A meteorological measurement campaign was held during March 2005 over the Venezuelan
llanos using pilot balloon and radiosonde observations. The objectives of these measurements were to:
1) determine the diurnal variation of the low level jet that
has been observed in pilot balloon data during the last 3 years at San Fernando de Apure,
2) determine the structure of the LLJ from a network of 7 pilot balloon sites,
3) validate the structure of the LLJ that has been shown from the recently available regional analyses, and
4) strengthen the interest in meteorological measurements in Venezuela as part of the educational component of PACS-SONET.
In 2001 a pilot balloon observation site was established at San Fernando de Apure
in the heart of the Venezuelan llanos, the very flat region along the drainage of
the Orinoco river, south of the cordillera de la costa (Fig 1). The purpose of this site, part
of the PACS-SONET project, was to make daily wind soundings to monitor the trade wind flow and
its variability over the interior of northern South America. No other wind soundings were
being made at the time in the region on a routine basis. After the San Fernando installation
was completed, other sites at Cd Bolivar and Maracay were also established.
The most striking aspect of the pilot balloon wind observations made at San Fernando during the past
four years (data set is not continuous during this period due to gas supply problems
at times) has been the nearly constant presence of a low-level jet in the
morning soundings during the dry season (November – April). Since the observations were
made at San Fernando only during the morning, this suggested that afternoon observation might
reveal different profiles, as with low level jets observed elsewhere. The identification of
possible variability to the low level flow was not of isolated importance. To compare
wind observations made at san Fernando with analyses produced by global or regional
data assimilation schemes used to produce the widely used “reanalysis” products for
climate research studies it is essential to be able to relate observations made at a
few observation times per day to a mean made over the entire diurnal cycle. This is
because most large-scale analyses cannot resolve details of the flow interaction with
topography, or land-sea breeze or mountain valley circulations that may be locally or
Thus, during February 2005 afternoon observation were started at San Fernando,
to identify the extent of the diurnal variation. Observation made during this
time showed significant diurnal changes (Fig 2). Thus the motivation was
established for a more extensive set of measurements planned for March 2005.
During early march 2005 higher spatial resolution analyses (32 km grid spacing Eta model)
produced as part of the regional reanalysis effort became available to us and were used
to produce maps of the mean winds over the llanos at 3-hourly time resolution (Fig 3).
Although this region is near the edge of the model domain, the regional reanalyses show a
strong low level jet with a very strong diurnal variation. This provided additional
motivation to carry out very frequent measurements at the San Fernando site.
In summary, a set of extensive measurements were planned for mid-late March 2005,
to take place prior to the onset of the rainy season. The climatology of the LLJ
at San Fernando, based on the 3+ years of soundings (Fig 4) showed the jet to be
strongest in February, but still quite evident in March.
Logistics of the experiment
We contacted the FAV (Venezuelan Air Force) in February 2005 regarding
the possibility of making additional measurements at several stations around
the llanos to better describe the LLJ. We prepared a brief web document
showing an idealized set of stations and the motivation for the activity.
The FAV agreed to the activity and by early march we were contacting other
individuals (Rigoberto Andressen, Jorge Lopez, Rafael Garcia) at Universities
in Venezuela, some of whom had participated in the 2001 PACS-SONET organized
course in Panama.
Four theodolites were shipped to Venezuela via Miami, where they were picked up by the
Venezuelan Air Force, as part of regular flights to Miami. Additional balloons
and other accessories were also sent. A radiosonde system from NSSL
(old Vaisala PP-11 system) and RS-80 radiosondes (RS-80 15N Omegasondes)
were taken to Venezuela with us on March 16th. This equipment was established
at San Fernando, while the other pilot balloon theodolites were sent to different
sites. The actual network as proposed by the FAV included the three operating sites
of San Fernando, Cd. Bolivar, and Maracay, and also temporary sites at Guasdualito,
Calabozo, Puerto Ayacucho, and Isla Orchila (about 100km offshore in the Caribbean).
These were sites where there were already FAV meteorological observers and where
some communications infrastructure already existed. All sites could also be
accessed by air transport. The Isla Orchila site had to be changed to Coro at
the last minute due to difficulties in obtaining permission from the Venezuelan
Navy to use the island facilities. The purpose of the sites were as follows:
San Fernando: describe diurnal variation of the winds and boundary layer near the suspected core of the jet.
Calabozo: help describe the wind gradient between the weak winds at maracay and the llj core near san Fernando.
Guasdualito and Cd. Bolivar: describe downstream and upstream winds from the suspect region of maximum winds near san Fernando
Coro: describe trade wind regime and its variations along the north coast of Venezuela to see if diurnal
cycle and synoptic variations were in phase with those over the llanos.
We arrived at Maiquetia international airport on the evening of March 16th, flights
from Houston arriving at approximately 10PM. The lines were lengthy through
immigration, but customs was greatly expedited with the help of Arnoldo Lugo and
other FAV personnel sent from Maracay. We arrived in Maracay at about 2 AM.
On Thursday, March 17th, we presented a summary talk to a combined group of experiment
participants, which included approximately 14 FAV personnel and about 10 students
from the Universidad Central in Barquisimeto and the Universidad Central de Venezuela.
The groups were established for each site, with both students and FAV personnel being
deployed to each site. Material was separated for each group and a lecture was presented
describing the scientific motivation for the measurements.
The actual deployment to the sites was organized by the FAV, and involved having
all personnel fly to the sites onboard FAV aircraft. Personnel for Cd Bolivar
flew on once aircraft, while all others were to deploy to the sites using one
C-130 that was to visit all of the other sites on Friday, March 18th. Given
the relatively large distances involved, and the poor state of some of the
road segments, this was viewed as the most efficient manner to deploy personnel
to the various widely separated sites. (Note: gasoline (and presumably AV gas
for the aircraft) is extremely inexpensive in Venezuela, about 20 US cents per
gallon, thus gas costs are a relatively small cost, even with aircraft usage).
The first site the aircraft reached (approximately on time) was Coro. The
aircraft then proceeded to Maracaibo, where it was to pick up equipment to
take elsewhere. There were delays on the ground at Maracaibo, and one engine
did not start. After a starter motor was changed it eventually was started
but during the flight to the next location, Guasdualito, the engine quit.
Thus, the pilot decided to return to Maracay directly. Thus on March 18th,
only the personnel for Guasdualito and Coro were deployed via the C-130 aircraft.
The Cd. Bolivar site personnel were also deployed by aircraft; the aircraft
made an emergency landing prior to reaching Cd Bolivar and the personnel made
the rest of the journey by bus that day.
The next day, the remaining personnel for Puerto Ayacucho, San Fernando and
Calabozo returned to the airport near Maracay to prepare for another flight.
Before take off it was decided that the Calabozo team and part of the larger San
Fernando team would travel by car to their sites. This they did, arriving in
late afternoon in San Fernando. The aircraft teams were successfully deployed
to Puerto Ayacucho and also San Fernando – the aircraft arrived a half hour
sooner than the vehicle sent via road, despite the plane having stopped at
three other sites prior to San Fernando. Thus, all site were operating by
late afternoon of Saturday, March 19th.
The sounding strategy
The need for hydrogen gas had been previously estimated and delivered
to each site before the additional observers for each site arrived. The
need for additional observers was because round the clock observations
were anticipated, including nighttime pilot balloon soundings. Three
gas cylinders had been deployed to each site (sufficient for about 60
soundings that were anticipated at each site) except for San Fernando where
there were additional cylinders for the planned radiosonde launches.
The sounding schedule as it finally evolved was as follows:
Pilot balloon soundings at all sites at: 0000, 0430, 0700, 1000,
1300, 1600, 1830 local time. The 0000, 0430 and 1830 observations
were made using pilot balloon lights, attached to the balloons by
line approximately 3 m long.
At San Fernando more observations were made as follows:
Radiosonde observations, using 100gm balloons, at 0000, 0700, 1300, 1800
LT initially, and then during the last two days radiosonde observations were also made at 1000 LT.
Pilot balloon observations were made at: 0430, 1000, 1600, and
2030 initially, with additional observations at 0200, 0830, 1130,
and 2200 during the last two days. Thus, between 8 and 12 observations
were made per day at San Fernando.
The observations began as soon as each site was established, with Coro,
Guasdualito and Cd. Bolivar beginning on Friday, March 18th, while the
other special sites began on Saturday. San Fernando had run out of
balloons on March 18th, and began soundings again on the afternoon of March 19th.
Longer period observations after the intensive observing period (IOP)
Because the IOP was relatively short, and because there was an
excess of gas available after the end of the IOP, it was decided to
continue additional observations at some sites for several additional
weeks. These sites include Calabozo, San Fernando, and Cd Bolivar, where
twice-daily observations will be made for a period of at least 2 weeks
after the IOP. This will improve estimates of the difference between AM
and PM winds, and identify any systematic differences between the winds
at the three sites.
Logistical implications of the activity
- Gas usage
- Set-up of the stations
- Real-time data processing and analysis
Scientific results (preliminary)
The results can be accessed through the scrollbar to the left.
A discussion of the result will be soon included.