1. Introduction
Global warming has recently become one of the major international social concerns. To address this problem, it is important to know how the greenhouse gases are distributed and how they affect the climate. To ascertain the global dynamics and chemistry of the atmosphere, fundamental data of the atmosphere, such as the density and the proportion of various atmospheric compounds, must be monitored in both the troposphere and stratosphere at various locations.
Since 1985, we have carried out a series of balloon-borne air sampling experiments in collaboration with several other institutes in Japan. We have developed whole air samplers to collect atmospheric samples at various altitudes, especially in the lower stratosphere, by using the cryogenic pumping technique (Honda 1990; Honda et al. 1996; Honda 2001; Morimoto et al. 2009). A large plastic balloon filled with helium gas is used to bring the sampler above aircraft altitude. The whole air sampling method has an advantage in high-resolution, high-precision measurements of local absolute concentration and is complementary with other methods, such as remote sensing (which is suitable for continuous broad-area measurement) and in situ real-time measurement (which is suitable for quick and high-frequent measurement).
In these 28 years, we succeeded in air sampling many times repeatedly, not only over the midlatitude region (19 flights over Sanriku or Taiki in Japan) but also over the north polar region (2 flights over Kiruna, Sweden) and the south polar region (9 flights over Syowa, Antarctica) (Aoki et al. 2003). As a result, concentration data of various greenhouse gas compounds (such as carbon dioxide, methane, dinitrogen monoxide, and sulfur hexafluoride) and isotopes (such as nitrogen, oxygen, carbon monoxide, hydrogen, and argon) have been monitored. High-quality data obtained with these repeated measurements using an identical apparatus (and thus small systematic errors) show certain long-term changes of greenhouse gas densities. The precise data obtained at various latitudes as well as at various altitudes is very useful to discuss the regional differences of proportion for both major and minor compounds. In this manner, our data have been utilized to elucidate the transportation process and chemical processes in the atmosphere. Precise analyses of the obtained air samples also led us to the discovery of the gravitational separation of atmospheric components in the vertical distributions in the stratosphere (Ishidoya et al. 2008).
However, we still have a data gap. It is important to investigate the air over the equator in the lower stratosphere. In the equatorial troposphere, the atmosphere rises from the surface and flows poleward in the upper troposphere, which is well known as the tropical atmosphere circulation of the Hadley cell. Above the upper troposphere, the equatorial region is known as the starting point of the Brewer–Dobson circulation, which transports the tropical atmosphere poleward in the stratosphere. By comparing the “fresh” air at the equatorial region with the data obtained at lower streams in midlatitude and higher latitude, it becomes possible to discuss in detail the transportational and chemical processes in the overall Brewer–Dobson circulation. Nevertheless, few air samplings have been carried out over the equatorial region. Therefore, in order to fill this data gap, we planned to sample the air over the equator. And to realize it, we developed a novel balloon flight operation method.
2. Operation outline
Figure 1 shows a conceptual diagram of the flight sequence. The balloon is launched from the deck of an ocean vessel. During the balloon ascent of 1~2 h, the sampler collects the air for 5–10 min at predefined altitudes in the stratosphere between around 15 and 30 km. The telemetry data are received on the vessel to monitor the balloon position and the sampler status. Completion of the sampling triggers the flight termination. The sampler parachutes down onto the sea and is recovered by the vessel itself.
A conceptual diagram of the flight sequence. The air sampler launched by a balloon from the deck of Hakuho Maru samples the air during the balloon ascent and then parachutes down onto the sea. Hakuho Maru receives the position data from the payload both in the sky and on the sea, enabling a quick recovery of the sampler.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
Regarding the large plastic balloon (as opposed to the small rubber balloon), few balloons have been launched from the ocean in these days since the antique Skyhook balloon project carried out by U.S. Navy Office of Naval Research in the late 1940s and 1950s (Pfotzer 1972; Nishimura 2002). The Skyhook project mainly used a naval aircraft carrier to launch balloons on the open sea. However, in present-day Japan, it is unrealistic to use the aircraft carrier.
Instead, we used Research Vessel Hakuho Maru, belonging to Japan Agency for Marine-Earth Science and Technology (JAMSTEC 2014). Hakuho Maru has 3991 gross tonnage with a body 100 m long and 16 m wide, and can run at a navigation speed of 16 kt (8.2 m s−1) at maximum. The crew capacity is 89, including 35 researchers. Since its first voyage in 1989, Hakuho Maru has been utilized for various academic activities on the sea.
As for the cryogenic sampler, we used the so-called Joule–Thomson (JT)-type compact air samplers (Morimoto et al. 2009). The JT sampler contains only one sample reservoir but is much lighter (22 kg) and smaller (0.35 m long, 0.35 m wide, 0.6 m high) than our conventional full-scale cryogenic sampler with a dozen reservoirs (Honda 1990). The portable JT sampler is thus appropriate for this ocean experiment. The JT sampler uses liquid neon as refrigerant to solidify or liquefy the atmosphere. Using the Joule–Thomson effect, neon is liquefied from onboard neon gas precooled by liquid nitrogen. Typically, around 0.0072 m3 atmosphere [standard temperature and pressure (STP) equivalent] can be obtained by 4 min of sampling at an ambient pressure of 200 hPa. To simplify the system, all functions are implemented to be controlled automatically by a controller on board the JT sampler. The opening and the closing of valves are electrically controlled by the controller on the basis of the monitored in-flight GPS altitude. The controller merges housekeeping data, such as balloon position data and sampler status data, and transmits it down to Hakuho Maru. A GPS-type radiosonde is hung below the sampler, mainly to obtain redundant payload position information but also to monitor meteorological conditions, such as pressure, temperature, humidity, and wind, of the air parcel where the sampler passes through. The radiosonde data are also transmitted down to Hakuho Maru. The flight termination is automatically executed by the sampler controller. After parachuting down onto the sea, a buoy mounted on the sampler functions as a radio beacon, emitting an additional signal to guide Hakuho Maru to the payload. The payload and the parachute are designed to float on water.
To minimize the size of the balloon, which had to be launched from a narrow vessel deck as described hereinafter, the payload was minimized to just one JT sampler. Four balloons were prepared to launch four samplers individually. To sample the air at various altitudes, two types of balloons with a volume of 2000 and 5000 m3 were used. Figure 2 shows a conceptual diagram of the flight train, the weight breakdown list, and the amount of helium gas filled in the balloon.
A conceptual diagram of the flight train using two types of balloon: “B2” with a volume of 2000 m3 and “B5” with a volume of 5000 m3. The weight breakdown lists and the amounts of filled helium gas for each flight train are also presented.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
The balloon operation was carried out in February 2012 during Hakuho Maru’s voyage on the eastern Pacific Ocean from Peru to Hawaii. The launch site area was defined in the equatorial area at latitudes between 2°N and 2°S and longitudes between 100° and 120°W.
3. New launch method
One of the most severe technical challenges of this experiment was to launch a balloon from the narrow deck of the vessel. Hakuho Maru was not designed on the assumption of use in balloon operations. The open space of the afterdeck is limited to only 20 m long × 7 m wide, which is much narrower than the aircraft carrier deck. In addition, much equipment, such as a big C-frame crane, poles, and antennas, is located on and around the Hakuho Maru deck. The number of crew (scientists) designated for this experiment is limited to only four, although some of the ship’s crew volunteers assisted us at times. The amount of cargo allowed for this experiment was also limited. Therefore, a key technical issue was how to launch a balloon from a limited space by a limited number of people with the maximum utilization of the existing equipment on the vessel.
From a narrow deck surrounded by much equipment, the balloon must out of necessity be launched straight upward. The major launch methods often used on land in these days (as well as in the old Skyhook project), such as the dynamic or the semidynamic launch methods (Fig. 3, top and middle), require a wide open space to stand the balloon after spool opening, and thus are not suitable for this case. On the other hand, the static launch method (Akiyama et al. 1983; Nishimura et al. 1993) (Fig. 3 bottom) does not need a wide open space and is intended to launch a balloon straight up. Consequently, we developed a new method by modifying the conventional static launch method.
A conceptual diagram of three types of conventional launch methods: (top) dynamic launch method, (middle) semidynamic launch method, and (bottom) static launch method.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
Figure 4 shows a conceptional diagram of this new launch procedure. In Fig. 4a, the balloon is inflated with helium gas from pressurized cylinders stacked on the deck. To simplify the procedure, the collar is attached to the balloon before starting the inflation. During the gas inflation, the balloon bottom is held (pulled) by a winch that is attached to the Hakuho Maru deck. A transportable launch spool is fixed to deck screw holes. In Fig. 4b, once the balloon is inflated, the winch rope is gradually wound off manually. In the original static method, the spool moves forward (toward the balloon bottom) during the gas inflation. Instead, in this method, the balloon bottom moves to the spool. And the balloon is stood up by further winding off the winch rope, so that the balloon bottom comes near the spool. Then, the balloon bottom is connected to a hook on the deck near the spool by an anchor rope. In Fig. 4c-1, the spool is opened. In Fig. 4c-2, the balloon bottom is also joined to the flight train (parachute and the sampler payload) and disconnected from the winch rope. In Fig. 4c-3, the collar is released by radio command. In Fig. 4c-4, finally the anchor rope is cut by remote control, and the flight train ascends into the sky. Figure 5 shows photos of equipment on the Hakuho Maru deck, such as the helium gas cylinders, the spool, and the winch, that was used for the launch operation. As for the helium gas, 80 cylinders, each filled with 
A conceptual diagram of the new launch method developed for the Hakuho Maru operation. (a) The balloon bottom is held by a winch during the gas inflation. (b) After the gas inflation, the balloon is stood up by winding off the winch rope. Then the balloon bottom is connected to an anchor rope. (c-1) The spool is opened. (c-2) The balloon bottom is connected to the flight train. (c-3) The collar is released. (c-4) The anchor rope is cut, and the whole flight train is launched.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
Photos of helium gas cylinders, spool, and winch on the Hakuho Maru deck, which were used for the launch operation. The red plastic film seen between the spool and the winch is a cover of the folded balloon.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
To minimize the risk that the flight train might hit various pieces of equipment on the deck and the deck erection, the open sky for the launch from afterdeck must be maximized, especially just above the deck and on the lee side. Therefore, during the launch operation, the vessel must be navigated to cancel the relative wind on the deck so as to stand and launch the balloon straight up. That is to say, the vessel must run forward at a speed nearly equal to (or slightly faster than) the surface wind speed in parallel to the wind direction toward lee. A captive balloon with streamers attached to its suspension cable (Fig. 7) is used to monitor the wind conditions: not only the wind speed and wind direction but also the surface wind shear and wind shift. The launch should be attempted only when the wind is sufficiently stable and calm for Hakuho Maru to cancel it with its full navigation speed of 8.2 m s−1.
The distance between the spool and the winch is around 20 m. The total length of a folded balloon with a volume of 5000 m3 is around 34 m. The meridian length of a balloon bubble containing helium gas corresponding to the total buoyancy of 75 kg is around 10 m. Therefore, the length of folded balloon between the collar and the balloon bottom is similar to the length between the spool and the winch. Which is to say, the 5000 m3-class balloon is almost the maximum balloon that can be launched by this method from Hakuho Maru. The portable spool, which was originally developed to launch thin-film balloons, can hold buoyancy up to around 200 kg and thus is well adaptable for this case.
Prior to the actual campaign on the sea, we carried out a ground rehearsal in September 2011 in Japan. In a large hangar at Taiki Aerospace Research Field (Fuke et al. 2010), a restricted area was defined to simulate the narrow deck. Equipment such as the spool, winch, anchor hook, and real-size balloons was handled by the scientists who actually served aboard the vessel. Details of the procedure were acquired and verified through this rehearsal.
4. Ocean campaign
Hakuho Maru departed Tokyo Bay, Japan, on 1 December 2011 and arrived at Callao, Peru, on 25 January 2012 via Honolulu, Hawaii. During its return voyage from Peru to Hawaii, between 4 and 8 February 2012, all four balloons were launched from the vessel as scheduled. Table 1 summarizes the specifications of four flights.
A summary of the air sampling balloon flights from Hakuho Maru.
Figures 6–8 show photos of the launch and the recovery of the fourth balloon on 8 February. The original plan was to tilt the C-frame crane to the bow side to maximize the open sky at the moment of launch. However, in practice, we tilted the C-frame crane to the stern side (as shown in Fig. 6) to maximize the open sky during the gas inflation. This was possible because the risk that the payload might hit the crane just after the launch was minimized because of the calm surface wind during the balloon campaign. At every launch, the released collar fell down on the deck, because of the calm wind, and could be reused. Typically, it has taken an hour for the launch operation, from starting the balloon deployment on the deck until the balloon launch. During the launch operation, the speed and the orientation of the vessel was navigated exactly as planned. In this way all four balloons were successfully launched.
After the gas inflation, a collared B5 balloon stands up on the afterdeck of Hakuho Maru. The big C-frame crane is tilted to the stern side.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
A balloon launch of the air sampler from Hakuho Maru. The yellow spherical rubber balloon seen near the B5 balloon is a captive balloon used to monitor the wind conditions by watching streamers attached to its suspension cable.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
A sampler recovered on the ocean by Hakuho Maru.
Citation: Journal of Atmospheric and Oceanic Technology 31, 7; 10.1175/JTECH-D-13-00248.1
Throughout the flights, from launch to splash down, telemetry data were received at Hakuho Maru. The flight trajectories were predicted at our affiliate, the Japan Aerospace Exploration Agency’s Institute of Space and Astronautical Science, in Japan, by using the latest global meteorological forecast, and were sent to Hakuho Maru via a satellite communication link. Trajectory prediction was utilized to assess flight opportunity in terms of the accessibility of the splashdown position (or the flight distance) and was also utilized to navigate Hakuho Maru to the splashdown point to make a quick recovery. The radio beacon signal also helped us to quickly access the payload on the ocean. As the result, all four payloads with their air samples were recovered successfully (Fig. 8).
After a short port call at Honolulu, Hawaii, on around 20 February, Hakuho Maru safely brought all four air samples back to Japan on 7 March 2012. The air samples are now under investigation at several institutes in Japan to analyze the concentration and the isotopic ratio of various atmospheric compounds. The weighted altitude mean of each sampling will be measured using a low pressure chamber by simulating the actual pressure profile during each flight. The result of these ongoing analyses will be presented elsewhere.
5. Summary
To sample the stratospheric atmosphere over the eastern Pacific equator, four flight operations of balloon-borne air sampling were carried out over the ocean using Hakuho Maru. To overcome the technical challenge of launching a balloon from a narrow vessel deck, we developed a newly modified static launch method. In February 2012 four balloons were successfully launched and all four payloads were recovered without damage. These successes verify the validity of both the new launch method and the shipboard flight operation, including the telemetry tracking and the safe recovery. Our new launch method can be adopted as a method to launch a balloon safely not only from the other ocean vessels but also from the ground, especially in cases where space is limited. Thus, balloon-borne air sampling will be enabled in various locations.
Acknowledgments
We thank the Japan Agency for Marine-Earth Science and Technology (JAMSTEC) and all the Hakuho Maru crews for their support in carrying out the balloon operation on the ocean.
REFERENCES
Akiyama, J., Nishimura J. , Namiki M. , Okabe Y. , Matsuzaka Y. , and Hirosawa H. , 1983: A new static-launch method for plastic balloons. Adv. Space Res., 3, 97–100, doi:10.1016/0273-1177(83)90084-4.
Aoki, S., and Coauthors, 2003: Carbon dioxide variations in the stratosphere over Japan, Scandinavia and Antarctica. Tellus, 55B, 178–186, doi:10.1034/j.1600-0889.2003.00059.x.
Fuke, H., and Coauthors, 2010: A new balloon base in Japan. Adv. Space Res., 45, 490–497, doi:10.1016/j.asr.2009.10.003.
Honda, H., 1990: Balloon-borne cryogenic whole air sampling system. ISAS Research Note 433, 39 pp.
Honda, H., 2001: Research on balloon-borne whole air sampling system for studying stratospheric minor constituents (in Japanese). ISAS Research Rep. 115, 93 pp.
Honda, H., Aoki S. , Nakazawa T. , Morimoto S. , and Yajima N. , 1996: Cryogenic air sampling system for measurements of the concentrations of stratospheric trace gases and their isotopic ratios over Antarctica. J. Geomagn. Geoelectr., 48, 1145–1155, doi:10.5636/jgg.48.1145.
Ishidoya, S., Sugawara S. , Morimoto S. , Aoki S. , and Nakazawa T. , 2008: Gravitational separation of major atmospheric components of nitrogen and oxygen in the stratosphere. Geophys. Res. Lett.,35, L03811, doi:10.1029/2007GL030456.
JAMSTEC, cited 2014: Research Vessel Hakuho Maru. [Available online at http://www.jamstec.go.jp/e/about/equipment/ships/hakuhomaru.html.]
Morimoto, S., and Coauthors, 2009: A new compact cryogenic air sampler and its application in stratospheric greenhouse gas observation at Syowa Station, Antarctica. J. Atmos. Oceanic Technol., 26, 2182–2191, doi:10.1175/2009JTECHA1283.1.
Nishimura, J., 2002: Scientific ballooning in the 20th century; A historical perspective. Adv. Space Res.,30, 1071–1085, doi:10.1016/S0273-1177(02)00517-3.
Nishimura, J., and Coauthors, 1993: The improvement of the static launch method in Japan. Adv. Space Res.,13, 63–66, doi:10.1016/0273-1177(93)90277-I.
Pfotzer, G., 1972: History of the use of balloons in scientific experiments. Space Sci. Rev.,13, 199–242, doi:10.1007/BF00175313.
