Submerged Landscapes of Port Phillip Bay – Publications

The submerged landscapes project was initially planned as a 3 month research sabbatical, but the surprising success of analysing old data, and the enthusiasm and interest from academics, government agencies and the local Traditional Owner groups meant that work on the project continued for over 18 months. In order to maintain interest, and keep people updated on what was a ground-breaking project in Australia, I wrote a number of short updates for the Australasian Institute for Maritime Archaeology (AIMA) Newsletter.

Unfortunately these Newsletters haven’t yet been transferred to the new AIMA website but thanks to the publications team at AIMA these updates are now available online via my page. The first dates from 2007 and provides an introduction to the project. The second is a halfway point update from 2008. The final one from 2009 provides a summary of the results, including background on the animation, the pollen diagram and radiocarbon date created as a spin off for the project, and a brief discussion on the validity of the methodology and it’s potential for use elsewhere in Australia. I hope they are of interest for those looking for references from this project whilst I try to find some time to finish the journal publication.

Just to get you in the mood, here is one of the 19th century images that helped inform and inspire our animation.



Merri Creek, Plenty Ranges 1864 from The Melbourne Album, published by Charles Troedel, Melbourne, 1863–64. National Gallery of Victoria, Melbourne. Online




Submerged Landscapes of Port Phillip Bay – The Animation

The following information was formerly hosted online by Heritage Victoria, Department of Planning and Community Development, but is no longer available. The information has been slightly edited for clarity.


The Submerged Landscapes of Port Phillip Bay project used published sources, and previously collected raw data to investigate the potential for the survival of ancient land surfaces beneath modern marine sediment and sea water in Port Phillip Bay. Based on sub-bottom profiler data, pollen data, archaeological and historical evidence it was possible to reconstruct the drowned landscape and ancient environment of Port Phillip Bay area around 10,000 years ago, before the Bay was flooded by rising sea levels. This project was the first of its kind in Australia, and demonstrated the value of this multi-disciplinary approach to investigating submerged ancient landscapes in Australian environments.

The landscape reconstruction aimed to provide an idea of what the ancient landscape might have looked like, and uses topographic information from the seismic data and vegetation information from pollen data. The activities of the people represented are based on a combination of archaeological information and historical illustrations.

The research was funded by Heritage Victoria through Victoria’s Heritage Strategy.

How do we know the Bay was dry land?

Since the earliest occupation of Australia, around 60,000BP (Flood 1994: 1), global changes in climate have had dramatic effects on sea levels, exposing and covering vast areas of land. Periods of intense cold (Ice Ages) have frozen sea water in glaciers and ice sheets, resulting in sea levels dropping up to 150m below present levels (Chappell et al 1996).

The last time the sea level dropped on this scale was at the peak of the last Ice Age around 20,000 years ago. It is estimated that Tasmania and the Bass Strait islands was joined to mainland Australia until around 14,000 years ago, when the sea level was approximately 50m below present levels (Lambeck & Chappell 2001). Using current seabed depths as a guide, it is thought that the coastline 14,000 years ago might have been about 7km south of Point Nepean, with the Yarra River running through the area now known as Port Phillip Bay. The area beyond Port Phillip Heads may have been like a river delta, as the freshwater ran southwards to meet the sea, but the effect that the narrow gorge at the Heads had on the water flow is uncertain. It is possible that a waterfall could have formed due to the height difference between the Bay and the land beyond the Heads. Without the assessment of additional geological data from this area, it is not possible to draw definite conclusions. It is estimated that Port Phillip Bay was flooded by rising sea levels around 8000 years ago (Holdgate et al 2001).

How do we know what the landscape was like?

Sub-bottom profile data provide a image like a slice (profile) through the seabed, and show us the layers of sediment beneath the sea. Data from Port Phillip Bay shows river valleys and later accumulation of sediments through time. Figure 1 shows one of the profiles through Port Phillip Bay where large valleys are visible, but now filled in with sediments.


Figure 1: Example of seismic sub-bottom profile data identifying the submerged and buried river channel in the northern part of Port Phillip Bay (Esso pipeline survey).

By combining position and depth information from a number of profiles, a 3D model  was created, shown in Figure 2, which provides an idea of what the topography would have been like prior to flooding. The sub-bottom profile survey lines are approximately 2-3km apart, making this model very rough, but figure 2 does suggest that the Bay area was generally low lying, with large shallow valleys up to 3km wide running through it in the northern part, and around 6 narrower, deeper valleys in the southern part.


Figure 2: A) The 3D representation of the river channels flowing through the central part of Port Phillip Bay and B) the location of the 3D reconstruction within the Bay (Source: Author)

The channels could not be traced any further south due to the type of seabed sediments, which prevented deeper penetration by the sub bottom profiler. It is thought that the smaller channels re-joined into one large channel, then flowed out through Port Phillip Heads in a deep channel or waterfall (Holdgate 1981: 128).

How do we know what trees & plants were here?

Generalisations about vegetation at the end of the last Ice Age suggest a thin and broken band of woodland along the eastern and south-eastern coast Australia, whilst the Bass Strait Islands were dominated by open scrub and heath, with some forested areas (D’Costa & Kershaw 1997). It has been suggested that the windy and dry climatic conditions of the period may have reduced woodlands to localised favourable sites, such as river valleys, and that grass and scrub covered much of the eastern coast (Adams & Faure 1997).

The vegetation reconstruction in this animation is based on pollen information collected from buried sediments in Port Phillip Bay. Sediments recovered within core samples contained pollen from ancient trees and plants. The pollen of each plant and tree type is unique and can be seen under a microscope, allowing the identification of plant and tree species that were alive in the area in the past.

Sediment and pollen from the top of the core is the most recent, whilst the sediment and pollen from the bottom of the core is the oldest. Core 7D, collected by Guy Holdgate in 1971 from the central part of Port Phillip Bay, was re-analysed and a new dates obtained. An oyster shell from 1.5m down core 7D was dated by radiocarbon to 6273 ± 36 BP (Wk-23494). The pollen from the lowest sample (3.75m) produced evidence of Sheoaks (Casuarinaceae), daisies (Asteraceae), a small amount of grass (Poaceae), saltbush (Chenopodiaceae) and some fern, moss and liverwort spores (Barbara Wagstaff pers comm.). As the pollen comes from a layer below the radiocarbon dated oyster shell, the pollen must be older than 6273 ± 36 BP.

How do we know that people were here?

Dreamtime stories passed down through the generations tell the stories of the Aboriginal groups who lived, and still live in the Port Phillip Bay area. The stories describe the formation of the landscape, the daily lives, relationships, customs and rituals of the people living here.

Archaeological excavations from around Port Phillip Bay have found evidence of people living in the area continuously from around 32,000 years ago. Sites at Keilor and Pakenham Lakeside provide information about daily life in temporary campsites, with hearths and stone tool working areas. Burials at Green Gully, Brimbank Park and on the Werribee River provide evidence of customs and rituals.

Historical records and traditional crafts can provide an indication of tool kits that might not survive archaeologically, such as basketry.

What are the people in the reconstructed campsite doing (Figure 3)?


Figure 3: Still image of the reconstructed campsite at the end of the animation (Source: Janet Saw and Hanna Steyne)

It is thought that the climate of the Melbourne area around 10,000 years ago would have been slightly damper and cooler than today. The only direct information we have about how people kept warm in this damper and colder climate are the archaeological remains of hearths, so we have shown one of these at the camp.Historical records and oral history from contemporary Traditional Owner groups, describe the local community using Possum skin cloaks to keep dry and warm, so we have dressed our people in these. In addition, a number of historical drawings and photographs show that different designs of bark huts were used as shelter, but it is possible that shelters could have been roofed with grasses, reeds or rushes or animal hides.

We have depicted people at the camp fishing and with a Kangaroo ready for butchering. This is based on both tools and animal bones that have been found archaeologically, and Aboriginal knowledge of food sources in the Port Phillip Area that has been passed down through the generations.

The animation

The animation can be found here on YouTube

The landscape visualisation and animation was created by Janet Saw and Tom Chandler at the Information Technology Unit (ITU) at Monash University. Through the combination of the DEM model and environmental information. It took around 120 hours to create, using four pieces of animation software; Autodesk 3ds Max, Forest Pack Pro 3, Adobe Illustrator, Adobe After Effects. The team converted digital elevation models, that were created using the seismic sub-bottom profile data, into a 3D mesh. Landscape colour, texture and vegetation was then added. The trees were mostly Casaurina and were added using random scattering algorithms. The limited budget did not allow for a full spectrum of accurate vegetation to be created, and therefore alternative, closest match models were used.



Adams J.M. & Faure H. 1997. (eds), QEN members. Review and Atlas of Palaeovegetation: Preliminary land ecosystem maps of the world since the Last Glacial Maximum. Oak Ridge National Laboratory, TN, USA.

Chappell et al. 1996. Reconciliation of late Quaternary sea levels derived from coral terraces at Huon peninsula with deep sea oxygen isotope records. Earth and Planetary Science Letters. 141: 227-236.

D’Costa, D.D. & Kershaw, P. 1997. An expanded Recent Pollen Database from South-eastern Australia and its Potential for refinement of Palaeoclimatic Estimates. Aust. J. Bot. 45: 583-605.

Flood, J. 1994. Archaeology of the Dreamtime. Marleston, South Australia: J.B. Publishing.

Holdgate, G.R., Thompson, B.R. & Guerin, B. 1981. Late Pleistocene Channels in Port Phillip. Proceedings of the Royal Society of Victoria. 92: 119-130

Holdgate et al. 2001. Marine geology of Port Phillip, Victoria. Australian Journal of Earth Sciences. 48: 429-455.

Lambeck, K. & Chappell, J. 2001. Sea Level Change Through the Last Glacial Cycle. Science. 292: 679-686.