Large Submarine Landslide Discovered on the Outer Shelf and Slope of the Great Barrier Reef; A New Local Mechanism
Capable of Generating Significant Tsunamis Along the Northeast Australian Coastline?
OS13E-1298
(1) NetSurvey Ltd, Banbury, UK (2) James Cook University, Townsville, Australia
(3)University of Sydney, Sydney, Austrlia (4) James Cook University, Cairns, Australia
N. George1,2*, J, M. Webster3, R. Beaman4, E, A. Abbey3, P, J.Davies3
ABSTRACT
1. INTRODUCTION & METHODOLOGY
3. FORMATION
We present the first submarine landslide ever discovered on the outer Great Barrier Reef
(GBR), Australia. The feature is 7 km wide and stretches out 5.5 km from the debris head
to the toe. Hard, rocky, limestone dominated, shelf edge material has fallen from 70 m to
depths of up to 225 m, resting on the fine sediment covered upper slope. The slide occurred at least 15,000 years ago, but is likely even older, occurring during a sea level high
stand. The area and thickness of the landslide were inputted into a Ward and Day (2003)
calculation and the results showed the potential to create a wave with a 1 m run up on
the Australian coastline up to 180m away.
Submarine landslides have not been well studied compared to their terrestrial counterparts. They can be on a massive scale and have been seen to destroy underwater infrastructure, dramatically change the geomorphology of a region and create devastating tsunamis (Masson et al. 2006). This paper describes a large scale landslide on the
outer Great Barrier Reef (GBR) (Fig.1). This is the first landslide ever discovered in the GBR. We will present the geomorphology of the submarine landslide (2), describe the
formation of the feature (3), suggest the age of the landslide (4) and then calculate the size of a tsunami created by the movement (5).
The slide debris is a contiguous, transparent facies. Although there are three
separate escarpments and associated debris, there is no evidence of settling
within the facies. This suggests that the mass movements were near simultaneous. These would be adjacent successive flows, where by the collapse of one escarpment leads to the weakening and fall of another (Mulder and Cochonat,
1996). The surface of the landslide is the current sea floor, but it lies on a previous
seafloor, as well as on a bedded deposit to the west. In places the landslide has
eroded the surface below (the rupture surface). This, along with the stepped features suggests there has been a rotational submarine landslide.
2. GEOMORPHOLOGY
All data was taken on the R/V Southern Surveyor between September and October 2007. Bathymetry and backscatter data was collected using the EM300 system. These
were processed in Caris HIPS/SIPS© and FM Geocoder© respectively and brought into Fledermaus© and ArcGIS© for analysis. Seven seismic lines were run parallel to the shelf
edge over the feature using the Topas PS-18 system. A seafloor sample was taken within the landslide debris and coralline algae on the sample was dated at ANSTO. Surface
complexity was calculated (Adron 2002) and analysed with backscatter to locate areas of hard rock and soft, fine sediment. Thickness and volume were determined using a
technique similar to that used by Völker (2010). A Ward and Day (2003) formula was used to estimate run up on the Australian coastline at the time of the mass movement.
Escarpments
10x vertical exaggeration
Fig.3
75
Fig.4A
75 m
125
0m
5m
12
2
125
Fig.7B
225
250
275
4
250
5
250
275
5
275
300
6
250
200 m
3
275
275
6
300
325
N
1
150
250
m
5k
225
275
225
3
100
250
3
4
200
175 m
75
125
175
200
225
150
175
100
1
2
200
225
250
150 m
10
Depth (m)
m
5k
100
100 m
125 m
Vertical reference is two way time (TWT)
125
175
2
Depth (m)
7 km
Fig.7A
Three escarpments cutting into the shelf edge are evident (Fig.3). The central escarpment covers the largest area, cutting back into the slope 400 m and is 1.2 km wide. It has
the steepest scarp face at a max 15o and a height of 19 m. The escarpment to the north
west is 1.6 km wide. The total width of the landslide is 7 km.
100
1
7
300
325
325
7
175
325
200
o
146oE
250
5.5
o
147 E
148 E
17oS
Cairns
A
km
22
Ba
N
Re
ef
19oS
Stepped Feature
Townsville
0
100 km
Fig.5
N
10x vertical exaggeration
The northwestern escarpment has a stepped formation downslope. A steep scarp
slope ranging between 7o and 10o is followed by a flat surface and another steep Figure 1. The submarine landslide. A collapse of the continental shelf edge, with coral reef limestones and unconsolidated sediments deposited onto the upper slope. At its widest point
slope. (Fig. 4B)
the landslide is 7 km and from the head to the toe it is 5.5 km. The inset map of the Queensland and Australia coastlines shows the location of the landslide 10 km from the modern
1
smooth, fine sediment
Viper Reef within the GBR. The position and the visual direction of the surrounding figure s are plotted on the main image.
A
100 m B
0.6
N
B
70 m
10
Figure 4. (A)Two soft fine sediment areas in the rocky debris (B)The stepped feature
terraces
shelf edge
soft, fine
sediment
landslide
debris extent
large, rocky debris
The length from the head to the toe of the slide debris is a maximum of 5.5 km. In
line with each escarpment is a V shaped debris deposit. These have merged together to form the final shape (Fig.5). The debris is spread out the greatest distance in line with the largest, centre escarpment. There is a general progression of
larger size debris with blocks up to 16,500 m2 and 17 m tall to finer sediment
below data resolution. This follows the general paradigm that submarine landslide evolve downslope (Masson et al. 2006).
10x vertical exaggeration
75
2
dredge
seismic line
hard rock
4
200 m
5
75
0
5 km
225 m
soft fine sediment
150 m
225 m
0
5 km
10x vertical exaggeration
175 m
10x vertical exaggeration
Figure 7. (A) Side on to the landslide looking along the shelf edge showing the
coral terrace at 70 m curved around the escarpments. (B) Coral reef features
growing on the landslide debris
The landslide has dramatically affected local coral reef development. Ancient
coral reef terraces likely to have formed in the last sea level transgression since
15,000ya (Beaman et al. 2008) which lie parallel along the shelf edge, curve
around the escarpment features at 80 m. In addition there is evidence that coral
reefs structures are present on the surface of the landslide debris. Features stand
taller that they do wide suggesting growth in situ rather than deposition.
3
Depth (m)
250
m
150 m
175 m
150
100 m
125 m
17 m
175 m
200 m
B
N
125 m
150 m
Slide Debris
75 m
A
100 m
0m
10x vertical exaggeration
10x vertical exaggeration
1
100 m
125 m
N
N
150 m
hard, rocky debris
A
16 m
N
125 m
N
2
The dated dredge samples show that the landslide occurred at least 15,000 years
ago. No other landslide has occurred since. The sea level on the GBR from transgression to regression has a range between current level to -120 m. Due to the
size and shape of the feature the landslide appears to be submarine rather than
sub aerial or partially sub aerial. The debris has travelled over a relatively large distance, it forms a V shape with a nose and has a progression of debris sorted from
large blocks near the source to fine sediment at the toe. 15,000 years ago the sea
level was around 100 m below the present state. Therefore, to have formed at
least 15,000 ya and at time when the current landlisde area was submarine it
would have occured up to 90,000 ya during the previous trangression.
200 m
at
Gre
rri
er
150 m
175 m
18oS
Two smooth, fine, sediment mounds in deeper indents are downslope and adjacent to the escarpments to the southeast. These sediment areas are surrounded
by hard, rocky debris which appears to trap the sediment (Fig.4A).
1.1 km
350
4. TIMING
Figure 3. Close up of the 3 escarpment features cutting into the shelf edge
Sediment Mounds
km
325
B
Figure 6. (A) The seven seismic images looking upslope. (B) Interpretation of the
seismic by identifying facies and areas of erosion.
5m
Fig.4A
350
N
225
6
6. CONCLUSIONS
225
m
20
0
m
Figure 2. (A) A plan view of the landslide coloured by depth. Boundaries of surface classification are labeled. (B) A plan view of the landslide with the surface coloured by the results of
100 m
bathymetry and surface complexity analysis. The location of the dredge and 6 of the 7 seismic lines are also shown. The 7th lines is parallel with line 6 downslope on undisturbed slope - A submarine landslide measuring 7 km wide and 5.5 km long is found on the
125 m
shelf edge of the GBR.
Thickness
5. TSUNAMI
- Coral reef limestone and unconsolidated sediment from the shelf edge is depos150 m
Gain 6- 28 m
10x vertical exaggeration
m
Gain 0.5 - 5.9 m
100 5 m
5 m3 of material was lost whilst 5.9 x 105 m3 was gained. This is
ited
onto the upper slope
The
volume
calculations
show
that
2.52
x
10
2
1
N
Loss 0.5 - 12 m
m
0
175 m
5
1
Loss 12 - 27 m
- The landslide is a culmination of three adjacent near simultaneous movements
comparable to the smaller submarine landslides measured by McAdoo et al. (2000) on the American continen5m
17
hard rock
- The feature is an example of a rotational slide with stepped formations and erotal shelf edge. Thickness and total area of the debris were also measured using the volume calculation. The
area covered by the landslide was 18.7 km² with a maximum thickness of 26.6 m and average of 6 m. These
sion of the rupture surface
200 m
1
3
variables were entered into the Ward and Day (2003) formula. Adopting the assumption that the slide oc- Due to the size and the timing of the landslide a tsunami could have been procurred at a high stand similar to current sea level a tsunami with a run up of 1 m on the Australian coastline up
duced with a run up of 1m on the Australian coastline within a distance of 180m.
to 180 km away could have been created. The ability of the GBR to anuate a wave has been described (Baba
Acknowledgements: We thank the crew of the R/V Southern Surveyor for their skilled operation
et al. 2008) and would warrant further investigation in relation to this event. However, tsunami deposits have
during data collection and processing. Also, thanks to NetSurvey for supporting this presentation.
previously been discovered inside the Great Barrier Reef (Bryant and Nott, 2001). No evidence as yet has been
supplied as to the source of these tsunamis. Landslides such as this one should be considered.
*Presenter: Nicholas George
Figure
8.
Pre
slide
model
of
the
shelf
225 m
0
5 km
soft fine sediment
2
Masson, D.G., Harbitz, C.B., Wynn, R.B., Pedersen, G., and Lovholt, F., (2006)Philosophical Transactions of the
References
NetSurvey, 2A Banbury Office Village,
edge and upper slope. The model as- Ardron, J.A., (2002) Marine Geography: GIS for the Oceans and Seas
Royal Society of London: Series A, Mathematical and Physical Sciences 364: 2009-2039
Figure
9.
Plan
view
showing
areas
Noral
Way, Banbury, OXON, X16 2SB.
Mulder,
T.,
and
Cochonat,
P.,
(1996)
Journal
of
Sedimentary
Research
66:
43-57
Figure 5. Surface showing debris mound from the shelf edge on to the upper slope. sumes a consistent depth of terraces Baba, T., Mleczko, R., Burbidge, D., Cummins, P. R,Thio, H. K., (2008). Pure and
McAdoo, B.G., Pratson, L.F., and Orange, D.L., (2000) Marine Geology 169: 103-136
Applied Geophysics165: 2003-2018
of volume loss and gain and the Tel: +44 1295 817740
Facing upslope, the debris contrasts to the fine sediment of upper slope.
Völker, D.J., (2010) Geo-Marine Letters 30: 541-547
Beaman, R.J., Weber, J.M., and Wust, R.A.J., (2008) Marine Geology 247:17-34
and contours parallel to the slope
thickness of the landslide debris
[email protected]
www.netsurvey.co.uk
N
Bryant, E.A., and Nott, J.,(2001). Natural Hazards 24: 231-249
Ward, S.N., and Day, S., (2003) Geophysics Journal International 154: 891-902
N