for the passenger payload. Per address, an Argos beacon can transmit to the ground at least 32 kbits at latitudes higher than 50°S.

An onboard supervisor card, using an Intel 186 range microprocessor programmed in C-language, is the base of the onboard automaton that fulfils the following tasks:

power management for GPS receiver, Argos receiver and scientific bus,

acquisition and computing of sensors signals,

acquisition and eventually processing of passenger payload measurements,

data transmission to Argos satellite.

An air safety device is also included. The onboard automaton compares the position provided by GPS to the allowed pre-programmed flight domain: if one limit is reached, an end-of-flight signal is generated and cutting device is electrically activated.

Dynamical payload :

This payload is composed of:

Air temperature sensor for a specified precision of 0.3°K,

Pressure sensor (Digiquartz Paroscientific barometer) with 0.1 hPa precision, after recalibration at LMD,

GPS receiver for 3 dimensional location, weighting 140 g and consuming 2W.

Every extra payload interfacing itself with the Stratéole dynamical payload is so called 'passenger'. In the nominal configuration, the place allowed onboard for a passenger is describe in Table 3.

Weight :	3 kg
Energy:	35 Wh per day at least
Telemetry:	32 kbits per day
Interface:	RS232, other possible
Flight duration:	3 months
Flight level:	50 or 70 hPa

Table 3

It must be noticed that an alternative solution consists in using the 10 m balloon at the lower level (70 hPa) to be able to carry an heavier passenger. The estimated possibilities are shown in Table 4.

Total load (kg)	Payload weight (kg)	Flight level (hPa)
15	10	50
20	15	56
25	20	62
30	25	68

Table 4

One specific electronic card is dedicated to the management of the interface with the passenger. Connected to the scientific bus, this card can provide the passenger with power, commands and data computing.

The flight train

The specified maximum weight for the flight train is 5 kg, including cables and air safety devices. Its length, between the balloon and the gondola, has to be minimised due to operation constraints (see below). The aerostats are equipped with the following devices:

a pressure sensor, mounted on the top valve measures the internal pressure,

a pyrotechnic cutting device, commanded from the gondola, coupled with a barometric system to put an end to the flight as soon as the lower latitude limit of flight is reach,

a passive radar reflector for detection of the aerostat by planes,

a flashlight for visual detection of the flight train during the ascent phase.

Tests of the aerostat

Due to the need of production and launch of 200 balloons, strict definition and management of the production has been realised.

Qualification tests will be carried out mainly from Aire-sur-l'Adour, in France, until mid 1998. The goals are to verify:

procedures for preparation before launch including helium tightness test,

launch technique,

ascent phase of the balloon with a high rate of free lift,

abilities of the balloon in function of the load at hook,

inflight functional test of the gondola.

Once the development tests will be achieved, the aerostats will undergo long duration flight tests (at least one month) to validate envelop strength to cycle of overpressure in day/night transition, helium tightness in very cold condition and functional qualification of gondola in the real experimental conditions. The validation of balloon packaging for transport will also be an aim of these tests because the material is fragile and micro-leaks can be formed during this phase.

For these tests, 3 to 5 balloons will be launch by the end of 1998 from South America or from Arctic. It is thought that a complementary specific campaign will be necessary to validate all the procedures and the aerostats in the extreme conditions of Stratéole. Later on qualification campaign will be made from the selected launching sites before the Stratéole campaign. It is mainly aimed to qualify the operations from these sites, the logistic supply and the ground section.

Launch operations

The balloon material, for being rigid, fears to be shaken on the launch pad and during the ascent. Therefore launch is allowed only in calm ground conditions, that means surface wind velocity smaller than 2 ms-1. Moreover, a specific launch technique will be implemented to avoid sail effects at launch: the bubble will always be loaded until the whole flight train is erected and the gondola is lifted by the balloon (see Figure).

An important point in Stratéole system definition was to find suitable places to launch the aerostats which combine both scientific purposes and technical feasibility. A search was done by CNES (Bidau, 1996) with a visit to the potential sites in Argentina. Its resulted in the nominal choice of Ushuaia, Marambio and McMurdo. The enquiry was based on the following criteria:

ground meteorological conditions,

logistics and telecommunications,

possibilities of access during winter,

possibility to built a large hangar used as shelter to prepare balloons in advance before launch.

Implementing a reliability coefficient, it is thought to launch 90 balloons from Marambio, 40 from Ushuaia and 20 from McMurdo.

Marambio (56°38 W, 64°14 S) is a small island belong to the Antarctic Argentina, close to the Antarctic Peninsula. The lift base is built on a table-land, 200 meters above the sea level. It is managed by Argentine Air Force (FAA) that offers logistic support to scientific teams. It can be reach all year long by C-130. Analysis of surface wind, provided by the Argentine meteorological service (SMN) shows that launches are possible from there. The location and preliminary plans of the preparation hangar has been studied in situ by Argentine and CNES experts. It will use specific technology and will be based on piling because of permafrost soil

Ushuaia (68°19 W, 54°48 S) is a town located at the extremity of Tierra del Fuego. It is reachable all year by plane and boat and offer all necessary logistic support. Analysis of surface wind shows that there is no problem to launch the required number of balloons in a 3 month period.

McMurdo (166°37 E, 77°51 S) is a US base managed by the National Science Foundation (NSF). Big stratospheric balloons are launched there is summer. It is not foreseen to build an hangar. Meteorological data indicates that opportunities of launching are similar to those of Marambio.

Schematic of the Stratéole Aerostat.

APPENDIX 2

The Envisioned Stratéole Passengers

This appendix describes the three passengers which are actually foreseen to be hosted in some of the Stratéole gondolas.

The Total Direct Diffuse Radiometer

The Total Direct Diffuse Radiometer (TDDR) has been developed at NASA Ames Research Center for aerosol measurements; it has been successfully deployed during field mission in the Arctic using aircraft as a platform (Valero et al., 1989).

The TDDR avoids the complexity of sun tracking devices by using an optical system which is capable of separating the contribution of the direct (parallel) solar beam from the total hemispheric radiative flux. This is achieved by incorporating an oscillating shadow ring in the front of an hemispherical field-of-view radiometer. At some point during the oscillation cycle, the ring projects a shadow which excludes the direct solar beam from the field of view of the radiometer; at this point only the scattered or diffuse component Fd of the total radiation filed reaches the aperture of the optical system. On the other hand, when the ring is out of the field of view, the total hemispherical radiation field Ft is detected. The resulting measurement of the direct solar flux

Fs = Ft - Fd

is then compared with exo-atmospheric values, acquired from Langley method calibrations, to determine the optical depth along the sun direction.

For the Stratéole project, the instrument has to be modified in the following way:

substitution of three O3 and NO2 channels to three of the present seven aerosols channels (Table1);

adaptation to balloon technology, power consumption and weight.

Because TDDR does not track the sun, there is a cosine dependence of the measured flux with the solar angle, demanding that the orientation of the balloon platform with respect to the sun be accurately determined. This can be achieved by adding a second shadow ring having its axis orthogonal to the first one, the two rings oscillating in an alternating fashion. In addition, 3 orthogonal flux-gate magnetometers will be available on the platform. TDDR is already under modifications to be adapted to Stratéole gondola. Two balloon flights were also successfully carried out in 1993, although not with its definitive Stratéole version (Valero and Pilewskie, 1994).

Channel number	Center l (nm)	Bandwidth (nm)	Measured Species
1	380	10	Aerosols, NO2 *
2	448	2	Aerosols, NO2, O3
3	453	2	Aerosols, NO2, O3
4	500	10	Aerosols, NO2, O3
5	600	10	Aerosols, NO2, O3 *
6	862	10	Aerosols
7	1064	10	Aerosols
(* Absorption maximum selected for each particular spectral band)

Table 1

In the selected configuration, the instrument will be able to measure:

the integrated column O3 content,

the integrated column NO2 content,

aerosol optical depths.

In addition it will be able to detect Polar Stratospheric Clouds above the balloon drift level.

Radiometric measurements are of importance for Stratéole because:

they are optimally suited for Stratéole balloons in term of weight and power budget;

they can serve as direct references for satellites observations, because they are of similar type, have comparable horizontal scale and have the advantages of better accuracy (looking upwards) as well as better sampling in time (10-20 minutes) and in space (20 to 100 km).

The Infra Red Broad Band Radiometer

Possible addition of the InfraRed Broad Band Radiometer provided by SCRIPPS Institution to measure the upwelling infrared radiation flux is under study. Such an addition, as explained above, would be most valuable for investigating the interaction between radiative processes and vortex dynamics.

This electrically calibrated pyroelectric radiometer differs from other radiometers in the use of optical chopping and null-balanced operation. Optical chopping is made possible by the fast response of the lithium tantale pyroelectric detector and is effective in eliminating drift. A gold-black coating of the detector gives it a spectral range from the ultraviolet to the far infrared. Null-balanced operation, in turn, renders the resulting measurement insensitive to detector responsivity and amplifier gain and hence to ambient temperature. An airborne model of this instrument is described by Valero et al. (1982, 1984).

The Lightweight Atmospheric Methane Analyser

Methane is a relatively good passive tracer in the lower stratosphere. Thus its measurements permit to observe origin and mixing of the air masses near the vortex edge. This allows also estimation of the diabatic descent inside the vortex during polar night. During daytime, in the lower stratosphere, methane is oxided by hydroxide radical, this reaction becoming more important at higher altitudes. But, simultaneous measurements of methane and water vapour allow to make a balance of hydrogen and permit to deduce the degree of deshydratation of the lower stratosphere induced by the formation of PSCs.

LAMA, whose feasibility is under study, is a proposed light-weight near infrared diode laser spectrometer which should be able to measure very low CH4 mixing ratio (of the order of 1 ppmv) at the pressure level of 70 hPa. The strongest spectral signatures of many atmospheric constituents appear in the mid-infrared (5-10 mm) where the fundamental vibration-rotation bands can be used for in-situ measurements. But this is only possible with cryogenic detectors which are inconceivable for operation in Stratéole payloads. LAMA takes advantages of progress made in the field of near-infrared telecommunication diodes at 1.3 mm and 1.55 mm, which could allow to measure CH4 by in-situ absorption spectrometry by this type of diodes. The proposal, more specifically, is to study the Q branch of the 2n3 band of CH4 around 1.66 mm and the n2+2n3 band around 1.33 mm. The absorption to be measured for a 10 m path (realistic on a small balloon) at 70 hPa and for a nominal 1.8 ppmv mixing ratio of CH4, should be of order of 10-5 for lines in the strongest 1.66 mm region and 10-6 for lines in the weakest 1.33 mm region. The latter band is of great interest because it includes a micro window where it is possible to monitor also H2O simultaneously with CH4.

The required level of precision on these values should be of the order of 5% which translate into an absorption noise level, at the edge of the most sensitive systems described up to now (Camy-Perret, 1994 and references therein). Nevertheless, the intrinsic small size and low power consumption of near-infrared laser diodes are very attractive characteristics for the Stratéole project. The feasibility of such a detector is now in industrial R & D.

REFERENCES

Babiano A., C. Basdevant, P. Leroy, and R. Sadourny, 1990: Relative dispersion in two-dimensional turbulence, J. Fluid Mech., 214, 535-557.

Bauer R., A. Engel, H. Franken, E. Klein, G. Kulessa, C. Schiller, U. Schmidt, R. Borchers, and J. Lee, 1994: Monitoring the vertical structure of the Arctic polar vortex over Northern Scandinavia during EASOE: Regular N2O profile observations, Geophys. Res. Lett., 21, 1211-1214.

Bidau M., 1996: Historique du choix des stations de lancement des ballons Stratéole, STR-NT-5-69-CNES.

Cadet D., H. Ovarlez, and G. Sommeria, 1981: The BALSAMINE Experiment during the summer MONEX, Bull. Amer. Meteor. Soc., 62, 381-388.

Camy-Perret C., 1994: .A proposal for measuring methane using near-infrared diode laser in-situ spectroscopy during Stratéole, Proceeding of the second Stratéole Workshop, CNES, Paris.

Cariolle D., A. Lessere-Bigorry, J. F. Royer, and J. F. Geleyn, 1990: A general circulation model simulation of the springtime Antarctic ozone decrease and its impact on midlatitudes, J. Geophys. Res., 95, 1883-1898.

Desbois M., 1975: Large-scale kinetic energy spectra from Eulerian analysis of EOLE wind data, J. Atmos. Sci., 32, 1838-1847.

Dubourg V., F. Nouel, M. Bidau, H. Laplace, A. Vecten, Ph. Coquerez, Ph. Mauroy and P. Malaterre, 1997: The Stratéole observing system, Proceedings 13th ESA Symposium on Rockets and Balloons Programmes & Related Research, ESA SP-370, in press.

Dubourg V., F. Nouel, M. Bidau, H. Laplace, A. Vecten, Ph. Coquerez, Ph. Mauroy and P. Malaterre, 1997: The Stratéole project status: 200 pressurized ballon for the polar vortex study, A.I A A 97-1410, in press.

Edouard S. and R. Vautard, 1994: Construction of Potential Vorticity maps using ECMWF analysis, Proceeding of the second Stratéole Workshop, CNES, Paris.

Farman J. C., B. G. Gardiner, and J. D. Shanklin, 1985: Large losses of ozone in Antarctica reveal seasonal ClOx/NOx interaction, Nature, 315, 207-210.

Hart T., 1994: : .Daily and interannual variations in the edge of the Southern Polar Vortex during late winter, Proceeding of the second Stratéole Workshop, CNES, Paris.

Juckes and M.E. McIntyre, 1987: A High resolution one-layer model of breaking planetary waves in the stratosphere, Nature, 328, 590-596.

Julian P. R., 1981: The tropical observing system in FGGE: an analysis of results, ICSU/WMO GARP Report of the International Conference on Early Results of FGGE and Large-Scale Aspects of its Monsoon Experiment, 2, 2-9.

Legras B., and D. G. Dritschel, 1993: Vortex stripping and the generation of high vorticity gradients in two-dimensional flows, Appl. Sci. Res., 51, 445-455.

McIntyre M. E., 1989: On the Antarctic ozone hole, J. Atmos. Terr. Phys., 51, 29-43.

McIntyre M. E., 1995: The stratospheric polar vortex and sub-vortex: fluid dynamics and midlatitude ozone loss, Phil. Trans. R. Soc. Lond., A 352, 227-240.

Mariotti A., M. Moustaoui, B. Legras and H. Teitelbaum, 1997: Comparisons between vertical ozone soundings and reconstructed potential vorticity maps by contour advection with surgery, J. Geophys. Res, 102, 6131-6142.

Mauroy Ph., 1995: Performance des ballons Stratéole entre août et décembre, Internal report, CNES.

Mauroy Ph., 1996: Stratéole: modèles d'ingénierie d'atmosphère; nouvelle présentation, Internal report, CNES.

Morel P., and W. Bandeen, 1973: The EOLE experiment, early results and current objectives, Bull. Amer. Meteor. Soc., 54, 298-306.

Morel P., and M. Larchevêque, 1974: Relative dispersion of constant-level balloons in the 200 mb general circulation, J. Atmos. Sci., 31, 2189-2196.

Ovarlez J., J. Capus, M. Forichon, and H. Ovarlez, 1989: Mesure in-situ d'humidité dans l'atmosphère moyenne, Proceedings, 9th ESA Symposium on European Rocket and Balloon Programmes and Related Research, ESA SP 281, 43-47.

Ovarlez J., and H. Ovarlez, 1994: Stratospheric water vapour content evolution during EASOE. Geophys. Res. Lett., 21, 1235-1238.

Paparella F., A. Babiano, C. Basdevant, A. Provenzale, et P. Tanga, 1997: A Lagrangian study of the Antarctic Polar Vortex, J. Geophys. Res., 102, 6765-6773.

Peter P., 1997: Microphysics and heterogeneous chemistry of polar stratospheric clouds, Ann. Rev. Phys. Chem., 48, 785-822.

Plumb R. A, D. W. Waugh, R. J. Atkinson, P. A. Newman, L. R. Lait, M. R. Schoeberl, E. V. Browell, A. J. Simmons, and M. Loewenstein, 1994: Intrusions into the lower stratospheric Arctic vortex during the winter of 1991-1992, J. Geophys. Res., 99, 1089-1105.

Pollack J.B., and C.P. McKay, 1985: The impact of polar stratospheric clouds on the heating rates in the winter polar stratosphere, J. Atmos. Sci., 42, 245ñ262.

Schoerbel M. R., R. S. Stolarski, A. J. Krueger, and P. A. Newman, 1989: The 1988 Antarctic ozone depletion: comparisons with previous years, Proceedings 7th Conference on the Meteorology of the Middle Atmosphere, 19.

Solomon S., 1988: The mystery of Antarctic ozone hole, Rev. Geophys., 26, 131-148.

Teitelbaum H., J. Ovarlez, H. Kelder, and F. Lott, 1994: Some observations of gravity wave-induced structures in ozone and water vapour during EASOE. Geophys. Res. Lett., 21, 1483-1486.

Teitelbaum H., M. Moustaoui, J. Ovarlez, and H. Kelder, 1996: The role of atmospheric waves in the laminated structure of ozone profiles at high latitude, Tellus, 48A, 442-455.

Teitelbaum H., M. Moustaoui, P.F.J. Van Velthoven and H. Kelder, 1998: Decrease of total ozone at low latitudes in the southern hemisphere by a combination of linear and non-linear processes,Q.J.R. Meteorol. Soc., 124, 2625-2644.

Teitelbaum H. and R. Sadourny, 1998: The role of planetary waves in the formation of polar stratospheric clouds, Tellus, 50A, 302-312.

Trounday B., L. Perthuis, S. Strebelle, J. D. Farrara, and C. R. Mechoso, 1995: Dispersion properties of the flow in southern stratosphere during winter and spring, J. Geophys. Res., 100, 13901-13917.

TWERLE Team, 1977: The TWERL Experiment, Bull. Amer. Meteor. Soc., 58, 936-948.

Valero F. P. J., W. J. Y. Gore, and L. P. M. Giver, 1982: Radiative flux measurements in the troposphere, Apll. Optics, 21, 831-840.

Valero F. P. J. T. P. Ackerman, and W. J. Y. Gore, 1984: Radiative effects of the Arctic haze, Geophys. Res. Lett., 11, 465-468.

Valero F. P., T. P. Ackerman, and W. G. Gore, 1989: The effects of the Arctic haze as determined from airborne radiometric measurements during AGASP II, J. Atmos. Chemistry, 9, 225-344.

Valero F. P. J., S. Platnick, S. Kinne, P. Pilewskie, and A. Bucholtz, 1993: Airborne brightness temperature measurements of the polar winter troposphere as part of the Airborne Arctic Stratosphere Experiment II and the effect of brighness temperature variations on the diabatic heating of the lower stratosphere, Geophys. Res. Lett., 20, 2575-2578.

Valero F. P., and P. Pilewskie, 1994: Total Direct Diffuse Radiometer (TDDR) test flight, Proceeding of the second Stratéole Workshop, CNES, Paris.

Wauben W. M. F., P. F. J. Van Velthoven, and H. Kelder, 1994: Dispersion of particles in and around the Antarctic vortex, Proceeding of the second Stratéole Workshop, CNES, Paris.

Wauben W. M. F., R. Bintanja, P. F. J. van Velthoven, et H. Kelder, 1997: On the magnitude of transport out the Antarctic Polar Vortex, J. Geophys. Res., 102, 1229-1238.