R. Manasseh Papers
R. Manasseh Home
The instrumentation at Sydney Airport
(figure 1) consists of five tower-mounted
anemometers on the airfield plus a sixth at Kurnell. The anemometers are about
13 m above ground level (and thus, effectively `at sea level') and well sited
away from obstructions, with the exception of sensor 6 at Kurnell, which
suffers from a higher wind variance owing to its proximity to a sand dune. The
wind speeds and directions and are logged every two seconds at each sensor. The
anemometers are cup-and-vane type instruments, for which the inherent
mechanical filtering can be expected to cut out gusts with frequencies higher
than about 1 Hz. The resolution of the data in speed is 1 knot (about
0.5 
) and is 1 degree in direction. Air pressure at the airport's
meteorological station is logged at ten-second intervals. The data is
transmitted in real time to the University of New South Wales (UNSW) for daily
analysis and archiving. The data files are each 24 hours long, and run from
midday to midday Eastern Standard Time (Universal Time + 10 hours). The
system was originally initiated with a midday changeover to permit UNSW staff
to make adjustments to the system during the working day. Pre-processing
and analysis codes check for and remove spurious records.
Figure 1: Sensors 1 to 5 are on the Sydney Airport airfield: Sensor 1 and Sensor
2 are at either end of the shorter runway. Sensor 4 and Sensor 5 are at either
end of the longer runway. Sensor 6 is at Kurnell and is currently of limited
use owing to poor siting. Met: the airport's meteorological station.
With regard to aviation safety, information on the surface wind gust regime
should give an indication of the likelihood of gusts of various strengths both
along and across major runways. To this end, the wind velocities from Sydney
Airport have been resolved into two orthogonal components 
and 
, based on
the directions of runway 16/34. A positive direction indicates winds
blowing from the Botany Bay end of runway 16/34 towards the terminal buildings
end, ie. on a bearing of 
48
14
from true North. A positive
thus represents an approximately southerly wind. The positive is
orthogonal to the positive direction and hence represents an approximately
westerly wind. Note that runway 07/25 is not exactly at right angles to runway
16/34 so that the component is not exactly aligned with it.
A gust is defined as the deviation from a local mean 
of wind velocity. Thus for measurements of the horizontal wind velocity the
gust components are simply
The local mean is taken over a relatively short period; since aircraft
descending or ascending near an airport can cover a kilometre in less than half
a minute, 20 seconds are used here for the averaging time. In contrast, for
wind-engineering applications the averaging period is typically one
hour
. For the purposes of this paper, changes in wind speed
occurring over timescales longer than 20 seconds are not considered as `gusts';
although many severe meteorological conditions, for example thunderstorm
microbursts, produce sustained wind shear for periods longer than 20 seconds,
the immediate concern of this paper is the `instantaneous' gusts. The aim is
simply to estimate probability distributions of gust strength.
Therefore, unlike in wind-engineering models, no statistical requirements (such
as stationarity) are placed on the nature of the data. Since the data files are
24 hours long, the statistics are calculated on a daily basis. It will be
demonstrated shortly that the relative scarcity of strong gusts means that
statistical significance would not be obtained by considering gusts over a much
shorter period, say one hour.
There are several ways of looking at gusts from the point of view of aircraft
operations. The simplest, true-direction gusts, involve calculations of
and 
as defined in (1)-(2),
their means, variances, and higher moments. These gust statistics are easy to
interpret if the wind was blowing in essentially the same direction throughout
the statistical period. For example, if the wind always had a westerly
component, a negative is always a drop (or reversal) in the westerly
component of wind; if the wind had an easterly component, a negative is
always an increase in the easterly component of wind. However, if the wind
direction varied throughout the statistical period, some of the meaning of
positive and negative gusts is lost - a negative could be either a
reduction in the local mean or an increase. An alternative system of gust
classification, operations-relevant gusts, has more meaning for aircraft
operations under a variety of wind directions. This system will be described in
§2.3.
R. Manasseh Papers
R. Manasseh Home
