Oceanography and circulation pattern of the Zeewijk Channel, Houtman Abrolhos Islands, Western Australia. MICHAEL MASLIN This dissertation is submitted as partial requirement for the degree of Bachelor of Engineering (Applied Ocean Science) Supervisor: Prof. Charitha Pattiaratchi
94
Embed
Oceanography and circulation pattern of the Zeewijk …...Oceanography and circulation pattern of the Zeewijk Channel, Houtman Abrolhos Islands, Western Australia. MICHAEL MASLIN This
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Oceanography and circulation pattern of the Zeewijk Channel, Houtman Abrolhos
Islands, Western Australia.
MICHAEL MASLIN
This dissertation is submitted as partial requirement for the degree of Bachelor of Engineering (Applied Ocean Science)
Supervisor: Prof. Charitha Pattiaratchi
i
ABSTRACT The Houtman Abrolhos Islands are a significant contributor to Australia’s western rock lobster industry and they are also a major tourist attraction for Geraldton and the central west coast. Understanding the physical and biological characteristics of the region has been critical for well-developed management in the past, and is essential for a future that may involve aquaculture developments within this Fish Habitat Protection Area. Oceanographic processes on a range of scales affect the channels between island groups, channels which comprise a large portion of the Abrolhos’ functionality as both a fishery and worthwhile tourist destination. Understanding the circulation patterns, with this thesis particularly interested in the Zeewijk Channel, can give valuable and practical insights that may be used in future assessment of the island chain. For the spring period in 2002 and 2003 wind, wave and climate data has been reviewed and correlates with a strong influence on surface dynamics in the channel. An ADCP measured surface and bottom currents during this time frame with current strengths and direction varying greatly between surface and bottom. Sustained periods of strong westerly currents at the surface indicate that water expulsion towards the continental shelf slope from the channel is significant and linked to the local wind pattern. Bottom currents exhibit a strongly cyclical trend related to the local tidal regime. Bathymetric, salinity and temperature information from the CSIRO vessel Southern Surveyor in 2003 aided in assessing the hydrodynamic properties of the channel as relatively well mixed. It also depicted the Leeuwin Current’s influence over the continental shelf slope during the testing period and the likely influence on channel circulation indirectly associated with its presence. The dynamics of the Zeewijk Channel currents give a high degree of support to the aquaculture trial proposed for its confines due to the water exchange that they represent at the surface. However further investigation into vertical particle transport is required since the re-circulation of particles at the bottom may present an environmental risk. Seasonal and inter-annual variability has also influenced the outcome of this thesis greatly.
ii
iii
ACKNOWLEDGEMENTS
This thesis has been produced with the general support and assistance of a number of people.
First and foremost, I would like to thank my supervisor, Professor Charitha Pattiaratchi for
his help and guidance throughout the year. His patience, time and assistance are greatly
appreciated.
I would like to thank the Australian Hydrographic Service, the Bureau of Meteorology, DPI
and the CSIRO for contributing the data to make this study possible. It was not necessarily
meant for me initially, but when I approached all parties there assistance was prompt and
friendly, and the information crucial.
To all the students at the Centre for Water Research, thanks for the most enjoyable and
challenging year of university I have ever had. You were great company through the long
hours spent in the lab and the occasional tavern visits. Cheers to the masters of MATLAB
and Excel, you know who you are and the invaluable aid you gave me on the computer front
is appreciated.
A special and sincere thank you must go out to “Team Ocean”. I look forward to the
consultancy we open in a decades time, and the fulfilment of the power and potential that
is…….”Team Ocean”.
To my mother and father, a huge thanks for putting up with me all year. Without your support
and direction I might not have made it here (I would have run out of petrol). To all my friends
that told me to ‘work like a dog’ this year, thanks for the inspiration.
Michael Maslin
October 2005
A man may write at any time, if he will set himself doggedly to it. -- Samuel Johnson
iv
Contents v
TABLE OF CONTENTS ABSTRACT .......................................................................................................................................................... I ACKNOWLEDGEMENTS .............................................................................................................................. III TABLE OF CONTENTS .................................................................................................................................... V LIST OF FIGURES..........................................................................................................................................VII LIST OF TABLES.......................................................................................................................................... VIII 1. INTRODUCTION....................................................................................................................................... 1 2. BACKGROUND ......................................................................................................................................... 4
2.2.3.1. Western Rock Lobster.........................................................................................................................12 2.2.3.2. Southern Saucer Scallop .....................................................................................................................13 2.2.3.3. Aquaculture Proposal..........................................................................................................................13
2.3. OCEANOGRAPHIC SETTING................................................................................................................. 15 2.3.1. The Leeuwin Current .................................................................................................................... 15
2.3.1.1. The Capes Current ..............................................................................................................................16 2.3.2. El Niño/Southern Oscillation........................................................................................................ 18 2.3.3. Physical Parameters ..................................................................................................................... 19
2.3.3.1. Sea Temperature .................................................................................................................................19 2.3.3.2. Salinity................................................................................................................................................20 2.3.3.3. Stratification and mixing.....................................................................................................................21
4.1.1. Surface Temperature..................................................................................................................... 42 4.1.2. Water Column Temperature ......................................................................................................... 43
4.3. WIND.................................................................................................................................................. 64 4.3.1. Spring 2003................................................................................................................................... 64 4.3.2. Spring 2002................................................................................................................................... 66
4.4. WAVES............................................................................................................................................... 67 4.4.1. Spring 2002................................................................................................................................... 67
4.5. EDDY FORMATION ............................................................................................................................. 69 4.5.1. Spring 2003................................................................................................................................... 69
5. DISCUSSION ............................................................................................................................................ 71 6. CONCLUSIONS ....................................................................................................................................... 76 7. RECOMMENDATIONS FOR FUTURE WORK ................................................................................. 77 REFERENCES ................................................................................................................................................... 79
Contents vii
LIST OF FIGURES Figure 1.1: the Houtman Abrolhos Islands from space. Image courtesy of Earth Sciences and Image Analysis Laboratory, NASA Johnson Space Centre.......................................................3 Figure 2.1: the Houtman Abrolhos Islands. The extensive reef system can be seen in green relative to the small island areas illustrated in yellow (Penn (ed)1999). ...................................5 Figure 2.2: focus on the Zeewijk Channel bathymetry looking in the along-channel direction. Half moon reef is represented by the dark red section on the middle right of the figure and the Easter Island reef system is on the left. The image was created using the digitisation method described in the methodology. ...................................................................................................7 Figure 2.3: typical winter synoptic weather chart for Australia in 2003 (B.O.M. 2005). .........8 Figure 2.4: typical summer synoptic weather charts for Australia in 2003 (B.O.M. 2005). .....9 Figure 2.5: map of the sea-cage site in the Zeewijk channel (Diver & Prince 2003)..............14 Figure 2.6: The progression of cooler, Capes Current water northwards can be seen in this satellite image. Also of note are the systems of eddies off the continental shelf that are comprised of Leeuwin Current water that has entrained the Capes water from the shallower regions (Pearce & Pattiaratchi 1999). ......................................................................................17 Figure 2.7: Southern Oscillation Index for the beginning of 2000 until mid-2005. Negative SOI for extended periods will mean El Niño conditions, positive means La Niña (B.O.M. 2005). .......................................................................................................................................18 Figure 2.8: Sea surface temperatures as measured by mercury thermometer on monthly visits to the puerulus collectors; estimated accuracy 0.3°C. The period of data analysis is from 1970 to 1976 and then 1984 to 1995. Sampling is ongoing (CSIRO 2005). ...........................20 Figure 2.9: Monthly mean salinities from the CSIRO hydrographic station near Geraldton. Salinity at the surface and at a depth of 40m is illustrated (adapted from Sukumaran 1997). 21 Figure 2.10: the processes driving stratification and mixing of a water column.....................21 Figure 2.11: Ekman transport in the Northern Hemisphere. The wind-induced surface current turns 45° to the direction of the wind (Pattiaratchi 2005)........................................................27 Figure 2.12: wind and current directions according to the recognised naming conventions...33 Figure 3.1: Southern Surveyor transects and approximate sampling station locations through the Zeewijk and Middle Channels, October 27th and 28th, 2003 (Australian Hydrographic Service, 2005). .........................................................................................................................38 Figure 3.2: the ADCP used in the Zeewijk Channel and its deployment location. .................39 Figure 3.3: digitised bathymetry of the Zeewijk Channel. Depth is in metres on the z-axis, latitude and longitude are on the y and x-axis respectively. Data courtesy of the Australian Hydrographic Service (2005)...................................................................................................41 Figure 4.1: the change in surface water temperature from the continental shelf slope, into the Zeewijk Channel. A noticeable temperature decrease is observed after 20km where the testing entered waters beyond the 60m contour line of the continental shelf. .........................43 Figure 4.2: temperature-depth profiles of stations 1 to 5 on the Zeewijk transect, 27/10/03. .44 Figure 4.3: temperature contour plot relative to the depth of the continental shelf.................44 Figure 4.4: temperature-depth profiles of stations 6 to 9, and a microscopic view of Figure 4.2, to compare the temperatures of the inner and outer stations in the surface 100m on the Zeewijk Channel transect, 27/10/03. .......................................................................................45 Figure 4.5: the change in surface water salinity from the continental shelf slope, into the Zeewijk Channel. A noticeable salinity increase is observed after 20km where the testing entered waters beyond the 60m contour line of the continental shelf. The same annotations as the surface temperature plot apply. ..........................................................................................46 Figure 4.6: salinity-depth profiles of stations 1 to 5 on the Zeewijk transect, 27/10/03. ........47 Figure 4.7: salinity contour plot relative to the depth of the continental shelf. .......................47 Figure 4.8: salinity-depth profiles of stations 6 to 9, and a microscopic view of Figure 4.6, to compare the temperatures of the inner and outer stations in the surface 100m on the Zeewijk Channel transect, 27/10/03.......................................................................................................48
Contents viii
Figure 4.9: TS diagrams to compare the water column properties of the Zeewijk and Middle Channels. The lower figures represent the water column properties of the ‘channel stations’, that is, stations 6 to 9. Station 1 has been omitted from all figures to aid interpretation due to the scale of its depth and properties. Station 9 of the Middle Channel is located at approximately (35.61, 19.2), and is hard to view as it is almost a point. ................................50 Figure 4.10: Temperature-salinity diagram highlighting the currents present in the water column on the Gascoyne continental shelf slope, particularly out at the 1000m contour. ......51 Figure 4.11: current rose of surface currents in the Zeewijk Channel during spring 2003. ....54 Figure 4.12: current rose of bottom currents in the Zeewijk Channel during spring 2003......54 Figure 4.13: surface current variability in the Zeewijk Channel for the cross- and along-shore directions during spring 2003. .................................................................................................55 Figure 4.14: bottom current variability in the Zeewijk Channel for the cross- and along-shore directions during spring 2003. .................................................................................................56 Figure 4.15: plot of the path travelled by a particle under the influence of the mean daily surface currents in spring 2003. The current path is elliptical in the along-channel direction...................................................................................................................................................58 Figure 4.16: progressive plot of the path travelled by a particle under the influence of the mean daily bottom currents in spring 2003. There is a strong north-west to south-easterly current trend. ............................................................................................................................59 Figure 4.17: the distances travelled by a particle at the surface under the daily mean current conditions for each day of the testing period. ..........................................................................60 Figure 4.18: the distances travelled by a particle near the bottom under the daily mean current conditions for each day of the testing period. ..........................................................................61 Figure 4.19: Spectral density versus frequency for surface currents in the Zeewijk Channel during spring 2003. ..................................................................................................................63 Figure 4.20: Spectral density versus frequency for bottom currents in the Zeewijk Channel during spring 2003. ..................................................................................................................63 Figure 4.21: wind influence on surface currents for 16 days of the testing period, between 1/9/03 and 17/9/03. ..................................................................................................................64 Figure 4.22: wind influence on surface currents for 10 days of the testing period, between 28/10/03 and 6/11/03. ..............................................................................................................65 Figure 4.23: current rose of winds recorded at the North Island recording station 1/10/02 and 31/12/02. ..................................................................................................................................66 Figure 4.24: depicts the wave height, period and direction of waves in the Zeewijk Channel for the period 12/10/02 to 16/11/02. ........................................................................................68 Figure 4.25: sea level anomaly, related geostrophic velocity and sea surface temperature for post-October 6th 2003 (CSIRO 2005). .....................................................................................70 Figure 4.26: sea level anomaly, related geostrophic velocity and sea surface temperature for post-November 9th 2003 (CSIRO 2005). .................................................................................70
LIST OF TABLES Table 2.1: tidal constituents for the Pelsaert Island Group 4. Adapted from the Australian National Tide Tables (2002). ...................................................................................................35 Table 3.1: exact locations, and the associated water depth, of the sampling stations for CTD testing on the Zeewijk and Middle Channel transects. ............................................................37 Table 4.1: temperature, salinity and related density of surface water inside the Zeewijk Channel and above the continental shelf slope. .......................................................................49
ix
x
Introduction 1
1. Introduction
As part of a large Dutch East India Company fleet carrying considerable wealth and trade
goods, the Zealand ship Zeewijk left Holland in November of 1726 under strict orders to
follow a prescribed course across the Indian Ocean and into safe anchorage at Batavia, Java.
Against these orders and into disaster, its ambitious captain took a heading to the east-north-
east, with the intent to explore the relatively undiscovered ‘Eendracht’, known today as
Western Australia. At 7:30pm on the 9th of June 1727 and less than 70km from the Western
Australian coastline, the ship crashed into the northern edge of Half Moon Reef at the
Houtman Abrolhos Islands. The lookout had perilously mistaken the line of breakers as the
moons reflection on the sea surface. Unbeknownst to the crew, they had come agonisingly
close to sailing smoothly through a safe channel fractionally to the north, and into
Unlike the ill-fated vessel Batavia, also shipwrecked at the Abrolhos Islands and famous for
its tales of mutiny, murder and cannibalism, the crew of the Zeewijk faced shipwreck and
stranding with firm resolve against the isolation of the Abrolhos and fierce surf on the
offending reef during salvage. Eventually they constructed a sloopy capable of carrying the
survivors to Java. Their journey lasted over a month and when they finally reached Batavia in
April of 1728, only 82 of the original 208 strong party that had set out from the Netherlands
remained. The channel that had offered tantalising access to the Abrolhos Islands and in turn
Western Australia’s coastal waters, is named in recognition of their plight (Ingleman-
Sundberg 1976).
The Zeewijk Channel is one of three breaks in the Houtman Abrolhos archipelago that
enables transition between the continental shelf slope and coastal shelf waters. Situated
between the Pelsaert and Easter Group Islands, it also provides access to the lagoons and
islands of these systems. The local current pattern is viewed as a major cause of the
geological evolution of the island chain. Research has shown it to be strongly affected by
mesoscale features such as the Leeuwin Current on the continental shelf slope side, as well as
more direct and localised parameters such as wind and the tidal regime. It is important to
understand these processes and the nature of the channel’s circulation as it has important
ramifications for the biota of the region and anthropogenic exponents of its fisheries.
Introduction 2
The Abrolhos Island’s have been classified a Fish Habitat Protection Area by Fisheries
Western Australia, such is the ecological value of the region. Juxtaposed against this
classification is the significant proportion of the states Western Rock Lobster catch, nearly
15% at an estimated value of $50 million, taken in the immediate vicinity. The Abrolhos’ role
in lobster recruitment for the rest of the state is still being investigated but early studies have
indicated its importance to the industry state-wide, as a lobster puerulus breeding ground with
significant oceanographic parameters distributing the puerulus over vast distances.
Diversification of the local fishery includes a proposed yellow fin tuna aquaculture project to
be established in the Zeewijk Channel. An understanding of the channel’s currents and
therefore the water circulation pattern could aid future environmental impact assessments in
the context of an aquaculture farm in the area. The choice of aquaculture sites on the basis of
the circulation or water exchange regime of the area makes sound environmental reasoning
considering the potential for benthic habitat damage associated with these endeavours.
This thesis aims to examine the oceanography and circulation pattern of the Zeewijk Channel,
specifically in terms of the spring conditions in 2002 and 2003. The mesoscale oceanography
of the Gascoyne continental shelf and shelf slope has also been examined, to assess the extent
of its influence on the Zeewijk Channel circulation pattern.
A background consisting of the Abrolhos Islands physical, biological and oceanographic
characteristics has been established through critical review of previous studies of the region
(chapter 2). The procedures employed to obtain useful and varied sets of data are outlined in
the methodology (chapter 3). A successful collation of data was paramount to achieving the
aims of this dissertation. Analysis and synthesis of this data gives insight into the pattern of
circulation in the Zeewijk Channel and is displayed in the results and discussion (chapters 4
and 5). These chapters also speculate on the mesoscale processes that may be driving
circulation in the channel.
Introduction 3
Figure 1.1: the Houtman Abrolhos Islands from space. Image courtesy of Earth Sciences and Image Analysis Laboratory, NASA Johnson Space Centre.
Background 4
2. Background
2.1. Physical Setting
The natural characteristics of the Zeewijk Channel, and greater Houtman Abrolhos Islands,
will be outlined in this chapter with particular consideration of the meteorological and
oceanic processes that shape them. The physical and biological characteristics of the region
are closely related to these processes.
2.1.1. The Houtman Abrolhos Islands
An archipelago of 122 small islands and a considerable reef system, the Houtman Abrolhos
Islands are located approximately 70 km west of Geraldton, Western Australia, and 10 km
inside the 200 metre depth contour of the nearby continental shelf. At latitude 28°15’ to 29°S,
and longitude 113°35’ to 114°05’E, it comprises the southern most coral reef complex in the
Indian Ocean with a general area of around 1100km2 to the 50m depth contour, and a low tide
island area of approximately 18.5 km2.
The Houtman Abrolhos can be divided geographically into three main groups of islands and
each group is roughly triangular in shape. The Pelsaert (Southern) Group is separated from
the Easter Group by the Zeewijk Channel and the Middle Channel divides the Easter and
North Island-Wallabi Groups. The channels are each approximately 40 m deep and there is an
extensive submerged reef platform connecting North Island to the rest of the Wallabi Group.
Background 5
Figure 2.1: the Houtman Abrolhos Islands. The extensive reef system can be seen in green relative to the small island areas illustrated in yellow (Penn (ed)1999).
Background 6
2.1.1.1. Bathymetry
The Gascoyne continental shelf is located along the northern central coastline of Western
Australia between latitudes 21°S and 29°S. Off Geraldton, the shelf is relatively flat and
shallow with a slope of approximately 1:750 (Pearce 1997). The Islands are encompassed by
a 50m isobar, despite the isobar extending less than half the distance offshore to the
immediate south of the Pelsaert Group and similarly to the north. The continental shelf slope
steepens drastically approximately 10km west of the Abrolhos, and this shelf break can be
seen in the relatively short distance over which water depth alters from 100m to 500m plus,
in Figure 2.1
2.1.1.2. Zeewijk Channel Bathymetry
This dissertation is particularly interested in the Zeewijk Channel; therefore its bathymetry
should be examined more closely. The channel is approximately 35m deep when entering
from the landward side but it is characterised by a drop in depth of around 20m close to its
western opening. This depth intrusion goes far enough to be in line with Gun Island to the
south and Wooded Island to the north, and is inside the fringing reef line. There is a
Pleistocene ridge nearly 12km offshore from the step feature, characterised by a narrow
opening or break approximately 1.5km in width on its northern extremity.
The channel length can be estimate as 30km from the ridge feature through to Snapper Bank,
shown as a mound at the top of the bathymetric map. The width of the western opening is
approximately 6km, although channel width increases significantly moving eastwards
through the channel. The along-channel path is orientated from south-west to north-east, in
other words it is tilted approximately 42° from an east/west orientation.
Background 7
(m) Figure 2.2: focus on the Zeewijk Channel bathymetry looking in the along-channel direction. Half moon reef is represented by the dark red section on the middle right of the figure and the Easter Island reef system is on the left. The image was created using the digitisation method described in the methodology.
Background 8
2.1.1.3. Geomorphology
Early surveys conducted on the Houtman Abrolhos Islands found the islands to be built of
Pleistocene coralline limestone topped with aggregated coral rubble and sand (Wilson 1978).
The more central islands found in the three groups are platform islands that can be aged to the
last Interglacial period, whereas the windward and leeward reefs are Holocene in age(Collins
et al. 1996). The various reef structures in the Abrolhos include platform, fringing, shore
platform and lagoon patch reefs in both the windward and leeward directions but all with
varying degrees of coral growth (Wilson 1978). The Islands are atoll-like due to their high
abundance of corals with relatively low species diversity, and yet there geomorphology
classes them an archipelago (Sukumaran 1997).
2.1.2. Climate
The Abrolhos Islands share the meteorological and climatic characteristics of the nearby mid-
west coast of Australia. The region is subtropical with hot dry summers and mild to cool
winters (B.O.M. 2005). Its weather patterns rely heavily on the eastward progression of high
pressure cells from the Indian Ocean, and the North-South movement of the subtropical
anticyclonic wind-belt (CSIRO 2005). Typical synoptic weather patterns that influence the
Western Australian coastline in summer and winter are illustrated in Figure 2.3 and Figure
2.4.
Figure 2.3: typical winter synoptic weather chart for Australia in 2003 (B.O.M. 2005).
Background 9
Figure 2.4: typical summer synoptic weather charts for Australia in 2003 (B.O.M. 2005).
2.1.2.1. Air Temperature
Monthly mean air temperatures at the Abrolhos vary from a low of 17.7°C in August to a
high of 23.5°C in February. The minimum temperature recorded between 1990 and 1995 was
10.7°C and the maximum temperature recorded was 37.7°C. Both the monthly means and
minimum and maximum temperatures in the Abrolhos are ameliorated by the ocean when
compared to the inshore area at Geraldton (Fisheries 2000). Consider how the difference in
monthly mean air temperature differs by more than 10°C at Geraldton and yet there is only a
difference of 6°C at the islands. Air temperature is appreciably warmer at the islands during
winter months compared to Geraldton (Pearce 1997).
2.1.2.2. Rainfall
The Abrolhos Islands archipelago has relatively low rainfall, with an average of 89 rain days
per year producing mean annual precipitation of 461mm. Rainfall is largely seasonal, and
nearly 100mm of the annual precipitation occurs in June. This seasonality is further
emphasised since 86% of the regions entire rainfall occurs in the six months between April
and September (Pearce 1997).
Background 10
2.1.2.3. Wind
Localised wind conditions are an important source of energy for water circulation and
mixing. The Houtman Abrolhos Islands wind pattern can change considerably over diurnal
and seasonal timescales. In the summer months, the stronger winds result from a combination
of the synoptic situation and strong sea breezes (Pattiaratchi 1993). In May, a weaker, more
variable winter wind pattern is re-established (Pearce 1997). The prevailing winds in summer
are moderately strong south-east to south-westerlies that can reach velocities in excess of
6ms-1 for 75% of their duration, with only short periods of calm (Wells 1997). In winter
months winds are more variable from the south-west to north-east, however storm events are
more frequent and often of a higher intensity producing the highest velocity winds
experienced by the region.
Three types of storms occur in the Abrolhos. Firstly the rare and infrequent tropical cyclone,
they usually reach this far south once in three years but they are potentially very destructive
with wind speeds over 30ms-1. Secondly there are summer squalls, usually between
December and April with wind speeds between 25 and 30 m/s that can occur from any
direction. Finally winter gales resulting from the passage of east bound storms south of the
Abrolhos may cause wind gusts and speeds of up to 35m/s (Sukumaran 1997).
The diurnal heating of land and water creates what is known as a sea breeze affect, where
cool seaward air flows onshore to replace hot dry air that has been heated over land during
the day and risen (Pattiaratchi & Imberger 1991). The Western Australian coastline
experiences a sea breeze, locally termed the ‘Fremantle Doctor’, an order of magnitude
stronger than that found at the majority of locations in the world whom experience the same
phenomena (Pattiaratchi 1993). The meteorological station on North Island in the Houtman
Abrolhos documents wind data for the region and has been used to collect the relevant data
for this report.
Background 11
2.2. Biological Setting
Western Australia is home to the only reef system to thrive in an eastern boundary region on
the globe. The most notable component of this reef system is the Ningaloo Reef in the states
north, but it is considered to stretch as far south as Cape Leeuwin. The system south of North
West Cape is recognized as having great significance as a zone of bio-geographic overlap,
with the oceanographic properties of the continental shelf a driving factor.
2.2.1. Biodiversity
The Houtman Abrolhos is home to a variety of biota whose species have migrated from
tropical and temperate regions of the Western Australian coastline, as well as a small number
of endemic species of flora and fauna. Research has shown that tropical species, corals and
fish in particular, dominate the area (Wilson 1978). It has been speculated that the abundance
of tropical fish is a result of the pole ward flowing Leeuwin current bringing water from the
north (Pearce 1997). The occurrence of temperate species is believed to be a product of the
Capes current and its progression northwards along the inner-continental shelf (Pearce &
Pattiaratchi 1999). These species most notably consist of macro algae and seagrasses
(Brearley 1997).
2.2.2. Seabird population
The Houtman Abrolhos Islands is an important nesting and breeding ground for a variety of
seabirds. It supports the largest breeding colonies in Western Australia of a number of bird
species including the White-faced Storm Petrel (Pelagodroma marina) and the White-
breasted Sea Eagle (Haliaeetus leucogaster) (A.I.T.F. 1988). It is also home to the rare Red-
tailed Tropic Bird (Paethon Rubricauda) that is close to extinction and has endemic species
such as the Painted Button-quail (Turnix varia) in residence (A.I.T.F. 1988).
2.2.3. Fisheries
Commercial fishing is the primary industry in the Abrolhos Islands. The three major
commercial fisheries operating at the Abrolhos include the Western Rock Lobster, Southern
Saucer Scallop and a selection of finfish. The finfish catch will not be addressed in detail as it
has less bearing on the Zeewijk Channel, but it targets species such as pink snapper (Pagras
auratus) and coral trout (Plectropomus leopardus) using hook and line (Fisheries 2000).
Background 12
2.2.3.1. Western Rock Lobster
One of the most important resident species is the Western Rock Lobster, Panulirus Cygnus.
In the last decade it has become Australia’s most valuable single species fishery, with a
seasonal gross value of production between $300 and $350 million, and around 15% of the
seasonal catch is produced in the Houtman Abrolhos region despite a shortened season
(Chubb & Barker 2005). The Abrolhos Islands fishery is restricted to the period 15 March to
30 June, unlike the rest of the coast that is fished from as early as November through to 30
June. The impact of the reef system and its circulation patterns on the lobster industry is not
easy to quantify but studies have established its importance as a nursery for young lobster or
puerulus (Phillips et al. 1991). In fact the Abrolhos’ importance to the Western Rock Lobster
industry goes beyond its catch strength, as 50 to 80% of the total larvae for the state may be
produced here (Chubb & Barker 2005).
The ability of the rock lobsters to move off and onshore is affected by localised sea
conditions and the oceanographic properties of the region. After hatching in summer near the
edge of the continental shelf, millions of larvae are transported to the order of hundreds of
kilometres from the West Australian coastline, via surface currents including the Leeuwin
and Capes Currents. They float free in the Indian Ocean for several months but eventually
return onshore.
Artificial seaweed is monitored monthly on the full moon so that the puerulus content of the
seaweed can be measured. The annual indices are then used to predict the strength of the
early season catch in 4 years time (CSIRO 2005). The catch consists of newly-moulted
lobsters or ‘whites’ heading off-shore to breed but the index can alternatively predict the late
season catch for three years time of mature coastal lobsters, or ‘reds’ that have resumed
residence on the reef system (Caputi 1995).
Studies have shown that annual fluctuations in the strength of the Leeuwin Current, and
subsequently the ENSO or El Nino effect, directly influence the yearly catch of the Western
Rock Lobster which means that the interaction between shelf break and Abrolhos water
bodies takes on additional, in particular financial, importance (Caputi 1995).
Background 13
2.2.3.2. Southern Saucer Scallop
The most prominent commercial species of scallop in Western Australia the Southern saucer
scallop, Amusium balloti, is found in harvestable quantities in the benthic habitats of the
Middle and Zeewijk Channels. This fishery was declared limited-entry at the Abrolhos in
1986 and 16 licensed vessels currently fish it (Penn (ed)1999). The scallop season runs
between April and June, therefore it coincides with the Abrolhos Island Rock Lobster
Fishery, and the aim is to take the maximum yield available as the scallops have finished
spawning by the time the season opens (Penn (ed)1999).
In 2003, a particularly good year for the fishery with high scallop abundance, there were 9
124 standardised trawl hours recorded during an extended season for an estimated catch value
of $19.6 million (Chubb & Barker 2005). This contrasts with the 2002 season when only 912
trawl hours were recorded, but highlights that a significant amount of bottom trawling can
occur annually and that it is an important factor when determining management of the
channels. While the relationship is not completely understood, the Leeuwin Current is
believed to affect scallop abundance and therefore catch sizes in a similar manner to that in
which it affects the Western Rock Lobster (Chubb & Barker 2005).
2.2.3.3. Aquaculture Proposal
The Zeewijk Channel has been speculated upon as a possible aquaculture site in an attempt to
further diversify fisheries in the Abrolhos region. The Aquaculture Plan for the Houtman
Abrolhos Islands (Fisheries 2000), outlines the relevant concerns regarding all types of likely
aquaculture ventures from ocean-ranching, for example stocking and harvesting scallops on
defined areas of the sea-floor, to the tuna farming now prevalent in South Australia.
A proposal has gone as far as approval by the Environmental Protection Authority, that
entails capturing via purse seine yellow fin tuna, Thunnus albacares, in waters between
Geraldton and Exmouth before transporting and culturing them in sea-cages in the Zeewijk
Channel (Diver & Prince 2003). This process would be closely scrutinised by the Australian
Fisheries Management Authority. Details of the proposal can be found in the relevant EPA
report (2003) or in the work done by Diver and Prince for Latitude Fisheries Pty Ltd.
Background 14
An initial biomass of up to 200 tonne of tuna is speculated to be caught and held in eight 40
metre diameter sea-cages for up to 7 months whilst supplementary feed increases their
biomass. A 30 hectare area would be chosen to house these cages within the greater licensed
area illustrated in Figure 2.5. The cages that will be used to hold the tuna are similar to those
used in the South Australian blue fin tuna industry.
Key environmental issues raised both in the EPA report and after public consultation include:
• the potential for impact on the Abrolhos Island bird populations
• the potential for impact on benthic habitat and water quality
• the disease risks associated with using bait fish and feeding methods
Figure 2.5: map of the sea-cage site in the Zeewijk channel (Diver & Prince 2003)
The circulation pattern of the Zeewijk Channel will directly affect two of these concerns, and
indirectly affect the third. Disease risks, water quality issues and concerns for the benthic
habitat can be assuage by confirmation that the Zeewijk Channel has sufficient water
exchange with the areas beyond its confines. Conversely, analysis indicating minimal water
exchange with outside sources or poor circulation would establish sound reasoning for
reassessment of the aquaculture proposal.
Background 15
2.3. Oceanographic Setting
This chapter outlines the dominant forces and physical processes affecting the Zeewijk
Channel, and the greater Houtman Abrolhos Islands. It begins with an in depth review of the
Leeuwin Current and associated features, due to this currents significance in terms of the
regional oceanography.
2.3.1. The Leeuwin Current
The Leeuwin Current’s biological and physical oceanographic influence on Western Australia
stretches from the North West Cape to as far south as the Great Australian Bight. It is an
anomalous eastern boundary current as it flow’s pole ward, rather than equator-ward like
other eastern boundary currents, and it travels against the prevailing south westerly winds
with a core velocity of up to 1ms-1 (Cresswell et al. 1989). The current originates from the
tropics and in particular the Pacific Ocean, where water is pushed westward by the south-east
trade winds and channelled via the Indonesian through-flow into the Eastern Indian Ocean.
The Coriolis force and geostrophic factors push the current onto the Western Australian
continental shelf and force it pole ward. During its passage down the shelf it is typically
50km wide and also relatively shallow, at approximately 250m deep (Smith et al. 1991).
There is a northwards moving counter current, the Leeuwin Under-current, recognised but not
investigated in detail, flowing below this 250m deep surface layer (Pattiaratchi 2005).
It is generally considered that the pole ward movement of the current is driven by a steric
height gradient that runs south from the North West Shelf as a result of temperature contrasts
from north to south (Smith et al. 1991). The steric height gradient causes geostrophic flow of
water eastwards in the Indian Ocean which is then blocked by the Western Australian coast
and causes the pole ward momentum. A relatively deep mixed layer at the surface of up to
50m, means that the momentum induced by equator ward wind is distributed over a relatively
deep layer, and hence the associated wind stress is too weak to overcome the steric height
forcing and the Leeuwin Current travels south (Smith et al. 1991).
Background 16
Research has found the Leeuwin Current to be stronger in winter then in summer months and
this seasonal variability is coupled with a strong inter-annual variability linked to the
southern oscillation index. El Niño events pre-empt a weakened Leeuwin Current and La
Niña events enhance it (CSIRO 2005). It is characterised by warmer water of low salinity and
depleted nutrient levels, in comparison to the more temperate waters surrounding it (Pearce
1997). Studies of the currents seasonal variability indicate it has volume transport of 5-7 Sv
in winter (Smith et al. 1991) and 1.4 Sv in summer (Pearce & Griffiths 1991). This seasonal
change in intensity has been attributed to the wind stress variability of localised regions.
Previous studies, including (Holloway 1995), indicate that there are systems of eddies,
meanders and alongshore jets associated with the Leeuwin Current as well as sections of a
single current core and these features can often be deduced from satellite imagery
(Pattiaratchi & Buchan 1991). Its impact on up welling and down welling of water along the
coastline is debatable due to the significant biological productivity of the regions it traverses
which would typically be attributed to high nutrient levels caused by localised up welling in
other eastern boundary locations. This biological productivity is not considered to be a direct
response to the low nutrient levels of the Leeuwin Current, but rather a by-product requiring
further investigation but possibly linked to the Capes Current (Cresswell et al. 1989).
The extent of the Leeuwin Currents intrusion into shallower shelf waters is an ongoing
research topic and is relevant in this dissertation with regards to its interaction with the waters
past the 50m depth contour surrounding the Abrolhos Islands.
2.3.1.1. The Capes Current
The dynamics of the Capes Current have been described by (Gersbach 1999). Essentially the
southerly wind stress overcomes the alongshore pressure gradient forcing surface water to
flow equator ward inside the 50m contour of the continental shelf (Pearce & Pattiaratchi
1999). This surface layer move tends to move offshore due to the Coriolis force, which in
turn results in up welling of colder bottom water onto the continental shelf. The Leeuwin
Current interacts with the Capes Current through edification and mixing and will often
migrate further offshore under the Capes Current’s influence (Gersbach 1999).
Background 17
(Pearce & Pattiaratchi 1999) demonstrated the seasonality of the Capes Current using satellite
images covering the period 1987 to 1994. The cool northwards flow started in October and
decayed significantly by April the following year as a result of the seasonal strengthening of
the northwards wind stress. The Capes Current is to a large degree displaced by the relatively
strong pole ward flowing Leeuwin Current in winter months (Pearce & Pattiaratchi 1999).
Figure 2.6: The progression of cooler, Capes Current water northwards can be seen in this satellite image. Also of note are the systems of eddies off the continental shelf that are comprised of Leeuwin Current water that has entrained the Capes water from the shallower regions (Pearce & Pattiaratchi 1999).
Background 18
2.3.2. El Niño/Southern Oscillation
One of the most important factors affecting Australia’s climate is the El Niño/Southern
Oscillation (ENSO) effect. This phenomenon, derived from coupled processes in the ocean
and atmosphere, occurs predominantly in the equatorial Pacific region although its influence
can be observed over most of the globe (CSIRO 2005). It has a largely irregular period of
between 3 and 7 years, although there are important longer period fluctuations as well. A
commonly accepted measure of the strength of ENSO is the Southern Oscillation Index
(SOI), which is the normalized sea-level pressure difference between Darwin and Tahiti
(B.O.M. 2005).
The SOI shown graphically in Figure 2.7, shows we were coming out of a strong El Niño
event from 2002 to mid-2003 and that the SOI was nearly negative 3 in October of 2003
when the CTD field testing occurred. A weighted average of approximately 0 SOI could be
calculated for the 3 month period in which ADCP testing took place in the Zeewijk Channel.
Due to time lag effects, the relatively large El Niño event in the year and a half leading up to
the field work is likely to be of greater importance than the SOI during the testing period
(B.O.M. 2005).
Figure 2.7: Southern Oscillation Index for the beginning of 2000 until mid-2005. Negative SOI for extended periods will mean El Niño conditions, positive means La Niña (B.O.M. 2005).
Background 19
2.3.3. Physical Parameters
Studies of water column properties such as temperature and salinity at the Abrolhos in the
past have primarily been concerned with water near the continental shelf, and its exchange
with landward sources. These studies have been endorsed by Fisheries Western Australia, to
fill knowledge gaps concerning the Western Rock Lobster’s puerulus stage and to further
scientific knowledge associated with continental shelf features such as the Leeuwin Current
and/or its interaction with coastal water such as the Capes Current as outlined earlier.
2.3.3.1. Sea Temperature
According to research by the (CSIRO 2005), there is a seasonally-reversing temperature
gradient across the Western Australian continental shelf. In summer, shallow near-coastal
waters warm because of heat input from the sun and the atmosphere, hence water temperature
fall slightly with increasing distance offshore. Conversely in winter, coastal waters cool
rapidly because of heat loss to the atmosphere, and at the same time the Leeuwin Current is
maintaining warm conditions offshore, so there can be a large increase in surface temperature
(up to 4°C) between the coast and the edge of the continental shelf. The Abrolhos is
significant as its relatively shallow bathymetry, and close proximity to the continental shelf
break, means it is effectively fronting this temperature gradient.
Sea surface temperatures have been recorded at the Fisheries Western Australia puerulus
collection site in the Abrolhos Islands since 1970, though with gaps in the period of testing.
The site is about 2 km northwest of Rat Island (28° 45’S) and the Easter Group which makes
it nearly 80km from the WA shoreline and hence near the edge of the continental shelf. The
site was chosen to determine the influence of the southward-flowing Leeuwin Current and
shows the seasonality of sea surface temperatures. Summer sea temperatures can be as high
as 26°C between February and April and the lowest values recorded where approximately
18°C. Sea temperatures below 18°C for short periods of time were found to occur during
instances of prolonged easterly winds (Wilson 1978).
Background 20
Figure 2.8: Sea surface temperatures as measured by mercury thermometer on monthly visits to the puerulus collectors; estimated accuracy 0.3°C. The period of data analysis is from 1970 to 1976 and then 1984 to 1995. Sampling is ongoing (CSIRO 2005).
2.3.3.2. Salinity
(Pearce 1997) determined two main influences on surface and sub-surface salinity across the
continental shelf region near the Abrolhos. High salinity coastal waters pushing westwards
and the relatively low salinity Leeuwin Current water transported from the north. The extent
of their interaction depends on the relative strengths of the currents as prescribed by their
seasonal and inter-annual variability (Pearce 1997).
The CSIRO source that provided sea temperature data near Rat Island was not available for
salinity, therefore monthly mean surface salinities from the Geraldton hydrographic station
up to 1997 can be found in Sukumaran (1997). They illustrate that higher mean salinity is
experienced by the region in summer, with little difference through the 40m water column.
Salinity can decrease by over 0.5ppt into the winter period, and through this drop the surface
and near-bottom salinity levels follow each other closely. Salinity steadily increases through
spring and into summer as the prevailing southerly winds increase in intensity. Sukumaran
(1997) speculates that the resumption of the northerly wind stress in summer forces the
relatively low salinity water back offshore with Leeuwin Current water, and that relatively
high salinity water replaces via up welling and northwards transport.
Background 21
Salinities at Geraldton Hydrographic Station
35.3
35.5
35.7
35.9
36.1
1 2 3 4 5 6 7 8 9 10 11 12
Month (Jan-Dec)
Salin
ity (p
pt)
Surface
Depth-40m
Figure 2.9: Monthly mean salinities from the CSIRO hydrographic station near Geraldton. Salinity at the surface and at a depth of 40m is illustrated (adapted from Sukumaran 1997).
2.3.3.3. Stratification and mixing
Stratification in a water column is effectively the input of buoyancy to a system as a result of
driving mechanisms such as radiative heating, evaporation and/or freshwater inputs
(Pattiaratchi 2005). De-stratification requires the input of energy. In the case of the Zeewijk
Channel this energy is contributed by wave motion and tides, however the dominant input is
wind.
Figure 2.10: the processes driving stratification and mixing of a water column.
Background 22
The stronger the pycnocline or change in density through the water column, the greater the
amount of energy required to breach the layer and in turn allow a significant degree of mixing
to occur. The source of this energy is turbulence which is generated both at the sea-surface
via wind stress and mesoscale current fields, and at the seabed due to currents that are often
tidally driven. Pearce (1997) found thermal stratification to be limited at the Abrolhos’ ‘Rat
Island’ testing site and deduced that this property was likely for the majority of the Abrolhos
region.
The waters are currently considered well-mixed, with relatively high temperature and salinity
throughout the Abrolhos’ water columns in summer (Fisheries 2000). Evaporation and
subsequently evaporative cooling effects, can contribute to the mixing process and these
temperature and salinity conditions (Pattiaratchi 2005). The lower temperatures and salinity
presiding in winter can be attributed to higher levels of precipitation during the winter period.
The easing and increased variability of the prevailing southerly winds can also be a factor
towards decreased salinity levels in winter, as the Leeuwin Current can more easily flood the
shelf waters under this less debilitating wind regime (Pearce 1997)
Background 23
2.3.4. Hydrodynamic Processes
This section will outline the dynamic processes occurring at the surface, and through the
water column, likely to have contributed to the aforementioned water properties. They are
particular important when considered as the driving mechanisms of the Zeewijk Channel
circulation pattern.
2.3.4.1. Baroclinic Circulation
Horizontal gradients in density are an important mechanism in driving ocean currents.
Variations in density are directly related to variations in temperature, salinity and pressure
(Mellor 1996). In situ sea water density can be calculated from the known values of these
parameters using mathematical programs such as MATLAB’s Sea Water toolbox. A density
difference of as little as 0.01kgm-3 between two locations can have important dynamic
consequences (Pattiaratchi 2005). The velocity, U, of a density driven surface current
between two points can be estimated using the following equation,
⎟⎠⎞
⎜⎝⎛
∂∂
=xK
hgUzm
ρρ
42
3201
Where, h is the average water depth between the two positions
ρm is the mean density
Kz = 0.01m2s-1
g is the acceleration due to gravity = 9.81ms-2
and x∂∂ρ is the difference in density over the distance between the two positions.
Horizontal density gradients contribute to vertical variations in horizontal pressure gradients.
Consequently, horizontal currents may also vary in the vertical and the result is baroclinic
flows. For example, a mass of low salinity water will be induced to flow towards a region of
high salinity water that is effectively sinking relative to the less dense water. After the two
masses have entered each others density fields then the resultant salinity of the system will
eventually be some median value between the original salinities due to circulation and
mixing. A subsequent result of baroclinic forcing is therefore a change in the density field of
the system (Pattiaratchi 2005).
Background 24
On the relatively shallow landward side of the Abrolhos Islands diurnal variations in heating
and cooling, and alternatively freshwater inputs, may also drive thermohaline circulation
within continental shelf waters. (Fahrner & Pattiaratchi 1994) illustrated this type of current
formation in the shallow waters of Geographe Bay, south-west Australia. A density gradient
resulting from freshwater inputs at the near shore, for example from the Spalding River in the
regional context, could have a similar affect.
2.3.4.2. Barotropic forcing
Barotropic forces are external forces such as wind stress, tides, atmospheric pressure
variations or the Coriolis force driving the circulation of a water body (Mellor 1996). The
main barotropic forces acting on the Zeewijk Channel would be the wind stress and wind-
related waves, Coriolis and tides. The movement of water due to geostrophic anomalies over
the continental shelf slope may also be a significant mode of barotropic forcing, depending
on how near to the Abrolhos they occur.
Barotropic forcings are characterised by little or no change in the density field of the water
body after it has been transported (Pattiaratchi 2005). Put simply there is no change in the
waters temperature or salinity after the forcing. This is one reason why cold water that has
been entrained off-shore by eddy formations can often still be seen inside the eddy as a
‘spiral’ via satellite imagery.
2.3.4.3. Wind-driven circulation
Wind blowing over the sea surface creates friction between the moving air mass and the sea
surface that induces currents (Pond & Pickard 1983). The speed and direction of wind
induced currents are controlled by factors such as the momentum of the wind, inertia of the
water surface and water column, pressure gradients due to wind induced motion, the Coriolis
force and bottom friction. The wind stress τ, when measured as the horizontal force per unit
area, is given for the sea surface by, 2
10UC AD ρτ =
Background 25
Where, CD is the drag coefficient (dimensionless)
U is the wind speed 10m above sea level
and ρA is the density of air.
Water at the surface of the ocean is driven at approximately 3% of the wind speed at 10m
(Fahrner & Pattiaratchi 1994). Below the surface, current speeds decrease with depth as the
stress is transferred from layer to layer (Pattiaratchi & Imberger 1991). (Pugh 1987),
neglected the affect of the Coriolis force and assumed that the water column had constant
density to give a velocity profile through the water column of,
⎟⎟⎠
⎞⎜⎜⎝
⎛−=
0
*0 ln
zzu
UU Z κ
Where, Κ is the von Karman co-efficient (~0.4)
U0 is the surface current speed (~3% of U0)
u* is the wind friction velocity = (τ/ρA)0.5
and z0 is the roughness coefficient, usually between 0.0005m and 0.0015m.
(Csanady 1973), described how this forcing induces currents in lakes. In shallow water the
direction of the wind induces a current in the same direction at the lakes surface. From the
above equation we know that this will decrease with depth, but in fact it may decrease to the
degree where the current reverses, relative to the surface direction, at the bottom. For
shallower water the dominant force balance is between surface wind stress and bottom
friction. For deeper water, current may flow in opposition to the wind, proportional to depth,
as the force balance is between the pressure gradient force and the bottom friction. So if we
have a basin whose bathymetry consists of shallow to deep water, we would generate a
topographic gyre.
(Fischer et al. 1979), elaborated on this further. An assumption is made that the wind is
inducing a uniform stress everywhere on the basin of water’s surface. The line of action of
the wind stress will be through the centroid of the water surface. The centre of mass of water
in the basin is towards the deeper end, since there is obviously a greater mass on that side. A
torque is induced since the line of action of the wind stress is to the shallow side of the centre
of mass, and the water mass will rotate in the windward direction at the shallow region and
against the wind at the deep region (Hunter & Hearn 1987).
Background 26
2.3.4.4. Coriolis and Rossby Number
The Coriolis force is an inertial oscillation created by the Earth’s rotation that forces currents
to the left of their original momentum in the Southern Hemisphere (Pugh 1987). The Coriolis
parameter is defined as,
φsin2Ω=f
Where, Ω is the Earth’s angular velocity, srad /10272.7 5−×=Ω
and φ is the latitude of the water body being affected.
The Zeewijk Channel is between 28.5° and 29°S, therefore the Coriolis parameter of the
channel at mean latitude of 28.75°S is approximately .10995.6 15 −−× s
The Rossby Number will indicate how significantly the Coriolis force will affect the
barotropic flow of a given water body at its designated latitude. As described in (Fischer et
al. 1979), the Rossby Number is the ratio of the period of rotation to the time of advection. It
can be calculated from the formula,
fLUR =
where, U is the characteristic velocity scale
L is the characteristic length scale of the water body
and f is the Coriolis parameter.
If the Rossby Number is significantly larger than one, then the fluids momentum would be
sufficient to most likely overcome the effects of the Coriolis force. A Rossby Number less
than one would mean that the water body could be expected to exhibit some degree of
rotational transport (Fischer et al. 1979), anticlockwise in the Zeewijk Channel’s case.
Calculating the Rossby Number for the Zeewijk Channel will determine if the Coriolis force
is an important factor in its circulation pattern.
Background 27
Assume the characteristic velocity scale, U, to be the average of the absolute values of the
along-channel currents recorded by the ADCP in Spring 2003, giving a value of 0.32ms-1.
The characteristic length scale of the channel has been determined from the latitudes and
longitudes of stations 7 and 9 on the Zeewijk Channel transect using the method of (Kirvan
1997). This calculation gave a length scale estimate of 26.7km.
An approximate value of the Rossby Number for this system will therefore be 0.17, well
below one and an indication that the Coriolis force may influence the circulation pattern. It is
important to note however, that the physical barriers that impede currents to the north and
south of the along-channel current may also negate this influence, and restrict rotational flow
physically.
2.3.4.5. Ekman Veering
Ekman veering is the next progression in the process of a barotropic surface current being
pushed to the left by the Coriolis force in the southern hemisphere. As the currents influence
extends deeper down the water column, the lower layers are deflected slightly more to the left
than the ones above and a distinct spiralling pattern develops (Pattiaratchi 2005). This
‘Ekman Spiral’ affect will usually occur in deeper offshore waters well away from the coast
and where the bottom friction will not affect the direction of the currents. In the southern
hemisphere the theoretical deflection of surface currents is 45° to the left of the direction of
the barotropic forcing, for example the wind stress, and this deflection increases with depth
but simultaneously reduces in terms of current strength (Pond & Pickard 1983).
Figure 2.11: Ekman transport in the Northern Hemisphere. The wind-induced surface current turns 45° to the direction of the wind (Pattiaratchi 2005).
Background 28
The depth of the Ekman layer ,DE, where water is moving in the opposite direction to the
original surface current due to the progressive rotation down the water column, can be
calculated with (Pugh 1987),
fAD Z
E ρπ
2=
where, f is the Coriolis parameter
ρ is the density of sea water, 31025 −= kgmρ
and AZ is the vertical eddy coefficient, assume 1140 −−= skgmAZ
The depth of the potential Ekman layer under the conditions stated for the Zeewijk Channel is
104.9m. Since the channel ranges from 30 to 60m in depth, a full Ekman spiral will not occur
due to the influence of the Coriolis force. This does not mean that Ekman transport will not
be significant through the water column, just that total reversal of the current direction from
surface to bottom is unlikely and it will be impeded also by bottom friction. Trends of
opposing current directions at the surface and bottom will need to be considered in terms of
other forcing mechanisms.
2.3.4.6. Atmospheric Pressure Changes
The inverse barometric effect occurs where the difference in atmospheric pressure between
two points has driven a change in water level also between those points (Pond & Pickard
1983). Static sea level response to air pressure dictates that a decrease in air pressure of 1hPa
will translate to a rise in sea level of approximately 1cm. This can be described
mathematically by,
atmPΔ−=Δ 993.0η
where ηΔ is the sea level change in centimetres, and atmPΔ the pressure change in hecta-
pascals.
Background 29
Pressure systems are not stationary in nature however, and since they will typically be
moving at some speed CA, the dynamic sea level response can be calculated using the static
response to the system, gravity ( g ) and water column depth ( h ),
⎟⎟⎠
⎞⎜⎜⎝
⎛−
=
ghC
responsesealevelstaticresponsesealeveldynamicA
2
1
Sea level changes due to meteorological influences such as atmospheric pressure and wind
stress should be considered in addition to localised tidal and wave effects. Sea surface
velocities produced by these meteorological conditions are usually relatively small (Mellor
1996), but it is important to remember that they can have devastating effects in terms of
storm surge levels and coastal flooding, especially on low lying regions relative to mean sea
level such as the Abrolhos Islands.
2.3.4.7. Continental Shelf Waves
Continental shelf waves are the result of storm events passing over a coastal boundary and
forcing a ‘set-up’ of coastal water due to variations in atmospheric pressure and the action of
associated wind stress (Gill & Schuman 1974). This stored energy will then propagate as
waves in the alongshore direction on the continental shelf (Robinson 1964). During the
Western Australian summer, these long period waves usually originate from the northwest
due to the incidence of cyclones in this season. Continental shelf waves have relatively large
periods of 10 to 20 days and can travel at up to 6ms-1 for distances of 1000 to 2000km
alongshore (Chua 2002). They have a time lag down the coast, and would be expected to
attenuate considerably away from the associated weather system (Gill & Schuman 1974).
Continental shelf waves have the potential to influence the surrounding water levels of the
Abrolhos and subsequently the passage of water through the Zeewijk Channel. If this
influence is occurring with some significance then it can be identified in spectral analysis of
the circulation pattern.
Background 30
Continental Shelf Seiche
Seiches are standing rather than progressive waves and are important in closed and semi-
enclosed basins, bays, marginal seas and Gulfs. Seiches usually develop when a strong
prevailing wind blows over a basin, or when there is an imbalance in barometric pressure at
opposing ends of a basin (Ingmanson & Wallace 1985). A standing wave is characterised by
nodes and anti-nodes. Nodes are stationary points on the wave and anti-nodes are the
maximum and minimum heights of the water.
The physics of a standing wave involves the wave entering the basin, or conversely being
generated, reflecting from the basin end and returning in exactly the opposite formation that it
arrived. This process creates a situation where the water oscillates back and forth with nodes
of no movement positioned between the anti-nodes of maximum and minimum wave height.
Seiches are not known for causing the damage and erosion to shorelines that progressive
waves are recognised for. Rather their importance is related to enclosed bays and harbours,
where seiche periods close to that of the local swell and natural forcing processes can disrupt
moorings and harbour functionality (Drake et al. 1978).
A seiche exists in the semi-enclosed basin formed by the coastline near Geraldton (the closed
end), and approximately the 100m contour line of the continental shelf. The period of this
seiche can be calculated using the following equation,
ghLT 4
=
where, T is the tidal period in s
L is the approximate length of the basin, L = 70km
g is the acceleration due to gravity, g = 9.81 ms-2
h is the average depth of the water column, h = 35m
The period of the continental shelf seiche near Geraldton is approximately 4.2 hours using the
assumed values. Its significance in terms of the circulation pattern of the Zeewijk Channel
will be determined through spectral analysis.
Background 31
2.3.4.8. Waves
Waves possess kinetic energy in the form of the orbital motion of the particles it displaces.
They have potential energy through the displacement of water above sea level (Ingmanson &
Wallace 1985). Wind is the major cause of waves although submarine earthquakes,
submarine landslides, submarine volcanic eruptions, landslides into the sea, ships and tidal
forces are also causes of waves. Wave period is the time for one wave to pass a specific point
(wave frequency is the inverse of this), wave amplitude is the height of the wave above or
below sea level and wavelength is the distance between equal points on adjacent waves
(Mellor 1996).
Wave conditions depend on a number of factors including fetch length, that is the area over
which the wind blows, the duration of the wave generating wind, wind speed, bathymetry and
the distance from the area of fetch. Wave velocities increase with increasing duration, fetch
length and wind speed however they decrease as the waves distance from the fetch area
increases ((Pattiaratchi 2005). Sea waves are choppy waves with short periods that form in
the vicinity of storms or due to localised wind affects. Swells are waves that are present on
calm days, away from the winds that generated them and these have longer periods and a
smoother appearance.
In deep water, swells can travel thousands of kilometres away from a storm system without
imparting significant energy, moving more rapidly than waves with shorter wavelengths.
Waves can further be classified as ‘shallow-water’ waves and ‘deep-water’ waves according
to the relation of their wavelength to water depth.
Wave heights in the open ocean near the Abrolhos Islands average approximately 2m.
However 10% of wave heights exceed 4m and these can be even greater during storms
(Fisheries 2000). In fact the Abrolhos has a reputation among sea farers in the region for its
large brutal seas during the winter months. The greatest wave energy is experienced on the
southwest facing reef margins and wave heights are substantially lower in the island groups
of the archipelago as a result of significant dampening by the shallow reefs and islands.
Refracted swell and wind waves will still enter the Zeewijk channel and a primarily westward
swell may pass relatively unimpeded along its course.
Background 32
2.3.4.9. Currents
The circulation features of currents near the continental shelf were tested in (Cresswell et al.
1989). These studies used current meters to record the strength and direction of surface
currents and then a comparison was made of the results with atmospheric and coastal tide
gauge data. They found that on a seasonal scale, the time frame was March to August and
therefore this was a winter assessment, the continental shelf currents near the Abrolhos had
directional persistence or means of ten days and that they were predominantly southward at
approximately 0.2ms-1. In instances when this prevailing direction did not occur, the current
had a heavy northerly component at 0.1ms-1. Alongshore currents had variability of up to
0.5ms-1, and had strong correlation to local sea level changes as well as the alongshore wind
pattern recorded at Geraldton after a phase lag was incorporated.
Currents are also generated by large scale wind fields that are associated with the passage of
weather systems and changes in mean sea levels. Southerly wind events in summer produce
sea level troughs at latitudes close to the Abrolhos and generate northward current pulses
along the Western Australian coastline (Pattiaratchi 2005). Conversely frontal systems, and
associated low pressure passing over Cape Leeuwin, produce sea level crests to the north and
southerly currents are pulsed down the coast (Pattiaratchi 2005).
Analysis of cross-shore currents near the Abrolhos Island’s have indicated that they occur for
approximately 33% of the year, with magnitudes less than 0.2 ms-1 (Pearce 1997). Cross-
shore currents take on particular importance when considering larvae or puerulus transport on
and off the continental shelf. They can also be analysed when considering pollutant fluxes,
for example nutrient exchange in the context of aquaculture cages within the Zeewijk
Channel. In this instance cross-shore currents must be responsible for channel flushing due to
the physical barriers, reefs and islands, in the alongshore directions. The circulation pattern of
the Houtman Abrolhos is considered responsible, to a large degree, for the triangular
evolution of the Pelsaert and Easter Group Islands (Wells 1997).
Background 33
Terminology regarding current direction must be clarified to avoid confusion during analysis
of results. Currents heading towards the south are generally referred to as southerly currents.
This is a direct contradiction to wind direction terminology but if we consider that a northerly
wind creates a southerly wind stress at the surface, then the natural progression will be a
current of heading south and excluding other oceanographic factors.
Figure 2.12: wind and current directions according to the recognised naming conventions.
In this project it was more appropriate to describe currents as being cross-channel and along-
channel, rather than along-shore and cross-shore respectively. Cross-channel currents refer to
currents with an approximately north-south orientation in the channel (to be defined more
clearly in the methodology), which would translate to alongshore currents on the nearby
Western Australian coastline. Along-channel currents effectively move ‘through’ the channel
and would be termed cross-shore if incident on the nearby coast.
2.3.4.10. Shelf Currents
We have earlier recognised the presence and influence of the Leeuwin Current on the
continental shelf waters of Western Australia. It may be helpful to understand the typical
shelf currents found at other eastern boundary regions around the globe. Currents such as the
California, Benguela and Canary Currents follow the dominant wind pattern equator-ward
and are a result of subtropical gyre systems in their respective oceans (Lass & Mohrholz
2005).
NortherlyWind
Southerly Current
Background 34
The mean longshore current for all three of these eastern boundary currents is characterized
by a two layer structure. The wind-driven surface current is observed in the upper 30m of the
water column. It is subject to variability but is generally equator-ward with the prevailing
wind, and the pole-ward current is relatively constant below 30m in the water column. The
measured mean velocity of the Benguela under-current is approximately 0.04ms-1 according
to (Lass & Mohrholz 2005). This highlights the anomalous nature of the current system found
along the central Western Australian coastline.
2.3.4.11. Tides
Tides are periodic movements of the ocean which are directly related in amplitude and phase
to some periodic geophysical force. Gravitational forcing, due to the simultaneously
occurring moon-earth revolution, earth-sun revolution and declination effect of the moon, is
the dominant influence on tidal regimes. Meteorological forces can also contribute to smaller
tidal movements within each solar day. The periodic nature of tides means that variations can
be represented by a number, ‘n’, of harmonic terms in the following equation, (Pattiaratchi
2005)
)cos()( nnnp gtHtu −= ω
where up is the surface level, Hn is an amplitude, gn is a phase lag from the equilibrium tide at
Greenwich and ωn is an angular speed (Pugh 1987). Consequently with Tn as the tidal period
the angular speed is given by,
nn T
πω 2=
Background 35
The nature of a tide at any location in the world can be determined using the form factor
equation and the dominant tidal constituent’s specific to that location. In the Abrolhos Islands
the dominant diurnal constituents are K1 and O1, and the dominant semi-diurnal constituents
are M2 and S2. Their respective values are shown in Table 2.1 with respect to the Pelsaert
Island Group as found in the Australian National Tide Tables (2002). The form factor, F,
equation is expressed as,
22
11
SMOKF
++
=
A form factor greater than 3 indicates that diurnal tides dominate the system whilst a value
less than 0.25 will mean semi-diurnal tides dominate. Between these two values the system is
mixed, however it is a sliding scale and the system may be mixed but predominantly diurnal
or semi-diurnal (Pattiaratchi 2005).
Using the tidal constituents for 2002, the calculated form factor is 2.3. The Abrolhos tidal
regime can therefore be classified as mixed, although the diurnal tidal components dominate.
This is expected since Pearce (1997) observed that the tidal range at the Abrolhos had a
strong correlation to that of Geraldton, where the tidal range is close to 0.6m and it has a form
factor of 2.4 (Australian National Tide Tables 2002).
Table 2.1: tidal constituents for the Pelsaert Island Group 4. Adapted from the Australian National Tide Tables (2002).
Pelsaert Island (28.97°S 113.97°E) LAT (m) Constituents M2 S2 K1 O1 0.59 H - Amplitude (m) 0.069 0.058 0.167 0.127 g - Phase (°) 284.1 297.5 300.8 288.7
Methodology 36
3. Methodology
The methods used to collect, source and analyse oceanographic data to describe the
circulation pattern of the Zeewijk Channel are outlined in this chapter.
3.1. Data Compilation
This thesis presented the opportunity to analyse past data collected by independent sources
for the purpose of understanding many facets of a regions properties and dynamics. This is in
contrast to a methodology that would implement field testing as a direct influence of the
formulation of a question or concept. Inherent problems emerge in the demonstrated
approach, in particular the lack of a means to rectify methodology that is not ideal for
answering the thesis question. Gaps in field reporting caused by equipment malfunction and
spatial or temporal necessities cannot therefore be re-assessed.
The assembled wind and current data included spring and Summer time periods for 2002 and
2003. Testing of water column properties by a CSIRO research vessel occurred in spring of
2003 and wave data was recorded in spring 2002. There were data shortages due to
equipment malfunction in the 2003 wind, and 2002 current, recordings.
Due to the importance of a full current data set when considering circulation patterns, and the
opportunity to relate this pattern with known water properties, the 2003 data has been
analysed. Deficiencies in the wind pattern for the identical time frame exist however. The
general faults in the 2002 data contributed to it not being used for comparison purposes,
although the 2002 wind and wave data can be useful as a description of the seasonal
characteristics and prevailing conditions.
Methodology 37
3.2. The Southern Surveyor
The CSIRO research vessel, Southern Surveyor, traversed the Western Australian coastline
from September to November of 2003 between latitudes 27° and 36° south. Its mission was
to test the oceanographic properties of this vast expanse of water and in particular to assess
the interaction of coastal and offshore waters across the continental shelf during this time
frame. The ship managed to track two separate paths through the Houtman Abrolhos Islands
that offer reasonable substance for analysis in this dissertation.
The Southern Surveyor wound its way through the Middle Channel on the 27th of October
after previous testing of waters to the north. A nine station transect was recorded as the vessel
travelled east to west. It headed out beyond the 1000m depth contour of the continental shelf
before turning back eastward to traverse a route through the Zeewijk Channel and recorded a
further 9 testing stations along this transect and into coastal waters. The positions of the
sampling stations reflect the interests of the CSIRO regarding the properties of water
specifically on the continental shelf slope, and the exact positions with the respective water
depth are recorded in Table 3.1.
Table 3.1: exact locations, and the associated water depth, of the sampling stations for CTD testing on the Zeewijk and Middle Channel transects.
Zeewijk Channel Middle Channel Station Latitude Longitude Depth (m) Latitude Longitude Depth (m)
The stations are not ideal for an accurate assessment of the properties within the Zeewijk
Channel, but still offer the opportunity for surmisal of the basic channel properties.
Comparisons of the water’s properties between the two transects can also be made.
Methodology 38
1 23
4 56
7 8
9
1
23
5
4
6 7
8
9
Figure 3.1: Southern Surveyor transects and approximate sampling station locations through the Zeewijk and Middle Channels, October 27th and 28th, 2003 (Australian Hydrographic Service, 2005).
The sampling station locations can be classified into two groups. ‘Shelf stations’, 1 to 5, and
‘channel stations’, which must therefore include stations 6 to 9 and these classifications apply
for both of the transects. The two groups of stations are separated by the 100m contour and
should enable the ability to contrast continental shelf water and that associated with the
Abrolhos Island Channels. Examining the groups separately at the appropriate stages will
prevent results from being obscured by effects such as scaling also.
3.3. CTD
The Conductivity, Temperature and Depth probe is a multi-parameter instrument which
measures and records conductivity, temperature and pressure as it is lowered through the
water column. This enables the analysis of salinity, temperature and depth and often a host of
other parameters. The CTD probe contains three internal sensors (one each for temperature,
conductivity and pressure), around which water can flow for measurement. The signals from
the three sensors are then transmitted to the surface to be recorded in a data logger. Across
the CSIRO vessels transects deployment of the CTD occurred at the stations illustrated in
Figure 3.1, with subsequent sampling at approximately 2 metre intervals down the water
column and beginning near the surface.
Methodology 39
Supplementing the water column testing was the pumping of surface water through an
onboard analyser to provide surface values of temperature and salinity. This data was
recorded at far more regular intervals than the prevalence of testing stations to highlight
surface water properties with greater accuracy. Recorded also was the water column depth at
each of these surface sampling positions, which can be used to give a bathymetry across the
continental shelf and through each of the channels relative to the alignment of the transect.
The surface water analysis begins from the coordinates (28.93°S, 113.62°E) and ends at
(28.28°S, 114.04°E). The water column depth range across the transect, in terms of where the
surface sampling occurred, is approximately 850m to 40m.
3.4. ADCP
The Acoustic Doppler Current Profiler (ADCP) is a eulerian, or fixed position device, used to
measure the current speeds and directions of the water moving through the Zeewijk Channel.
It was deployed north of Gun Island in the Zeewijk Channel at coordinate (28.8°S 113.86°E)
for two periods of record, firstly from October to December in 2002, and secondly from
September 1st through to November 28th in 2003.
Figure 3.2: the ADCP used in the Zeewijk Channel and its deployment location.
Methodology 40
The profiler illustrated in Figure 3.2 is a 614.4 kHz upward looking instrument that was fixed
to the bottom in approximately 33m of water. The ADCP has a blanking distance of 2m,
therefore no data was recorded within 2m of the seabed and bin placements, were at 1m
intervals through the water column. Acoustic beams are emitted from a transducer and then
scattered by small particles and zooplankton moving with the currents. The reflected beams
are measured by a sensor on the instrument. The bins used for current analysis were located
close to the sea bed and also near the top of the water column.
Current roses will be used to illustrate the prevailing current directions through the Zeewijk
Channel for the 88 day period of testing, at the surface and near the sea bed. To highlight
current variability, the data has been converted into cross-channel and along-channel
components. Channel orientation is predominantly east to west but from observation of the
topography, a tilt of approximately 42° towards the north has been factored into the
calculations. Cross channel is therefore across the y-axis, north-westerly being positive and
south-easterly negative, with these directions characterised by the barriers at their extremities.
Along channel effectively means ‘through the channel’, therefore north-easterly currents are
positive and south-westerly currents are negative. MATLAB (MathsWorks Inc.),
programming has been used to create these figures after manipulation of the data sets in
Microsoft Excel.
3.5. Wind Analysis
Wind data recorded at the Bureau of Meteorology’s North Island testing station (28°18’S,
113°36’E) was obtained for the periods of testing. Significant gaps exist in the spring 2003
wind data as a result equipment malfunction. The spring 2002 data offers a complete period
of record therefore it has been analysed as a reference.
Wind speed and direction values are actually an averaged figure over the 10 minutes prior to
the hourly observation time. Wind speeds are in kilometres per hour, and although wind
direction could be output as either degrees or points of the compass, degrees were chosen
with values rounded to the nearest 10 degrees. Feather plots, again created in MATLAB,
have been used to illustrate the incidence of wind-induced currents. The wind data has been
converted into its component form and re-orientated so that the feather plots point in the
direction the wind is heading, rather than where it has come from to assist interpreting of
wind stress on currents. Wind roses demonstrate the prevailing seasonal conditions.
Methodology 41
3.6. Digitisation of Bathymetry
Since the resolution of bathymetry for the region was at 1000m from available sources,
enquiries were made through the DPI and Australian Navy to determine if finer resolution
data existed. This was eventually obtained from the Australian Hydrographic Service. The
data was stored as a column matrix consisting of longitude, latitude and the depth recorded at
each of these coordinates by the testing vessel. This data was then loaded into MATLAB and
the following script used to create the digitised bathymetry:
>>surf(XI,YI,ZI)
>>shading flat
>>caxis([-150 0])
>>colorbar(‘vert’)
Note that area’s where the testing vessel was physically unable to proceed encompasses both
shallow reefs and actual islands and this default (zero) surface representation is plotted as
dark red on the bathymetric map. The colour bar depicts a 0 to 150m depth scale, so that the
channel bathymetry is clearly shown.
Figure 3.3: digitised bathymetry of the Zeewijk Channel. Depth is in metres on the z-axis, latitude and longitude are on the y and x-axis respectively. Data courtesy of the Australian Hydrographic Service (2005).
Results 42
4. Results
Field results from the respective surveys will be used to assess the oceanographic properties
of the Zeewijk Channel and to determine a general circulation pattern for the period of
testing. The dominant forcing mechanisms driving these currents will be identified in the
synthesis of results.
4.1. CTD
The conductivity, temperature and depth analysis will identify if discrete water bodies are
present at the surface, and progressing through the water column of the two transects tested.
The temperature and salinity properties of the water columns should indicate the degree of
mixing occurring in the channels relative to that outside of it. The surface temperature and
salinity recordings are for the Zeewijk Channel transect. Middle Channel data was not
obtained for analysis.
The annotations to the surface temperature and surface salinity plots indicate the relative
water depth to the surface water properties at significant locations along the transect. The
depth of the water column at the first recorded sample is 850m and the depth of the last, 41m.
4.1.1. Surface Temperature
The temperature of surface water decreased considerably, after the Southern Surveyor entered
waters inside the 60m contour. The difference in temperature is nearly 0.5°C between the
shelf break water and continental shelf water. This trend suggests the presence of a relatively
warm mass of surface water above the continental shelf slope near the Abrolhos at the time of
testing.
Results 43
Figure 4.1: the change in surface water temperature from the continental shelf slope, into the Zeewijk Channel. A noticeable temperature decrease is observed after 20km where the testing entered waters beyond the 60m contour line of the continental shelf.
4.1.2. Water Column Temperature
There is a steady decrease in temperature with depth for all of the ‘shelf stations’. Station 1,
the deepest and western-most station, is colder than the other stations at the same depths
through the water column until after 350m. Stations 2 to 5 have very similar temperature
profiles beginning at surface temperatures very close to 19.7°C for each station, and
decreasing at a slightly faster rate after the 100 to 150m mark.
A temperature profile relative to the continental shelf contour is illustrated also, although the
scaling affect of the relative length scale is not conducive to accurate visualisation of the
temperature profile above the 200m contour. The characteristic increase in temperature with
depth of the offshore stations is evident however.
Surface Water Temperature Change Across Transect
19.2 19.3 19.4 19.5 19.6 19.7 19.8 19.9
20 20.1
0 10 20 30 40 50 60 Distance of Transect (km)
Tem
pera
ture
(C) Annotation Depth (m)
400 60 40
Results 44
Temperature (C)
-1000
-900
-800
-700
-600
-500
-400
-300
-200
-100
00 10 20
Dep
th (m
)Station 1
Station 2
Station 3
Statoin 4
Station 5
Station 6
Staton 7
Station 8
Station 9
Figure 4.2: temperature-depth profiles of stations 1 to 5 on the Zeewijk transect, 27/10/03.
Figure 4.3: temperature contour plot relative to the depth of the continental shelf.
Results 45
4.1.2.1. Zeewijk Channel
The ‘channel stations’ profiles indicate two separate bodies of water according to temperature
with depth. Stations 6 and 7, situated west of the ‘step’ feature in the Zeewijk Channel
recorded higher temperature with depth than the landward stations. Stations 8 and 9 also
demonstrated relatively constant temperature with depth after the initial few metres.
Figure 4.4: temperature-depth profiles of stations 6 to 9, and a microscopic view of Figure 4.2, to compare the temperatures of the inner and outer stations in the surface 100m on the Zeewijk Channel transect, 27/10/03.
Comparing the channel and shelf stations over the same depth range indicates that stations 2
to 7 have similar temperature profiles including a shared surface temperature of
approximately 19.7°C and a moderate thermocline between the 5 and 10m interval. Station 1
has cooler water in the upper 100m as evidenced earlier, but it shares this property, including
a surface temperature of approximately 19.25°C with stations 8 and 9 from inside the
channel. Stations 8 and 9 have almost uniform temperature through the relatively short depth
of their water columns however.
Temperature (C)
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
019 19.5 20
Station 2
Station 3
Station 4
Station 5
Station 1
Temperature (C)
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
019 19.5 20
Dep
th (m
)
Station 6
Staton 7
Station 8
Station 9
Results 46
4.1.3. Surface Salinity
The salinity of surface water increased after the Southern Surveyor entered waters inside the
60m shelf contour. The salinity increase was approximately 0.05ppt. The region of testing
above the continental shelf slope has low salinity, relative to the adjacent water inside the
60m contour.
Surface Water Salinity Across Transect
35.5235.5435.5635.5835.6
35.6235.6435.6635.68
0 10 20 30 40 50 60
Distance (km)
Salin
ity (p
pt)
Figure 4.5: the change in surface water salinity from the continental shelf slope, into the Zeewijk Channel. A noticeable salinity increase is observed after 20km where the testing entered waters beyond the 60m contour line of the continental shelf. The same annotations as the surface temperature plot apply.
4.1.4. Water Column Salinity
Salinity initially increased with depth by approximately 0.2ppt for the first 200 to 250m of
the water column before decreasing fairly rapidly below this mark. Station 1 has the highest
salinity level at the surface at 35.68ppt. Station 1’s salinity increases with depth at a similar
rate to the other stations but it is the first station to reverse this trend and have decreasing
salinity beginning from approximately the 100m mark.
The contour plot for salinity depicts the increasing salinity in the first 200 metres of the water
column as a bubble of lower salinity water positioned on the upper shelf break.
Results 47
Salinity (ppt)
-1000
-900
-800
-700
-600
-500
-400
-300
-200
-100
034.00 35.00 36.00
Dep
th (m
)
Station 1
Station 2
Station 3
Station 4
Station 5
Figure 4.6: salinity-depth profiles of stations 1 to 5 on the Zeewijk transect, 27/10/03.
Figure 4.7: salinity contour plot relative to the depth of the continental shelf.
Results 48
4.1.4.1. Zeewijk Channel
The ‘channel stations’ profiles indicate very little difference in salinity levels between the
four stations, especially in the first 40m of the water column. Stations 6 and 7 have increasing
salinity with depth. The increase is only slight but is particularly evident in the first 10m and
below the 40m mark of the water column. Stations 8 and 9 had relatively uniform salinity
with depth.
Figure 4.8: salinity-depth profiles of stations 6 to 9, and a microscopic view of Figure 4.6, to compare the temperatures of the inner and outer stations in the surface 100m on the Zeewijk Channel transect, 27/10/03.
Comparing the shelf and channel stations for the first 100m of the water columns highlights
that Station 1 is markedly more saline at the surface, by nearly 0.1ppt, than all other stations.
Stations 8 and 9 are distinct due to their uniform salinity between 35.6 and 35.62ppt. Stations
6 and 7 tend towards the trend exhibited by the intermediate stations.
Salinity (ppt)
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
035.55 35.60 35.65 35.70
Dep
th (m
)
Station 1
Station 2
Station 3
Station 4
Station 5
Salinity (ppt)
-100
-90
-80
-70
-60
-50
-40
-30
-20
-10
035.55 35.60 35.65 35.70
Dep
th (m
) Station 7
Station 8
Station 9
Station 6
Results 49
4.1.5. Sea water Density
An estimate of the density difference between the 2 discrete surface water masses illustrated
in the surface temperature and salinity plots is necessary to determine if currents in the
channel are being driven by baroclinic forces. The discrete water bodies can be classified as
being either offshore of the 60m depth contour and therefore over the shelf slope (in context
the 60m contour line is near the Pleistocene ridge formation), or inshore of the 60m contour
and therefore within the Zeewijk Channel. The temperature, salinity and related density of the
surface water is described in Table 4.1.
Table 4.1: temperature, salinity and related density of surface water inside the Zeewijk Channel and above the continental shelf slope.
Temperature (°C) Salinity (ppt) Density (kgm-3) Shelf Slope 19.75 35.58 1025.3
Channel 19.25 35.63 1025.4
The calculation of density was made using MATLAB’s Sea Water tool box. Since assessment
of the surface water density was required, pressure was assumed to be atmospheric and
temperature and salinity were inputted in the units shown in Table 4.1. The density difference
between the two water masses is 0.1kgm-3. According to (Pattiaratchi 2005) this is a
sufficient density difference to impact on the hydrodynamics of the area.
An estimate can be made of the baroclinic velocity induced by these different sea surface
properties. Using the equation and constants outlined in Chapter 2.3.4.1, and stations 4 and 8
as the geographic estimates of the water body positions, gives an estimate of U, the surface
velocity, as being approximately 0.00058ms-1 under these conditions.
Results 50
4.1.6. Middle Channel vs. Zeewijk Channel
The Middle and Zeewijk Channels can be compared in terms of water column properties to
determine if these properties are localised or typical of the greater system.
Figure 4.9: TS diagrams to compare the water column properties of the Zeewijk and Middle Channels. The lower figures represent the water column properties of the ‘channel stations’, that is, stations 6 to 9. Station 1 has been omitted from all figures to aid interpretation due to the scale of its depth and properties. Station 9 of the Middle Channel is located at approximately (35.61, 19.2), and is hard to view as it is almost a point.
There is close correlation between the trends in water properties of the two channels despite
the significant differences in depths between stations 2 to 5. The Middle Channel stations are
slightly warmer, surface temperatures are over 20°C for the intermediate stations, and salinity
is also lower by approximately 0.05ppt in the shelf slope region. The correlation is strongest
in terms of temperature decreasing with depth, and salinity increasing for a period and then
decreasing after the 200m mark, for the shelf stations on both transects.
TS Diagram Zeewijk Channel
19
19.2
19.4
19.6
19.8
20
20.2
20.4
35.45 35.55 35.65 35.75
Salinity (ppt)
Tem
pera
ture
(C)
Station 6
Station 7
Station 8
Station 9
TS Diagram Middle Channel
19
19.2
19.4
19.6
19.8
20
20.2
20.4
35.45 35.55 35.65 35.75
Salinity (ppt)
Tem
pera
ture
(C)
Station 6
Station 7
Station 8
Station 9
TS Diagram Zeewijk Channel
10
12
14
16
18
20
22
35.4 35.5 35.6 35.7 35.8 35.9
Salinity (ppt)
Tem
pera
ture
(C)
Station 2
Station 3
Station 4
Station 5
Station 6
Station 7
Station 8
Station 9
TS Diagram Middle Channel
10
12
14
16
18
20
22
35.4 35.5 35.6 35.7 35.8 35.9
Salinity (ppt)
Tem
pera
ture
(C)
Station 2
Station 3
Station 4
Station 5
Station 6
Station 7
Station 8
Station 9
Results 51
The uniformity of properties throughout the water column’s of stations 8 and 9, is shared by
the Middle and Zeewijk Channels. This is evidenced in the TS diagram by the properties
profiles represented as points, or very small lines, rather than curves. Station 6, situated just
inside the 100m contour on both transects, has the same TS profile of decreasing temperature
and increasing salinity with depth for both channels. Station 9 has very similar water column
properties on both transects, with approximate temperature of 19.2°C and salinity 35.61ppt,
for their entire depth’s of less than 40m.
The Zeewijk Channel T-S diagram with station 1 included is notable for its visualisation of
the three major current systems off the continental shelf. The Leeuwin Current, Leeuwin
Undercurrent and the presence of Antarctic Intermediate water can be identified from the T-S
diagram in Figure 4.10, distinguished by the gradient changes in the figure. This is a typical
T-S diagram for the region in 1000m of water (Pattiaratchi 2005).
TS Diagram
0
5
10
15
20
25
34 34.5 35 35.5 36
Salinity (ppt)
Tem
pera
ture
(C)
Figure 4.10: Temperature-salinity diagram highlighting the currents present in the water column on the Gascoyne continental shelf slope, particularly out at the 1000m contour.
Results 52
Results 53
4.2. ADCP
The ADCP data was recorded in the Zeewijk Channel from the 1st of September through to
the 28th of November in 2003 at the location described in the methodology. This spring
period of 2003 yielded a complete data set from the ADCP and has been selected over the
incomplete 2002 data set. It has been analysed for the surface and bottom currents in terms of
the prevailing current directions and current variability. Spectrum analysis will be employed
to determine the factors influencing current variability.
4.2.1. Current Roses
Current roses will demonstrate the net, or prevailing, current directions at the surface and
bottom of the Zeewijk Channel during spring of 2003. Using the speed and direction recorded
by the ADCP, the roses illustrate the percentage incidence of currents.
Surface
The dominant surface current direction is westerly, contributing to over a third of the surface
current pattern. A significant west to south-westerly component exists as well which means
that net transport at the surface is offshore. The influence of barriers in the form of Island
groups and reefs to the north and south is evident from the east to west dominance of the
current rose.
Bottom
The bottom currents are more evenly distributed compared to the surface currents, with
orientation between north-westerly and south-easterly. There is a higher incidence of cross-
channel currents at the bottom, especially towards the north, and it should also be noted that
peak bottom current speeds are approximately half the magnitude of those at the surface.
Results 54
10 %
20 %
30 %
40 % N
NE
E
SE
S
SW
W
NW
Speed scale in cm/s
0 12 24 36 48 60
Figure 4.11: current rose of surface currents in the Zeewijk Channel during spring 2003.
10 %
20 %
30 %
40 % N
NE
E
SE
S
SW
W
NW
Speed scale in cm/s
0 5 10 15 20 25
Figure 4.12: current rose of bottom currents in the Zeewijk Channel during spring 2003.
Results 55
4.2.2. Current Variability
Analysis of the cross- and along-shore variability will determine periodic features of the
current pattern. In relating the current variability to the current rose results, the along-shore
and cross-shore currents are to be viewed as the respective components of the actual current
direction at any point in the time series. For example stages of positive cross-shore current
that correlate to negative along-shore currents at the same time, are actually periods of
westerly current flow.
Surface
Surface current variability is characterised by a high incidence of ocean-ward or westerly
current direction. The major alternative current direction is easterly. The current fluctuations
vary from a relatively large 5 day oscillation to daily changes in current strength. The 5 day
oscillation is interrupted by an extended period of westerly currents, lasting 10 to 15 days and
with diurnal fluctuations within the time frame. It returns to the 5 day oscillation pattern at
the end of this phenomenon and persists until the end of the period of record.
250 260 270 280 290 300 310 320 330
-50
0
50
Along-channel Surface Currents
250 260 270 280 290 300 310 320 330
-50
0
50
Cur
rent
Spe
ed (c
m/s
)
Day
Cross-channel Surface Currents
Figure 4.13: surface current variability in the Zeewijk Channel for the cross- and along-shore directions during spring 2003.
Results 56
The maximum current speed at the surface is approximately 0.7ms-1 in the cross-channel
direction. When combined with a relatively high but negative along-channel component the
resulting vector is of westerly orientation and also has magnitude greater than 0.7ms-1.
Bottom
The short, sharp oscillations in the along- and cross-shore bottom currents illustrate a high
incidence of diurnal factors influencing the bottom current direction and magnitude.
Maximum magnitudes are nearly five times less at the bottom compared to the surface, and
this is true in both current directions. Only one incidence of bottom currents exceeding
0.15ms-1 can be seen and this occurred as a negative (south-easterly) in the cross-channel
direction. Since the maximum along-channel current correlates with this component the
maximum bottom current speed reached is approximately 0.22ms-1 and in an easterly
direction. An oscillation is also evident for an approximately 10 to 15 day period in both the
along- and cross-channel directions.
250 260 270 280 290 300 310 320 330-20
-10
0
10
20Along-channel Bottom Currents
250 260 270 280 290 300 310 320 330-20
-10
0
10
20Cross-channel Bottom Currents
Cur
rent
Spe
ed (c
m/s
)
Day Figure 4.14: bottom current variability in the Zeewijk Channel for the cross- and along-shore directions
during spring 2003.
Results 57
The cross-channel currents demonstrate a shift from predominantly north-westerly (positive)
to predominantly south-easterly (negative) over the 88 days of testing. The along-channel
currents do not share this trend and are relatively constant in their oscillation between
positive and negative values. The resulting bottom current is relatively balanced between
north-westerly and south-easterly. However we now know that the contribution of the
northerly component was primarily in the first half of the testing period, and the southerly
component was therefore predominantly in the second half of the testing period.
Results 58
4.2.3. Particle Progression
The particle progression plots were created in MATLAB to illustrate the progressive motion
of a particle at the surface or bottom for the duration of ADCP recording. For clarity in the
resulting plot, the mean of the 48 current vectors that contribute to each 24 hour period was
calculated. The plot is therefore a representation of the progressive travel of a particle using
the daily mean current vectors for the 88 day period of record. A standard Cartesian plane has
been used, with north/south and east/west orientation of the axis.
Surface
The important feature of the surface particle progression plot is that the current pattern
follows the calculated 42 degree orientation of the channel in terms of the hypothetical
movement of a particle past this point. There is generally a north-east to south-west direction
of currents at the surface. This significantly elliptical circulation of surface currents in the
‘along-channel’ direction, as specified in the ADCP methodology, is evident in Figure
4.15Error! Reference source not found..
-100 -80 -60 -40 -20 0 20 40 60 80 100-40
-30
-20
-10
0
10
20
30
40Particle Progression Plot - Surface
West-East(km)
Sou
th-N
orth
(km
)
Figure 4.15: plot of the path travelled by a particle under the influence of the mean daily surface currents in spring 2003. The current path is elliptical in the along-channel direction.
A particle is unlikely to travel more than 90km either side of the ADCP position due to the
mean daily currents before a direction change would force it to return channel-ward. The
implications of these particle progressions on the circulation pattern are explored in the
discussion.
Results 59
Bottom
The bottom progressive vector plot suggests tendencies towards north-west and south-east
migration of particles due to bottom currents. This may be linked to the channel bathymetry
and the gap in the ‘ridge’ at the northern end of the ocean-ward channel opening. The
magnitude of particle movement at the bottom is significantly smaller than the magnitude of
particle movement at the surface.
-15 -10 -5 0 5 10 15
-10
-5
0
5
10
West-East (km)
Sou
th-N
orth
(km
)
Particle Progression Plot - Bottom
Figure 4.16: progressive plot of the path travelled by a particle under the influence of the mean daily bottom currents in spring 2003. There is a strong north-west to south-easterly current trend.
The range of the particle circulation is approximately 30km in comparison to the surface
range that is over 150km. Since the length scale of the channel is approximately 30km also, it
appears that bottom circulation is largely confined to the channel region according to this
result.
Results 60
4.2.4. Particle Excursion
The particle excursion plots were also created in MATLAB, and illustrate the possible
distance a particle could travel at the surface or bottom in a day using the daily mean current
vectors for the period of record. A standard Cartesian plane has again been used, with north-
south and east-west orientation of the axis.
Surface
As expected, the surface particle excursion plot follows the 42 degree orientation of the
channel that was illustrated in the surface particle progression analysis. The particle
excursions are therefore typically in the east to north-east and west to south-west directions.
The maximum particle excursion in one day was nearly 50km and it occurred for both
directions. This plot provides a better representation then the progressive figure, of the range
a particle could be expected to travel through the Zeewijk Channel in one day.
-50 -40 -30 -20 -10 0 10 20 30 40 50-20
-15
-10
-5
0
5
10
15
20Particle Excursion Plot - Surface
West-East (km)
Sou
th-N
orth
(km
)
Figure 4.17: the distances travelled by a particle at the surface under the daily mean current conditions for each day of the testing period.
Results 61
Bottom
The bottom particle excursion plot depicts the tendency towards north-west and south to
south-east excursions. The maximum distanced travelled by a particle in one day is
approximately 8km. Particle excursion near the bottom is over 5 times less than the
maximum particle excursions at the surface. The excursion range of nearly 16km, 8km in
either direction, is further evidence that bottom circulation is highly channel-centric.
-8 -6 -4 -2 0 2 4 6 8-8
-6
-4
-2
0
2
4
6
8Particle Excursion Plot - Bottom
West-East (km)
Sou
th-N
orth
(km
)
Figure 4.18: the distances travelled by a particle near the bottom under the daily mean current conditions for each day of the testing period.
Results 62
4.2.5. Spectral Analysis
The spectral density analysis illustrates the spectrum of energy in the system for the time
frame of testing. Energy fluxes that correspond to barotropic forcing mechanisms are
distinguished by the frequency, and therefore period, at which these processes are known to
oscillate. The energy spectrum for periods greater than approximately 20 days is inherently
large as a result of the time scale of the field data, around 90 days, and this makes it
statistically weak as a result.
The energy in the system is nearly 102 times greater at the surface than at the bottom. Also,
the along-shore component has greater energy input to the system at the surface than its
related cross-shore component. This trend is not depicted at the bottom. Following the colour
convention used in ‘current variability’, the red line represents along-channel currents and the
blue represents cross-channel currents for the surface and bottom graphs.
Surface
By far the most significant proportion of the surface current’s energy spectrum correlates to
periods between 5 and 15 days. As mentioned earlier this is expected, but the spectral density
of these longer period frequencies is still a significant factor for understanding of the
channels circulation as they are associated with mesoscale oceanography and climatology.
There are definite spikes in spectral density at 24 hours and to a lesser degree close to 12
hours, associated with the tidal components of the channel. The 2.2 days or 52 hours
frequency represents a greater energy input than the tidal component, and the implications of
this will be discussed later. The last significant spike in spectral density is at the frequency
for a period of 4 hours. This value corresponds to the period of the continental shelf seiche
for the region.
Bottom
The feature of spectral analysis of the bottom currents is the greater significance in the
relative energy of the tidal components. Evidence of this includes the 5 day frequency
containing less spectral density than the diurnal tidal regime. There is still some 10 to 15 day
controlling factor that can be associated with the majority of energy found in the system. This
can be linked to the spring/neap tidal cycle, further evidence that the tidal regime is far more
important for bottom circulation relative to its importance at the surface. The 52 hour
frequency also has less significance for the bottom circulation. The continental shelf seiche
exhibits a small degree of influence on bottom currents also.
Results 63
Figure 4.19: Spectral density versus frequency for surface currents in the Zeewijk Channel during spring
2003.
Figure 4.20: Spectral density versus frequency for bottom currents in the Zeewijk Channel during spring
2003.
Results 64
4.3. Wind
Wind data for the periods 1/9/03 to 16/9/03, and 28/10/03 to 6/11/03, has been used for
analysis as these periods represented the largest sections of complete recording by the North
Island station. It has been plotted relative to the surface currents at the same time to depict the
incidence of wind-induced currents in the system. The spring 2002 data will be represented as
a wind rose to illustrate the prevailing wind conditions.
4.3.1. Spring 2003
The wind data from 1/9/03 to 16/9/03 shows a correlation with the direction of surface
currents. The feather plot illustrates a particularly strong change in wind pattern. Fluctuating
north-west to north-easterlies interrupt the incidence of southerly winds and induce a south-
easterly surface current in the channel for a short time. The return to prevailing wind
conditions negates the southerly wind stress and the surface current pattern returns to what
seems to be the prevailing oscillatory conditions, including the significant westerly direction.
The strongest winds experienced in this period of record were the north-westerlies. They
reached 19ms-1 approximately at there peak.
244 246 248 250 252 254 256 258 260
-50
0
50
Along-channel Surface Currents
244 246 248 250 252 254 256 258 260
-50
0
50
Cross-channel Surface Currents
Cur
rent
Spe
ed (c
m/s
)
Day
-20
-10
0
10
Wind for days 243 to 260Spe
ed (m
/s) a
nd D
irect
ion
Figure 4.21: wind influence on surface currents for 16 days of the testing period, between 1/9/03 and 17/9/03.
Results 65
The second section of spring wind data, 28/10/03 to 6/11/03, is shorter than the first by 6
days and consists of a steady stream of easterly to southerly winds. In fact, the wind swings
from east to south repeatedly and the period of each swing is in the order of 1 to 2 days. Wind
strength increases as its orientation becomes increasingly southerly also.
301 302 303 304 305 306 307 308 309 310
-50
0
50
Along-channel Surface Currents
301 302 303 304 305 306 307 308 309 310
-50
0
50
Cross-channel Surface Currents
Cur
rent
Spe
ed (c
m/s
)
Day
-5
0
5
10
15
Wind for days 301 to 310Spe
ed (m
/s) a
nd D
irect
ion
Figure 4.22: wind influence on surface currents for 10 days of the testing period, between 28/10/03 and 6/11/03.
The surface currents represent a nine day period of largely westerly flow that has periodic
oscillations in magnitude. These oscillations correlate to the east-south variations in wind
speed and direction, although other diurnal processes including the tidal regime cannot be
discounted in terms of importance. The strongest southerly wind peaked at approximately
13ms-1, and correlates with a period of strong westerly surface current at the end of day 303.
The strongest current during this time frame occurred during day 305 under the influence of
moderate southerly winds, and reached an estimated 0.70ms-1.
Results 66
4.3.2. Spring 2002
The wind rose of percentage wind incidence and magnitude for spring of 2002 highlights the
prevailing southerly wind pattern in the region during this season. The wind rose depicts data
from the 1st of October to the 31st of December. This is slightly longer than the data sets
analysed for spring 2003, however the rose is still useful as a representation of the seasonal
wind pattern. Over 70% of the recorded wind directions had strong southerly components.
Figure 4.23: current rose of winds recorded at the North Island recording station 1/10/02 and 31/12/02.
Results 67
4.4. Waves
Wave data was recorded at the ADCP location north of Gun Island for the period 12/10/02 to
16/11/02. Wave heights are relatively small due to the sheltering aspect of Half moon Reef
and the Pelsaert Island Group.
4.4.1. Spring 2002
Wave heights appear periodic with peaks in magnitude at approximately 5 day intervals. The
maximum and minimum wave heights experienced at the testing location were 1.25 and 0.3m
respectively. Between days 306 and 317 there was relatively constant wave height between
0.5 and 1 metre with diurnal fluctuations evident over this time frame.
The period of recorded waves ranges between 10 and 20 seconds for the majority of the
testing record. However there are 3 distinct drops in period. They are at days 298, where its
rise from the low of nearly 3 seconds correlates with a rise in wave height to the maximum of
1.25m, and at days 316 and 318 which could be associated with the apparatus failure that
caused the loss of data at this time.
Wave direction stays in the range between approximately 270° and 330° for the majority of
the testing period which correlates to a predominantly west to south-westerly wave direction.
Waves follow the same naming conventions as winds therefore they are basically travelling
into the western entrance of the channel after being generated in the Indian Ocean. The
apparent direction change at day 318 may have been induced by a period of calm or the
occurrence of very strong easterly winds forcing offshore wind waves at the sea surface for a
short period of time.
Results 68
Figure 4.24: depicts the wave height, period and direction of waves in the Zeewijk Channel for the period 12/10/02 to 16/11/02.
Results 69
4.5. Eddy Formation
Using plots of sea level anomaly obtained from the CSIRO Marine Division (2005), eddy
formation near the Abrolhos Islands during the period of testing can be examined. There are
at least two instances in the spring of 2003 where the geostrophic anomaly induced eddy
formations that impacted upon water circulation around the archipelago. They are both the
result of a ‘mound’ of higher sea level to the south of the Pelsaert Group.
4.5.1. Spring 2003
The first eddy formation identified is from the 6th of October. The geostrophic anomaly
represents a difference in sea level of approximately 0.3m. The anti-clockwise motion of the
eddy is illustrated by the directional markers. They also quantify the geostrophic velocity
outside the Zeewijk Channel as between 0.1 and 0.3ms-1 (some magnification of the figure
may be required for clarity) with an associated westerly or offshore heading.
The sea surface temperature is approximately 20°C at the centre of the eddy and slightly
cooler at the position identified as in front of the Zeewijk Channel opening. The southern side
of the eddy is forcing relatively cold water towards the Western Australian coastline. The
northern side is entraining coastal water from the Abrolhos’ landward side, through the
channels due to the bathymetry of the archipelago, and offshore.
The second eddy formation is on the 9th of November. This time the geostrophic anomaly is
slightly greater, with a sea level difference closer to 0.4m. In turn the geostrophic velocity of
the resulting current pattern is also greater, between 0.2 and 0.4ms-1, and the direction of flow
in front of the Zeewijk Channel is south and westerly. The circulatory influence and presence
of this eddy is clearer than the eddy on October 6th, its size and influence is easily discernible
from the image of sea surface temperature.
There is transport of water northwards inside the 100m contour of the continental shelf from
as far south as Cape Leeuwin, although the geostrophic velocity of this transport fluctuates in
intensity between 0 and 0.3ms-1. The November 9th eddy pushes offshore water into the
northward coastal stream at its southern margin, effectively strengthening the northward
geostrophic velocity that had weakened approximately 200km north of Perth. The eddy
entrains coastal water offshore at its northern end and since this is aligned with the Abrolhos
Islands, the water is brought through the channels.
Results 70
Figure 4.25: sea level anomaly, related geostrophic velocity and sea surface temperature for post-October 6th 2003 (CSIRO 2005).
Figure 4.26: sea level anomaly, related geostrophic velocity and sea surface temperature for post-November 9th 2003 (CSIRO 2005).
Discussion 71
5. Discussion
Beginning with the broader patterns concerning water properties and circulation in the
Zeewijk Channel requires consideration of climatic features. The Houtman Abrolhos Islands
were emerging from winter and the regions annual period of heavy rainfall during the time
frame of field testing. This has contributed to relatively low surface salinity and temperature
values being recorded in comparison to possible values if sampling had of occurred in
summer. Seasonality is also an integral factor when explaining the significance of the
Leeuwin Current’s influence for the duration of field testing.
In the same context inter-annual factors become important. El Niño events generally
debilitate the strength and influence of the Leeuwin current on the Western Australian
coastline. Spring of 2003 is characterised by the end of a strong 18 month El Niño
phenomenon, therefore we may expect to see evidence of the Leeuwin Current’s presence
near the Abrolhos Islands, but not necessarily large scale inundation of the field site as
observed in the past (Pearce 1997).
Topography and bathymetric features of the Houtman Abrolhos and Zeewijk Channel are
important for the current circulation pattern as they can channel the direction and indirectly
the magnitude of the water’s movement. The barriers formed by the island and reef chains
means access to the Indian Ocean for continental shelf water behind the Abrolhos that is
compelled to move offshore must be through the channel systems. Features of the bottom
topography including the ‘step’ and ‘ridge’ can be controlling factors for bottom water
transport and will be discussed later.
Warmer water is evident near the surface on both the surface recording and the depth profiles
for the region outside the immediate channel vicinity and primarily over the continental shelf
slope. The inner channel stations, 8 and 9, were generally colder than the intermediate, or
shelf slope, stations. Lower salinity water is evident near the surface on both the surface
recording and the depth profiles for the intermediate stations over the continental shelf slope.
The inner channel stations have higher salinity water at the surface compared to these shelf
slope stations.
Discussion 72
The mass of relatively warm and low salinity water identified is evidence that the Leeuwin
Current was present above the major continental shelf slope during the period of testing. The
colder water on the Geelvink Channel end of the system is a result of the Capes Current’s
influence and establishes that the Leeuwin Current was not inundating the Abrolhos Islands
during the spring of 2003.
The extent of the Leeuwin Currents indirect influence on the circulation pattern of the
Zeewijk Channel depends on the significance of baroclinic and barotropic flows associated
with its presence. In terms of the baroclinic processes that were occurring, water of lesser
density should flow toward and over higher density water at the surface depending on the size
of the density difference between the two masses. The density gradient between the channel
and shelf slope stations, or Capes Current and Leeuwin Current water respectively, was
calculated to be 0.1kgm-3 and the direction of associated flow under these circumstances
would be from the shelf slope towards the channel. The velocity of horizontal baroclinic flow
at the time of testing has been estimated as 0.00058ms-1.
Baroclinic flow was therefore a relatively minor component of the Zeewijk Channels
circulation pattern for the examined time frame. In fact, it is likely that baroclinic effects
were made redundant by the scale of barotropic flow in the opposing direction during the
spring of 2003.
The surface current rose depicted net offshore, or westerly, transport at the location of field
testing in the channel. Water from inside the Geelvink Channel was therefore predominantly
induced through the narrow channel opening to mix with water above the continental shelf
slope during the testing period. Currents inside the Zeewijk Channel were significantly
greater than those observed by Pearce (1997) at the surface outside the Abrolhos Islands
channels, suggesting that there is a magnification effect within the channels. Spectral analysis
indicated that these currents are driven by processes with relatively long time scales. The
influence of frequencies with over 5 day periods is particularly significant and can be linked
to mesoscale affects, such as the passage of weather systems that intensify the localised wind
stress for relatively long periods. This is also an appropriate time scale to apply to flows
related to large eddies driven by geostrophic anomaly.
Discussion 73
The comparison of wind patterns against surface current variability suggests that there are
long periods where wind-stress is the major circulation driver at the surface in the Zeewijk
Channel. Further evidence of this includes the well mixed water columns found in the
immediate channel vicinity. The surface temperature at station 1 is relatively cold also, but
where the outer station has decreasing temperature with depth even within the first 40m of
the water column, the two channel stations have uniform temperature with depth. This
indicates a high degree of mixing in the channel relative to the outer station. The distinct lack
of haloclines at the inner channel stations emphasises that the water columns are well mixed
in the channel vicinity.
This vertical mixing is most likely a result of wind-induced turbulence and Ekman transport
down the water column. The prevailing southerly winds induce a northerly wind stress which
in turn is forced to the left at the surface, under the influence of the Coriolis force, creating a
predominantly westerly current pattern in the channel. As outlined in the background, this
rotation continues with depth but with decreasing current strength through the water column
and contributes through turbulence to the mixing of layers down the water column.
The shifts in wind pattern between relatively strong easterly and southerly winds over the
duration of two days could be an explanation for the considerable energy associated with the
frequency of the 52 hour period. It is essentially then, a periodic shift in wind stress
associated with the prevailing south-easterly wind field. The independence from wind pattern
shown by the bottom currents follows on from this. They are less affected by the frequency
associated with the 52 hour period then the surface currents, which is clearly shown in the
spectrum analysis and are influenced more heavily by other factors.
The bottom currents show a greater degree of duality than currents at the surface. There is a
relatively balanced current pattern between currents in the north-westerly and south-easterly
directions which correlates to the assumed cross-channel current direction. The bottom
current is more evenly oscillatory, or ‘back and forwards’, and this is due to the affects of the
tidal regime. Evidence of this is can be found in the spectral analysis where the energy for
frequencies corresponding to the 24 and 12 hour periods are relatively higher at the bottom
than the surface. They are still nowhere near the significance, in terms of energy input, as the
frequencies corresponding to periods over 10 days. Considering this to be the result of the 14
day neap/spring tidal cycle however, it can be assumed then that the major influence on
bottom currents is the local tidal regime.
Discussion 74
The magnitudes of bottom current and particle excursions suggest that the circulation pattern
near the bottom is very channel-centric and that particles entering the channel near the
bottom, or reaching the benthic zones after transport from the surface, are not likely to be
exported from the system in any short period of time. It would take 3 to 5 days of relatively
strong bottom current with uniform direction to transport particles from the 30km channel
region. Considering the oscillations in bottom current direction illustrated in the current
variability analysis, this is unlikely to occur. Bottom water exiting the channel at the western
end will be influenced by the channel bathymetry, in particular the gap to the north of the
20m high Pleistocene ridge, which is directing the westerly bottom flow further north.
For the particle progression and excursion patterns at the surface, the expected trends may
have been increasingly westerly transport and significant westerly excursions considering the
current rose result at the surface. Instead there is an elliptical progression with a relatively
even north-east to south-west distribution. It is possible that the sorting and manipulation of
data to establish mean currents from the 48 current vectors in each day has influenced the
impressions gained from this result. However the particle excursions do follow the almost
45° from north alignment assigned as the along-channel direction.
The distances involved in the particle excursion and progression analysis indicate that a
particle can easily be transported out of the channel in a day at the surface. A high proportion
of particles were estimated as travelling over 25km in a day, implying that with at least 24
hours of constant flow direction, the currents are sufficiently strong to exchange a high
proportion of surface water with new water from outside the channel. Once outside the
channel the Leeuwin Current meander on the western side, or equally other oceanographic
factors over the continental shelf slope and in the Geelvink Channel direction, will dominate
transport.
The Leeuwin Current meander on the Abrolhos’ western side is also responsible for the
formation of pockets or ‘mounds’ of relatively warm, low salinity water over the continental
shelf slope (Pearce & Griffiths 1991). The wind-induced westerly currents exiting the
Houtman Abrolhos channels push the meander further offshore adjacent to the islands.
Easing of the westerly current allows the Leeuwin Current meander to re-establish itself
alongside the archipelago depending on its initial strength, whilst leaving a pocket of
Leeuwin water over the continental shelf slope.
Discussion 75
The relatively low density water is a geostrophic anomaly, with higher sea level relative to
the water surrounding it and therefore resembling a mound of water in the regional setting.
The images of sea level anomaly obtained from the CSIRO for the testing period illustrate
that the mounds establish themselves to the south-west of the Pelsaert Group and that they
can impose a significant influence on the circulation pattern of the area even under relatively
weak Leeuwin Current conditions.
Eddy formation due to water being transported down these mounds and subsequently left
under the influence of the Coriolis force, enhanced the transport of relatively cool, Capes
Current surface water through the Zeewijk Channel and out onto the continental shelf slope
during the spring of 2003. Although these eddies may only form once or twice per month,
they can drive the barotropic circulation of the region for 5 to 15 days at a time whilst
dissipating in energy with time. The spectral energy distribution for this time scale, whilst it
should be remembered that this period encapsulates a large range of possible events, supports
the relative importance of eddy formation as a channel circulation driver.
Waves incident on the Abrolhos Islands typically come from a west to south-westerly
orientation. The recorded wave heights are actually a sheltered version of the possible wave
heights reaching Half Moon Reef due to refraction and dampening by the reef system. This
implies that swell and waves will have larger wave heights and energy in the open channel.
Since the prevailing current at the surface is in the opposite direction to wave passage
through the channel the effect of waves can be assumed to have little bearing on the
circulation pattern of the channel.
Conclusions 76
6. Conclusions The circulation pattern of the Zeewijk Channel has significantly different characteristics
between its surface and bottom flows with regards to the spring period in 2003. Surface
currents are predominantly driven by the local wind conditions resulting in bulk westerly
flow through the Zeewijk Channel. The currents measured in the Zeewijk Channel showed
greater current velocity at the surface than previous current studies from outside the channel
systems. Mesoscale processes are also of relatively high importance to the circulation pattern,
affecting the wind field via the passage of weather systems and contributing to the westerly
surface currents through geostrophic anomaly induced flows and entrainment.
The significance of baroclinic flow induced by the density difference between the Leeuwin
Current and Zeewijk Channel water is negligible, as it is overshadowed by the opposing
westerly barotropic flows. However, the Leeuwin Current features strongly in many
processes driving the oceanography and circulation pattern of the channel and its indirect
effects must be considered as thoroughly as its direct effects.
The features of bottom circulation in the Zeewijk Channel include its cyclic nature that can
be linked to the relatively greater influence of the tidal regime, and the significantly
decreased current strength at the bottom relative to the surface. Topographic barriers and
bathymetry of the channel are important controls on circulation, directing a large proportion
of potential north-west surface flow further west and also guiding the movement of bottom
water at the western margins of the channel. The topography was found to have a magnifying
effect on the strength of currents within the Zeewijk Channel relative to currents outside of its
confines.
The extent of particle transport at the surface in the Zeewijk Channel suggests that it has
sufficient current strength to support proposed aquaculture trials. The surface current
magnitudes indicate that exchanges of surface water are relatively constant. Considering the
prevailing westerly current orientation, pollution and nutrients associated with aquaculture
cages would be dispersed out into continental shelf slope waters and pose no threat to the
Fish Habitat Protection Area. Questions remain regarding the vertical rate of transport of
waste and nutrients through the water column, since vertical mixing was a feature of the
channel water properties. In particular if pollution was to reach the benthic regions of the
channel, findings indicate that they are then unlikely to be transported out of the system for
long periods of time due to the re-circulation of bottom currents.
Recommendations 77
7. Recommendations for Future Work
A vast amount of research remains to be conducted on the oceanography of the Houtman
Abrolhos Islands and the circulation pattern of its channels. Indeed, a number of projects
focused on the greater oceanography of the region are on-going within the Centre for Water
Research at this point in time. Since this thesis focused on spring conditions due to field work
and data gathering restrictions, the potential to examine the circulation pattern under summer
and winter conditions remains. Furthermore, seasonality and inter-annual variability greatly
affected the results regarding the Leeuwin Current. It is important for groups such as
Fisheries W.A. and those pursuing aquaculture ventures to understand the circulation pattern
under Leeuwin Current enhancing conditions.
A study conducted in the winter of a well-established La Niña period could answer questions
concerning the extent of Leeuwin Current inundation of the islands and eddy formation from
a strong Leeuwin meander. Changes to the circulation regime resulting from these new
conditions may affect the relative importance of the wind induced westerly current field
through the Zeewijk Channel. The significance of baroclinic flow could also be investigated
with regard to the new sea surface temperature and salinity conditions.
Drifter deployment and analysis would provide further insight into the circulation pattern of
surface water in the Zeewijk Channel. Quantitative evidence of particle progression and
excursion would validate the results estimated in this study. The progression and destination
of flows that have left the channel confines may also be obtained through this field work and
have application in the accurate assessment of the environmental impacts of an aquaculture
farm in the region. Vertical migration of nutrients and detritus associated with farming cages
also provides an avenue for further research, particularly considering the vastly different
surface and bottom current conditions. Environmental impact assessments must give grave
consideration to seasonal and inter-annual current variability in the Zeewijk Channel, if it is
to be an economic and environmental success as a future aquaculture site.
Recommendations 78
References 79
References
2002, Australian National Tide Tables. A.I.T.F. 1988, Abrolhos Islands Planning Strategy, Geraldton. B.O.M., Bureau of Meteorology [Online], Available: www.bom.gov.au/ [30/6/05-30/10/05]. Brearley, A. 1997, 'Seagrasses and isopod borers from the Wallabi Islands, Houtman
Abrolhos Islands,WA', in The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australian Museum, Perth, pp. 63-74.
Caputi, N., Brown, R.S., Chubb, C.F. 1995, 'Regional prediction of the Western rock lobster, Panulirus cygnus, commercial catch in Western Australia', Crustaceana, vol. 68, pp. 245-256.
Chua, J. 2002, Oceanographic Modelling of Jurien Bay, Western Australia in Centre for Water Research, University of Western Australia, Perth, p. Honours Thesis.
Chubb, C. F. & Barker, E. H. 2005, The western rock lobster fishery 2001/2002 to 2002/2003, Fisheries Research Division, Perth.
Collins, L. B., Wyrwoll, K.-H. & Zhu, Z. R. 1996, 'The Structure of the Easter Platform, Houtman Abrolhos Reefs- Pleistocene foundations and Holocene reef growth', Marine Geology, vol. 115, no. 1-2, pp. 29-46.
Cresswell, G., Boland, F. M., Peterson, J. L. & Wells, G. S. 1989, 'Continental Shelf current near Abrolhos Islands, Western Australia', Australian Journal of Marine Freshwater Research, vol. 40, pp. 113-128.
Csanady, G. T. 1973, 'Wind Induced Barotropic Motions in Long Lakes', Journal of Physical Oceanography, vol. 3, pp. 429-438.
CSIRO, Marine and Atmospheric Research Remote Sensing Facility [Online], Available: www.marine.csiro.au/remotesensing/ [30/6/05-30/10/05].
Diver & Prince 2003, Yellowfin tuna aquaculture trial, Zeewijk Channel, Abrolhos Islands, Environmental Protection Authority, Perth, Western Australia.
Drake, C. L., Imbrie, J., Knauss, J. A. & Turekian, K. K. 1978, Oceanography, Holt, Rinehart and Winston, United States of America.
Fahrner, C. & Pattiaratchi, C. B. 1994, The physical oceanography of Geographe Bay, Western Australia, Centre for Water Research, Perth.
Fischer, H. B., List, E. J., Koh, R. C. Y., Imberger, J. & Brooks, N. H. 1979, Mixing in Inland and
Coastal Waters, Academic Press, San Diego. Fisheries 2000, Aquaculture Plan for the Houtman Abrolhos Islands, Perth WA. Gersbach, G. H., Pattiaratchi, C.B., Ivey G.N. and Cresswell, G.R. 1999, 'Upwelling on the
south-west coast of Australia source of the Capes Current?' Continental Shelf Research, vol. 19, pp. 363-400.
Gill, A. E. & Schuman, E. H. 1974, 'The Generation of Long Shelf Waves by the Wind', J. Phys. Ocean., vol. 4, pp. 83-90.
Holloway, P. E. 1995, 'Leeuwin current observations on the Australian North West Shelf, May-June 1993', Deep Sea Research Part I: Oceanographic Research Papers, vol. 42, no. 3, pp. 285-305.
Hunter, J. R. & Hearn, C. J. 1987, 'Lateral and Vertical Variations in the wind driven circulaton in long shallow lakes', Journal of Geophysical Research, vol. 92, pp. 13106-13114.
Ingleman-Sundberg, C. 1976, The V.O.C. Ship Zeewijk: A report on the 1976 Survey of the Site., Maritime Archaeological Department, W.A. Museum, Fremantle.
Ingmanson, D. E. & Wallace, W. J. 1985, Oceanography: An Introduction, 3rd edn, Wadsworth Publishing Company, California.
References 80
Kirvan, A. P., (20/5/05), Latitude/Longitude, [Online], Available: http://www.ncgia.ucsb.edu/giscc/units/u014/u014.html [30/6/05].
Lass, H. U. & Mohrholz, V. 2005, 'On the fluctuations and vertical structure of the shelf circulation off Walvis Bay, Namibia', Continental Shelf Research, vol. 25, no. 12-13, pp. 1473-1497.
Mellor, G. 1996, Introduction to Physical Oceanography, American Institute of Physics, New York.
Pattiaratchi & Buchan 1991, 'Implications of long-term climate change for the Leeuwin Current', Journal of the Royal Society of Western Australia, vol. 74, pp. 133-140.
Pattiaratchi, C. B. 1993, 'Effect of sea breeze activity on nearshore and foreshore coastal processes', in Proceedings of the 11th Australiasian Conference on Coastal and Ocean Engineering, Townsville, Queensland, pp. 161-166.
Pattiaratchi, C. B. 2005, Physical Oceanography, M. Maslin, Perth, p. Lecture Material. Pattiaratchi, C. B. & Imberger, J. 1991, Physical Processes Along the Western Australian
Continental Shelf - a Review, Centre for Water Research, Perth. Pearce, A. & Griffiths, R. W. 1991, 'The mesoscale structure of the Leeuwin Current',
Journal of Geophysical Research, vol. C 96, pp. 16739-16757. Pearce, A. & Pattiaratchi, C. 1999, 'The Capes Current: a summer countercurrent flowing
past Cape Leeuwin and Cape Naturaliste, Western Australia', Continental Shelf Research, vol. 19, no. 3, pp. 401-420.
Pearce, A. F. 1997, 'The Leeuwin Current and the Houtman Abrolhos Islands.' in The Marine Flora and Fauna of the Houtman Abrolhos Islands, Western Australia, Proceedings of the 7th Int. Marine Biology Workshop, Western Australian Museum, pp. 11-46.
Penn, J. W. (ed)1999, State of the Fisheries Report 1998-1999, Perth WA. Phillips, B. F., Pearce, A. F. & Litchfield, R. J. 1991, 'The Leeuwin Current and larval
recruitment to the rock(spiny)lobster fishery off WA', Journal of the Royal Society of Western Australia, vol. 74, pp. 93-100.
Pond, S. & Pickard, G. L. 1983, Introductory Dynamical Oceanography, 2nd edn, Pergamon Press, UK.
Pugh, D. T. 1987, Tides, Surges and Mean Sea Level: A Handbook for Engineers and Scientists, Wiley and Sons, Chichester.
Robinson, A. R. 1964, 'Continental shelf waves and the response of sea level to weather systems', J. Geophys. Res., vol. 69, pp. 367-368.
Smith, R. L., Huyer, A., Godfrey, J. S. & Church, J. A. 1991, 'The Leeuwin Current Off Western Australia, 1986-1987', Journal of Physical Oceanography, vol. 21, no. 2, pp. 323-345.
Sukumaran, A. 1997, Circulation and Flushing Characteristics of the Easter Lagoon, Houtman Abrolhos Islands, Honours Thesis, UWA.
Wells, F. E. 1997, 'Introduction to the marine environment of the Houtman Abrolhos Islands, Western Australia', in The marine flora and fauna of the Houtman Abrolhos Islands Western Australia, WA Museum, Perth.
Wilson, B. R., Marsh, L.M. 1978, 'Coral reef communities at the Houtman Abrolhos, Western Australia, in a zone of bio-geographic overlap.' in Proceedings of the International Symposium on Marine Biogeography in the Southern Hemisphere., New Zealand Department of Scientific and Industrial Research Information Series, pp. 259-278.