Tenerife, Spain LLWAS Site Visit Report

航空

游客，如果您要查看本帖隐藏内容请回复

航空

Tenerife, Spain LLWAS Site Visit Report
Larry Cornman
National Center for Atmospheric Research
Introduction
A site visit to the Tenerife Reina Sofia (Tenerife-Sur) Airport was conducted on
9-10 February, 2005. Three issues regarding the LLWAS system were to be addressed: (1)
Evaluate the sensor siting, (2) look into the over-alerting problem, and (3) hold
discussions with airport meteorological and air traffic staff. Each of these will be
discussed in detail the following sections.
Sensor Siting Evaluation
On 9 February, a visit was made to all of the LLWAS sites at the Tenerife-Sur
airport. Figure 1 shows the island of Tenerife, and the location of the Tenerife-Sur airport.
A brief discussion of each site will be given below, but as can be seen from Figure 2, the
overall the sensors are laid out in a close to optimal fashion. This network should provide
good wind shear protection for the runway and one nautical mile to either side. There is a
single runway, oriented 08/26. Due to the prevailing winds, the normal airport operations
are landing on runway 08 and take off on runway 26. On the larger-scale, the terrain is
gently sloping from North to South, and hence sites 1-5 are above the runway height and
sites 6-10 are below it. There are small terrain features (typically small ditches or terrain
depressions) in close proximity to many of the sites; however due to the sensor heights,
these should not present any problems in providing unbiased wind measurements. There
are some buildings in close proximity to the south of site number 10; however the tower
height for this site is adequate, and the prevailing winds (ENE) do not come from this
direction.
Figure 1. View of the island of Tenerife, indicating the Reina Sofia Airport in the South.
Figure 2. Aerial view of Tenerife-Sur, with approximate locations of the sensors and runway ends.
Sensor Site #1
This site is situated to the North-West of the West end of the runway, and
approximately 20 meters higher. There are a number of small ditches in close proximity
to Site 1, especially to the South-West. Since the tower is reasonably high (15m), the
terrain indentations should not pose any problems in making wind measurements. Figure
3 shows a view East-South-East from this site and Figure 4 is looking to the North. One
of the ditches can be seen on the left-hand side of this latter figure.
Figure 3. Looking East-South-East from Site 1.
Figure 4. Looking North from Site 1.
Sensor Site #2
Site 2 is reasonably unobstructed; it is to the North of the runway and
approximately 40 meters higher. There is small ditch to the North-East and small ridge to
the North. Part of the ditch can be seen in Figure 5 and Figure 6. The 15 m tower is high
enough, so that these terrain features should not pose any problems.
Figure 5. Looking South from Site 2.
Figure 6. Looking East from site 2.
Sensor Site #3
Site 3 is just to the East of the main road heading into the airport, and just to the
North of the main airport complex. This site is reasonably unobstructed, with small
bushes and small trees nearby. The tower is 25 meters tall, so there should be no problem
in getting accurate winds from this site. The location is approximately 60 meters above
the runway height. Figure 7 is a view looking South-South-East from the sensor location.
The control tower and the terminal building can be seen in the middle of the picture.
Figure 8 is a view looking West-South-West from the site. The lighting fixtures and
bushes and small trees along the airport road can be seen in this photo.
Figure 7. Looking South-South-East from Site 3.
Figure 8. Looking West-South-West from Site 3.
Sensor Site #4
Site 4 is an unobstructed site on a 15 meter tower, to the North of the runway. It is
approximately 60 meters above the runway height. There is a small ditch to the West of
the site, but it should not pose any problem in getting accurate wind measurements.
Figure 9 is a view looking South-South-East from this site, and Figure 10 is a view
looking to the North. The small ditch to the West of the site can be seen in the middle-left
of Figure 10.
Figure 9. View south-south-east from site 4.
Figure 10. Looking North from Site 4.
Sensor Site #5
Site 5 is to the North-East of the East end of the runway, at approximately 40
meters above the runway height. The tower is 20 meters tall. There are some agricultural
hot-houses in close proximity to the South-East, as can be seen in Figure 11. There is a
large ditch to the West of the site, as can be seen in Figure 14. Neither of these items
should pose any problems in getting good wind measurements from this site.
Figure 11. Looking South-South-East from Site 5.
Figure 12. Looking West from Site 4.
Sensor Site #6
Site 6 is to the South-East of the East end of the runway, and is approximately 20
meters below the runway height. The tower is 15 meters tall. This is an unobstructed site.
Figure 13 is a view looking to the South-East of the site, and Figure 14 is a view looking
to the West.
Figure 13. Looking South-East from Site 6.
Figure 14. Looking West from Site 6.
Sensor Site #7
Site 7 is South of the runway, approximately 15 meters below the runway height.
The tower is 15 meters tall. This is an unobstructed site, with a small ditch to the East.
This ditch can be seen in both Figure 15, which looks to the South-East, and Figure 16,
which looks to the East.
Figure 15. Looking South-East from Site 7.
Figure 16. Looking East from Site 7.
Sensor Site #8
Site 8 is to the South of the mid-point of the runway, at an elevation of
approximately 15 meters below the runway height. The tower is 15 meters tall. The
terrain slopes down and then up going to the East, as can be seen in Figure 17, and is
relatively flat to the West. There are some agricultural hot-houses in close proximity to
the South-East, as can be seen in Figure 18. Neither of these features should pose any
problems in making good wind measurements.
Figure 17. Looking East from site 8.
Figure 18. Looking south-east from site 8
Sensor Site #9
Site 9 is located to the South of the runway, towards its West end. It is situated
approximately 16 meters below the runway height. The tower is 20 meters tall. This is an
unobstructed site with relatively flat ground surrounding it. Figure 19 is a view looking to
the South of the site, and Figure 20 looks to the West-North-West. The dirt mound that
can be seen in Figure 20 is due to some road work, and does not pose any problems with
the site.
Figure 19. Looking South from site 9.
Figure 20. Looking West-North-West from site 9.
Sensor Site #10
Site 10 is to the South-West of the West end of the runway. It is approximately 30
meters below the runway height. The tower is 25 meters tall. There is a new apartment
complex with three storey buildings just to the South of the site – as can be seen in Figure
21. The only potential problem would be with winds from the south, but as the sensor site
is slightly above the first floor of the buildings, and with height of the tower, there should
not be any significant degradation in the wind measurements. Furthermore, the prevailing
winds are from the Easterly directions. Figure 22 is a view looking to the West-North-
West of the site, showing the unobstructed terrain in that direction.
Figure 21. Looking South from site 10.
Figure 22. Looking West-North-West from site 10.
The Over-alerting Problem
One of the items that UCAR was asked to look at during the Tenerife site visit
was a persistent over-alerting with the LLWAS system. It was quite clear on inspection
that the problem was not of a meteorological nature, but rather, something to do with the
system hardware or software. Almos subsequently discovered a database error in the
Airport Configuration File (ACF). Almos then generated and loaded a new ACF that
should solve the problem.
Another problem that was observed during the visit was intermittency in some of
the sensor readings. Telvent personnel believed that the problem was due to a lack of
solar battery power at some of the sites, which in turn was due to a lack of maintenance at
the sites.
Discussions with Airport Meteorological and Air Traffic Staff
On 10 February, a meeting was held with airport meteorological staff and air
traffic controllers. One interesting wind shear condition was discussed: a persistent
summertime condition wherein pilots lose airspeed when descending through 1000 feet
on landings from the West. Air traffic controllers indicated that airspeed loses on the
order of 30 knots had been encountered. These are significant values, and have resulted in
go-around procedures on occasion.
Without further investigation, it is unclear what the specific mechanisms are that
could be generating this vertical shear of the horizontal wind phenomenon. Nevertheless,
a few potential causes can be postulated. From Figure 23, it is clear that the island of
Tenerife is dominated by the Pico de Teide, a 3718 meter volcanic peak. Vortices shed
off of this large terrain feature are one potential cause. It is well-known that verticallyaligned
vortices, known as a von Karman street, can be shed by oragraphic features such
as the Teide. However, there is also a secondary terrain feature which is located in close
proximity to the approach path for Runway 08. This is a 430 meter hill, (indicted by the
arrow on Figure 23), approximately 10 km to the West-North-West of the airport.
Orographic wind effects from this terrain feature, by themselves, or in combination with
the larger scale vortices could also be the source. A brief discussion of these two
mechanisms is presented below, along with a discussion of measurement devices that
could be used to investigate the phenomenon – or even be used as part of an operational
warning system.
Figure 23. Topographic map of the island of Tenerife.
A brief discussion of the synoptic trade wind patterns over the Canary Islands can
be found in Varela et al.1 (see also the references sited in that paper). Figure 24 from that
reference illustrates a typical synoptic wind pattern: cooler maritime air flowing from the
North-East at the surface, with cooler dryer air from the North-West aloft. A thermal
inversion layer forms between 1000-1500 meters.
Figure 24. Synoptic trade wind behavior in the Canary Islands. (from Varela et al.)
von Karman vortex street
It is well-known from fluid mechanics that as a fluid flows around an obstacle,
such as a cylinder, the boundary layers separate from each side of the cylinder surface
and form two shear layers that trail aft in the flow and bound the wake. Since the
innermost portion of the shear layers, which is in contact with the cylinder, moves much
more slowly than the outermost portion of the shear layers, which is in contact with the
free flow, the shear layers roll into the near wake, where they fold on each other and
coalesce into discrete swirling vortices. A regular pattern of vortices, called a von
Karman vortex street, trails aft in the wake. Figure 25 illustrates this phenomenon. For a
vertically aligned obstacle, in this case a cylinder, the vortices are aligned vertically.
Figure 25. A view from above of a von Karman vortex street forming behind a cylinder.
1 A.M. Varela, et al. 2004: Non-correlation between atmospheric extinction coefficient and TOMS aerosol
index at the Canarian Observatories. Remote Sensing of Clouds and the Atmosphere IX, ed. Schafer et al.,
Proceedings of the SPIE Vol 5571, pp. 105-115.
This vortex shedding behavior is often observed with isolated mountains. A
satellite view of this phenomenon associated with the Canary Islands is shown in Figure
26. Tenerife is indicated by the arrow. The lack of clouds just downwind of Tenerife is
due to subsidence of the air mass as it flows down the slopes of the Teide. This does not
mean that the vortices are absent, rather there are no clouds there to mark them.
Figure 26. von Karman street vortices formed downstream of the Canary Islands. The island
of Tenerife is indicated by the arrow.
Figure 27 and Figure 28 illustrate results from a numerical modeling simulation of
the wind field in the wake of the Hawaiian of Kauai, performed by NCAR for NASA. As
can be seen from Figure 27, the island of Kauai is very similar in structure to Tenerife.
Figure 28 illustrates the simulated wind field at 1000 m, with the left-hand image
showing contours of horizontal wind velocities (turbulence is indicated by the blue
shading), and the right-hand image showing the horizontal wind vectors. In this case, the
flow pattern was reasonably consistent in the vertical, so that the winds at 1000 feet are
probably not too different than those shown here. Notice the sharp gradients in horizontal
velocities as the air flows around the island (left-hand image). These simulations also
captured von Karman vortices in the downstream flow. This flow pattern close to the
island, as well as downstream is probably similar to patterns that would be encountered at
Tenerife.
The persistent trade wind field and the flow around and downstream of Tenerife,
as indicated above, is certainly a potential source of the vertical wind shear that is
experienced by pilots approaching Tenerife-Sur from the West.
Figure 27. Model domain used for wind flow simulation around the Hawaiian island of
Kauai. The contour lines indicate the terrain.
Figure 28. Fine-scale model winds at 1000 meter altitude from Kauai simulation. Left is a contour of
the wind field with turbulence indicated with the blue shading. On the right are the wind vectors.
Smaller-scale orographic effects.
As mentioned above, (Figure 23), there is a 430m terrain feature approximately
10 km to the West-North-West of the airport. Localized flow around and over this hill
could also affect aircraft as they approach the airport from the West. Figure 29 shows
another situation of von Karman vortices formed downstream of the Canary Islands.
Figure 30 is a blow-up of the region surrounding Tenerife. The box in this figure is
approximately centered over the 430m terrain feature. Note the sharp discontinuity of the
(presumed) low-level clouds (translucent grey) along the South-West corner of the island.
Note also, the small set of convective clouds that lie along the discontinuity, just
downstream of the terrain feature. It is possible that these convective clouds are being
formed by air being lifted by the terrain feature.
Figure 29. Satellite image of the Canary Islands, showing von Karman vortices downstream.
Figure 30. Blow-up of Figure 29, showing Tenerife. The box is centered approximately over the 430
terrain feature.
Sensors for investigating/detecting the wind shear phenomenon.
The discussion presented above, is not intended to answer the question as to the source of
the vertical wind shear phenomenon that was mentioned by the Tenerife-Sur air traffic
controllers. Rather, it was intended to indicate some potential factors that could be related
to the condition. In order to further investigate the situation, and to perhaps be used in a
warning system, there are two sensors that could be of use. The first is a scanning eyesafe
Doppler lidar and the second is a vertically pointing Doppler radar.
The Doppler lidars provides very accurate radial velocities by reflecting off of
aerosols in the atmosphere. Depending on the level of aerosols, the range of these devices
can reach 15 km, with a range resolution of 60-100 meters. Such a device, placed on the
airport property and scanning in a vertical plane above and below the approach glide path
of Runway 08, would most likely be able to see the wind shear phenomenon discussed
above. Automated algorithms could be developed to detect the wind shear and provide
alerts to the air traffic controllers. On of the downsides to these devices is their high cost
(on the order of $1 million USD).
The vertically pointing Doppler radars, also known as wind profilers, can provide
a vertical profile of the horizontal wind above the device. So-called boundary-layer wind
profilers can generate accurate winds up to a height of 1-2 km, with a 60-100 meter range
resolution. These devices operate at microwave frequencies, and measure radial velocities
by scattering off of index of refraction variations as well as Rayleigh scattering off of
hydrometeors. If the wind shear phenomenon is relatively homogeneous in space, one of
these devices placed on the airport property could detect the wind shear that is affecting
the aircraft. These devices are far less expensive than the Doppler lidars (on the order of
$2-300 K USD), on the other hand, there are more data quality control issues with
Doppler wind profilers.
UCAR/NCAR has a great deal of experience with both of these devices.
Furthermore, UCAR/NCAR has developed wind shear detection algorithms similar to
those just mentioned. A data collection campaign could be performed using either or both
of these devices, to investigate their capabilities in detecting the wind shear and providing
warnings. A decision could then be made as to whether such devices and warning
algorithms should be deployed.
It should be noted that addressing the abovementioned vertical wind shear
problem does not mean that the LLWAS system is not needed. The LLWAS system is
designed for, and does an excellent job of detecting low-level wind shear due to
microbursts and gust fronts. Therefore, the use of other sensors to detect the vertical wind
shear condition is viewed as complimentary to the LLWAS system. Furthermore, if other
sensors and warning algorithms are deployed for the vertical wind shear problem, the
alerts should be integrated with the LLWAS alerts. UCAR/NCAR would work with
Almos, Telvent and INM as needed to assist in this process.

brightjm

thank   you.

ingooooo

谢谢分享，学习一下。