Clear air turbulence over South Africa

Document technical information

Format pdf
Size 987.0 kB
First found May 22, 2018

Document content analysis

Category Also themed
not defined
no text concepts found


Andrei Linde
Andrei Linde

wikipedia, lookup




Clear air turbulence over South Africa
M P de Villiers,
South African Weather Bureau, Private Bag X97, Pretoria, South Africa,
J van Heerden,
University of Pretoria, Department of Civil Engineering, Pretoria, South Africa
Clear air turbulence (CAT) at high altitude remains a hazard to aviation which can result in passenger
injury and aircraft damage. Two limited surveys of CAT events over South Africa, 1993-1995 (inclusive)
and 1998, are used to illustrate the most likely synoptic conditions under which CAT can be expected. A
case study of CAT associated with an upper-air trough and a mountain wave is presented. The study also
evaluates the effectiveness of the Ellrod Turbulence Index (ETI) derived from model data provided by
the UK Met. Office. A forecast of ETI derived from the Global Spectral Model of the United States
National Center for Environmental Prediction (NCEP) is also reviewed.
1. Introduction
High altitude clear air turbulence (CAT) has been the
cause of numerous incidents in which commercial aircraft passengers have been injured and deaths have
sometimes occurred. On rare occasions aircraft suffer
structural damage and temporary loss of pilot control.
Other effects of CAT are increased fuel consumption
and late arrival due to reduced airspeed (Ellrod, 1992).
CAT is turbulence encountered by aircraft when flying
through air space devoid of clouds and is caused by
marked changes in wind speed and/or direction, either
vertically or horizontally (American Meteorological
Society, 1989; Met. Office, 1991).
According to the US Department of Commerce (1966)
and Met. Office (1994), CAT is divided into three categories: light, moderate and severe.
During light turbulence passengers may be
required to use seat belts, but loose objects within
the aircraft remain at rest.
Moderate turbulence results in passengers being
required to wear seat belts and occasionally being
thrown against the seat belt. Loose objects in the
aircraft move about. Frequent rolling occurs and
there is difficulty in walking about in the aircraft.
Occurrences of severe turbulence may cause the
aircraft to be momentarily out of control and it is
difficult to maintain flight altitude. Passengers are
thrown violently against the seat belt and back into
the seat and loose objects are tossed about. Under
extremely severe turbulent conditions, which fortunately rarely occurs, the aircraft is violently
tossed about and is almost impossible to control.
Structural damage is also likely.
In April 1993 a flight from Shanghai to Los Angeles
encountered severe CAT over the northern Pacific
Ocean at 33,000 ft. Of the 265 passengers on board, 169
were injured, several were severely injured and one pas-
senger died. The aircraft suffered no external damage
but the interior was badly torn up. In November of the
same year another aircraft, flying a similar route,
encountered severe turbulence which damaged its elevators. In South Africa, during January 1994, two stewardesses were injured when a flight between Durban
and Cape Town flew into CAT.
According to Godson (1970) for every crash or incident of severe structural damage and injury caused by
CAT, ‘it is believed that there have been at least a hundred near encounters’. In addition, Stack (1991) states
that there were 96 CAT events from 1985 to 1991,
which ‘included passenger death, major and minor
injuries and aircraft damage’. He also adds that nearly
half of the injuries occurred to flight attendants.
Given the potential for serious harm, it seemed surprising that no research into the occurrence of CAT over
South Africa had been undertaken, especially considering the marked increase in commercial air traffic over
the country in recent years. The number of airlines
operating into South Africa, for example, has risen
from approximately 20 in 1993 and to 80 in 1997. With
this in mind it was considered essential to know more
about CAT over South Africa so that it could be predicted more accurately.
To this end, a survey was made of all pilot reports of
CAT between 1993 and 1995, and later, during 1998, a
further more detailed survey was conducted during the
winter months of May to September inclusive, to determine the conditions under which CAT is most likely to
occur over South Africa. A summary of the results of
the earlier survey is presented in this paper along with
a more detailed analysis of the 1998 survey.
A forecast CAT index was introduced by the South
African Weather Bureau (SAWB) to assist the aviation
forecaster in forecasting CAT over the country. This
took the form of the dimensionless Ellrod Turbulence
Index (ETI) which is derived from the Global Spectral
Model (GSM) of the United States National Center for
Environmental Prediction (NCEP). An example of a
weather system that produced high altitude CAT associated with an upper-air trough over central South
Africa and mountain wave turbulence is presented. The
role played by the Ellrod Turbulence Index (Ellrod &
Knapp, 1992) in predicting the turbulence, derived
from numerical weather prediction models, is also
Without going into detail, the ETI was selected because
two independent evaluations found it to be the most
suitable (McCann, 1993; Smith et al., 1995). It was
already in use by NCEP, and trial use over South
Africa produced favourable results.
A drawback of all indices is that they cannot successfully predict all cases of CAT and its severity.
However, they often alert forecasters to areas that normally would not be considered high threat regions.
Ellrod & Knapp (1992) also state that there is a tendency for the ETI to predict CAT through too deep a
Note that places mentioned in the text are given in
Figure 1.
2. The 1993–1995 survey
The meteorological conditions at the time of all available reports of CAT over South Africa, from 1993 to
1995 inclusive, were studied in order to determine local
characteristics (de Villiers, 1997). This was by no means
a comprehensive survey and consisted of 15 events.
Only those that came to the attention of the author
were researched and there were doubtless other CAT
events that went unreported or unrecorded. For example, a mountain wave event on 10 January 1994 was discovered by chance when it was reported in a Cape
Town newspaper. Nevertheless, the samples helped to
give an idea of the local conditions suitable for CAT.
A jet stream was present in all of the 15 incidents when
CAT was reported (Table 1), although in the mountain
wave events the reports of CAT were at a level well
below the jet stream. They were also sometimes well
ahead of the jet stream and consequently on the warm
or high pressure side. In the case of the one cut-off low,
when a turbulence report was received, the jet was well
to the north around the northern edge of the low. A
trough was present most of the time, including in the
two mountain wave instances (80%). Of these, 58% of
the CAT observations (including the two mountain
wave cases) were in the north-westerly flow east of the
trough, 25% at the trough axis and 17% in the posttrough southerly flow. Two events were associated
with an anticyclone, with one on the warm, or high
pressure side of the jet and also at the same level as the
jet. Including one mountain wave event, this meant that
only 20% of the CAT incidents were in the warm sector. By far the most occurrences were below the jet on
the cold and low pressure side (67%) and 13% directly
below. Most of the reports were within 5,000 ft of the
jet (53%) and 67% within 10,000 ft. These results were
consistent with the findings of Briggs (1961).
With respect to vertical shear, most events were in the
lower vertical wind shear range of 4 to 5 kn/1000 ft,
while only one exceeded 8 kn/1000 ft. This is supported
by the fact that no CAT reports in excess of moderate
were received, although with shear greater than 8
kn/1000 ft it is surprising that there was not at least one
report of severe CAT.
Horizontal velocity (speed and/or directional) shear
was noted on nine occasions (60%), but it was difficult
to classify due to the distances between observations,
which are at least 3° latitude apart and often about 7°.
Therefore the results are not to be trusted. In assessing
the horizontal shear the overall synoptic pattern was
considered. There were also instances where significant
shear was evident, but not where the CAT was experienced. A significant horizontal shear yardstick of
100 kn/100 miles (Colson, 1963) was used and under
these conditions shear was noted on four out of the
nine occasions. If the UK Meteorological Office
(UKMO) limits of 20 kn/° latitude for moderate CAT
and 30 kn/° latitude for severe CAT (from Starr, 1996)
had been used, the figures would no doubt have been
Most of the 15 events were associated with moderate
stability, i.e. a positive potential temperature change of
1.5–4.5 °K/1000 ft. Two were associated with high stability, one of which was 10 °K/1000 ft (de Villiers,
3. The 1998 survey
Figure 1. Map of South Africa showing the places mentioned
in the text.
A more detailed survey was conducted during the winter months of 1998 from May to September, inclusive.
Table 2. Analysis of the synoptic conditions at the time
Table 1. Analysis of the synoptic conditions at the time
of 15 CAT events during the 1993–1995 survey.
of 34 CAT events during the 1998 survey.
CAT east of upper-air trough
CAT at trough axis
CAT in post-trough southerly flow
CAT due to an upper-air cut-off low
CAT due to mountain wave
CAT in an anticyclonic flow
CAT below the jet on the cold/low pressure side
CAT below the jet on the warm/high pressure side
CAT below the jet
CAT above the jet on the cold/low pressure side
CAT above the jet on the warm/high pressure side
Distance of CAT from jet <5 000 ft
Distance of CAT from jet 5 000–10 000 ft
Distance of CAT from jet >10 000 ft
Prior to beginning the survey of 1998, a special appeal
was made to airline pilots, air traffic controllers and
SAWB personnel to observe and pass on information
to the author. This had a positive effect, because during
the three-year period from 1993 to 1995 only 15
reports were received on 15 different days, whereas
during the 1998 five-month long survey, reports
totalling 49 were received on 34 days. On 28 May 1998
as many as six reports of CAT were received (de
Villiers, 1997).
A high number of CAT reports occurred at a trough
axis and downwind east of the axis. The high number of
CAT reports at the trough axis suggests that directional
shear is an important factor in producing CAT (Table
2). Unfortunately, the SAWB, because of financial
constraints, has had to cut the number of upper-air
soundings made at stations around the country, with
some stations carrying out only one ascent a day. This
made the computation of horizontal shear nearly
impossible. Hence, the data on horizontal shear in
Table 2 should be treated with caution.
CAT occurred on one occasion on 11 July 1998 (Table
2) in the throat of an upper-air cut-off low. Another
incident occurred where a cut-off low was present to
the west of the country over the Atlantic Ocean, but
the CAT occurred in mountain wave activity in a
north-westerly flow, well in advance of the low, over
the Western Cape mountains in the south-western part
of the country.
Consistent with the findings reported in Starr (1996)
and the research of Ellrod (1993), Hopkins (1977) and
Endlich (1964), by far the highest number of CAT
reports were received below the core and to the south
below the core on the cold side (Table 1). Nevertheless,
a high number of CAT reports were in the warm air
below the jet stream. Five of them were associated with
CAT east of upper-air trough
CAT at trough axis
CAT in post-trough southerly flow
CAT due to an upper-air cut-off low
CAT due to mountain wave
CAT in an anticyclonic flow
CAT below the jet on the cold/low pressure side
CAT below the jet on the warm/high pressure side
CAT below the jet
CAT above the jet on the cold/low pressure side
CAT above the jet on the warm/high pressure side
Distance of CAT from jet <5 000 ft
Distance of CAT from jet 5 000–10 000 ft
Distance of CAT from jet >10 000 ft
mountain wave activity well ahead of an approaching
trough. Three reports were in developing anticyclonic
flow. That is, where the circulation was in the process
of changing from a post trough southerly circulation to
westerly and north-westerly.
As in the 1993–1995 survey, the distance of CAT from
the jet stream core was mostly within 5,000 ft.
However, inclusion of mountain wave events meant
that the report was often much greater than 10,000 ft
below the core (Table 2).
Most (six) of the lower flight level reports of turbulence
came from the lee of the Drakensberg mountains (in
the east), with three from the Eastern Cape mountains
in the south and one in the vicinity of the Western Cape
mountains. However, with respect to the last, there
seems to be a connection between the passage of trough
systems and the mountains in the area because numerous reports of CAT at high altitude were received in
this area (nine reports). Shutts (1997), quoting Smith
(1979), states that a steep lee slope relative to the windward side of a mountain range is also a criteria for
strong lee waves. This may help to explain the high
number of reports over KwaZulu-Natal in the lee of
the Drakensberg.
Most of the CAT reports occurred with 4 to 8 kn/1000
ft vertical wind shear (Table 3). What is interesting is
that although a 4 kn/1000 ft vertical wind shear is usually considered the cut-off point for moderate turbulence there were seven occasions when the shear was
below this threshold and yet light to moderate CAT
was reported (Table 4). All seven of the CAT reports
were associated with light to moderate stability. Three
of the events were associated with mountain waves and
two with a trough axis (which suggests horizontal
directional shear). The remaining reports were of
light/moderate CAT east of the trough axis.
In contrast to the 1993 to 1995 survey, where vertical
wind shear >8 kn/1000 ft occurred but no reports of
CAT were received, during the 1998 survey there were
five reports of severe CAT, but vertical wind shear >8
kn/1000 ft was observed on only one occasion (Table
4). However, a report of severe CAT was received
when the air was very stable (>+4.5 °K/1000 ft potential temperature change) with vertical wind shear of 4
to 5 kn/1000 ft.
Low to moderate stability was present during most
CAT reports (Table 3). This corresponds to the preponderance of vertical wind shear <8 kn/1000 ft. In
other words, lower stability is necessary where large
vertical shear is absent (Colson, 1961).
By and large, pilot reports of moderate turbulence and
moderate to severe turbulence coincide with moderate
vertical wind shear and moderate stability (Table 4).
However, there were occasions when vertical wind
shear and stability were less favourable for CAT development and yet severe CAT was reported (Table 4).
For example, on 7 May 1998 a pilot reported severe turbulence with a 4 to 5 kn/1000 ft vertical wind shear and
a low potential temperature change of <+1.5 °K/1000
ft. On 12 June 1998 a stewardess was injured when the
pilot reported no more than moderate CAT in the lee
of the Drakensberg. The aircraft, on route from
Durban to Johannesburg, was ascending through
FL210 to FL250 in winds of 40 to 50 kn, while the
vertical wind shear was less than 4 kn/1000 ft with low
stability. The conclusion is that mountain waves were
the cause of the turbulence. The author has noted similar turbulence on numerous commercial flights when
climbing or descending on the sector of the flight
between Durban and Ladysmith, particularly with
Berg wind conditions. A Berg wind is the local name
for the hot and dry Fohn wind which descends from
the mountains to the coast. These incidences support
the assumption that a steep lee slope is a factor in producing mountain wave CAT.
From the above it is apparent that hard and fast rules
cannot be made about when and how differing levels of
severity of CAT occur. Confirmation of this is apparent from research by Colson (1963) where reports of
moderate to severe CAT were received with vertical
wind shear as low as 4 kn/1000 ft. However, Ellrod
(1990) states that moderate CAT usually occurs with
vertical wind shear of 6 to 9 kn/1000 ft and severe CAT
at higher levels of shear, but that the severity of CAT is
compounded by the presence of directional shear.
4. High altitude CAT and mountain wave CAT
4.1. Method
Shortly after GSM aviation products were introduced
in the SAWB Central Forecast Office (CFO), an
opportunity presented itself to view the effectiveness of
the ETI (Ellrod & Knapp, 1992) using the GSM model.
The situation arose when a baroclinic system with a
marked cold front and negative vorticity centre associated with a pronounced upper-air trough passed over
the country on 18 September 1996.
Table 3. Analysis of the vertical and horizontal wind
shear and atmospheric stability at the time of the 34
CAT events during the 1998 survey. PPTC is the
positive potential temperature change.
Three CAT reports were received. Two reports were of
moderate to severe CAT in the lee of the Drakensberg
over the eastern part of the country. The first at FL170
at 0805 UTC (10:05 SAST), followed by a report at a
similar level at 0900 UTC (11:00 SAST). The last report
was of moderate to severe CAT over Victoria West
between FL310 to FL330 at 1030 UTC (12:30 SAST).
Vertical shear <4 kn/1000 ft
Vertical shear 4–5 kn/1000 ft
Vertical shear 6–8 kn/1000 ft
Vertical shear >8 kn/1000 ft
Horizontal (directional and/or speed) shear
Horizontal shear >100 kn/100 miles
Wind shift ≥75°
PPTC <1.5°K/1000 ft, low stability
PPTC 1.5–4.5°K/1000 ft, moderate stability
PPTC >4.5°K/1000 ft, high stability
Observed upper-air data from atmospheric soundings
and numerical model wind analysis from the UK
Meteorological Office (UKMO) from 1200 UTC on 18
September 1996 were used to analyse the upper-air
conditions nearest to the time of the CAT incidents.
Table 4. Severity of CAT reports in relation to vertical wind shear and stability during the 1998 survey.
Severity of CAT
Vertical wind shear (kn/1000 ft)
Stability(°K/1000 ft)
The ETI was then calculated, from the wind analysis
fields, to determine if the index would have given
advance warning of the CAT conditions. This was
compared with the same ETI forecast derived from the
NCEP GSM data received on the Global Transmission
Service (GTS). An example of the model field is given
in Figure 2.
The ETI is based on horizontal stretching deformation
(DST) and shearing deformation (DSH) at the 300, 250
and 200 hPa levels (collectively the DEF) with vertical
wind shear (VWS) in the 400–300, 300–250 and
250–200 hPa layers (see equations (1) and (2) below).
The index has been used by NCEP in Washington
since 1988 (Bakker, 1993). It originates from
Pettersen’s frontogenetic intensity equation which
relates frontogenesis with increased vertical shear and
therefore the likelihood of turbulence (Ellrod &
Knapp, 1992):
DEF = DST 2 + DSH 2
∂u ∂v
∂x ∂y
∂v ∂u
∂x ∂y
with u and v the east/west and north/south horizontal
wind components.
VWS uses the resultant layer difference in u and v wind
components from model forecast data so that:
∆u2 + ∆v2
with ∆u and ∆v changing in east/west and north/south
direction in the vertical layer z.
4.2. Results
Figure 3 reveals that the CAT report over Victoria
West is near the trough axis with cold air advection at
FL300 (approximately 300 hPa) evident from the fact
that the FL300 winds cross the isotherms at a large
angle. Comparison with the winds 5,000 feet lower, at
FL250, shows that the vertical shear was about 8
kn/1000 ft with directional shear. The report is also
close to the jet stream and tropopause shear (Figure 4).
Increased stability is also apparent in the form of a temperature inversion at the level of CAT at De Aar
(Figure 5). According to Hopkins (1977) and Colson
(1961) these are all favourable CAT conditions.
Figure 2. Ellrod Turbulence Index (ETI) for the 24-hour
forecast from the GSM model valid at 1200 UTC on 18
September 1996.
The other two turbulence reports were much further to
the east in the lee of the Drakensberg mountains
(FL150 in Figure 3) with a strong north-westerly wind
and more-or-less under the jet stream (Figure 4). The
wind speed increased with altitude and it was virtually
90° to the Drakensberg mountain range. The air was
also stable, as indicated by the sounding at Durban
(Figure 5). In other words the criteria for the generation of mountain waves were met (Alaka, 1958) and this
was the cause of the turbulence experienced in the lee
of the Drakensberg.
The ETI values, using UKMO analysis data (Figure 3),
produced a band of moderate to severe levels of ETI,
which straddled the report of turbulence over the western interior. Figure 6 gives the corresponding 12-hour
GSM forecast, valid at 0000 UTC on 18 September, of
ETI values at 250 and 300 hPa (approximately FL350
and FL300); these compare favourably with the
UKMO analysis. The GSM prognosis, therefore, presented a more than adequate indication of CAT. As a
result of this CAT was included in local aviation significant weather charts. By 1800 UTC (T+18) the GSM
CAT prognosis showed a marked decrease, with maximum values of 2 over the south-western part of the
country at 250 hPa and it was assumed that the system
had passed its peak. This also led to the exclusion of
CAT on the aviation forecast significant weather charts
in the evening and underlines the transitory nature of
CAT. It is worth noting in Figure 2, that CAT was
already indicated nearly 24 hours in advance, albeit a
bit too far to the east.
GSM ETI values are not calculated for levels below 300
hPa, but they were calculated using UKMO analysis
data (Figure 3). These showed turbulence well to the
west in the vicinity of wind shear, but gave poor results
in the vicinity of the mountain range where horizontal
and vertical wind shear was less evident. The conclusion is that orographic influence in a steady flow cannot be detected by the ETI.
Figure 3. Analysis of upper winds and temperature (left) and ETI (right) for 1200 UTC on 18 September 1996.
5. Conclusions
Analysis of the two surveys reveals that CAT over
South Africa occurs under conditions similar to those
elsewhere in the world.
The jet stream is a dominant factor in the occurrence of CAT and the probability of CAT in association with a jet stream is increased when it passes
over a mountain range. This is particularly evident
in a pre-trough north-westerly flow over the
Drakensberg and the Eastern Cape mountains.
CAT is most likely within 5,000 to 10,000 feet
below the jet stream core and on the cold or low
pressure side of the core. However, mountain wave
CAT is often likely to occur sufficiently far ahead
of the trough for it to occur in the warm air of the
jet stream.
In general increased vertical wind shear and greater
stability are indicative of more severe CAT, but
this is not a strict rule. For example, reports of
moderate to severe and severe mountain wave
CAT were received with weak vertical wind shear
and low stability. This emphasises the increased
effect of mountains in the path of the air flow.
An upper-air trough is the dominant synoptic feature and the high number of reports at the axis of
the trough emphasises the importance of directional shear. However, it is difficult to determine
the effect of horizontal shear on CAT reports due
to the large distances between observing points.
This is particularly so during the 1998 survey.
The results of the limited surveys are in agreement with
overseas research in that most CAT occurred with
curved segments of jet streams associated with troughs
and ridges and vertical wind shear. In accord with the
research of Ellrod & Knapp (1992), moderate CAT is
associated with ETI values of 4–8 and severe CAT with
values above 8.
The limited number of events during the three-year
period is believed to be in part due to the fact that the
incidents are not always passed on to the SAWB. The
other consideration is that, because of South Africa’s
sub-tropical position, occasions of moderate to severe
CAT are not as prevalent as would be the case in more
temperate latitudes nearer to the polar jet stream.
In the study presented, the GSM ETI performed well as
an indicator of CAT where vertical and horizontal
wind shear were present. In practice, in the CFO, the
ETI has been found to be a reliable indicator of CAT
areas. This must be qualified by stating that no detailed
evaluation has been made. Confidence in the ETI is
simply because reported CAT events (when received)
have supported the ETI prognosis. However, the ETI
does not appear to be a reliable indicator of mountain
wave turbulence and in this respect it would be best to
use traditional forecasting methods.
Figure 4. Cross-section of wind (knots, solid lines), air temperature (°C, dashed lines), tropopause (dark solid line), jet
stream core (widely spaced hashed lines) and CAT locations
(closely spaced hashed lines) for 1200 UTC on 18 September
In the introduction it was pointed out that high altitude
CAT has been the cause of numerous incidents in
which aircraft passengers have been injured and sometimes even died. On rare occasions aircraft have suffered structural damage and temporary loss of pilot
control. It is the responsibility of the aviation forecaster
to be conscientious in providing pilots with a warning
of potential areas of clear air turbulence, especially
severe turbulence, at the flight planning stage. This
enables suitable advance changes in flight route to be
made. Furthermore, forewarned is forearmed and a
pilot who encounters CAT is more likely to immediately recognise it for what it is and take suitable corrective action.
The authors are indebted to Mr G. Schulze, Chief
Director of the SAWB, without whose support this
Figure 5. Atmospheric soundings from De Aar (left) and Durban (right) showing the air temperature and dew-point temperature for 0000 UTC (dashed lines) and 1200 UTC (solid lines) on 18 September 1996, plus the location of CAT reports.
Figure 6. 12-hour forecast of ETI from the GSM valid for 1200 UTC on 18 September 1996.
paper would not have been possible. Appreciative
thanks also go to South African Airways for providing
the aircraft data recorder technical information concerning the CAT incident on 10 January 1994; to the
pilots of British Airways (Comair), SA Airlink,
Sabena/Nationwide, SA Express, South African
Airways and Sun Air for their observation and reports
of CAT; and to Air Traffic Navigation Services for
their efforts in relaying information from pilots to
SAWB Weather Offices for the transmission of CAT
reports to the Central Forecast Office.
Alaka, M. A. (1958). Aviation aspects of mountain waves.
W.M.O. No. 68 Technical Note No. 18. Geneva: World
Meteorological Organization, pp. 11, 14, 16, 23.
American Meteorological Society (1989). Glossary of
Meteorology. American Meteorological Society, Boston, p.
Bakker, H. (1993). Drie objectieve indices voor clear-air turbulence nader bekeken. Technische rapporten; TR-160. De
Bilt: Koninklijk Nederlands Meteorologisch Instituut.
Briggs, J. (1961). Severe turbulence near the British Isles.
Meteorol. Mag., 90: 245–254.
Colson, D. (1961). Analysis of clear air turbulence during
April 1960. Mon. Wea. Rev., 89: 94–98.
Colson, D. (1963). Analysis of clear air turbulence data for
March 1962. Mon. Wea. Rev., 91: 73–98.
de Villiers, M. P. (1997). Some examples of clear air turbulence over the Republic of South Africa. In Proc. of Fifth
International Conference on Southern Hemisphere
Meteorology and Oceanography, 7- 11 April 1997, Pretoria,
South Africa. American Meteorological Society, Boston,
Ellrod, G. (1990). Satellite Imagery Indicators of Turbulence.
A Script-Slide Programme. National Oceanic and
Atmospheric Administration/National Environmental
Satellite Data and Information Service. Temple Hills:
National Weather Association.
Ellrod, G. P. (1993). A northern hemisphere clear air turbulence climatology. In Proc. of Fifth International
Conference on Aviation Weather Systems, 2–6 August
1993, Vienna, Virginia, USA, American Meteorological
Society, Boston, 444–448.
Ellrod, G. P. & Knapp, D. I. (1992). An objective clear-air
turbulence forecasting technique: Verification and operational use. Weather and Forecasting, 7: 150–165.
Endlich, R. M. (1964). The mesoscale structure of some
regions of clear air-turbulence. J. Appl. Meteorol., 3:
Godson, J. (1970). Unsafe At Any Height. New York: Simon
and Schuster.
Hopkins, R. H. (1977). Forecasting techniques of C.A.T.
including that associated with mountain waves. W.M.O.
No. 482 Technical Note 155, World Meteorological
Organization, Geneva.
McCann, D. W. (1993). An evaluation of clear-air turbulence
indices. In Proc. of Fifth International Conference on
Aviation Weather Systems, 2–6 August 1993, American
Meteorological Society, Boston, 449- 453.
Met. Office (1991). Meteorology Glossary. Sixth edition.
HMSO, London, 56.
Met. Office (1994). Handbook of Aviation Meteorology.
Third edition. HMSO, London, 81–83.
Shutts, G. (1997). Operational lee wave forecasting. Meteorol.
Appl., 4: 23–35.
Smith, R. B. (1979) The influence of mountains on the atmosphere. Adv. Geophys., 21: 87–230.
Smith, T. L. Ramer, J. E. & Benjamin, S. G. (1995). Map forecasts of aviation-impact variables. In Proc. of Sixth
Conference on Aviation Weather Systems, 15–20 January
1995. American Meteorological Society, Boston, 51–56.
Stack, D. T. (1991). Turbulence avoidance. In Proc. of Fourth
International Conference on Aviation Weather Systems,
24–28 June 1991. American Meteorological Society,
Boston, 283–286.
Starr, J. R. (1996). Source Book to the Forecasters Reference
Book. Met. Office College, Reading.
US Department of Commerce (1966). Report of the National
Committee for Clear Air Turbulence. Environmental
Science Services Administration, Appendix C: 43.

Similar documents


Report this document