Content uploaded by Graeme Hubbert
Author content
All content in this area was uploaded by Graeme Hubbert on Jul 04, 2014
Content may be subject to copyright.
1
Comparisons between topographically surveyed debris lines and modelled
inundation levels from severe tropical cyclones Vance and Chris, and their
geomorphic impact on the sand coast
Jonathan Nott, James Cook University, Cairns, Queensland
Graeme Hubbert, Global Environmental Modelling Systems, Warrandyte, Victoria
Abstract
Comparisons of topographically surveyed debris lines and modelled
inundation levels from two category 5 cyclones in Western Australia show close
agreement. The largest differences between model predictions and surveyed debris
occurred where wave run-up did not overtop the frontal sand dunes and was registered
in the coastal landscape. Excellent agreement was found where dune overtopping
occurred, as the debris largely represented the height of the storm tide.
The post-event field surveys also showed a clear relationship between the
height of the inundation, the height of the coastal sand dunes, and the extent of
erosion of those dunes. Dunes were eroded vertically, and completely removed, where
they were overtopped by the marine inundation (including wave run-up). Horizontal
erosion or scarping and landward retreat of the dune occurred where the inundation
was well below the dune crest. These observations are significant for policies on set-
back distances and minimum habitable floor levels of coastal developments. The
results show that these standards need to be above the level of the marine inundation,
not the storm tide or storm ‘still water’ level, otherwise, buildings erected on filled
land behind frontal dunes could be undermined through wave attack and experience
structural failure.
Introduction
Surveys of the physical aspects of landfalling tropical cyclones are important
for at least three reasons;
1. To test the veracity of numerical surge models by comparison with topographically
surveyed debris lines that mark the level of marine inundation, along with records
from tide gauges,
2. To determine the relative contributions of surge, tide, wave action and wave run-up
to the total marine inundation, and,
3. To elucidate the nature and impact of marine inundations on sandy coasts.
Post-cyclone surveys are common in the United States, where much emphasis
is placed on examining the impacts of tropical cyclones (hurricanes) on coasts. This is
because the U.S. Flood Disaster Protection Act of 1973 requires the Federal
Emergency Management Authority (FEMA) to map areas of coastal land susceptible
to wave attack during hurricanes. This is not the case in Australia. With the exception
of Hopley (1974) few, if any, detailed studies of the nature and impact on the coast of
cyclone induced marine inundations have occurred in this country. As a consequence,
little is known of the nature and extent of erosion along Australian sandy coasts
during intense tropical cyclones, and the relative contributions of the various
components of the total marine inundation. This paper aims to partly redress this
situation by presenting the results of topographic surveys of the heights, and
2
descriptions of the impacts, of marine inundations on the sandy coast of Western
Australia during landfall of Tropical Cyclones Vance, in March 1999, and Chris, in
February, 2002. The relationship between the field evidence of marine inundation
heights, numerical storm surge simulations, the extent and nature of erosion, and the
implications of these results for assessing risks associated with coastal development in
tropical cyclone prone areas, are discussed.
TC Vance and Chris characteristics
TC Vance had an estimated central pressure of 910 hPa as it approached and
entered Exmouth Gulf on the 22nd March 1999 (Fig. 1). TC Vance crossed the coast
with a translational velocity of approximately 25 km/h and radius of maximum winds
of 30 km/h. TC Vance was the most intense cyclone, historically, to cross the
Australian coast. It generated winds in excess of 300 km/hr around its eye and
produced the strongest ever-recorded wind gust in Australia of 267 km/hr (Bureau of
Meteorology, 2000). A storm surge 3.6 m above Australian Height Datum (AHD) was
recorded at Exmouth and a surge of 3.3 m was recorded at Onslow during the cyclone
(Bureau of Meteorology, 2000). The waves during this event were not measured, but
they must have been of considerable size inside Exmouth marina as they hurled large
boats over 2 m high jetties to impale and sink other moored boats. Five metre high
rock walls surround the marina on its ocean side. At Onslow a large fishing boat was
lifted onto the levy surrounding the Onslow salt ponds.
Tropical cyclone Chris, category 5, with central pressure 915 hPa, crossed the
coast of Western Australia 160 km northeast of Port Hedland between Pardoo and
Wallal station homesteads on the 6th February, 2002 (Fig. 2). Chris had a radius of
maximum winds of 30 km and translational velocity of 20 km/hr as it crossed the
coast. The zone of maximum winds came within approximately 10 km of the Pardoo
Roadhouse which incurred significant damage. All persons sheltering within the
roadhouse were physically unharmed.
TC Vance – comparisons between the survey and model data
A field survey of the impact of TC Vance on the coastal landscape was
undertaken within a few days of the cyclone making landfall. Topographic surveys of
debris lines were undertaken using an electronic theodolite and electronic distance
measurer and the heights of total marine inundation relative to the tide at the time of
landfall of TC Vance were determined. The height of the marine inundation was
determined from the position of debris lines which were composed of grasses eroded
from the back of the beach and foredunes, seaweed, shells and coral clasts, and
marine fauna such as juvenile sea turtles and fish. These debris lines normally occur
at a consistent height along a stretch of coast following a storm induced marine
inundation. They usually, but not always, represent the total extent of the marine
inundation including wave run-up. Where the marine inundation breaches or overtops
dunes, however, the surging seawater can often flow within the confines of a dune
swale. Here, debris lines provide a more accurate reflection of the height of the storm
tide or storm ‘still water’ level because the dune swales are more protected from the
wave energy.
Surveys were undertaken at 7 locations including Onslow, Locker Point near
Tubridgi Point on the eastern side of the entrance to Exmouth Gulf, Giralia Landing
and Wanns Beach at the head of Exmouth Gulf, Pebbly Beach near Exmouth and
3
within Exmouth Marina. The total inundation heights during TC Vance, relative to the
event tide, are presented in Table 1 and Figure 1.
The surveyed data, and records of the surge from tide gauges at Onslow and
Exmouth show close agreement. The surveyed inundation level at Onslow was 4.3 m
above the event tide (8.30 am 22/3/99) and the surge level was measured at 4 m (3.3
m AHD) on the tide gauge. The surveyed inundation level at Pebbly Beach, close to
Exmouth Marina, was 4.2 m above the event tide and the tide gauge registered a 3.6
m surge at Exmouth. The differences of 0.3 m at Onslow and 0.6 m at Exmouth
between surveyed and tide gauge data are very likely to be due to wave action and
wave run-up, respectively. The surveyed level at Onslow probably did not incorporate
wave run-up as the area is fronted by a low gradient foreshore and the survey level
was taken against a toilet block in the Onslow caravan park where water stains, fresh
chip marks on the bricks and interview with an eyewitness confirmed the level of
inundation here with an accuracy of less than 10 cm. The difference between survey
and tide gauge data at Pebbly Beach was likely due to wave run-up as the surveyed
debris line was at the rear of a moderately sloping (~ 50 ) beach where the inundation
did not overtop the foredune.
Surge model runs, using the GCOM2D surge model, were undertaken for
comparison with the survey data at both Onslow and near Tubridgi Point. The
GCOM2D model solves a set of mathematical equations over an equally spaced grid
to determine water depths and currents over varying bathymetry and inundated
topography and incorporates wind stresses and atmospheric pressure gradients acting
on the ocean surface, and friction on the ocean floor. The GCOM2D modelled surge
at Onslow was 4 m (3.3 m AHD), which corresponds exactly with the tide gauge
measured surge and differs by only 0.3 m from the survey data. The modelled surge
for Tubridgi Point was 7. 3 m (6.6 m AHD), which also corresponds precisely with
the survey data. The zone of maximum winds during TC Vance struck the coast near
Tubridgi Point. Here, debris trapped in a small tree on the crest of the foredune was
7.3 m above the event tide. Wave run-up does not appear to have played a role in the
elevation of the debris at this site because the debris, in the tree, is above the dune
crest; if wave run-up had played a role, then the debris would be most likely on the
dune itself. To be lodged in the tree, suggests the debris was emplaced by either the
storm tide and / or wave action on top of the storm tide. Hence, the storm tide appears
to have been able to overtop the 6-7 m high dunes.
TC Vance – geomorphic impacts of the storm tide
TC Vance had a dramatic impact on the sandy coast near Tubridgi Point. In
most locations along this section of coast, the first two rows of dunes were completely
eroded and the sand transported away, presumably offshore, and the third or most
inland dune row was eroded to form a steep scarp. Precisely where the zone of
maximum winds struck, however, all three rows of 6- 7 m high dunes (Fig. 3) were
removed and the sand transported inland as an extensive splay approximately 400 -
500 wide and 200 - 250 m deep (extending inland from the position of the former first
dune row) (Figs. 4 and 5). This sand splay decreased in thickness from 1.5 m at the
rear of the position of the former third row of dunes, to 0.75 m thick at its most inland
extent. The splay terminates abruptly at a salt marsh where it is marked by a steep
fronted (~300 angle) toe slope. Sediments within the splay were deposited as steep
(~300) tabular cross-beds (Fig. 6). Medium to coarse-grained sand occurs at the base
of the unit, along with clasts of coral and shells, and grades upwards into medium to
fine-grained sand. Lithic gravel measuring up to 25 cm across, were deposited in an
4
imbricated position (imbrication is the depositional alignment of gravel or clasts
where the particles lie against each other in a shingle like pattern) within the sand
sheet inland of the position of the former dunes. These gravel dip towards the ocean
showing they were deposited by an onshore flow. The marine inundation also
generated scour pits within the sand splay. These pits measure up to 3 m wide and 1 m
deep and occur on the lee (inland) side of tree stumps. The onshore flow was also able
to carve channels 1 m deep and several metres wide over distances of several tens of
metres (Fig. 7). These channels were carved by the onshore flow of water rather than
water draining from the land as demonstrated by the imbricated gravels in the sand
floor of these channels. All of the vegetation including clumps of spinifex grasses,
which dominated the low lying salt marsh plain behind the former dunes, small trees
(up to 3 m high), and bushes, were stripped from the land surface up to 300 m inland
by this flow. Only occasionally were the dead stumps of trees remaining, such as on
the crest of an eroded knoll of the former foredune, or inland of the former dunes.
Debris in these dead stumps show that flow depths were still at least 2 m, 150-200 m
inland, and flow velocities also high as gravel clasts were embedded within the debris
at an elevation of 1.5 m above the present ground surface (Fig. 7).
The extent of the sand splay, the tabular cross-beds, scour pits, and imbricated
gravel suggest that the marine inundation generated by TC Vance near Tubridgi Point,
struck the coast with considerable force and moved inland as a reasonably high
velocity bore. The presence of similar scour pits and erosional features in beach and
dune sands have been determined empirically to occur when flow velocities exceed 4
– 5 m / sec, during the passage of tsunamis across sand barriers. Also, the fact that the
flow was transporting lithic gravel clasts as suspended, or at least saltating
(bouncing), load 1.5 m high over 200 m inland shows that the flow velocity must have
been high. Hence, the flow depth was between 6 - 7 m high (above the event tide)
across the beach here and at least 2 m high 200 m inland. Such flows and the impacts
upon the coast are more reminiscent of tsunamis than that normally ascribed to storm
surge.
TC Chris field survey
An initial field reconnaissance of the area impacted by TC Chris was
undertaken by helicopter with staff of the Bureau of Meteorology and Fire and
Emergency Services Western Australia. Detailed topographic surveys were then
undertaken where the storm surge impact appeared to have been greatest. The results
of these surveys are presented in Table 2 and Figure 8. As with TC Vance,
topographic surveys were undertaken using an electronic theodolite and electronic
distance measurer. At only one location, site 3 (Fig. 3b) was debris surveyed in a hind
dune swale. This debris was emplaced following inundation of the foredune. It is this
type of evidence that provides probably the most accurate representation of the height
of the storm tide. Debris lines seaward of foredunes often reflect the extent of wave
run-up. The surveyed height of the debris line at Site 3 therefore, represents the most
reliable indicator of the level of the storm tide or still water level during the event
The height of the debris lines varied between survey sites. In Figure 8, these
heights are presented relative to the tide level at the time of landfall of TC Chris. The
debris at survey site 1 was 5.4 m above the level of the tide at the time of cyclone
crossing, 4.1 m at site 2 and 4.2 m at site 3. It is likely that wave run-up has
contributed to these heights at sites 1 and 2. The extent of the wave run-up
5
contribution can be estimated by making comparisons between the height of the
surveyed debris lines and the predicted surge heights from numerical models.
TC Chris - comparisons between field survey data and numerically
modelled estimates of the storm tide.
The storm surge and offshore significant wave height during TC Chris were
modelled using the GEMS Coastal Ocean Model (GCOM2D) (Hubbert et al, 1990,
Hubbert and McInnes, 1999) and the SWAN shallow water wave model. The highest
storm surge predicted by the model for TC Chris was 3.7 m (Fig. 9). This figure refers
only to storm surge and does not take into account the effects of the tide which is a
non-linear addition to the surge. Wave set-up also adds to the height of the water
column and is often found to be equivalent to approximately 10% of offshore
significant wave height (average of highest one-third of waves). The SWAN shallow
water wave model estimated the highest offshore significant wave height at 2.8 m.
Hence, the highest numerically modelled storm surge plus wave set-up level was less
than 4 m.
The modelled and surveyed heights of marine inundation agree well. The
highest predicted (modelled) surge was 30 km east of the location of cyclone crossing
(eye). This was approximately 2 km east of survey site 3. No survey was undertaken
here because of difficulties in gaining access. The location of predicted maximum
surge was approximately where the eastern or outer limit of the zone of maximum
winds occurred as suggested by radar imagery; the radius of maximum winds was
approximately 30 km (A. Burton, BOM, pers. comm). The next highest predicted
surge was only 10 km east of the location of cyclone landfall and this site (site 1
Table 2) corresponded to the highest surveyed debris height. Sites 2 and 3 both
showed lower predicted and surveyed surge heights. Interestingly, both surveyed
debris lines and the modelled surge heights showed that the surge and marine
inundation did not increase progressively from the point of cyclone crossing to the
location of highest surge.
The surveyed debris lines lie at slightly higher elevations than the level of the
predicted or numerically modelled estimates of surge height plus wave set-up. Table 2
shows these differences vary from 1.52 m to 0.82 m. The lowest difference occurred
at survey site 3 which is regarded, based upon the field setting, to be the most reliable
indicator of the actual storm surge plus wave set-up height. The most likely reasons
for the differences between the modelled and surveyed heights are errors in the
coastal bathymetry and the interpretation of the surveyed debris lines. Unlike the
modelling of TC Vance, where very accurate coastal bathymetry was available, the
bathymetry used in modelling TC Chris was quite coarse and may not have
represented the coastal bathymetry very accurately. It is almost certainly the case that
a substantial part of the difference in heights at sites 1 and 2 is because the debris
lines here are also a function of wave run-up. Wave run-up is likely to be less
significant at site 3 for the reasons stated earlier, but it is possible that it has
contributed somewhat to the elevation of the debris line here. Another possible source
of error is the lack of a precise datum to which the debris lines were surveyed. Datum
was taken as the level of the tide at the time of surveying and this requires a minor
calculation to relate this to Australian Height Datum (AHD). Normally, if surveying
close to a standard tide port, this method should be accurate to less than 20 cm.
However, the survey sites lie between standard tide gauge ports (Port Hedland and
6
Broome) and the time difference in tides between the survey sites and Port Hedland
has not been determined by the National Tidal Facility, Flinders University. Hence, it
is reasonable to expect a possible error margin of several tens of centimetres (and
possibly as high as 50 cm) in the estimation of the datum height.
TC Chris - geomorphic impacts of the storm tide
Based upon the extent of erosion of foredunes, the maximum surge appears to
have occurred along a 500 – 700 m stretch of coast located adjacent to survey site 3.
As discussed, this section of coast lies close to the modelled estimate of the location
of likely maximum surge, and also the likely location of maximum winds from
interpretation of the cyclone crossing point as shown by BOM radar imagery. Dune
erosion appears to have been greater here than elsewhere along the coast. At each of
the other survey locations (sites 1 & 2) dune erosion occurred in the form of scarping
i.e. there was erosion of the seaward face of the dunes resulting in a more vertical
profile and landward recession of the dune face. At survey site 3 the frontal dunes
were overtopped by the storm tide and wave run-up, and there has been vertical
erosion i.e. reduction in the height of the dune, and in some instances complete
removal of the frontal dune (Fig. 10). Here also, dune swales behind the frontal dune
have channelled the storm tide and this has probably exacerbated the erosion of the
second and third rows of dunes (Figs 11 & 12). The dune swales here appear to have
been colonised previously by mangroves for dead mangrove stumps occur in the
swale between the first and second rows of dunes. It is unlikely that these mangroves
were killed during TC Chris for the dead stumps appear to have been in their present
form for at least several years.
Dead cows littered the beach along this entire section of coast. At all locations,
except survey site 3, these cows lay seaward of the first row of dunes. At survey site 3
some of the cows were deposited amongst both the second and third rows of dunes. In
one instance a cow was deposited in a narrow opening through the third row of dunes
(Fig. 12). This opening appears to have been scoured, or at least enlarged, during the
cyclone storm tide event, for an oblate shaped depression has been eroded into the
sand floor at the mouth of the opening. Mud was eroded and transported from the
swale between the first and second rows of dunes and deposited at the head of this
opening well inland of the third row of dunes. The storm tide in this instance
penetrated approximately 150 m inland beyond the position of the first row of dunes.
Discussion
Tropical cyclones Vance and Chris displayed both differences and similarities
in their respective impacts on the coast. The major difference between the two was the
much greater extent of erosion caused by Vance, resulting in deposition of the sand
splay and development of erosional features (scour pits). Presumably, this more
extensive impact is related to the size of the surge (>6.5 m) generated by Vance which
was almost double that generated by TC Chris. Local bathymetry and coastal
configuration must have played a substantial role in producing such a dramatic
difference in surge heights between the two cyclones for both systems were of similar
intensity, spatial extent (radius of maximum winds) and had similar forward
velocities.
7
The most striking similarity between the two inundations was the nature of
erosion to the sandy coast. Both events showed a relationship between the extent and
type of erosion, the height of inundation and height of coastal sand dunes. Extensive
erosion occurred where the inundation was able to overtop the sand dunes. These
dunes experienced vertical erosion resulting in virtually complete removal of that
dune. In the case of TC Vance, three rows of dunes near Tubridgi Point standing
approximately 6-7 m high were overtopped by the 6.6 m AHD inundation and these
dunes were removed. Elsewhere, along this coast, where the inundation did not
overtop the dunes, those dunes were scarped and experienced horizontal erosion or
retreat landwards. In these instances the dunes halted the inland penetration of the
storm inundation. The same was true of TC Chris. The dunes along the impacted
section of coast east of Port Hedland during TC Chris, were generally lower in height
than those along the coast impacted by TC Vance. And the marine inundation
generated by TC Chris appears to have been lower than that produced by TC Vance.
But where the inundation from TC Chris was able to overtop the dunes, those dunes
were largely removed and the inundation surged inland to erode the second and, to a
lesser extent, the third row of dunes. Where the inundation from TC Chris was
considerable lower than the dunes, those dunes experienced horizontal or landward
retreat.
The nature of erosion to the sandy coasts impacted by TCs Vance and Chris is
very similar to the erosional impacts of hurricanes in the United States. Based upon
observations and measurements of numerous hurricane impacts on US sandy coasts,
FEMA has noted that where the dune reservoir (area of dune above the storm ‘still
water’ level) is greater than 540 sq feet (~20 sq m. per metre length of beach) that
dune is likely to experience horizontal erosion only (FEMA, 2003). Where the dune
reservoir is less than 540 sq feet that dune is likely to erode vertically and be largely
removed. In the latter situation, FEMA note that the marine inundation will penetrate
inland beyond the former dune position. The USGS has also made similar
observations (USGS, 2003). While dune reservoirs have not been calculated for the
former dunes along the WA coast impacted by TCs Vance and Chris, the resulting
styles of erosion and relationship to inundation height were essentially the same.
These observations of the impacts of TCs Vance and Chris have important
implications for coastal planning in cyclone prone regions of Australia, particularly in
Queensland. Legislation and policies relating to coastal management and protection in
Queensland recommend that the elevation of the storm tide is the most appropriate
level in setting safe set-back distances and minimum habitable floor levels for coastal
developments in cyclone prone regions (Fig. 13). No acknowledgement of the impact
of erosion of dunes and inland penetration of the marine inundation, including high
velocity currents and waves is made in these policies. As shown by the inundations
during TCs Vance and Chris, once the frontal dune is eroded, substantial erosion of
hind dune areas occurs. In Queensland, low lying hind dunes areas are often filled
with sediment to raise the elevation of the land to the level of the 1% AEP storm tide
to allow construction of dwellings at the recommended minimum habitable floor
level. In an inundation that overtops the frontal dunes it is very likely that any infilled
land could be eroded, following removal of that foredune, causing potential serious
structural impacts to buildings. Many of the frontal dunes in Nth Queensland are low
(2-3 m AHD), so it would not take a substantial inundation for them to be overtopped.
But Queensland has not experienced a severe marine inundation since 1918 when a
8
category 5 cyclone made landfall at Innisfail. Hence, there is a lack of experience
with, and observations of, severe cyclones and their impact upon sandy coasts in
Queensland. This has led to a certain level of naivety in formulating policies for
coastal development. This does not appear to be the case in Western Australia where
set back distances for new developments consider the total ocean inundation
occurring during a category 5 cyclone (WA Planning Commission, 2003).
Comparisons between the surveyed and modelled inundation heights of TCs
Vance and Chris also provide useful data on the extent of wave run-up during these
types of events. Registered wave run-up values in the coastal landscape were greater
for TC Chris (0.8-1.5 m) because the inundation was not able to overtop the dunes in
many locations. Because the inundation from TC Vance easily overtopped dunes, the
debris lines in the landscape tended to record the storm tide height rather than wave
run-up. Near Pebbly Beach, on the eastern side of Exmouth Gulf, the difference
between the tide gauge measured storm tide and the surveyed debris was 0.6 m. This
value gives a measure of wave run-up during TC Vance where the surge was smaller
than that occurring near Tubridgi Point. Wave run-up therefore varied between
approximately 23 – 43 % of storm surge height for TC Chris and by approximately 10
% during TC Vance.
Determining wave run-up values relative to surge levels, and offshore wave
heights, provides a calibration for models estimating the intensity of prehistoric
tropical cyclones. Nott and Hayne (2001) and Nott (2003) used wave run-up values
between 40-70% of surge to determine prehistoric tropical cyclone intensities in
Queensland. These values were determined for much rougher and more permeable
surfaces (gravel and coral shingle beaches) than sandy beaches, where the latter
would be expected to produce much larger run-up values compared to the former.
Wave run-up values on the sandy beaches impacted by TCs Vance and Chris
therefore are considerably smaller than those adopted by Nott and Hayne (2001) and
Nott (2003). Hence, in comparison, it would appear that these authors have used
conservative values of wave run-up in their studies. This suggests that Nott and
Haynes estimates of prehistoric cyclone intensities in Queensland are also probably
conservative. Further post cyclone surveys are needed, however, before such
conclusions can be fully substantiated.
Conclusion
Close agreement was found between the GCOM2D storm surge model and the
heights of surveyed debris marking the level of marine inundation during tropical
cyclones Vance and Chris in Western Australia. The largest differences between
model estimates and survey debris occurred at those locations where the inundation
(wave run-up) did not overtop the frontal sand dunes. Excellent agreement between
the modelled and surveyed debris heights was found where dune overtopping did
occur, and the debris largely represented the height of the storm tide.
The impact of the marine inundations from both cyclones on the sandy coasts
of Western Australia is very similar to that observed in the United States following
hurricane impacts. Dunes were eroded vertically and completely removed where those
dunes were overtopped by the inundation. Where the inundation was well below the
crest of the dune, horizontal erosion or scarping and landward retreat of the dune
occurred. These observations are significant for policies governing set-back distances
9
and minimum habitable floor levels for coastal developments. They show that
aligning the minimum habitable floor level and set-back distance to the storm tide or
storm ‘still water’ level is inappropriate, for substantial erosion to both fore and hind
dune areas can occur when the foredune is overtopped by a marine inundation. The
results of this study suggest that a much safer minimum habitable floor level is one
that coincides with the level of total marine inundation where wave run-up and wave
action are considered, along with the likely extent of erosion during these extreme
events.
References
Bureau of Meteorology, 2000. Report on Tropical Cyclone Vance, March , 1999
FEMA, “Guidelines and Specifications for Flood Hazard Mapping Partners,
Appendix D: Guidance for Coastal Flooding Analyses and Mapping” (2002)
U.S. Federal Emergency Management Agency
(http://www.fema.gov/fhm/en_cfhtr.shtm – accessed June 17th, 2003); 2002.
Hopley, D. 1974. Coastal changes produced by tropical cyclone Althea in
Queensland; December 1971. Australian Geographer, 12, 445-456.
Hubbert, G.D., Leslie, L.M. and Manton, M.J. 1990. A storm surge model for the
Australian region. Quart. J. R. Met. Soc., 116, 1005-1020.
Hubbert, G. & McInnes, K. 1999. A storm surge inundation model for coastal
planning and impact studies. Journ. Coast. Res. 15, 168-185.
Nott, J. & Hayne, M., 2001. High frequency of ‘super-cyclones’ along the Great
Barrier Reef over the past 5,000 years. Nature, 413, 508-512.
Nott, J., “Intensity of prehistoric tropical cyclones” 2003. Journal of Geophysical
Research, 108 D7 p. 4212.
USGS, Mapping coastal change hazards. Coastal change hazard scale 2003.
United States Geological Survey
http://coastal.er.usgs.gov/hurricanes/mappingchange/scale.html accessed
July 2003.
Western Australian Planning Commission, 2003. State Coastal Planning Policy,
Statement of Planning Policy No. 26. WA State Law Publisher, Perth.
10
Figure Captions
Fig. 1 Location map of TC Vance coastal crossing and surveyed debris heights (above
event tide) near and within Exmouth Gulf
Fig. 2 Location map of TC Chris coastal crossing
Fig. 3 Cross-section of remnants of three parallel rows of dunes near Tubridgi Point
where the zone of maximum winds struck the coast during TC Vance.
Fig. 4 Oblique aerial photo of area shown in Fig. 3 near Tubridgi Point. Note virtual
complete removal of all three rows of dunes and extent inland of sand splay and
complete removal of al dune and hind dune vegetation.
Fig. 5. Remnant of dune near Tubridgi Point with debris in tree on crest of dune. Note
depth of marine inundation. Dune here is located at rear of beach, so actual depth of
marine inundation here is several metres greater than height of debris above
immediate ground level.
Fig. 6. Tabular cross-beds in toe of sand splay/sheet 300 m inland. Cross-beds such as
these indicate. Such sedimentary structures are typical of water lain deposits by a
unidirectional flow of reasonably high velocity.
Fig. 7. Scour pit on lee (inland) side of tree remnant. Pit measures approximately 3 m
long by 2 m wide and 1 m deep. Note also lithic clast and debris in tree indicating
depth velocity of flow here (180 m inland).
Fig. 8 Cross sections of beach and dunes impacted and surveyed debris heights
generated by TC Chris
Fig. 9 Model surge heights generated by TC Chris for different locations along
affected coast
Fig. 10 Eroded foredunes at survey site 3. Marine inundation overtopped foredunes
here dunes resulting in their vertical erosion and removal.
Fig. 11 Erosion of second row of dunes by waters surging through interdune swale.
Fig. 12 Erosion of third row of dunes. Note dead cow wedged in opening scoured by
storm surge/tide.
Fig. 13. Schematic model of relationship between storm tide and marine inundation
height and erosion of foredunes. Note that once marine inundation overtops foredune
that dune erodes vertically resulting in its complete removal. This allows waves and
high velocity currents to penetrate inland to areas of land often filled for urban and
tourist developments.
11
Table 1 Modelled and surveyed debris line heights between Onslow to Exmouth
for TC Vance
Location Est. time
inundation Surveyed
inundation
height
(m AET)
Surveyed
inundation
height
(m AHD)
Tide
gauge
residual
(m AHD)
Surge
model plus
set-up
(m AHD)
Comments
Onslow 8.30 am 4.3 3.6 3.3 3.3 Likely storm
‘still water’
level
Urala 9.00 am 5.1 4.4 N/A N/A Debris line
2 km east
Locker Pt 9.00 am 7.3 6.6 N/A N/A Debris line –
dunes
knocked out
500 m east
of Locker
Pt
9.00 am 4.6 3.9 N/A N/A Debris line
3rd row
dunes – 1st
& 2 rows
removed
Locker Pt 9.00 am 7.3 6.6 N/A 6.6 3 dune rows
removed
Giralia
Landing Midday 3.7 4.3 N/A N/A Debris line
Wanns
Beach Midday 3.6 4.2 N/A N/A Debris line
Pebbly
Beach 10 am 4.2 3.9 3.6 3.6 Debris line
AET = above event tide, AHD = Australian height datum
Table 2 Modelled and surveyed heights of marine inundation during TC Chris
Location Approx.
distance
from
crossing
(km)
Lat.
(degrees
south)
Long.
(degrees
east)
Modeled
surge and
wave set-up
(m AET)
Surveyed
height of
debris
(m AET)
Diff. in
height
(m)
Survey
site 1 10 19.92 120.178 3.88 5.4 1.52
Survey
site 2 20 19.9 120.236 3.18 4.1 0.92
Survey
site 3 27.5 19.89 120.273 3.38 4.2 0.82
Predicted
site of
max.
surge
30 19.87 120.30 3.98 N/A N/A
AET = above event tide
12
13
14
D
e
b
r
i
s
li
ne
V.E. = 10x
Saltmarsh
CD
0
2
4
6
7
M
etres (AHD)
0150
Distance from shore (metres)
Tr
a
n
sect
15
16
16
17
Survey site 1
A
HD
Tide level
Inundation level
010203040506070
0
1
2
3
4
5
6
7
8
9
-1
-2
Distance
(
m
)
Height
above
tide (m
)
Survey site 2
A
HD
Tide level
Inundation level
010203040506070
0
1
2
3
4
5
6
7
8
9
-1
-2
Distance
(
m
)
Height
above
tide (m)
Survey site 3
A
HD
Tide level
0 10203040506070
0
1
2
3
4
5
6
7
8
9
-1
-2
Distance
(
m
)
80 90 100
Inundation Height
above
tide (m
)
18
3.9
3.5
3.1
2.7
2.3
1.9
1.5
1.1
0.7
0.3
-0.1
-0.5012345678
Storm surge (m)
Time (hrs)
Figure 7
Results of numerical model simulations
19
20
Second row of dunes
21
