You are on page 1of 12

IDENTIFICATION OF VON KARMAN VORTICES IN THE SURFACE

WINDS OF HEARD ISLAND


Research Note
PAUL J. BEGGS1, , PATRICIA M. SELKIRK2 and DENNA L. KINGDOM1
Departments of 1 Physical Geography and 2 Biological Sciences, Macquarie University, Sydney,
NSW 2109, Australia

(Received in final form 12 September 2003)

Abstract. Heard Island (73◦ 30 E, 53◦ 05 S) is an isolated island in the sub-Antarctic located in
a region of strong westerly winds. The dominant topographic feature is a large glaciated volcano,
which peaks at 2,745 m in altitude. Only limited meteorological information exists for the island, with
incomplete records from a station at Atlas Cove covering the period 1948 to 1954, and from a station
at The Spit from 1992 onwards. We present the results of wind observations that were conducted
on Dovers Moraine at the eastern end of Heard Island over the 2000/2001 summer, with the aim of
characterising local winds at this location on the island. Wind was measured using a Woelfle type
mechanical wind recorder. Wind speed was comparable, and on occasions stronger, at this location
compared to simultaneous records at the western end of the island. Winds were predominantly from
the south-south-west, or north-north-west to north. A number of periods of relatively low wind speed
coincided with repeated wind direction patterns, which we associate with the surface signature of
von Karman vortices. Results from this study suggest there are significant topographically-generated
differences in wind characteristics between the eastern and western ends of the island.

Keywords: Heard Island, Sub-Antarctic, Surface, Topography, von Karman vortices, Wind.

1. Introduction

Sub-Antarctic Heard Island (73◦ 30 E, 53◦ 05 S) is approximately 4,850 km south-
east of southern Africa, 4,350 km south-west of Western Australia, and 1,650 km
north of the Antarctic continent (Figure 1). The dominant topographic feature is a
large glaciated volcano, Big Ben, which peaks at 2,745 m in altitude. The island
is 42 km long and 20 km wide, is oriented on a north-west to east-south-east axis,
and covers an area of 385 km2 .
Heard Island’s latitude places it in what are known as the ‘Furious Fifties’. The
predominant wind direction is from the westerly quarter. Only limited meteorolo-
gical information exists for the island. An Australian National Antarctic Research
Expedition (ANARE) station at the western end of the island (Atlas Cove) provided
routine meteorological observations between February 1948 and December 1954
(Site No.: 300005, WMO No.: 95997). The Australian Bureau of Meteorology has
a weather station at Spit Bay (The Spit) at the eastern end of the island (Site No.:
 E-mail: paul.beggs@mq.edu.au

Boundary-Layer Meteorology 113: 287–297, 2004.


© 2004 Kluwer Academic Publishers. Printed in the Netherlands.
288 PAUL J. BEGGS ET AL.

Figure 1. Map of Heard Island, including wind measurement sites [Source: modified from Figure 3,
Australian Antarctic Division, 1995].

300028, WMO No.: 94997) that has been open since April 1992. No long-term
measurements of wind speed or direction have been made at this site.
A number of authors have noted that local topography affects local climate sig-
nificantly (Allison and Keage, 1986; Bergstrom, 1993). For example, winds from
the south are deflected by Big Ben to appear south-west at the Atlas Cove station,
while westerly winds are influenced by Laurens Peninsula and may arrive at the
station from a more northerly direction (Allison and Keage, 1986). Observations
of wind speed and direction were made during the 1985 ANARE to the island (the
first since 1963). Continuous recordings were made at both the Atlas Cove and Spit
Bay ends of the island using Woelfle anemometers. Observations at the Spit Bay
end of the island showed that the area to the north of Dovers Moraine ‘was largely
sheltered from winds in the west-south-west to south to east-south-east sector by
Dovers Moraine and the lowlying hills running to the coast’ and that ‘winds from
these directions generally passed over the area and created areas of turbulence and
VON KARMAN VORTICES ON HEARD ISLAND 289

eddys [sic] which appeared to be accurately reflected on the chart paper record’
(Burton and Williams, 1985, p. 34).
It is important that research into wind on the island, and its interactions with
the island’s biota, be pursued. Previous observations have shown that abrasion
of plants by wind-blown particles, including sand, ash, snow and ice, is a major
environmental disturbance on Heard Island (Bergstrom and Selkirk, 2000). Fur-
ther, it has been stated recently in a review of the island’s climate, that there is
little quantitative information on the regional variability or microclimate (Thost
and Allison, 2003). This paper presents the results of wind observations that were
conducted on the eastern end of Heard Island over the 2000/2001 summer, with the
aim of further characterising local winds at this location on the island.

2. Methods

Data were collected between 1200 local time on 28 December 2000 and 1900 local
time on 23 February 2001. Wind was measured at the eastern end of the island
on Dovers Moraine at 65 m above sea level (73◦ 42.599 E, 53◦ 06.575 S), and at
the western end of the island at Atlas Cove at 21 m above sea level (73◦ 23.437 E,
53◦ 01.031 S). Coordinates were obtained using Global Positioning System (GPS)
in the field. The surface type at the Dovers Moraine site was extremely sparse
vegetation, with moss and low growing plants, and that at the Atlas Cove site was
open cushion carpet (Azorella selago) with a height of about 0.4 m.
Wind run and wind direction were measured using Woelfle type mechanical
wind recorders (Wilh. Lambrecht KG, Gröttingen). The instruments were mounted
such that the sensors were at a height of about 1 m above ground level, and were
calibrated prior to transport to the island.
Wind velocity was manually determined from wind run using two different
transparent evaluation rules supplied with the instrument. These enabled both in-
dividual observations and hourly means to be determined. A wind rose for Dovers
Moraine was produced using the RoseWorks program (Version 1.0, release 25
October 2001, UAI Environmental, Inc.). Two paper charts were used over the
measurement period, and the period between charts when no data were collected
and the chart was changed (2200 on 30 January 2001 to 1400 on 17 February 2001)
was removed for the purposes of producing the wind rose, so that percentage fre-
quencies would be based on periods of measurement (39 days and 15 hours), rather
than the period from first measurement to last measurement. Missing data within
a chart (1.5% of the record) were retained, and therefore the sum of percentage
frequencies for the wind rose is less than 100. For some hours (33.26% of the
record) the wind direction was so variable that it was not possible to assign a single
wind direction. Such hours have been retained for the purposes of producing the
wind rose but have been classed as missing data.
290 PAUL J. BEGGS ET AL.

Figure 2. Mean hourly wind speed at Dovers Moraine and Atlas Cove between 28 December 2000
and 23 February 2001. Dates on the x axis are expressed as days after 1 January 1900.

3. Results

Mean (and standard deviation of) hourly wind speed at Dovers Moraine was 9.3
(6.0) m s−1 , while the minimum and maximum mean hourly wind speeds were 1.1
and 32.6 m s−1 respectively. Mean hourly wind speed at Dovers Moraine over the
measurement period is shown in Figure 2; for comparison with wind speed at the
western end of the island, plots with simultaneous measurements that were made
at Atlas Cove are shown.
Dovers Moraine experienced high wind speeds, with mean hourly speeds of
approximately 15–35 m s−1 often recorded at this location. Only a relatively short
period of overlap between this location and measurements at the western end of
the island (Atlas Cove) was obtained. This suggests generally higher speeds at the
eastern end of the island, with several periods when the Dovers Moraine site wind
speed far exceeded that at the western end of the island.
Wind direction at Dovers Moraine was predominantly from the south-south-
west or north-north-west to north (Figure 3). All winds greater than 25 m s−1 were
from the south-south-west (including the peaks at and around 0900 7 January 2001,
2100 19 January 2001, 0500 25 January 2001, and 1400 20 February 2001) with
much of the wind between 20–25 m s−1 also from this direction. A number of
pronounced oscillations in wind speed with a period of about 20 hours (e.g., around
the peaks on 12, 13/14, 16, 19, and 30 January) were related to changes in wind
direction, with variable direction before and after, and north-north-west or south-
south-west during the stronger period. A number of periods during which the wind
direction was very constant were identified. For example, during much of 7 and 25
January 2001, gale force winds blew constantly from the south-south-west. Wind
direction was rarely from the west or east to south-east.
Hours during which wind direction was too variable to classify were common
at this location. There was often a coincidence between such variable wind direc-
tion and relatively low wind speeds (<10 m s−1 ). During some of these periods,
winds at this location exhibited backing or veering (Figure 4), and appeared to be
VON KARMAN VORTICES ON HEARD ISLAND 291

Figure 3. Wind rose for hourly mean wind at Dovers Moraine. Calms included at centre. Rings
drawn at 5% intervals. Wind flow is from the directions shown. The length of each petal represents
the actual frequency of the direction (and is indicated by the numeric label at the end of each petal
as a percentage), and the length of each segment within a petal represents the actual frequency of the
direction/speed category combination.

periodic, with a period of about one hour. A number of patterns were observed
including:

(i) even veering through a full 360◦ (e.g., Figure 4a, 1700 and 1800 on 8 January;
Figure 4c, 0600 and 0700);
(ii) even backing through a full 360◦ ;
292 PAUL J. BEGGS ET AL.

(iii) north then rapid veering to south, then rapid veering to north again (e.g., Figure
4a, 1900–0100; Figure 4c, 0800–1300);
(iv) north then rapid veering to south, then rapid backing to north again;
(v) north then rapid veering to south, then gradual backing to north again (e.g.,
Figure 4a, 1200–1700);
(vi) veering then backing through increasing ranges from north to east, to north to
east to south, such that the wind direction trace appears like a sine wave with
increasing amplitude (e.g., Figure 4b, 2300–0700).
Most events included a combination of these patterns. Such events lasted from
about 4 hrs to 43 hrs, although most were no longer than about 25 hrs. Such
events were reasonably common during the period of measurements. About ten
such events were identified, making an average frequency of about once every four
days.

4. Discussion

Results from this study suggest there are significant topographically-generated dif-
ferences in wind characteristics between the eastern and western ends of the island.
For example, the predominant south-south-west and north-north-west to north wind
directions (and infrequence of west-south-west to west-north-west) at the eastern
end of the island would appear to result from the upwind airflow moving around
the mountain, with very little flowing up and over the mountain. There would also
seem to be little evidence of katabatic flows off the glaciers adjacent to Dovers
Moraine (which would come from the west), although if such winds were present
on occasions, they may be diverted around the raised eastern portion of Dovers
Moraine (upon which the current measurements were made).
Height above sea level might be one explanatory factor for the high wind speeds
at Dovers Moraine. It has been noted by Thost and Allison (2003), for example,
that occasional wind speed and direction records from Spit Bay in 1985, 1990 and
1992, are all ‘affected by a wind shadow caused by Dovers Moraine, such that a
10 m s−1 wind recorded at the camp was often a 35 m s−1 wind only 10 m higher
on the moraine’ (Thost and Allison, 2003, p. 7). Lack of vegetation, and therefore
a lower roughness length, might be another explanatory factor for the high wind
speeds at Dovers Moraine. A lower roughness length may be part of a ‘positive
feedback’, whereby wind speed is high because the surface is relatively smooth,
and the surface is relatively smooth because the high wind speed retards the growth
of vegetation.
Observations on the island suggest the frequent strong and gale force winds
result in excessive production of aeolian sediment (Burton and Williams, 1985;
Kiernan and McConnell, 1999). Other observations of vegetation have suggested
that this aeolian sediment may contribute to significant plant damage in certain
locations (Bergstrom and Selkirk, 2000; Bergstrom et al., 2002). The concentration
VON KARMAN VORTICES ON HEARD ISLAND 293

Figure 4. Wind direction (top) and run (bottom) at Dovers Moraine from about (a) 1400 on 8 January
2001 to 1900 on 9 January 2001, (b) 1600 on 18 January 2001 to 1900 on 19 January 2001, and (c)
1800 on 21 January 2001 to 1700 on 22 January 2001, showing wind direction pattern representative
of von Karman vortices. The numbers (1–24) in the middle of the chart are the hours of the day. The
top of the chart indicates the direction from which the wind blows. Each trace across the bottom of
the chart represents a wind run of 10,000 m, and the closer together and more vertical the traces, the
faster the wind speed.
294 PAUL J. BEGGS ET AL.

of wind direction from the north and south may mean that plants in the area are
exposed to abrasion from two directions, rather than one, which has been observed
on other parts of the island where the wind direction is less variable. Indeed, this
may be another factor in the previously mentioned lack of vegetation at this site.
The wind direction patterns shown in Figure 4 and described in the Results are
interpreted as the surface signature of von Karman vortices, where a high isolated
three-dimensional mountain and a shallow boundary layer with overlying inversion
(whose base is below the height of the mountain) forces airflow around, rather
than over, the mountain (Big Ben). Although radiosonde observations were not
possible during the 2000/2001 summer research season on the island, analysis
of such observations for the month of January from the ANARE station (1949–
1954) suggests such conditions to be reasonably common. For example, inversions
between 900 and 850 hPa (geopotential height typically 810 m and 1,264 m re-
spectively), and 850 and 800 hPa (geopotential height typically 1,264 m and 1,741
m respectively), are common (observations obtained from ANARE (1953, 1955,
and 1957). Vortices form alternately on each side and to the lee of the mountain,
with vortices on each side rotating in opposite directions. Vorticity is generated by
friction with the mountain sides (Stull, 1988; Sturman and Tapper, 1996). Although
von Karman vortex streets have been observed downwind of the island in remotely
sensed images (NASA, 2000), which have been commented on briefly elsewhere
(Thost and Allison, 2003), this is the first known report from surface observations
on the island.
Investigations of von Karman vortex streets, both in the atmosphere and ocean,
have appeared in the literature since the first observation from space of a von Kar-
man vortex street, from Madeira Island in 1962 (Caldeira et al., 2002). Such studies
have included laboratory experiments (Barnett, 1972; Brighton, 1978; Pao et al.,
1982; Kourta et al., 1987; Horváth et al., 2000), numerical simulations (Potylitsin
and Peltier, 1998; Sohankar et al., 1998), and remote sensing (Jensen and Agee,
1978; Caldeira et al., 2002). Few field studies of von Karman vortices have been
conducted. One such study is that of Moll (1971), who analysed meteorological
data taken at the time of a particular satellite photograph of Madeira. No other
study examining the influence of von Karman vortices on surface winds at the
leeward end of oceanic islands is known to the authors.
Although measurements of wind speed and direction upwind of the island for
comparison with measurements on the island could not be conducted over the
2000/2001 summer, it is hypothesised that the occasions on which the wind direc-
tion patterns, related here to the surface signature of von Karman vortices, occurred
were during upwind westerly flow, making the eastern end of the island (includ-
ing Dovers Moraine) at the downwind end of the island. In such a situation, the
eastern end of the island would alternately encounter the influence of the vortices
formed on each side of the island (the northern and southern sides of the island
in this case). This is worth noting because at least one satellite image (Figure 2 in
Thost and Allison, 2003) shows von Karman vortex streets persisting about 440
VON KARMAN VORTICES ON HEARD ISLAND 295

km to the north of Heard Island, having been formed under southerly winds. In this
situation, it may be the northern side of the island that encounters wind direction
patterns of the sort described in this paper, while the western and eastern ends of
the island might encounter much more constant winds of the sort that have also
been described herein for the eastern end.
The von Karman vortices may have important implications for both terrestrial
and marine biota. There is potential that the cloud-free centres of the vortices
would allow increased exposure of the sea-surface layer downwind of the island
to solar radiation, which may result in the formation of a warm water wake. Recent
investigations around the Hawaiian Islands (Xie et al., 2001) and Madeira Island
(Caldeira et al., 2002) support the potential for air-sea interaction involving wind
wakes such as von Karman vortex streets. For example, Caldeira et al. (2002)
proposed that weaker winds and lower cloud coverage to the lee of Madeira Island
decrease sea-surface mixing and increase surface solar heating, and warming of
the sea surface takes place. The land area to the lee of Big Ben also appears to
be relatively cloud-free compared to the western end of the island at certain times
(Thost and Allison, 2003), including during von Karman vortex formation, and this
may also expose terrestrial biota to increased levels of total solar and ultraviolet
radiation. This could have important impacts. For example, studies from other
high latitude locations have found effects of enhanced UV-B on both terrestrial
and marine organisms (Newman et al., 1999; Solheim et al., 2002). Heard Island
lies in a region susceptible to the receipt of such enhanced UV-B.

5. Conclusion

Wind speed and direction characteristics exhibit considerable spatial variability on


Heard Island. Wind varies between the two ends of the island as well as between
different locations at each end of the island. Results from this and previous work
suggest the eastern end of the island is just as harsh, and in some ways harsher, than
the western end of the island in terms of excessive wind speed. The identification of
wind patterns associated with von Karman vortices on the eastern end of the island
warrant further investigation, as does the existence of a von Karman vortex surface
signature on other isolated mountainous oceanic islands. Future work is also to be
done on the implications of the wind climate on insect and pollen dispersal on the
island.

Acknowledgements

The authors wish to acknowledge the assistance of Robert Hyde, Jenny Potter, and
Don Adamson (all of Macquarie University), and Dana Bergstrom (of the Aus-
tralian Antarctic Division). We thank three anonymous referees for their valuable
comments on the manuscript.
296 PAUL J. BEGGS ET AL.

References

Allison, I. F. and Keage, P. L.: 1986, ‘Recent Changes in the Glaciers of Heard Island’, Polar Rec.
23(144), 255–271.
ANARE: 1953, 1955, 1957, ‘Heard and Macquarie Islands, 1949–1953’ and ‘Mawson, Heard and
Macquarie Islands, 1954’, A.N.A.R.E. Reports, Series D, Meteorology II–VII.
Australian Antarctic Division: 1995, ‘Heard Island Wilderness Reserve Management Plan’, Depart-
ment of the Environment, Sport and Territories, Canberra, ACT, 72 pp.
Barnett, K. M.: 1972, ‘A Wind-Tunnel Experiment Concerning Atmospheric Vortex Streets’,
Boundary-Layer Meteorol. 2, 427–443.
Bergstrom, D. M.: 1993, Bryophyte Diversity and Terrestrial Plant Ecology in the Subantarctic:
Heard Island, A Case Study, Ph.D. Thesis, Macquarie University, Sydney, NSW, 173 pp.
Bergstrom, D. M. and Selkirk, P. M.: 2000, ‘Terrestrial Vegetation and Environments on Heard
Island’, Papers and Proceedings of the Royal Society of Tasmania 133(2), 33–46.
Bergstrom, D. M., Whinam, J., and Belbin, L.: 2002, ‘A Classification of Subantarctic Heard Island
Vegetation’, Arct. Antarct. Alp. Res. 34, 169–177.
Brighton, P. W. M.: 1978, ‘Strongly Stratified Flow Past Three-Dimensional Obstacles’, Quart. J.
Roy. Meteorol. Soc. 104(440), 289–307.
Burton, H. R. and Williams, D. L.: 1985, Heard Island ANARE 1985 Report, Antarctic Division,
Kingston, Tasmania, 135 pp.
Caldeira, R. M. A., Groom, S., Miller, P., Pilgrim, D., and Nezlin, N. P.: 2002, ‘Sea-Surface Signa-
tures of the Island Mass Effect Phenomena around Madeira Island, Northeast Atlantic’, Remote
Sensing Environ. 80, 336–360.
Horváth, V. K., Cressman, J. R., Goldburg, W. I., and Wu, X. L.: 2000, ‘Hysteresis at Low Reynolds
Number: Onset of Two-Dimensional Vortex Shedding’, Phys. Rev. E 61(5-A), R4702–R4705.
Jensen, N. O. and Agee, E. M.: 1978, ‘Vortex Cloud Street during AMTEX 75’, Tellus 30, 517–523.
Kiernan, K. and McConnell, A.: 1999, ‘Geomorphology of the Sub-Antarctic Australian Territory of
Heard Island-McDonald Island’, Aust. Geographer 30, 159–195.
Kourta, A., Boisson, H. C., Chassaing, P., and Ha Minh, H.: 1987, ‘Nonlinear Interaction and the
Transition to Turbulence in the Wake of a Circular Cylinder’, J. Fluid Mech. 181, 141–161.
Moll, H.-G.:1971, ‘Die Atmosphärische Umströmung Madeiras’, Beitr. Phys. Atmos. 44, 227–244.
NASA: 2000, Various Views of von Karman Vortices, Goddard Space Flight Center, Distributed
Active Archive Center.
URL: http://daac.gsfc.nasa.gov/CAMPAIGN_DOCS/OCDST/vonKarman_vortices.html. Last
Revision Date: 08:16:55 EDT 1 September 2000. Last Access Date: 15 August 2002.
Newman, S. J., Nicol, S., Ritz, D., and Marchant, H.: 1999, ‘Susceptibility of Antarctic Krill
(Euphausia superba Dana) to Ultraviolet Radiation’, Polar Biol. 22, 50–55.
Pao, H.-P., Lai, R. Y., and Schemm, C. E.: 1982, ‘Vortex Trails in Stratified Fluids’, Johns Hopkins
Univ. Appl. Phys. Lab. Tech. Dig. 3, 12–18.
Potylitsin, P. G. and Peltier, W. R.: 1998, ‘Stratification Effects on the Stability of Columnar Vortices
on the f -Plane’, J. Fluid Mech. 355, 45–79.
Sohankar, A., Norberg, C., and Davidson, L.: 1998, ‘Low-Reynolds-Number Flow around a Square
Cylinder at Incidence: Study of Blockage, Onset of Vortex Shedding and Outlet Boundary
Condition’, Int. J. Numer. Meth. Fluids 26: 39–56.
Solheim, B., Johanson, U., Callaghan, T. V., Lee, J. A., Gwynn-Jones, D., and Bjorn, L. O.: 2002,
‘The Nitrogen Fixation Potential of Arctic Cryptogram Species Is Influenced by Enhanced UV-B
Radiation’, Oecologia 133, 90–93.
Stull, R. B.: 1988, An Introduction to Boundary Layer Meteorology, Kluwer Academic Publishers,
Dordrecht, 666 pp.
Sturman, A. and Tapper, N.: 1996, The Weather and Climate of Australia and New Zealand, Oxford
University Press, Melbourne, 476 pp.
VON KARMAN VORTICES ON HEARD ISLAND 297

Thost, D. and Allison, I.: 2003, ‘The Climate of Heard Island’, in K. Green and E. Woehler (eds.),
Heard Island: Southern Ocean Sentinel.
Xie, S.-P., Liu, W. T., Liu, Q., and Nonaka, M.: 2001, ‘Far-Reaching Effects of the Hawaiian Islands
on the Pacific Ocean-Atmosphere System’, Science 292, 2057–2060.

You might also like