1. Introduction

Stratospheric ozone depletion has been one of the major environmental issues for the past two decades and will require monitoring for many decades to come (WMO 1995). After the discovery of the ozone hole over the Antarctic in the mid-1980s, evidence of ozone destruction has also been found in the Northern Hemisphere, particularly since 1995, with ozone losses reaching almost 40% in the polar lower stratosphere in early spring (e.g., Isaksen et al. 1997; Rex et al. 1997; Muller et al. 1997; Goutail et al. 1999).

There are several ways to make the measurements that are necessary to monitor and understand ozone depletion. Remote sensing, from the ground or from space, is a powerful technique to study many atmospheric parameters, but one that has important limitations. From the ground, observers must wait for interesting atmospheric phenomena to pass overhead; from space, the horizontal and vertical resolution of measurements is often insufficient to capture cloud processes and mixing processes, which have a large impact on the chemical evolution of the stratosphere. Balloon-borne measurements can probe the stratosphere, but have only a few launch sites and require calm ground conditions for launch. The use of stratospheric aircraft makes it possible to fly instruments where and when the interesting phenomena occur, to a much greater extent. An aircraft undertaking scientific missions in the ozone layer should be capable of reaching altitudes of at least 20–21 km, capable of carrying payloads of more than one metric tonne, and capable of supplying several tens of kilowatts of energy to instruments.

In Europe such a plane has never before been available. Presently, the United States has two research aircraft that can carry heavy payloads to the stratosphere:the ER-2 and the WB-57. The former has a long and distinguished record of missions to study stratospheric chemistry and physics; the latter has recently carried out its first mission. Various versions of the unmanned aircraft of the Aurora Corporation are planned but are still under development. The Altus UAV (unmanned aircraft) has already been used by the Atmospheric Radiation Measurement (ARM) Program (see results at www.gat.com/asi/aero.html). Unmanned aircraft are an interesting and promising development, but today they can carry only light and small payloads.

The Russian M-55 Geophysica (Stefanutti et al. 1995), on the other hand, is able to meet all the above requirements. The Geophysica can carry heavy payloads, it has a very generous power supply on board, and it can operate in the lower stratosphere for several hours. This aircraft was selected for the Airborne Polar Experiment (APE; APE 1), and will continue as the main platform for the Airborne Platform for Earth observation program.

APE 1 may be divided into two main fields: science and technology. The first scientific results will be reported in a forthcoming special section of the Journal of Geophysical Research. Validation and technical issues, related to individual instruments, are reported in other papers appearing in this special issue. This paperoutlines technical issues related to the use of the Geophysica as a platform for research in the polar lower stratosphere.

The mission specifically aimed to fly in regions where lee wave clouds are likely to be found (i.e., the Scandinavian mountains and the Russian Urals) and in regions where synoptic-scale polar stratospheric clouds (PSCs) occurrence was most probable. The combination of high-frequency in situ measurements, new flight paths, meteorological analyses, and detailed modeling aimed to improve understanding of the spatial and temporal extent of PSCs, and hence ozone loss. These scientific objectives demanded high performance from the mission platform. Below, we describe first the aircraft and the instrument location, then the most significant flight paths accomplished during APE 1, focusing on novel elements.

2. The M-55 Geophysica aircraft

The Geophysica (Fig. 1) is a dual turbofan aircraft designed by Myasishchev Design Bureau (MDB). The aircraft can carry payloads of up to 1500 kg and can operate at an altitude up to 21 km for about 5–6 h. The cruise ground speed is about 750 km h⁻¹. The Geophysica has various bays where the instruments can be installed: in the fuselage, on the wings, and in the tail booms (Fig. 2). The main bay is longer than 5 m, so that very large instruments can be installed. In APE 1, this bay was occupied by the Airborne Lidar Experiment (ABLE; 350 kg, 1.8 m), Spectroscopy of the Atmosphere using Far-Infrared Emission (SAFIRE; 405 kg, 1.8 m), and Multiwavelength Aersol Scatterometer (MAS; 25 kg, 0.65 m) (see Fiocco et al. 1999; Carli et al. 1999; Adriani et al. 1999; respectively). The aircraft has a large reserve of electrical power for the instrumentation (60 KVA at 115 VAC, 400 Hz and 3 KW at 27.5 VDC). Flight parameters, necessary for the science mission, are regularly recorded by the aircraft and distributed to each instrument. Takeoff and landing distances are less than 1000 m, but a runway of at least 2000 m is required for safe operation. With its two engines and triangular undercarriage, the Geophysica is a very robust design and so can operate in a wide range of ground weather conditions, notably tolerating a maximum crosswind of 10 m s⁻¹.

Comparing the Geophysica to the workhorse of stratospheric research of the last 15 years, the ER-2, the M-55 Geophysica can be said to exceed the ER-2’s capabilities in payload, power supply, maneuverability, and is said to be less dependent on ground conditions. However, the ER-2 exceeds the Geophysica in range: about 4070 km compared to the Geophysica’s 3700 km. The two aircraft have almost the same maximum altitude, reaching a pressure altitude of about 21 500 m.

3. Some details of the current M-55 Geophysica payload

The location of each instrument is indicated in Fig. 2. The payload can be broken down into a number of different measurement types: in situ aerosol instruments, remote sensing aerosol instruments, in situ chemical instruments, and remote sensing chemical instruments (Table 1).

4. Technological tests on the instrumentation

All scientific instruments had to meet design specifications agreed between the PIs and MDB, the designers of the aircraft. These design specifications defined all the tests that mock-ups and instruments had to undergo in order to be mounted on the aircraft. The tests were subdivided into various phases:

1. vibration and shock tests on the mock-ups;

2. mock-up test flights;

3. electro magnetic interference (EMI) tests on the instruments;

4. baroclimatic tests on the instruments; and

5. instrument test flights.

The chief result of phase 1 was to highlight the importance of accurate finite element analysis in the design of instruments, particularly large and heavy instruments like SAFIRE and ABLE. For phase 2, MDB carried out all the necessary modifications to install the scientific instruments, and organized mock-up flights at their base in Zhukovsky, Russia. The mock-ups of ABLE, SAFIRE-A, GASCOD-A, MAS, and FSSP-300 were flown in three different flights during the second part of August 1996. During the mock-up flights, temperatures and vibrations were measured in critical points of each instrument.

EMI tests (phase 3) started on May 1996 at Ente per le Nuove Technologie, l’Enerigia e l’Ambiente (ENEA), and continued until the end of September 1996. No instrument passed the very stringent RTCA–DO-160C specifications at the first attempt. In September, baroclimatic tests (phase 4) were carried out at the military airport of Pratica di Mare.

Three instruments test flights, at Pratica di Mare, Italy, of approximately 3 h each, were performed. During these test flights, ozone sondes were launched from L’Aquila. Pressure, temperature, humidity (PTU) sondes were regularly launched from Pratica di Mare. An office was established at the airport where stratospheric analyses and forecasts from the European Centre for Medium-Range Weather Forecasting were delivered daily via the NADIR database at Norsk Institutt for Luftforskning (NILU), Norway. The sonde data and meteorological analyses were used to check instrument data for general consistency. Forecast data were used for flight planning. Mission management structures were tested, specific flight patterns tried, instrument behavior checked, modifications made to some instruments, and valuable scientific information collected.

One additional in-flight test was carried out during the mission itself. On 23 December 1996, the temperature sensors on board the Geophysica were compared with those on board the Falcon. The intercomparison occurred during a period of approximately 15 min, from 0925 UTC to 0940 UTC at 10 600- to 10 900-m altitude, over northern Sweden, with temperatures ranging between −54° and −56°C. The aircraft were in sight of each other during the intercomparison period. The difference in the recorded temperatures was smaller than one degree (Balestri 1998, personal communication). This intercomparison could be performed only at the Falcon’s maximum altitude, of course, and does not guarantee that the Geophysica’s sensors are reliable at higher altitudes and much lower temperatures. Their performance at high altitude was tested during the mock-up flights in August 1996 in Moscow, when several radiosondes were launched close to the airport. Good agreement was obtained at stratospheric levels between the aircraft temperature sensors and the PTU data. Although in this case the spatial match of temperature data is not as good as in the aircraft intercomparison, the quiescent conditions of the summer stratosphere mean that temperatures measured at one point are representative of a much larger area than in the wintertime lower stratosphere. The combination of these two intercomparisons gives us some confidence in the reliability of the temperature and pressure measurements on board Geophysica.

5. Flight planning

Conflicting measurement constraints arose between different instruments: MAL and MAS could operate only during nighttime, while GASCOD-A required sunlight to make measurements. To measure photochemical radical species, measurements by SAFIRE-A were best performed during the day. Moreover, FSSP-300, MAS, MAL, FOZAN, ECOC, ACH, and FLASH stood to benefit from slow ascents and descents, in order to maximize vertical profile resolution, while the remote sensing instruments—ABLE and SAFIRE—stood to benefit from rapid ascents that would maximize time at cruise altitudes. In the event most flights included short periods of daylight (Table 2).

The orientation of the flight path, with respect to the prevailing stratospheric flow, was also an important consideration. The maximum gradients in long-lived tracer fields occur across the flow, particularly across the polar vortex edge (McIntyre 1992; Plumb and Ko 1992). Therefore, in order to sample the greatest range of“equivalent latitudes” (equivalent latitude is vortex-centered, rather than pole-centered; see, e.g., Norton 1993), flights across the flow are preferred. However, the most useful information in aerosol microphysics is that recorded as closely to Lagrangian as possible, that is, along the direction of the mean flow. This is particularly important were the size of the aerosol perturbation (a PSC, say) is smaller than the aircraft’s range, so that both onset and end of the episode can be sampled. Mountain-induced waves are the most important example of such perturbations. Two possible flight paths were investigated: to fly with the wind, or to fly perpendicularly to the wind, turning back in such a way as to cross the same air parcel several times. In the first case stationary conditions are assumed along the wind pattern, since the aircraft was flying at much higher speed than the wind itself, and hence could not follow physically an air parcel. The second case requires rapid calculation of airmass trajectories and assumes that changes occur on timescales close to that of the encounter frequency. For mountain-wave clouds, changes have been observed on a much smaller scale than could be sampled using this second technique (e.g., Fiocco et al. 1997), and so the first solution was adopted.

6. The inaugural mission—APE-1 in Rovaniemi, Finland, 1996–97

The Arctic campaign started in the second part of December 1996. In a late addition to the plan, parts of the German Polar stratospheric clouds, Lee waves, Chemistry, Aerosols and Transport (POLECAT) program were coordinated with APE 1. The parts of POLECAT included in the joint mission were the Ozone Lidar Experiment (OLEX) aerosol lidar, flying on the DLR Falcon (Wirth and Renger 1996), ground-based lidar activities in Andoya, Sodankylä, and Ny Ålesund (e.g., Brogniez et al. 1997; Dahlback et al. 1994), and mesoscale modeling of mountain wave propagation (Dörnback et al. 1997). The coordinated activity of the Geophysica and the Falcon consisted in using the Falcon as a PSC pathfinder for the Geophysica. Using real-time, in-flight, analysis of the OLEX signal, the M-55 pilot was informed of the presence of PSCs by air-to-air communication. The M-55 was then able to modify his path, within predetermined limits, according to the information received from the Falcon. This is the first time such pathfinding has been attempted in stratospheric research.

During the campaign the Geophysica performed seven scientific flights in 23 days (an average of about one flight every 3 days), in addition to the transfer flights from Pratica di Mare, Italy, to Rovaniemi at the beginning of the campaign and from Rovaniemi to Moscow, Russia, at the end. Total flight time during the campaign was 29 h and 26 min (corresponding to about 19 000 km) plus 8 h and 15 min for the transfer flights (about 5500 km). The last scientific flight (14 January 1997) was the longest (3418 km in 5 h and 38 min) and the one in which the M-55 reached the highest pressure altitude (21 200 m). More details on the flights are given in Fig. 3 and Table 2.

During the first part of the campaign, up to the end of 1996, no PSC events occurred. However, the arctic stratosphere was very active dynamically. Therefore, the first three flights were devoted to the study of stratospheric dynamics and mixing processes. From the beginning of 1997, the middle stratosphere started to cool, and the threshold temperature for PSCs on the 550 K isentropic level was reached. From 5 January, sporadic areas of low temperatures were forecast, generally in the region between Spitzbergen and Greenland. On 9 January, strong lee wave activity was predicted over the Norwegian mountains. Sorties were flown to investigate all these features (Stefanutti et al. 1998). In order to demonstrate the capability of the Geophysica, two particular flights will be described here: the flight of 29 December 1996 and the flight of 9 January 1997.

On 29 December 1996 the vortex was strongly deformed from zonal symmetry by a strong intrusion of subtropical air into the polar region at about 0°. The cold center of the vortex was close to Novaja Zemlia, although temperatures were still above the PSC threshold. It was therefore decided to fly toward Novaja Zemlia to investigate different air masses and air mixing processes, and to perform a dive at the farthest end of the flight, to investigate the aerosol structure all the way down to the tropopause near the centre of the vortex. This flight was also intended to investigate the behavior of the various instruments during a deep dive.

The flight started at 1230 UTC and ended at 1640 UTC. Hence, part of the flight took place during the day and part during the night, enabling some sampling time for all instruments. A dive was performed from an altitude of 19 to 9 km, followed by reascent to maximum altitude, above 20 km. The dive, which is the deepest ever performed by a stratospheric research aircraft, was carried out close to Novaja Zemlia. The results of MAL, FSSP-300, ECOC, FLASH, and ACH show very low levels of background aerosols for all levels sampled (S. Borrmann 1997, personal communication). Figure 4 shows the altitude and temperature measured during the flight. The flight path was from the border of the vortex to the centre of the vortex (cf. Fig. 3 and Fig. 5).

On 9 January 1997 the vortex was more nearly zonal, as can be seen in Fig. 6. Mesoscale forecasts using The Pennsylvania State University/National Center for Atmospheric Research MM5 model indicated high probability of strong lee wave activity over the Norwegian mountains and northern Sweden (Fig. 7). Therefore, a “quasi-Lagrangian” flight through this wave activity was planned. The Falcon, equipped with the OLEX lidar acted as pathfinder. Along with the scientific goal of observing local PSC structure through the entire length of a cloud using both remote and in situ techniques, two technical objectives were defined: (i) to modify the Geophysica flight plan during the flight itself, if requested by the pathfinding aircraft; and (ii) to fly the Geophysica safely in the presence of strong mountain wave activity.

Although the PSCs observed during this flight were above the Geophysica flight altitude, and so could not be sampled using in situ instruments, remote sensing measurements were made from both the Falcon and the Geophysica. In addition, both technical goals were achieved. The Falcon preceded the Geophysica by about 30 minutes and, by means of direct radio contact, the flight path of the Geophysica was modified. For example, in the original flight plan, the possibility of the Geophysica flying farther out into the Arctic Ocean was considered. When the OLEX lidar did not detect any PSCs west of the mountains, the Geophysica shortened its flight in this direction and returned to the region of greater lee wave activity (see Fig. 8).

Strong lee wave activity was encountered downstream of the Norwegian mountains, as expected by the mesoscale forecast. The pilot reported a very bumpy ride, and the flight data recorded on the Geophysica bear him out. Rapid temperature and altitude oscillations were recorded by the aircraft sensors; these are probably due to mountain waves. The Geophysica flew through this area without difficulty. High-altitude PSCs were measured downwind of the Norwegian mountains both by the M-55 Geophysica and the Falcon during the flight. ABLE measured PSCs above the aircraft during two different legs of the flight between 1659 and 1705 UTC and 1806 and 1810 UTC (G. Fiocco 1997, personal communication). These PSC measurements occurred simultaneously with the temperature oscillations recorded by the aircraft (see Fig. 9). Major changes from the original plan were performed during the flight. This required manual piloting of the Geophysica for an extended period. This flight, we believe, demonstrated that mesoscale stratospheric forecasting and pathfinding aircraft can significantly enhance the scientific payback of missions in the wintertime polar lower stratosphere.

7. Summary

The early indications of the amount of scientific data of value retrieved from the different flights, the close spacing between one flight and the next, and the experimental flight paths all indicate that the APE 1 was a success, technically and scientifically. The mission was a key stage in the continuing development of the Geophysica for scientific research—a successful “proof of concept.” Much experience has been gained on the aircraft’s capabilities, on instrument development and integration, and in campaign management. The aircraft had never flown before in a cold stratosphere and had never flown over mountain waves. To our knowledge, this is the first report of such a deep dive and of quasi-Lagrangian flying guided by a pathfinder aircraft.