Technicians from the French space agency, CNES, prepare one of the pioneering driftsondes for launch at Zinder, Niger. (Photo by Terry Hock.)
by Bob Henson
Marching slowly from Africa across the Atlantic, a caravan of balloons made meteorological history earlier this year. It marked the first-ever research deployment of a new observing platform called the driftsonde. Long a dream of scientists and technicians at NCAR, the driftsonde became reality through years of planning and testing, plus a little help from ever-more-miniaturized technology.
The first driftsonde ever used for research took flight from the airport in Zinder, the second-largest city in Niger, on 28 August. Seven more driftsondes were released from Zinder over the following month, ascending to the lower stratosphere and moving westward into the tropical Atlantic. Dropping parachute-borne instrument packages along their leisurely paths, the driftsondes stayed airborne for periods ranging from 6 to 18 days.
The driftsondes were built and deployed through a collaboration among NCAR's Earth Observing Laboratory, the French space agency (CNES), and France's National Center for Scientific Research (CNRS). Each driftsonde's gondola held about 35 NCAR-designed instrument packages, or dropsondes. The gondolas were carried on ballooning systems designed by CNES. Development of the gondola and instrument packages was supported by NSF and NOAA.
In all, nearly 300 dropsondes gathered data from critical regions of the atmosphere that serve as breeding grounds for some of the worst U.S. hurricanes. Though the driftsondes moved
David Parsons (Photo by Carlye Calvin.)
generally westward as expected, their paths showed a good deal of small-scale variation, even looping eastward on occasion (see graphic). The surprisingly nonlinear paths highlighted the current undersampling of mid-Atlantic winds by standard observational tools.
Small, light, and tough
To build the driftsonde system, scientists, engineers, and machinists at NCAR and in France had to overcome many hurdles. The entire setup had to be robust enough to endure days of extreme stratospheric cold (averaging –62°C or –80°F) as well as the intense sunlight of the high, thin atmosphere. "Try letting your car sit at minus 80 for 14 days, and then try to start it," says David Parsons, the NCAR lead on the driftsonde project.
NCAR first tested the driftsonde concept many years ago. As far back as the 1970s, the late Vin Lally and his group of NCAR ballooning experts had developed instrument packages able to survive the harsh stratospheric conditions. However, the gondolas were large and heavy, requiring expensive balloons to loft them. Battery technology and communications links were also problematic.
The more recent impetus for driftsondes came from discussions between Lally and Mel Shapiro (NOAA). The idea was to support THORPEX, a ten-year global program to accelerate improvements in the prediction of high-impact weather.
For the balloon deployment to be affordable and practical, the system required low-cost, lightweight, off-the-shelf instruments capable of operating reliably in low pressure and in temperature extremes with very low power. They've been assisted by the Global Positioning System (GPS), which enables highly precise tracking, and by the continuing miniaturization of electronics. "The reduced size, weight, and power requirements of the new integrated circuits have made it possible to reduce every part of the driftsonde system—size, weight, and power," says NCAR's Hal Cole.
NCAR engineers and machinists worked together to produce a highly compact instrument package, roughly the size of a small bottle of water but weighing only about 140 grams (5 ounces). Called MIST (Miniature In-situ Sounding Technology), it weighs less than half as much as its predecessor, the dropsonde, which was designed at NCAR in the 1990s. Hundreds of dropsondes based on the original design are deposited by NOAA and Air Force aircraft into hurricanes each year.
Despite the straightforward westerly tracks depicted by numerical models (top), the actual tracks taken by this year's eight driftsondes included many loops and deviations (bottom). Dots along the tracks show where dropsondes were deployed. (Courtesy David Parsons.)
Suspense and success
The Niger launches this year were in conjunction with the African Monsoon Multidisciplinary Analysis (AMMA), a $50 million, long-term research effort initiated by French scientists to study the weather and climate over West Africa.
After being launched from Niger, each four-story-high balloon drifted from Africa toward the Caribbean at heights of around 20,000–22,000 meters (65,000–70,000 feet), where light easterly winds prevail. Twice a day, each gondola released a dropsonde that fell by parachute, sensing the weather conditions during its 20-minute descent and radioing data back to the gondola and then, by satellite, to the researchers.
At the AMMA operations center in Paris, a team from CNRS and NCAR kept track of each driftsonde on its westward trek. When a promising weather system developed, they could signal the gondola to release additional dropsondes as often as once per hour.
The first driftsonde's voyage proved dramatic. "The surface winds were very calm, so the launch went very smoothly," says NCAR engineer Terry Hock. Minutes later, though, NCAR's Joe VanAndel found himself unable to log onto the remote system for operating the driftsonde from Paris. "Everyone here in Zinder got very depressed," Hock says.
It took five long hours before a master timer reset itself and communications were finally established. "I could hear the sonde's signal on my hand-held radio—music to my ears," recalls Hock. About three days later, the driftsonde moved into the Atlantic (see map), observing clusters of showers and thunderstorms along its way.
The Niger site was selected in order to study the weak weather systems called easterly waves that serve as seedlings for hurricanes. Dozens of these waves move across Africa into the Atlantic between about 10° and 20°N. A small number develop into tropical storms and hurricanes, some of which reach the U.S. Atlantic and Gulf coasts.
The eastern tropical Atlantic is out of range for U.S. hurricane-hunter aircraft, and forecasters have little skill at predicting which systems will develop into hurricanes as they emerge off Africa. This year's driftsondes collected measurements from hurricanes-to-be Florence and Gordon just as each system reached tropical storm strength. Another driftsonde came close to Helene, sampling the near-storm environment but not making it into the cyclone itself.
In addition to tracking easterly waves and their progeny, the driftsondes gathered bird's-eye data on surges of hot, dry air that cascade into the Atlantic from the Sahel region of Africa. These surges carry huge amounts of dust as far west as Florida, influencing air chemistry, upper-ocean biology, and Atlantic weather systems. Other dry intrusions flowed south toward the driftsondes from midlatitudes. Dry air appeared to be one factor in helping bring this year's Atlantic hurricane activity from its red-hot 2005 pace to a near-average level. There were nine tropical storms and five hurricanes in 2006.
The driftsonde setup includes (a) a superpressure balloon that slowly inflates as it reaches the stratosphere; (b) a parachute that allows the entire unit to descend safely when needed; (c) a corner reflector, which helps aircraft-based radar to spot the unit; (d) the balloon gondola, with communications equipment that enables CNES to track the driftsonde; and (e) the NCAR gondola, which holds 35 or more dropsondes, each about the size of a small water bottle, as well as a computer and communications equipment that relays data from the dropsondes to an operations center. (Photos by Terry Hock.)
||During tests in Oregon, this driftsonde's balloon dwarfs the aircraft and automobile nearby as researchers stretch out the tether lines holding the gondola and communications devices.
On to Asia and Antarctica
Because of their flexible and relatively inexpensive nature, driftsondes may soon become a popular way to monitor and study many types of weather across the world's oceans and other remote regions, including Antarctica and the western Pacific.
"It would take a fleet of research aircraft to gather the same data," says Philippe Drobinski, the project's scientific co-lead from the CNRS.
Over the next year or
so, NCAR and its partners
will refine the driftsonde
technology, hoping to
maximize the number of
successful launches. (An
electrical storm corrupted
transmissions from one
of the eight driftsondes
launched from Niger;
another suffered a
balloon failure.) By 2008,
they hope to use
driftsondes in two
- a test of data
for the Antarctic region,
which would be part of
the International Polar
Year research effort; and
- the THORPEX
Campaign, examining the
evolution of weather conditions across east Asia and their impact on downstream weather in North America. Up to 50 driftsondes could be launched in two field campaigns for THORPEX between August and December 2008, depositing 1,500 dropsondes or more.
All in all, says Parsons, "this year was a successful proof of concept in many ways for the driftsondes as well as for the new CNES ballooning system. The partnerships we've developed will open up a new realm of observations."